chest1
1
Chest X-Ray - Basic Interpretation
by Robin Smithuis and Otto van Delden
The chest x-ray is the most frequently requested radiologic examination. In fact every radiologst should be an expert in chest film reading. The interpretation of a chest film requires the understanding of basic principles. In this article we will focus on: On the PA chest-film it is important to examine all the areas where the lung borders the diaphragm, the heart and other mediastinal structures. At these borders lung-soft tissue interfaces are seen resulting in a: These lines and silhouettes are useful localizers of disease, because they can be displaced or obscured with loss of the normal silhouette. This is called the silhouette sign, which we will discuss later. The paraspinal line may be displaced by a paravertebral abscess, hemorrhage due to a fracture or extravertebral extension of a neoplasm. Widening of the paratracheal line (> 2-3mm) may be due to lymphadenopathy, pleural thickening, hemorrhage or fluid overload and heart failure. Displacement of the para-aortic line can be due to elongation of the aorta, aneurysm, dissection and rupture. The anterior and posterior junction lines are formed where the upper lobes join anteriorly and posteriorly. These are usely not well seen and we will not discuss them. An important mediastinal-lung interface to look for is the azygoesophageal line or recess (arrow). The azygoesophageal recess is the region inferior to the level of the azygos vein arch in which the right lung forms an interface with the mediastinum between the heart anteriorly and vertebral column posteriorly. It is bordered on the left by the esophagus. Deviation of the azygoesophageal line is caused by (5): Notice the deviation of the azygoesophageal line on the PA-film. It is caused by a hiatal hernia. A common normal variant is the azygos lobe. The azygos lobe is created when a laterally displaced azygos vein makes a deep fissure in the upper part of the lung. On a chest film it is seen as a fine line that crosses the apex of the right lung. Here another patient with an azygos lobe. The azygos vein is seen as a thick structure within the azygos fissure. In some patients an extra joint is seen in the anterior part of the first rib at the point where the bone meets the calcified cartilageneous part (arrow). This may simulate a lung mass. In patients with a pectus excavatum the right heart border can be ill-defined, but this is normal. It produces a silhouette sign and thus simulating a consolidation or atelectasis of the right middle lobe. The lateral view is helpful in such cases. Pectus excavatum is a congenital deformity of the ribs and the sternum producing a concave appearance of the anterior chest wall. On a normal lateral view the contours of the heart are visible and the IVC is seen entering the right atrium. The retrosternal space should be radiolucent, since it only contains air. Any radiopacity in this area is suspective of a proces in the anterior mediastinum or upper lobes of the lung. As you go from superior to inferior over the vertebral bodies they should get darker, because usually there will be less soft tissue and more radiolucent lung tissue (red arrow). If this is not the case, look carefully for pathology in the lower lobes. The contours of the left and right diaphragm should be visible. The right diaphragm should be visible all the way to the anterior chest wall (red arrow). Actually we see the interface between the air in the lungs and the soft tissue structures in the abdomen. The left diaphragm can only be seen to a point where it borders the heart (blue arrow). Here the interface is lost, since the heart has the same density as the structures below the diaphragm. The left main pulmonary artery (in purple) passes over the left main bronchus and is higher than the right pulmonary artery (in blue) which passes in front of the right main bronchus. Once you know how the normal hilar structures look like on a lateral view, it is easier to detect abnormalities. In this case on the PA-view there is hilar enlargement. On the PA-view it is not clear whether this is due to dilated vessels or enlarged lymph nodes. On the lateral view there are round structures in areas where you don't expect any vessels. So we can conclude that we are dealing with enlarged lymph nodes. This patient has sarcoidosis. Notice also the widening of the paratracheal line (or stripe) as a result of enlarged lymph nodes. On the lateral view spondylosis may mimick a lung mass. Any density in the area of the vertebral bodies should lead you to the PA-film to look for spondylosis, which is usually located on the right side (arrows). On the left side the formation of osteophytes is hampered by the pulsations of the aorta. On the PA-view the superior mediastinum is widened. The lateral view is helpful in this case because it demonstrates a density in the retrosternal space. Now the differential diagnosis is limited to a mass in the anterior mediastinum (4 T's). This was a Hodgkins lymphoma. A common incidental finding in adults is a Bochdalek hernia, which is due to a congenital defect in the posterior diaphragm (arrows). In most cases it only contains retroperitoneal fat and is asymptomatic, but occasionally it may contain abdominal organs. Large hernias are sometimes seen in neonates and can be complicated by pulmonary hypoplasia. A hernia of Morgagni is also a congenital diaphragmatic hernia, but is less common. It is located anteriorly. Whenever you review a chest x-ray, always use a systematic approach. We use an inside-out approach from central to peripheral. First the heart figure is evaluated, followed by mediastinum and hili. Subsequently the lungs, lungborders and finally the chest wall and abdomen are examined. You have to know the normal anatomy and variants. Find subtle abnormalities by using the sihouette sign and mediastinal lines. Once you see an abnormality use a pattern approach to come up with the most likely diagnosis and differential diagnosis. It is extremely important to always compare with old films, as we will demonstrate in this case. Actually someone said that the most important radiograph is the old film, since it gives you so much information. For instance a lung mass, which hasn't changed in many years is not a lung cancer. First study the chest films. Then continue. Based on the CXR that you just saw, you could have made the diagnosis of congestive heart failure, but the findings are very subtle. However once you compare it to the old film, things become more obvious and you will be much more confident in your diagnosis: All these findings indicate the presence of heart failure. This is a very important sign. It enables us to find subtle pathology and to locate it within the chest. The loss of the normal silhouette of a structure is called the silhouette sign. Here an example to explain the silhouette sign: The heart is located anteriorly in the chest and it is bordered by the lingula of the left lung. The difference in density between the heart and the air in the lung enables us to see the silhouette of the left ventricle. When there is something in the lingula with the same 'water density' as the heart, the normal silhouette will be lost (blue arrow). When there is a pneumonia in the left lower lobe, which is located more posteriorly in the chest, the left ventricle will still be bordered by air in the lingula and we will still see the silhouette of the heart (red arrow). The PA-film shows a silhouette sign of the left heart border. Even without looking at the lateral film, we know, that the pathology must be located anteriorly in the left lung. This was a consolidation due to a pneumonia caused by Sterptococcus pneumoniae. Here we see a consolidation which is located in the left lower lobe. There is a normal silhouette of the left heart border. On this lateral film there is too much density over the lower part of the spine. By only looking at the interfaces of the left and right diaphragm on the lateral film, it is possible to tell on which side the pathology is located. First study the lateral film. Then continue. On a normal lateral chest film the silhouette of the left diaphragm 2- can be seen from posterior up to where it is bordered by the heart, which has the same density (blue arrow). One should be able to follow the contour of the right diaphragm -1- from posterior all the way to anterior, because it is only bordered by the lung. Here we cannot follow the contour of the right diaphragm all the way to posterior, which indicates that there is something of water-density in the right lower lobe (red arrow). On the PA-film there is a normal silhouette of the heart border, so the pathology is not in the anterior part of the chest, which we already suspected by studying the lateral view. Why do we still see the silhouette of the right diaphragm on the PA-film? What we see is actually the highest point of the right diaphragm, which is anterior to the pneumonia in the right lower lobe. The pneumonia does not border the highest point of the diaphragm. There are some areas that need special attention, because pathology in these areas can easily be overlooked: These areas are also known as the hidden areas. Notice that there is quite some lung volume below the dome of the diaphragm, which will need your attention (arrow). Here an example of a large lesion in the right lower lobe, which is difficult to detect on the PA-film, unless when you give special attention to the hidden areas. Here a pneumonia which was hidden in the right lower lobe mainly below the level of the dome of the diaphragm (red arrow). Notice the increase in density on the lateral film in the lower vertebral region. You may have to enlarge the image to get a better view. First study the CXR. Notice the subtle increased density in the area behind the heart that needs special attention (blue arrow). This was a lower lobe pneumonia. First study the CXR. We know that in some cases there is an extra joint in the anterior part of the first rib which may simulate a mass. However this is also a hidden area where it can be difficult to detect a mass. In this case a small lung cancer is seen behind the left first rib. Notice that is is also seen on the lateral view in the retrosternal area. Continue with the PET-CT. The PET-CT demonstrates the tumor (arrow) which has already spread to the bone and liver. The diagnosis was made by a biopsy of an osteeolytic metastasis in the iliac bone. First study the CXRs. There is a subtle consolidation in the left lower lobe in the hidden area behind the heart. Again there is increased density over the lower vertrebral region. On a chest film only the outer contours of the heart are seen. In many cases we can only tell whether the heart figure is normal or enlarged and it will be difficult to say anything about the different heart compartments. However it can be helpful to know where the different compartments are situated. Left Atrium Right Atrium Left Ventricle Right Ventricle Left Atrium Left Ventricle Right Ventricle Left Atrium enlargement This is a patient with longstanding mitral valve disease and mitral valve replacement. Extreme dilatation of the left atrium has resulted in bulging of the contours (blue and black arrows). Right ventricle enlargement First study the PA and lateral chest film and then continue reading. On these chest films the heart is extremely dilated. Notice that it is especially the right ventricle that is dilated. This is well seen on the lateral film (yellow arrow). There is a small aortic knob (blue arrow), while the pulmonary trunk and the right lower pulmonary artery are dilated. All these findings are probably the result of a left-to-right shunt with subsequent development of pulmonary hypertension. The location of the cardiac valves is best determined on the lateral radiograph. A line is drawn on the lateral radiograph from the carina to the cardiac apex. The pulmonic and aortic valves generally sit above this line and the tricuspid and mitral valves sit below this line (4). On this lateral view you can get a good impression of the enlargement of the left atrium. On the right side of the chest the lung will lie against the anterior chest wall. On the left however the inferior part of the lung may not reach the anterior chest wall, since the heart or pericardial fat or effusion is situated there. This causes a density on the anteroinferior side on the lateral view which can have many forms. It is a normal finding, which can be seen on many chest x-rays and should not be mistaken for pathology in the lingula or middle lobe. The explanation for the cardiac incisura is seen on this CT-image. At the level of the inferior part of the heart we can appreciate that the lower lobe of the right lung is seen more anteriorly compared to the left lower lobe. Pacemaker There are different types of cardiac pacemakers. Here we see a pacemaker with one lead in the right atrium and another in the right ventricle. A third lead is seen, which is guided through the coronary sinus towards the left ventricle. This is done in patients with asynchrone ventricular contractions. Pacing both ventricles at the same time will lead to synchrone contractions and a better cardiac output. Whenever we encounter a large heart figure, we should always be aware of the possibility of pericardial effusion simulating a large heart. On the chest x-ray it looks as if this patient has a dilated heart while on the CT it is clear, that it is the pericardial effusion that is responsible for the enlarged heart figure. Especially in patients who had recent cardiac surgery an enlargement of the heart figure can indicate pericardial bleeding. This patient had a change in the heart configuration and pericardial bleeding was suspected. Ultrasound demonstrated only a minimal pericardial effusion. Continue with the CT. There is a large pericardial effusion, which is located posteriorly to the left ventricle (blue arrow). The left ventricle id filled with contrast and is compressed (red arrow). At surgery a large hematoma in the posterior part of the pericardium was found. Notice that on the anterior side there is only a minimal collection of pericardial fluid, which explains why the ultrasound examination underestimated the amount of pericardial fluid. Here another patient who had valve-replacement. Notice the large heart size. There is redistribution of the pulmonary vessels which indicates heart failure. Continue with the CT. The CT-image shows a large pericardial effusion. Always compare these post-operative chest films with the pre-operative ones. Detection of calcifications within the heart is quite common. The most common are coronary artery calcifications and valve calcifications. Here we see pericardial calcifications which can be associated with constrictive pericarditis. In this case there are calcifications that look like pericardial calcifications, but these are myocardial calcifications in an infarcted area of the left ventricle. Notice that they follow the contour of the left ventricle. Pericardial fat depositions are common. Sometimes a large fat pad can be seen (figure). Necrosis of the fat pad has pathologic features similar to fat necrosis in epiploic appendagitis. It is an uncommon benign condition, that manifests as acute pleuritic chest pain in previously healthy persons (10). Pericardial cysts are connected to the pericardium and usually contain clear fluid. The majority of pericardial cysts arise in the anterior cardiophrenic angle, more frequently on the rightside, but they can be seen as high as the pericardial recesses at the level of the proximal aorta and pulmonary arteries (11). Most patients are asymptomatic. On the chest x-ray it seems as if there is a elevated left hemidiaphragm. On CT however there is a cyst connected to the pericardium. The normal hilar shadow is for 99% composed of vessels - pulmonary arteries and to a lesser extent veins (1). The vessel margins are smooth and the vessels have branches. The left pulmonary artery runs over the left main bronchus, while the right pulmonary artery runs in front of the right main bronchus, which is usually lower in position than the left main bronchus. Hence the left hilum is higher than the right. Only in a minority of cases the right hilus is at the same level as the left, but never higher. In this illustration the lower lobe arteries are coloured blue because they contain oxygen-poor blood. They have a more vertical orientation, while the pulmonary veins run more horizontally towards the left atrium, which is located below the level of the main pulmonary arteries. Both pulmonary arteries and veins can be identified on a lateral view and should not be mistaken for lymphadenopathy. Sometimes the pulmonary veins can be very prominent. The left main pulmonary artery passes over the left main bronchus and is higher than the right pulmonary artery which passes in front of the right main bronchus. These images are thick slab sagittal reconstructions of a chest-ct to get a better view of the hilar structures. The lower lobe pulmonary arteries extend inferiorly from the hilum. They are described as little fingers, because each has the size of a little finger (1). On the right side the little finger will be visible in 94% of normal CXRs and on the left side in 62% of normals (1). Study the CXR of a 70-year old male who fell from the stairs and has severe pain on the right flank.. Notice on the PA-film the absence of the little finger on the right and on the lateral view the increased density over the lower vertebral column. What is your diagnosis? There is a right lower lobe atelectasis. Notice the abnormal right border of the heart. The right interlobar artery is not visible, because it is not surrounded by aerated lung but by the collapsed lower lobe, which is adjacent to the right atrium. On a follow-up chest film the atelectasis has resolved. We assume that the atelectasis was a result of post-traumatic poor ventilation with mucus plugging. Notice the reappearance of the right little finger (red arrow) and the normal right heart border (blue arrow). The table summarizes the causes of hilar enlargement. Normal hili are: Enlargement of the hili is usually due to lymphadenopathy or enlarged vessels. In this case there is an enlarged hilar shadow on both sides. This could be the result of enlarged vessels or enlarged lymph nodes. A very helpful finding in this case is the mass on the right of the trachea. This is known as the 1-2-3 sign in sarcoidosis, i.e. enlargement of left hilum, right hilum and paratracheal. Here some more examples of sarcoidosis. Click to enlarge. Mediastinal masses are discussed in more detail in the article 'Mediastinal masses'. Here is just a brief overview. The mediastinum can be divided into an anterior, middle and posterior compartment, each with it's own pathology. Mediastinal lines or stripes are interfaces between the soft tissue of mediastinal structures and the lung. Displacement of these lines is helpful in finding mediastinal pathology, as we have discussed above. The most important mediastinal line to look for is the azygoesophageal line, which borders the azygoesophageal recess. This line is visible on most frontal CXRs. The causes of displacement of this line are summarized in the table. A hiatal hernia is the most common cause of displacement of the azygoesophageal line. Notice the air within the hernia on the lateral view. Another common cause of displacement of the azygoesophageal line is subcarinal lymphadenopathy. Notice the displacement of the upper part of the azygoesophageal line on the chest x-ray in the area below the carina. This is the result of massive lymphadenopathy in the subcarinal region (station 7). There are also nodes on the right of the trachea displacing the right paratracheal line. On the PET we can appreciate the massive lymphadenopathy far better than on the CXR. There are also lymphomas in the neck. this is an important finding, since these nodes are accessible for biopsy. Continue with images of CT and ultrasound. Here we see a CT-image. The azygoesophageal recess is displaced by lymph nodes that compress the left atrium. The final diagnosis of small cel lungcancer was made through a biopsy of a lymphnode in the neck. First study the chest x-ray. Then continue reading. Notice the following: Here we have a prior CXR of this patient. The AP-film shows a right paratracheal mass. The azygoesophageal recess is not identified, because it is displaced and parallels the border of the right atrium. The large round density in the left lung is the result of aspiration. Notice the massive dilatation of the esophagus on the CT. The aortopulmonary window is the interface below the aorta and above the pulmonary trunk and is concave or straight laterally. Here the AP-window is convex laterally due to a mass that fills the retrosternal space on the lateral view. On the CT-images a mass in the anterior mediastinum is seen. Final diagnosis: Hodgkins lymphoma. Here another case. On the PA-film a mass is seen that fills the aortopulmonary window. The PET better demonstrates the extent of the lymphnode metastases in this patient. Final diagnosis: small cell lungcarcinoma. Lung abnormalities mostly present as areas of increaseddensity, which can be divided into the following patterns: Less frequently areas of decreased density are seen as in emphysema or lungcysts. These lungpatterns are discussed in more detail in the article 'Chest X-Ray: Patterns of Lung Disease'. Consolidation is discussed the article 'Chest X-Ray: Patterns of Lung Disease'. Atelectasis is discussed in the article 'Chest X-Ray: Patterns of Lung Disease'. Nodules and masses are discussed in the article 'Chest X-ray: Patterns of Lung Disease'. Solitary pulmonary node - SPN is discussed in the article 'Solitary pulmonary nodule: benign versus malignant'. Click on the table to enlarge. Interstitial lung diseases are discussed in the article 'Chest X-Ray - Patterns of Lung Disease'. It takes about 200-300 ml of fluid before it comes visible on an CXR (figure). About 5 liters of pleural fluid are present when there is total opacification of the hemithorax. Total opacification of the right hemithorax in a patient with pleuritis carcinomatosa on both sides. On the right there is only some air visible in the major bronchi creating an air bronchogram within the compressed lung. Pleural fluid may become encysted. Here we see fluid entrapped within the fissure. This can sometimes give the impression of a mass and is called 'vanishing tumor'. The table lists the most common causes of a pneumothorax. The other cystic lungdisease which causes pneumothorax is Langerhans cell histiocytosis (LCH) which is seen in smokers. Study the CXR. There are two important findings. The retracted visceral pleura is seen (blue arrow) which indicates that there is a pneumothorax. There is a horizontal line visible (yellow arrow). Normally there are no straight lines in the human body unless when there is an air-fluid level. This means that there is a hydro-pneumothorax. When a pneumothorax is small, this air-fluid level can be the only key to the diagnosis of a pneumothorax. Study the CXR. There are 3 important findings. Notice that the mediastinum is slightly displaced to the left. Does this mean that there is a tension pneumothorax? Do you have an idea about the cause of the pneumothorax? There is a hydropneumothorax. Notice the air-fluid level (blue arrow). The upper lobe is still attached to the chest wall by adhesions. Maybe this patient was treated for a prior pneumothorax. There is a lung cyst in the upper lobe (red arrow). So we can assume that the pneumothorax has something to do with a cystic lung disease. Since this patient is a woman, lymphangioleiomyomatosis (LAM) is a possible diagnosis. LAM is a rare lung disease that results in a proliferation of smooth muscle throughout the lungs resulting in the obstruction of small airways leading to pulmonary cyst formation and pneumothorax. LAM also occurs in patients who have tuberous sclerosis. Study the CXR. What is your diagnosis? This is not a pneumothorax but a skin fold. The radiography was performed supine with a CR cassette inserted underneath the patient, which resulted in a skinfold. Notice that there are lung markings beyond the apparent pneumothorax. Here two CXRs of another patient with obvious skinfolds. Recognition of a pneumothorax depends on the volume of air in the pleural space and the position of the body. On a supine radiograph a pneumothorax can be subtle and approximately 30% of pneumothoraces are undetected. A sign to look for is the 'deep sulcus sign'. It represents lucency of the lateral costophrenic angle extending toward the hypochondrium (Figure). The image is of a patient in the ICU who is on mechanical ventilation. There was an acute exacerbation of the dyspnoe. There is a deep sulcus sign on the left. Notice that the left hemidiaphragm is depressed. This is an important finding since it indicates a tension pneumothorax. The image on the right is after insertion of an intercostal drain. Notice that the diaphragm has regained its normal appearance. The table lists the most common causes of pleural opacities. Pleural plaques The CXR shows multiple opacities. They have irregular shapes and do not look like a lung masses or consolidations. Some of these opacities are clearly bordering the chest wall (red arrows). All these findings indicate that we are dealing asbestos related pleural plaques. Asbestos related pleural plaques are usually: Unilateral pleural calcifications are usually due to: Pleural hematoma These images are of a patient, who had a pleural opacity after a chest trauma. It was believed to be a hematoma and resolved spontaneously. Ribfractures The most common identified chest wall abnormalities are old ribfractures. The CXR shows many rib deformities due to old fracturees. When a rib fracture heals, the callus formation may create a mass-like appearance (blue arrow). Sometimes a CT is necessary to differentiate a healing fracture from a lung mass. Notice the large lung volume and the enlarged pulmonary vessels. Probably we are dealing with pulmonary arterial hypertension in a patient with COPD. The second most common chest wall abnormalities that we see on a CXR are metastases in vertebral bodies and ribs. Notice the expansile mass in the posterior rib on the right. The most obvious finding on this CXR is free air under the diaphragm. This finding indicates a bowel perforation, unless when the patient had recent abdominal surgery and there is still some air left in the abdomen, which can stay there for several days. There is another subtle finding in the left upper lobe. A subtle density projecting over the first rib - hidden area - proved to be a lungcarcinoma. Here another patient with free abdominal air. Notice the very thin regular line which is the diaphragm (arrow). At first impression one might think that this is just some plate-like atelectasis due to poor inspiration. by Gerald de Lacey et al. Introduction to chest radiology by David M. Hansell et al Radiology 2008;246:697 by Jerry M. Gibbs et al RadioGraphics 2007;27:33-48 on RadDaily.com by Camilla R. Whitten May 2007 RadioGraphics, 27,657-671. Radiology 2003; 228:415-416 by James C. Reed by Richard Webb and Charles Higgins Víctor Pineda et al. January 2007 RadioGraphics,27, 19-32. By Mi-Young Jeung, et al. October 2002 RadioGraphics,22, S79-S93.Robin Smithuis and Otto van Delden PA view Vena azygos lobe Pectus excavatum Lateral view Old films Silhouette sign Hidden areas Cardiac incisura Pericardial effusion Calcifications Pericardial fatpad Pericardial cyst Hilar enlargement Mediastinal lines Azygoesophageal recess Aortopulmonary window Consolidation Atelectasis Nodule - Masses Interstitial pattern Pleural fluid Pneumothorax Pleural opacitiesChest X-Ray - Basic InterpretationRadiology Department of the Rijnland Hospital, Leiderdorp and the Academical Medical Centre, Amsterdam, the Netherlands
chest2
1
Chest X-Ray - Heart Failure
by Simone Cremers, Jennifer Bradshaw and Freek Herfkens
In this article we will discuss the radiographic signs of congestive heart failure on the chest X-ray. by Simone Cremers, Jennifer Bradshaw and Freek Herfkens Congestive heart failure (CHF) is the result of insufficient output because of cardiac failure, high resistance in the circulation or fluid overload. Left ventricle (LV) failure is the most common and results in decreased cardiac output and increased pulmonary venous pressure. In the lungs LV failure will lead to dilatation of pulmonary vessels, leakage of fluid into the interstitium and the pleural space and finally into the alveoli resulting in pulmonary edema. Right ventricle (RV) failure is usually the result of long standing LV failure or pulmonary disease and causes increased systemic venous pressure resulting in edema in dependent tissues and abdominal viscera. In the illustration on the left some of the features, that can be seen on a chest-film in a patient with CHF. Increased pulmonary venous pressure is related to the pulmonary capillary wedge pressure (PCWP) and can be graded into stages, each with its own radiographic features on the chest film (Table). This grading system provides a logical sequence of signs in congestive heart failure. In daily clinical practice however some of these features are not seen in this sequence and sometimes may not be present at all. This can be seen in patients with chronic heart failure, mitral valve disease and in chronic obstructive lung disease. In a normal chest film with the patient standing erect, the pulmonary vessels supplying the upper lung fields are smaller and fewer in number than those supplying the lung bases. The pulmonary vascular bed has a significant reserve capacity and recruitment may open previously non-perfused vessels and causes distension of already perfused vessels. This results in redistribution of pulmonary blood flow. First there is equalisation of blood flow and subsequently redistribution of flow from the lower to the upper lobes. The term redistribution applies to chest x-rays taken in full inspiration in the erect position. In daily clinical practice many chest films are taken in a supine or semi-erect position and the gravitational difference between the apex and the lung bases will be less. In the supine position, there will be equalisation of blood flow, which may give the false impression of redistribution. In these cases comparison with old fims can be helpful. Artery-to-bronchus ratio Normally the vessels in the upper lobes are smaller than the accompanying bronchus with a ratio of 0.85 (3). At the level of the hilum they are equal and in the lower lobes the arteries are larger with a ratio of 1.35. When there is redistribution of pulmonary blood flow there will be an increased artery-to-bronchus ratio in the upper and middle lobes. This is best visible in the perihilar region. On the left a patient with cardiomegaly and redistribution. The upper lobe vessels have a diameter > 3 mm (normal 1-2 mm). Notice the increased artery-to-bronchus ratio at hilar level (arrows). Stage II of CHF is characterized by fluid leakage into the interlobular and peribronchial interstitium as a result of the increased pressure in the capillaries. When fluid leaks into the peripheral interlobular septa it is seen as Kerley B or septal lines. Kerley-B lines are seen as peripheral short 1-2 cm horizontal lines near the costophrenic angles. These lines run perpendicular to the pleura. When fluid leaks into the peribronchovascular interstitium it is seen as thickening of the bronchial walls (peribronchial cuffing) and as loss of definition of these vessels (perihilar haze). On the left a patient with congestive heart failure. There is an increase in the caliber of the pulmonary vessels and they have lost their definition because they are surrounded by edema. On the left another patient with congestive heart failure. The lateral view nicely demonstrates the increased diameter of the pulmonary vessels and the hazy contours. Notice also the septal lines and the accentuated interstitium. Furthermore the fissura major is markedly thickened. CT will also demonstrate signs of congestive heart failure. On the image on the left notice the following: In a patient with a known malignancy lymphangitic carcinomatosis would be high in the differential diagnostic list. Ground glass opacity is the first presentation of alveolar edema and a precursor of consolidation. This stage is characterized by continued fluid leakage into the interstitium, which cannot be compensated by lymphatic drainage. This eventually leads to fluid leakage in the alveoli (alveolar edema) and to leakage into the pleural space (pleural effusion). The distribution of the alveolar edema can be influenced by: On the left a patient who was admitted with severe dyspnoe due to acute heart failure. The following signs indicate heart failure: alveolar edema with perihilar consolidations and air bronchograms (yellow arrows); pleural fluid (blue arrow); prominent azygos vein and increased width of the vascular pedicle (red arrow) and an enlarged cardiac silhouette (arrow heads). After treatment we can still see an enlarged cardiac silhouette, pleural fluid and redistribution of the pulmonary blood flow, but the edema has resolved. On the left another patient with alveolar edema at admission, which resolved after treatment. When you scroll through the images and go back and forth, you will notice the difference in vascular pedicle width and distribution of pulmonary flow. Both on the chest x-ray and on the CT the edema is gravity dependent and differences in density can be measured. Notice that even within each lobe there is a gravity dependent difference in density. This is only seen when the consolidations are the result of transudate like in CHF. This is not seen when the consolidations are the result of exsudate due to infection, blood due to hemorrhage or when there is a capillary leak like in ARDS. On the left a patient who first had a chest film in a supine position. Notice the pulmonary edema, which is almost exclusively seen in the right lung. A possible explanation for this phenomenon could be, that the patient had been lying on his right side for a while before the x-ray was taken. The cardiothoracic ratio (CTR) is the ratio of the transverse diameter of the heart to the internal diameter of the chest at its widest point just above the dome of the diaphragm as measured on a PA chest film. An increased cardiac silhouette is almost always the result of cardiomegaly, but occasionally it is due to pericardial effusion or even fat deposition. The heart size is considered too large when the CTR is > 50% on a PA chest x-ray. A CTR of > 50% has a sensitivity of 50% for CHF and a specificity of 75-80%. An increase in left ventricular volume of at least 66% is necessary before it is noticeable on a chest x-ray. On the left a patient with CHF. There is an increase in heart size compared to the old film. Other signs of CHF are visible, such as redistribution of pulmonary flow, interstitial edema and some pleural fluid. On a supine film the cardiac silhouette will be larger due to magnification and high position of the hemidiafragms. Exact measurements are not that helpful, but comparison to old supine films can be of value. On the left a patient, who recently underwent a valve replacement. There is a large cardiac silhouette, which could be the result of cardiomegaly. Because of the recent cardiac surgery, the possibility of pericardial effusion was taken into account, which is nicely demonstrated on the CT-image. On the left another patient with a large cardiac silhouette on the chest x-ray due to pericardial effusion. Pericardial effusion is demonstrated on the coronal CT-reconstruction. Pleural effusion is bilateral in 70% of cases of CHF. When unilateral, it is slightly more often on the right side than on the left side. There has to be at least 175 ml of pleural fluid, before it will be visible on a PA image as a meniscus in the costophrenic angle. On a lateral image effusion of > 75 ml can be visible. If pleural effusion is seen on a supine chest film, it means that there is at least 500 ml present. On the left images of a patient who has bilateral pleural effusions. Notice that it is more evident on the lateral view. Pleural effusion is not always visible as a meniscus in the costophrenic angle. A subpulmonic effusion may follow the contour of the diaphragm making it tricky to discern. In these cases, the only way to detect pleural effusion, is when you notice that there is an increased distance between the stomach bubble and the lung. The stomach is normally located directly under the diaphragm, so, on an erect PA radiograph, the stomach bubble should always appear in close proximity to the diaphragm and the lung. On the left images of a patient with signs of CHF. At first glance you might get the impression that there is a high position of the diaphragm. However when you notice the increased distance of the stomach air bubble to the lung base, you realize that there is a large amount of pleural fluid on both sides (arrow). The vascular pedicle is bordered on the right by the superior vena cava and on the left by the left subclavian artery origin (6). The vascular pedicle is an indicator of the intravascular volume. A vascular pedicle width less than 60 mm on a PA chest radiograph is seen in 90% of normal chest x-rays. A vascular pedicle width of more than 85 mm is pathologic in 80% of cases. 5 mm increase in diameter corresponds to 1 liter increase of intravascular fluid. An increase in width of the vascular pedicle is accompanied by an increased width of the azygos vein. There are three principal varieties of pulmonary edema: cardiac, overhydration and increased capillary permeability (ARDS). The vascular pedicle width (VPW) can help in differentiating these different forms of pulmonary edema (6):  On the left a patient with ARDS. There is alveolar edema in both lungs. Notice that the VPW is normal. The vessels in the upper lobes are not dilated and the cardiac silhouette is not enlarged. The VPW is best used as a measure to compare serial chest x-rays of the same patient, as there is a wide range of values for the VPW. The VPW may increase due to rotation to the right. On an AP-view the VPW will increase 20% compared to a PA-view. On the left a patient with subtle signs of congestive heart failure on the initial chest x-ray (image 1/2). There is a slightly enlarged vascular pedicle, which becomes more obvious when you compare to the chest film after diuretic therapy (image 2/2). Dilation of the azygos vein is a sign of increased right atrial pressure and is usually seen when there is also an increase in the width of the vascular pedicle. The diameter of the azygos vein varies according to the positioning. In the standing position a diameter > 7 mm is most likely abnormal and a diameter > 10 mm is definitely abnormal. In a supine patient > 15 mm is abnormal. An increase of 3 mm in comparison to previous films is suggestive of fluid overload. The difference of the azygos diameter on an inspiration film compared to an expiration film is only 1mm. This means that the diameter of the azygos is a valuable tool whether or not there is good inspiration. RV failure is most commonly caused by longstanding LV failure, which increases the pulmonary venous pressure and leads to pulmonary arterial hypertension, thus overloading the RV. Other less common causes of RV failure are: Radiographic signs of RV failure: Sonographic signs of RV failure: The indication for ultrasound examination in many of these patients is abnormal liver function tests. It is therefore important to consider the possibility of RV failure when a patient presents with liver enzyme abnormalities. Under normal conditions dynamic ultrasound will demonstrate changes in caliber of the IVC. These changes in caliber can be attributed to variations in blood flow in the IVC in accordance with the respiratory and cardiac cycles. in Merck manual by Carle Ravin Presented as part of a Conference on Chest Radiology 1982 by J H Woodring April 1991 Radiology, 179, 115-122. by Eric Milne et al American Journal of Roentgenology, Vol 144, Issue 5, 879-894 by M Pistolesi, E N Milne, M Miniati and C Giuntini July 1984 Radiology, 152, 9-17. by Au VW, Jones DN, Slavotinek JP. British Journal of Radiology. 2001 Jan; 74(877) 86-88. by M. Simon Circulation 1961; 24:185-190 by McHugh, T. J., Forrester, J. S., Adler, L., et al. Ann. Intern. Med., 1972; 76:29-33 Harrison M.O., Conte P.J., Heitzman E.R. Br J Radiol. 1971; 44:265-272 by Gamsu G,Kaufman K ,Swann S ,Brito A. Invest Radiol 1979 ; 14: 261-269 Bonomo L., et al. Radiologic Clinics of North America, vol. 46, nr 4, July 2008: 685-702Simone Cremers, Jennifer Bradshaw and Freek Herfkens Stage I - Redistribution Stage II - Interstitial edema Stage III - Alveolar edemaChest X-Ray - Heart FailureRadiology department of the Albert Schweitzer Hospital in Dordrecht and the Medical Centre Alkmaar, the Netherlands, the Netherlands
chest3
1
Chest X-Ray - Patterns of Lung disease
by Robin Smithuis and Otto van Delden
This article is still work in progress. When it is ready it will be announced in the newsletter. Lung diseases can present in various ways on a chest x-ray. Basically lung abnormalities will either present as areas of increased density, which are the most common or as areas of decreased density. In this article we will focus on a systematic pattern approach to lung diseases. The illustration summarizes the basic patterns of lung disease with increased density: Here we see chest x-ray examples of these patterns. You can click on the blue links below to go directly to the differential diagnosis. Consolidation is the result of replacement of air in the alveoli by transudate, pus, blood, cells or other substances. Pneumonia is by far the most common cause of consolidation. The key-findings on the X-ray are: An area of consolidation usually has ill-defined borders unless when it is bordered by a fissure, which will result in a sharp delineation, since consolidation will not cross a fissure. As the alveoli that surround the bronchi become more dense, the bronchi will become more visible, resulting in an air-bronchogram (figure). In consolidation there should be no or only minimal volume loss, which differentiates consolidation from atelectasis. Expansion of a consolidated lobe is not so common and seen in Klebsiella pneumoniae, Streptococcus pneumoniae, TB and lung cancer with obstructive pneumonia. The most common presentation of consolidation is lobar or segmental. The differential diagnostic list is long, but the most common diagnosis are listed in the table. On the chest x-ray there is an ill-defined area of increased density in the right upper lobe without volume loss. The right hilus is in a normal position. Notice the air-bronchogram (arrow). In the proper clinical setting this is most likely a lobar or segmental pneumonia. However if this patient had weight loss or long standing symptoms, we would include the list of causes of chronic consolidation. This was an acute lobar pneumonia caused by Streptcoccus pneumoniae. Based on the images alone, it can be difficult to determine the cause of the consolidation. Other things need to be considered, like acute or chronic illness, clinical data, other pulmonary and non-pulmonary findings. Here we have a number of x-rays with consolidation. Notice the similarity between these chest x-rays. The image on the right is almost identical to the previous image. In this case there was a solitary nodule in the right upper lobe and a biopsy was performed. The lobar consolidation is the result of hemorrhage as a complication of the procedure. The radiographic features of acute pulmonary thromboembolism are insensitive and nonspecific. The most common radiographic findings in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study were atelectasis and patchy pulmonary opacity. However in most cases of pulmonary emboli the chest x-ray is normal. This patient had pulmonary emboli, which were seen on a CTA. The peripheral consolidation is seen in the region of the emboli and can be attributed to hemorrhage in the infarcted area. The most common cause of diffuse consolidation is pulmonary edema due to heart failure. This is also called cardiogenic edema, to differentiate it from the various causes of non-cardiogenic edema. The increased heart size is usually what distinguishes between cardiogenic and non-cardiogenic. Some patients who suffer an acute cardiac infarction may still have a normal heart size, while some patients with a chronic heart disease may have non-cardiac pulmonary edema due to a superimposed pulmonay infection, ARDS, near-drowning etc. Usually it has a perihilar distribution, which is also called a batwing-pattern. In this patient there are many signs that indicate , that we are dealing with heart failure, like the increased interstitial markings, the large heart size and probably also the subtle increase of the vascular pedicle. Here another case of diffuse consolidation. This patient had fever and cough. Many micro-organismscan cause diffuse bronchopneumonia. This proved to be legionella pneumonia. The chest x-ray shows diffuse consolidation with 'white out' of the left lung with persistent air-bronchogram. This patient had a chronic disease with progressive consolidation. The disease started as a persitent consolidation in the left lung and finally spread to the right lung. This is typical for bronchoalveolar carcinoma. This is a difficult case. Based on the x-ray it is not sure whether we are dealing with masses or consolidation. Continue with the CT. The CT-image is not very helpful in the differentiation. There are hypodense areas, which could be masses. On the other hand this also could be areas of consolidation with hypodense areas due to necrosis. Finally the diagnosis non Hodgkin's disease was made based on biopsy. Batwing A perihilar distribution of consolidation is also called a Batwing distribution. The sparing of the periphery of the lung is attributed to a better lymphatic drainage in this area.It is most typical of pulmonary edema (both cardiogenic and non-cardiogenic), but is also seen in pneumonias. Reverse Batwing Peripheral or subpleural consolidation is called reverse Batwing distribution. It is rather speciphic for chronic lung disease. This is also described as multifocal ill-defined opacities or densities. In most cases these are the result of airspace-consolidations. In some cases however the underlying pathology is interstitial disease, like in the alveolar form of sarcoidosis in which the granulomas are very small and seem to fill up the alveoli. First study the chest x-ray. What are the findings and what is the differential diagnosis? Notice that there are multiple densities in both lungs. The larger ones are ill-defined and maybe there is an air-bronchogram in the right lower lobe. Probably we are dealing with multifocal consolidations, but one might also consider the possibility of multiple ill-defined masses. There is a peripheral distribution. This patient had several month history of chronic non-productive cough, that didn't respond to antibiotics. So we are dealing with the differential diagnosis of chronic consolidation, which we will discuss in a moment. The lab-findings were normal which makes bronchoalveolar carcinoma and lymphoma less likely. There was no eosinophilia, which excludes eosinophilic pneumonia. Biopsy revealed the diagnosis of organizing pneumonia (OP) also known as BOOP. It is important to differentiate between acute consolidation and chronic consolidation, because it will limit the differential diagnosis. In chronic disease we think of: On a chest x-ray it can be very difficult to determine whether there is interstitial lung disease. In many cases HRCT will be needed to determine if there is interstitial disease and what kind of pattern we are dealing with. The most common interstitial pattern is fine reticulation due to interstitial edema in heart failure and volume overload, that presents as fine septal lines. The image on the right shows interstitial edema in a patient with congestive heart failure. The old film is normal. Interstitial edema usually presents as fine reticulation. Sometimes Kerley B lines are visible. Kerley B lines are 1-2 cm long horizontal lines near the lateral pleura. The differential diagnosis of Kerley B lines is: Here another chest x-ray with interstitial edema and Kerley B lines. The CT shows the septal thickening. In this case of heart failure the reticulation is more coarse. In this case the chest x-ray shows subtle findings that could be described as fine reticulation. In many cases a HRCT is needed to determine the exact nature of the findings. The HRCT - not shown - demonstrated a fine nodular appearance as a result of sarcoidosis. Notice the subtle thickening of the minor fissure. ... ... Atelectasis or lung-collapse is the result of loss of air in a lung or part of the lung with subsequent volume loss due to airway obstruction or compression of the lung by pleural fluid or a pneumothorax. The key-findings on the X-ray are: Lobar atelectasis or lobar collaps is an important finding on a chest x-ray and has a limited differential diagnosis. The most common causes of atelectasis are: Sometimes lobar atelectasis produces only mild volume loss due to overinflation of the other lungparts. The illustration summarizes the findings of the different types of lobar atelectases. Right upper lobe atelectasis First study the images, then continue reading. Findings: On the PET-CT a lungneoplasm is seen with subsequent atelectasis of the right upper lobe due to obstruction of the upper lobe bronchus. A common finding in atelectasis of the right upper lobe is 'tenting' of the diafphragm (blue arrow). This patient had a centrally located lungcarcinoma with metastases in both lungs (red arrows). Right middle lobe atelectasis First study the x-rays and then continue reading. What are the findings? Usually right middle lobe atelectasis does not result in noticable elevation of the right diaphragm. A pectus excavatum (see article 'Chest X-Ray - Basic Interpretation') can mimick a middle lobe atelectasis on a frontal view, but the lateral view should solve this problem. Right lower lobe atelectasis Chest x-rays of a 70-year old male who fell from the stairs and has severe pain on the right flank. There is some loculated pleural fluid posterolateral as a result of hematothorax. What are the pulmonary findings? First study the images, then continue reading. There is a right lower lobe atelectasis. Notice the abnormal right border of the heart. The right interlobar artery is not visible, because it is not surrounded by aerated lung but by the collapsed lower lobe, which is adjacent to the right atrium. On a follow-up chest film the atelectasis has resolved. We assume that the atelectasis was a result of post-traumatic poor ventilation with mucus plugging. Notice the reappearance of the right interlobar artery (red arrow) and the normal right heart border (blue arrow). Left upper lobe atelectasis First study the x-rays, then continue reading. What are the findings? The CT-images demonstrate the atelectasis of the left upper lobe. There is a centrally located mass which obstructs the left upper lobe bronchus. This mass also causes the hilus overlay sign (red arrow). First study the x-rays then continue reading. What are the findings and what sign is seen here? There is an atelectasis of the left upper lobe. You would not expect the apical region to be this dark, but in fact this is caused by overinflation of the lower lobe, which causes the superior segment to creep all the way up to the apical region. This is called the luft sichel sign. Luft sichel means a sickle of air (blue arrow). Notice the bulging of the fissure on the lateral view. This is comparable to the golden-S sign in right upper lobe atelectasis and is suspective of a centrally obstructing mass. Left lower lobe atelectasis First study the x-rays then continue reading. Where is the abnormality located? There is a triangular density seen through the cardiac shadow. This must be an abnormality located posterior to the heart. This is confirmed on the lateral view. The contour of the left diaphragm is lost when you go from anterior to posterior. As the title suggests this is lower lobe atelectasis. We cannot see the lower lobe vessels, because they are surrounded by the atelectatic lobe. Normally when you follow the thoracic spine form top to bottom, the lower region becomes less opaque. Here we have the opposite (blue arrow). Study the images and then continue reading. There is a total collaps of the left upper lobe. Notice the high position of the left hilum. There is only a subtle band of density projecting behind the sternum. This is the collapsed upper lobe. In this case there is compensatory overinflation of the left lower lobe resulting in a normal position of the diaphragm and the mediastinum. Thechest x-ray shows total atelectasis of the right lung due to mucus plugging. Notice the displacement of the mediastinum to the right. Re-aeration on follow-up chest film after treatment with a suction catheter. The mediastinum has regained its normal position. A common cause of total atelectasis of a lung is a ventilation tube that is positioned too deep and thus obstructing one of the main bronchi. These images are of a patient who had widespread bronchopneumonia and was on ventilation. During follow up a white out on the left was seen. This was caused by a large mucus plug. After suction of the mucus plug the left lung was re-aerated. The chest x-ray shows a nearly total opacification of the left hemithorax. This patient was known to have pleuritic carcinomatosis. The left lung is almost completely compressed by the pleural fluid. Unlike most of the above cases, which were caused by obstruction, in this case the atelectasis is a result of compression. The compression of the lung by the loculated fluid collections is best seen on the CT-image (blue arrow). The CT-scan was performed, because the patient was suspected of having pulmonary emboli (red arrow). The typical findings of rounded atelectasis on CT are pleural thickening, pleural-based mass and comet tail sign. The theory is that a local pleuritis causes the pleura to thicken and contract. The underlying lung shrinks and atelectasis develops in a round configuration. The distorted vessels appear to be pulled into the mass and resemble a comet tail (4). First study the images and then continue reading. On the lateral view there is a mass-like lesion that is pleural-based. The first impresson is, that this is a pleural lesion. A CT was performed - see next images. The CT shows a lesion that originates in the lung. Many would have a lungcancer on the top of their differential diagnostic list. However there is also some pleural thickening (red arrow) and vessels seem to swirl around the mass (blue arrows). This is also described as the comet tail sign (4). Whenever you see a pleural-based lesion that looks like a lungcancer, also consider the possibility of rounded atelectasis. Rounded atelectasis is a benign lesion and when the findings are convincing, then biopsy is not needed. During follow up these lesions usually do not change in configuration. Rounded atelectasis is frequently seen in patients with a history of asbest exposure. The images show a density posteriorly in the left lower lobe. On the PA-film this looks like a mass or possibly a consolidation. On the lateral film however the boundaries seem to be sharp, which is in favor of a mass. Also notice that the pleura is thickened (red arrow). Although a peripheral lungcancer is on top of our list, we now also consider the possibility of rounded atelectasis. The CT-images show the typical features of a rounded atelectasis. There is an oval mass, pleural thickening and a comet tail sign (arrow). This lesion did not change in a two-year follow up. Platelike atelectasis is a common finding on chest x-rays and detected on an every day basis. They are characterized by linear shadows of increased density at the lung bases. They are usually horizontal, measure 1-3 mm in thickness and are only a few cm long. In most cases these findings have no clinical significance and are seen in smokers and elderly. They are seen in patients, that are in a poor condition and who breathe superficially, for instance after abdominal surgery (figure). Plate atelectasis is frequently seen in patients in the ICU and respond to increased ventilation. Platelike atelectasis is also frequently seen in pulmonary embolism, but since it is so non-specific, it is not a helpful sign in making the diagnosis of pulmonary embolism. Atelectasis can be the result of fibrosis of lungtissue. This is seen after radiotherapy and in chronic infection, especially TB. Here we have a patient who was treated with radiotherapy for lungcancer. Notice the increased density of the lung tissue and the volume loss. Here we have a patient with atelectasis of the right upper lobe as a result of TB. Notice the deviation of the trachea. There is also some atelectasis of the left upper lobe, which results in a high position of the left pulmonary artery as seen on the lateral view (red arrow) A solitary pulmonary nodule or SPN is defined as a discrete, well-marginated, rounded opacity less than or equal to 3 cm in diameter. It has to be completely surrounded by lung parenchyma, does not touch the hilum or mediastinum and is not associated with adenopathy, atelectasis or pleural effusion. The differential diagnosis of SPN is basically the same as of a mass except that the chance of malignancy increases with the size of a lesion. Lesions maller than 3 cm, i.e. SPN's are most commonly benign granulomas, while lesions larger than 3 cm are treated as malignancies until proven otherwise and are called masses. The table is adapted from the book: chest x-ray - a survival guide. The good thing about this list, is that it differentiates in common and uncommon diagnoses. There are many ways to differentiate a benign from a malignant SPN or mass as we have discussed in Solitary pulmonary nodule: benign versus malignant - Differentiation with CT and PET-CT In lesions that do not respond to antibiotics, probably the most important non-invasive diagnostic tool is nowadays the PET-CT. PET-CT can detect malignancy in focal pulmonary lesions of greater than 1 cm with a sensitivity of about 97% and a specificity of 78%. False-positive findings in the lung are seen in granulomatous disease and rheumatoid disease. False negatives are seen in low grade malignant tumors like carcinoid and alveolar cell carcinoma and lesions of less than 1 cm. The differential diagnostic list of multiple masses is very long. The most important diagnoses are listed in the table. Sometimes it is difficult to differentiate multifocal consolidations from masses and you have to add this list (see table on Multifocal Consolidation) to the possible diagnoses. Metastases Metastases are the most common cause of multiple pulmonary masses. Usually they vary in size and are well-defined. They predominate in the lower lobes and in the subpleural region. HRCT will demonstrate the random distribution (See 'Lung HRCT: Basic Interpretation') unlike other diseases that have a perilymphatic or centrilobular distribution. The images show a renal cell carcinoma that has invaded the inferior vena cava with subsequent spread of disease to the lungs. Here another patient with widespread pulmonary metastases of a cancer, that was located in the tongue. Mucoid impaction Mucus plugs or mucoid impaction can mimick the appearance of lung nodules. It is seen in bronchial obstruction (obstructing tumor or bronchial atresia), asthma, allergic bronchopulmonary aspergillosis and other causes of bronchiectasis. In this case there are some mass-like structures in the right lung. CT demonstrated bronchiectasis with mucoid impaction. A more common presentation of mucoid impaction in seen here. This is the typical 'finger-in-glove' appearance of mucoid impaction. The mucus in the dilated bronchi look like the fingers in a glove. The differential diagnostic list of consolidation is long. The table summarizes the most common diseases (Click to enlarge). A way to think of the differential diagnosis is to think of the possible content of the alveoli: Another way to think of consolidation, is to look at the pattern of distribution: Finally it is important to differentiate acute versus chronic disease. There are numerous interstitial lung diseases, but in clinical practice only about ten diseases account for approximately 90% of cases. The table shows the differential diagnosis of interstitial lung disease that we use in HRCT. by Richard Webb and Charles Higgins by James C. Reed by Gerald De Lacey, Simon Morley and Laurence Berman by Vince A. Partap November 1999 Radiology,213, 553-554. by Sudhakar N. J. Pipavath1 and J. David Godwin. AJR September 2008 vol. 191 no. 3 639-641. by Heber MacMahon et al. Radiology 2005; 237:395-400Robin Smithuis and Otto van Delden Pattern Approach Lobar consolidation Diffuse consolidation Multifocal ill-defined Acute vs chronic consolidation Fine reticulation Coarse reticulation Fine nodular Lobar atelectasis Total atelectasis Rounded atelectasis Platelike atelectasis Cicacitration atelectasis Solitary Pulmonary Nodule Multiple massesChest X-Ray - Patterns of Lung diseaseRadiology Department of the Rijnland Hospital, Leiderdorp and the Academical Medical Centre, Amsterdam, the Netherlands
chest4
1
Esophagus - Part I
by Terrence C. Demos, MD, Harold V. Posniak, MD, Wayde Nagamine, MD and Mary Olson, MD
In Esophagus part I we will discuss: Common structures that we can visualize are: The esophageal wall is composed of: At the gastroesophageal junction smooth, uniform folds in gastric fundus converge on very distal esophagus (arrow). Image next to it shows abnormal gastroesophageal junction: Barium outlines thick, irregular mucosal folds (asterisks). Fundal adenocarcinoma invades esophagus (arrows)   Spontaneous gastroesophageal reflux has been demonstrated in up to 1/3 of patients with reflux esophagitis. Various maneuvers during the examination have been used to increase sensitivity, but these are generally discredited as not being physiologic. In addition many asymptomatic patients have spontaneous reflux so that reflux during an esophagram is not sensitive or specific for relating symptoms to reflux. Normal: Abnormal: On the left tertiary contractions on first swallow (left). Normal primary contraction on next swallow (right). These tertiary contractions are non-propulsive, transient, and intermittent contractions that are inconstant in location and not accompanied by symptoms, usually in older patients. Sometimes transient tertiary contractions may simulate diverticula. On the left images of a patient with tertiary contractions, that during the examination look like diverticula. Diffuse esophageal spasm produces intermittent contractions of the mid and distal esophageal smooth muscle, associated with chest symptoms. Manometry shows simultaneous nonpropulsive contractions on at least 10% of swallows. Diagnosis is based on imaging, manometry, and symptoms. Nutcracker esophagus is a non-cardiac cause of chest pain attributed to high amplitude distal esophageal peristalsis. This is a controversial diagnosis that is made by manometry and does not have imaging manifestations  On the left another patient with achalasia. On the left another patient with achalasia. During fluoroscopy some peristalsis was seen with typical smooth, tapered narrowing just above diaphragm (arrows).  On the left a patient with a ring due to muscular contraction. Notice incidental gastric diverticulum (asterisk). On the left another patient with a non-persistent ring at the apex of a sliding hiatus hernia. The esophageal B-ring is located at the squamocolumnar junction, also termed the 'Z' line. The appearance does not change during the examination. On the left a patient with a 'B' ring (arrows) several cm above diaphragm at the apex of sliding hiatus hernia. Note unchanged appearance on these two images. On the left a 52-year-old man with episodic dysphagia. The image on the far left does not show a abnormality, but distal esophagus not distended . With dilation of the distal esophagus, a 13 mm wide Schatzki B-ring (arrows) that caused intermittent obstruction is demonstrated at the apex of a hiatus hernia (arrowhead). On the left a 71-year-old man with chest pain after fast food lunch. Distal obstructing filling defect (arrow) is a piece of meat that passed into stomach during study. Follow-up esophagram shows Schatzki B-ring (arrows) that caused obstruction.  On the left images of an asymptomatic 52-year-old man. AP and Lateral views show short, thin web (arrows) with minimal intraluminal extension. On the left images of a 42-year-old woman with dysphagia due to web. There is > 50% luminal narrowing Pulsion diverticula are due to increased intraluminal pressure. There are many pulsion diverticula: On the left a patient with a Zenker's diverticulum as a result of premature closure of the cricopharyngeal muscle. Traction diverticula are secondary to adjacent disease. Most located in mid-esophagus. A Zenker's diverticulum is a pulsion hypopharyngeal false diverticulum with only mucosa and submucosa protruding through triangular posterior wall weak site (Killian's dehiscence) between horizontal and oblique components of cricopharyngeus muscle. The etiology is controversial and is probably due to elevated upper esophageal pressure, cricopharyngeus dysfunction and reflux. The clinical presentation can be dysphagia, regurgitation, aspiration or a mass or air-fluid level on neck or chest radiographs. The esophagram shows collection with midline posterior origin just above cricopharyngeus protruding lateral, usually to left, and caudal with enlargement. Killian-Jamieson diverticulum is a pulsion diverticulum, that protrudes through a lateral anatomic weak site of the cervical esophagus below the cricopharyngeus muscle, unlike the posterior, midline origin of a Zenker's diverticulum. AP view shows diverticulum (arrow) originating laterally. Lateral view confirms diverticulum does not originate posteriorly as a Zenkers diverticulum would. Epiphrenic diverticulum These pulsion diverticula are classified by their location near the diaphragm. /ul> If large they can narrow the esophagus or lead to aspiration. On the left another example of an epiphrenic diverticulum. The CT demonstrates a large diverticulum (arrow) extends to the right just above diaphragm. This patient was asymptomatic Aortopulmonary window diverticulum The normal esophagus transiently protrudes into the aortopulmonary window. Fixed protrusion is an inconsequential diverticulum. On the left small aortopulmonary diverticula (arrows), that are incidental findings in two patients. On the far left a traction diverticulum (arrow) due to hilar granulomatous disease. Calcified adenopathy (asterisk). In the middle a pulsion diverticulum (arrow) due to high intraluminal pressure. On the right multiple pulsion diverticula (arrows) that preceded Heller myotomy for achalasia. On the left a traction diverticulum (arrows) secondary to post primary TB. It simulates a cavitary lung lesion on the chest radiograph. Pseudodiverticula can be seen in reflux esophagitis. On the left a patient with a hiatus hernia, reflux esophagitis, and pseudodiverticula (arrows) at site of proximal stricture Other pathologic conditions can simulate diverticula. On the left two patients with a iatrogenic perforation and a patient with a communicating duplication cyst. The types of hiatus hernia are listed in the table on the left. The relationship between hiatus hernia, reflux and reflux esophagitis is controversial and poorly understood. Most patients with gastroesophageal reflux disease (GERD) have hernias. Many patients with hiatus hernias do not have reflux. Many patients with reflux do not have hiatus hernias. Presence of reflux correlates poorly with GERD. A sliding hiatus hernia is of doubtful significance when an isolated finding in the absence of clinical or imaging findings of esophagitis. Diagnosis of GERD is based on imaging or endoscopic findings of esophagitis, not presence of a hiatus hernia. Sliding hernia On the left initially, GE junction is below the esophageal hiatus. Later, stomach protrudes through hiatus. Neither the hernia or stricture (arrow) due to reflux esophagitis were visible early in the examination. Paraesophageal hernia Large hernias can cause symptoms, and with progressive hiatal widening, increasing protrusion and rotation of the stomach can lead to gastric volvulus that can be complicated by hemorrhage, obstruction, strangulation, perforation. On the left two examples. On the far left gas filled gastric fundus (asterisk) protrudes through hiatus but GE junction (arrow) is below diaphragm. Next to it a paraesophageal hernia with most of 'upside down' stomach in chest with greater curvature (arrows) flipped up. On the left a mixed hernia. Distal esophagus is adjacent to the herniated gastric fundus, but unlike a paraesophageal hernia, the gastroesophageal junction (arrow) is above rather than below the diaphragm. Gastroesophageal reflux (GERD) is the most common cause of esophagitis. Other causes of esophagitis are listed in the table on the left. The findings on barium studies are listed in the table on the left. Air-contrast esophagram shows thick esophageal mucosal folds (arrows) and an ulcer (arrowhead) due to GERD. Single contrast esophagram shows stricture (arrow) and sliding hiatus hernia On the left Irregular stricture (arrowhead) and erosions (arrows) due to GERD. Barrett's esophagus (columnar metaplasia) is the result of long-standing reflux esophagitis. Most patients have reflux and a hiatus hernia. The diagnosis is strongly suggested by: On the left a patient with a Barrett's esophagus. The reticular mucosa is characteristic of Barrett's columnar metaplasia, especially with the associated web-like (arrow) stricture. On the left a patient with a Barrett's esophagus with an adenocarcinoma. There are abnormal distal mucosal folds. The upper margin of adenocarcinoma makes right angle with esophageal wall (arrow) indicating a mural lesion in patient with GERD and Barrett's esophagus. Candida esophagitis On the left a patient with an infectious esophagitis due to candida. The barium stury shows numerous fine erosions and small plaques due to Candida albicans in immunocompromised patient. Cytomegalovirus esophagitis On the left an AIDS patient with an infectious esophagitis due to Cytomegalovirus. Such giant ulcers can also be due to HIV alone. Crohn's esophagitis On the left a patient with Crohn's disease. There is a granulomatous esophagitis with aphthous ulcers (arrows). This is an uncommon manifestation of Crohn's disease. The figure on ther right shows the more common colonic aphthous ulcers. TB esophagitis On the left a patient with an infectious esophagitis due to primary TB. There is an irregular sinus tract from proximal esophagus (arrow). Chest radiograph shows enlarged lymph nodes widening mediastinum due to primary tuberculosis. Dilated mural glands or pseudodiverticulosis, is usually associated with histologic or endoscopic signs of inflammation, and many patients have strictures due to GERD. On the left a patient with esophageal pseudodiverticulosis. Eosinophilic esophagitis This diagnosis may be suggested by peripheral eosinophilia and confirmed by > 20 eosinophils per HPF on biopsy. Patients often have dysphagia and allergies. Imaging finding include diffuse narrowing, strictures, and a ringed appearance similar to transverse (feline esophagus) folds that are transient or associated with reflux. Steroid therapy is often curative. On the left a patient with eosinophilic esophagitis. There is diffuse distal narrowing and corrugated margins (arrows) due to ring-like indentations, that are characteristic of eosinophilic esophagitis. Glycogen acanthosis Glycogen plaques are frequently seen at endoscopy. The reported incidence at endoscopy is 5 to 15% of all patients. These benign epithelial collections of glycogen produce small mucosal nodules. Nodules are smooth and well-defined. This may be a degenerative process and produces no symptoms. Feline esophagus The delicate, concentric and transiently appearing folds of a feline esophagus should be distinguished from the thicker, interrupted, fixed folds indicative of longitudinal scarring from reflux esophagitis. The characteristics of a feline esophagus are: Textbook of Gastrointestinal Radiology. 2nd ed. Philadelphia, PA:W.B. Saunders, 2000:190-257, 316-509 by Gore RM, Levine MS. Levine MS, Rubesin SE, Laufer I. Double Contrast Gastrointestinal Radiology 3rd ed. Philadelphia, PA:W.B. Saunders, 2000:61-148 Levine MS. Radiology of the Esophagus Philadelphia, PA:W.B. Saunders, 1989 Eckberg O. Radiology of the Pharynx and the Esophagus. Berlin,Germany: Springer-Verlag, 2003Terrence C. Demos, MD, Harold V. Posniak, MD, Wayde Nagamine, MD and Mary Olson, MD Hypopharynx Upper esophageal sphincter Lower esophageal sphincter Gastroesophageal reflux Esophageal peristalsis Diffuse esophageal spasm Nutcracker esophagus Achalasia Esophageal web Diverticula Zenker's diverticulum Reflux esophagitis Barrett's esophagus Infectious esophagitis PseudodiverticulosisEsophagus - Part IDepartment of Radiology of the Loyola University Medical Center, USA
chest5
1
Esophagus II : Strictures, Acute syndromes, Neoplasms and Vascular impressions
by Terrence C. Demos, MD, Harold V. Posniak, MD, Wayde Nagamine, MD and Mary Olson, MD
In Esophagus II we will discuss: In the table on the left common and uncommon causes of esophageal strictures. On the far left a stricture (arrow) with irregular mucosal folds at stricture site on air-contrast view. This patient had Barrett's esophagus. Mid esophageal strictures and ulcers are suspicious for Barrett's esophagus. The two images on the right show a Barrett's esophagus with an irregular stricture due to adenocarcinoma. On the left a long, symmetric tapered benign stricture months after radiotherapy. On the left images of a patient with a benign stricture high in the esophagus (arrow). There is bilateral lower lobe lung consolidation due to repeated aspiration. Approximately 5,000-15,000 cases of caustic congestion occur in the US every year. About 50%-80% occur in the pediatric population. On the left a high stricture (arrow) following caustic ingestion Osteophytes (arrow) can impinge on the esophagus and hypopharynx. However they rarely cause symptoms. Multiple structures are uncommon. On the left a table with diseases that may present with multiple esophageal strictures. On the left a patient with benign pemphigoid. Mucosal bullae have led to multiple strictures (arrows). On the left a patient with benign epidermolysis bullosa. Multiple strictures (arrows) are a residual of mucosal bullous disease. Extensive bullous skin disease has led to webbed fingers and contractions. Corrosive ingestion can result in multiple strictures. In the table on the left are etiologies of an acute esophageal syndrome. Boerhaave syndrome is rupture of the esophageal wall. It is most often caused by excessive vomiting in eating disorders such as bulimia although it may rarely occur in extremely forceful coughing or other situations, such as obstruction by food. Boerhaave syndrome is a transmural or full-thickness perforation of the esophagus, distinct from Mallory-Weiss syndrome, a nontransmural esophageal tear also associated with vomiting. These syndromes are distinct from iatrogenic perforation, which accounts for 85-90% of cases of esophageal rupture, typically as a complication of an endoscopic procedure, feeding tube, or unrelated surgery. On the left a patient with Boerhaave syndrome. Chest radiographs show pneumomediastinum (arrows). Esophagram with extravasated water soluble contrast material in left hemithorax (asterisk) Perforation is almost always on left side of distal esophagus. Radiographs show mediastinal gas, effusion, and later pneumothorax. Esophagram is used to confirm leak, first with water-soluble contrast, then barium if no leak demonstrated. On the left a patient with Boerhaave syndrome. The barium study shows extraluminal gas (arrow) without contrast extravasation. CT shows extraluminal gas (arrows). Rent of distal left esophagus confirmed at surgery. CT can show small amounts of extraluminal gas or extravasation not visible on radiographs or esophagram. A Mallory-Weiss tear results from prolonged and forceful vomiting, coughing or convulsions. Typically the mucous membrane at the junction of the esophagus and the stomach develops lacerations which bleed, evident by bright red blood in vomitus, or bloody stools. It may occur as a result of excessive alcohol ingestion. This is an acute condition which usually resolves within 10 days without special treatment. On the left a patient with a Mallory-Weiss tear. Spot films show barium (arrows) in linear mucosal tear near gastroesophageal junction. Tears may be in distal esophagus, gastric fundus, or extend across the GE junction. These unusual lesions have been associated with increased esophageal intraluminal pressure, most often vomiting, instrumentation, and anticoagulation or bleeding disorders. Some are spontaneous. Blunt trauma is a rare cause. Hematomas are self-limited and almost never progress to perforation. Most esophageal hematomas resolve in 1-2 weeks with conservative treatment. On the left a patient with an esophagus hematoma. He presented with chest pain and dysphagia after vomiting. Aside from tortuous aorta chest radiograph is normal. The barium study shows a narrowed lumen (arrows) on AP view and flattened lumen on lateral view (arrowheads) suggestive of a intramural hematoma. On CT the diagnosis of an intramural hematoma was confirmed. A high density mural hematoma (arrowhead) is seen next to NG tube (arrow). Following conservative treatment, six months later the barium study was normal. On the left a patient who had a complicated endoscopy. Instrumentation caused a mucosal tear and dissecting intramural hematoma resulting in double lumen with separating stripe of mucosa (arrows). On the far left an intramural extravasation (arrow) after distal dilation for achalasia. In the middle an intramural extravasation (arrow) after complicated endoscopy. On the right a perforation after biopsy with extravasation of contrast material (arrow). On the left a list of benign esophageal masses. Leiomyomas are the most common benign esophageal neoplasm and are often large yet nonobstructive. Gastrointestinal stromal tumors (GIST) are least common in the esophagus. On the left an asymptomatic patient with a leiomyoma. On the chest film an abnormal opacity is seen behind the heart (arrow). The barium study demonstrates a lobulated mass (arrow) that does not obstruct despite its large size. Mucosal lesions are indicated by mucosal irregularities. Submucosal intramural lesions produce smooth filling defects, and in profile, the margins often form close to a right angle with the esophageal wall. Extrinsic lesions tend to form longer obtuse angles if not fixed to the esophageal wall, and their epicenter may be outside the esophagus. In practice, the location of a lesion may be difficult to determine. On radiograph, tumor (arrows) protrudes into azygoesophageal recess. On esophagram, the inferior margin of this intramural lesion forms close to a right angle (arrow) with esophageal wall. A calcified esophageal mass is almost always a leiomyoma. On the left a patient with a calcified esophageal lesion (arrows) protrudes into azygoesophageal recess on radiograph. Lesion (arrow) on CT and surgical specimen radiograph showing calcification. On the left a patient with granular cell myoblastomas, an uncommon benign tumor. These two lesions (arrows) are nonspecific in appearance, but the proximal lesion does demonstrate overhanging and right angle margins indicating mural location. Pedunculated fibrovascular polyps are rare lesions, that are difficult to diagnose on esophagrams. Their movement during the examination producing an inconstant position and shape may be suggestive as in this patient. The stalk is often difficult to identify. On the left a patient with an esophageal duplication. The findings on the barium study are non-specific. Lesion (arrows) is visible behind the heart on radiograph. Esophageal narrowing (arrows) is caused by duplication. A foregut duplication cyst is a congenital cyst. In the case on the left it displaces hypopharynx and opacified esophagus (arrow) posteriorly and trachea and larynx (asterisk) anteriorly. On the left a list of malignant esophageal masses. Early and small esophageal carcinoma are not synonymous. Early esophageal carcinoma is limited to the mucosa, submucosa with no lymph node metastases. Most are small ( Small esophageal carcinoma is defined by the size of the lesion, a diameter So an early carcinoma may be small, but a small carcinoma may be invasive or metastatic and thus not an early carcinoma. On the left a patient with an early esophageal carcinoma. Lesion is not visible on single contrast esophagram. Air-contrast esophagram shows surface irregularity (arrows) indicating a mucosal lesion. This was both a small lesion and a pathologically early squamous carcinoma. Advanced carcinoma has many gross appearances: On the left two cases of polypoid carcinoma. On the left a patient with an infiltrative ulcerated carcinoma. This lesion has an abrupt transition forming an acute angle and overhanging edge. This indicates mural involvement and is different than obtuse angles usually produced by extrinsic lesions that are not fixed to the esophagus. On the left a patient with a varicoid carcinoma. Unchanging appearance of filling defects indicate tumor rather than varices. Note sharp upper margin of lesion and ulceration (arrows) On the far left a patient with a varicoid carcinoma. Long lobulations simulate varices but did not vary during fluoroscopy. Note large irregular folds and soft tissue mass (arrow) of gastric fundus Next to it a patient with a superficial spreading carcinoma. Extensive superficial spread involves distal esophagus. This appearance can be seen with both early and advanced lesions. On the far left a patient with a carcinoma with stricture. An irregular, asymmetric stricture is highly suggestive of carcinoma. Smoothly tapered, symmetric strictures are characteristic of a benign etiology, but malignant strictures can have similar characteristics and mimic benign lesions. Next to it a patient with a carcinoma with stricture resembling achalasia. Distal esophageal malignancy may closely resemble achalasia. If esophageal motility is normal, achalasia can be excluded. If abnormal, however, subtle imaging features; asymmetric, irregular, abrupt, or high narrowing, mucosal abnormality, or fixed abnormality suggest diagnosis. On the left another case of pseudoachalasia. Distal narrowing simulates achalasia, but narrowing is eccentric, shoulders (arrows) asymmetric, mucosa irregular at tip of narrowing. CT shows gastric fundus thickening (arrows) due to adenocarcinoma. Tracheoesophageal stripe Width of the juxtaposed posterior tracheal and anterior esophageal walls > 5 mm on a lateral chest radiograph is suspicious for pathology, most often esophageal carcinoma or achalasia. On the left a patient with a widened 1 cm stripe (arrows). Esophagram shows widened stripe (arrows) and irregular margins of midesophageal carcinoma. CT shows abnormal soft tissue dorsal to trachea. The tumor invades mediastinum adjacent to aortic arch (arrow) Barrett's esophagus is a proven risk factor for the development of an adenocarcinoma. The incidence of cancer in Barrett's however is controversial. Who, how, and when individuals should be screened is unresolved. Adenocarcinoma was 10% of esophageal malignancies in 1960s. Since 1960s, incidence increasing in USA greater than any other carcinoma. Incidence now approaching or exceeding squamous carcinoma in Caucasian men in the USA and Europe. On the left a patient with an ulcerated (arrow) plaque like adenocarcinoma in a Barrett's esophagus. Primary gastric fundus adenocarcinoma can invade the esophagus, but means of differentiating invasion from a primary esophageal tumor are a subject of debate. On the left a patient with a gastric fundus adenocarcinoma. The barium study demonstrates marked irregular thickening of distal esophagus and folds at gastroesophageal junction. CT shows thickened irregular lesser curvature wall (arrows) near gastroesophageal junction. Spindle cell carcinomas are rare neoplasms, also called carcinosarcomas. They are often bulky but nonobstructive as in the case on the left. Leiomyosarcomas and rare primary melanomas of the esophagus also tend to be bulky but not cause significant obstruction. On the left a patient with a leiomyosarcoma of the esophagus. Margin (arrows) of bulky lesion visible on chest radiograph. Lateral view of esophagram shows marked irregularity and esophageal narrowing (arrows). On the left another patient with a leiomyosarcoma of the esophagus. Large lesion distorts esophageal lumen. CT shows lesion distorting but not obstructing esophageal lumen (arrow). On the left three patients with esophageal narrowing as a result of metastatic mediastinal lymphnodes. On the far left a bronchogenic carcinoma. Extrinsic mediastinal lymph nodes produce long obtuse angles at the interface with esophagus. In the middle another bronchogenic carcinoma. Irregular distal esophageal wall due to invasion of esophagus. In the right a patient with a breast carcinoma. There is mediastinal lymphadenopathy with esophageal invasion and obstruction. Due not confuse normal esophageal irregularities for impressions by lymphnodes. On the left a normal esophagus. The esophagus (arrow) protrudes under aortic arch into right side of AP window. Next to it mediastinal nodes (arrows) that displace the esophagus to right in a patient with bronchogenic carcinoma. On the left a list of vascular structures that may cause impressions on the esophagus. With portal hypertension, elevated portal venous pressure leads to reversed (hepatofugal) flow bypassing the liver through the left gastric vein to dilated esophageal and periesophageal veins that anastamose with the azygos and hemiazygos veins which drain uphill into the superior vena cava. Filling defects due to varices are characterized by change in appearance during the examination related to breath holding and thoracic pressure. On the left are CT images of a patient with large Uphill varices secondary to cirrhosis with portal hypertension. On the left CT images of a patient with uphill varices. Uphill varices can be mass-like as seen in the case on the left. Continue with next image. The CT shows mass-like mediastinal and esophageal varices (arrows). Varices have to be differentiated from varicoid carcinoma. On the left the fixed appearance of filling defects indicates tumor rather than varices. Note sharp upper margin of lesion (arrows) Downhill Varices With superior vena caval obstruction, upper body venous blood flows caudally downhill through esophageal veins to the azygos vein which empties into the superior vena cava caudal to the obstruction. If the obstruction is at or below the azygos, the blood flow extends further caudally to the portal system and then the hepatic veins to the inferior vena cava and the right atrium. On the left downhill varices in a patient with a superior vena cava obstruction due to histoplasmosis. On the barium study inconstant filling defects (arrows) represent downhill varices in upper esophagus. The angiogram demonstrates collateral vessels including a dilated left superior intercostal vein (arrow). The barium study demonstrates inconstant filling defects (blue arrows) due to downhill varices in upper esophagus. CT demonstrates esophageal (red arrow) and mediastinal varices. Continue with venogram. Upper arm venograms show SVC obstruction. This is the most common thoracic arterial anomaly and rarely causes symptoms. The artery extends up and to the right producing a dorsal diagonal impression on the esophagus (arrows). The CT demonstrates that the aberrant artery (arrow) is last vessel from arch and extends dorsal to trachea and esophagus. A right aortic arch with an aberrant left subclavian artery is most often an incidental finding. A right aortic arch with mirror-image branching however is almost always associated with congenital heart disease. CT shows right arch (R) and aberrant left subclavian artery (arrow) arising low off arch and extending to left dorsal to esophagus and trachea. On the left the esophagram of a patient with a right arch that produces a dorsal indentation on this lateral view (blue arrow). The diagram shows the aberrant left subclavian artery (L SCA) dorsal to the trachea and esophagus. Double arch most often presents with airway obstruction, dysphagia, aspiration in children. The arches indent esophagus at different levels. On the left another double arch. Chest radiograph with right lung consolidation due to aspiration in 6-year-old. Right and left arch indent esophagus (arrows) at different levels The aberrant left pulmonary artery indents the trachea dorsally and esophagus ventrally as it extends between them. Narrowing of right bronchus can cause air trapping or atelectasis. Tortuous aorta A tortous descending aorta is a common cause of extrinsic impression on the esophagus. The image on the far left shows a narrowed distal esophagus. Oblique view shows esophageal indentation by aorta with obtuse margins (arrows) characteristic of extrinsic compression. Aortic pathology On the far left the normal aortic arch impression on the esophagus. This impression can be enlarged if there is dilatation of the aorta as seen in this patient with a mycotic aortic arch aneurysm (arrows). On the left 3 images of a patient with a coarctation. On the chest film the 'Figure 3' shape of aortic knob due pre and post stenotic dilatation (arrows). The barium study demonstrates the 'Reverse 3 figure' indention of esophagus by pre and post stenotic aortic dilatation (arrows). An angiogram demonstrates a coarctation with pre and post stenotic dilatation in another patient. Textbook of Gastrointestinal Radiology. 2nd ed. Philadelphia, PA:W.B. Saunders, 2000:190-257, 316-509 by Gore RM, Levine MS. Levine MS, Rubesin SE, Laufer I. Double Contrast Gastrointestinal Radiology 3rd ed. Philadelphia, PA:W.B. Saunders, 2000:61-148 Levine MS. Radiology of the Esophagus Philadelphia, PA:W.B. Saunders, 1989 Eckberg O. Radiology of the Pharynx and the Esophagus. Berlin,Germany: Springer-Verlag, 2003Terrence C. Demos, MD, Harold V. Posniak, MD, Wayde Nagamine, MD and Mary Olson, MD Boerhaave syndrome Mallory-Weiss tear Esophageal hematoma Leiomyomas Fibrovascular polyp Duplication Barrett's esophagus and Adenocarcinoma Uphill varices Aberrant right subclavian artery Right aortic arch with aberrant left subclavian artery Double Arch Aberrant left pulmonary artery CoarctationEsophagus II : Strictures, Acute syndromes, Neoplasms and Vascular impressionsDepartment of Radiology of the Loyola University Medical Center, USA
chest6
1
Lung - Cancer New TNM
by IJsbrand Zijlstra, Otto van Delden, Cornelia Schaefer-Prokop and Robin Smithuis
This is an updated version of the 2005 article 'Lung cancer - staging'. It is adapted to the new guidelines in the 7th Edition of TNM in Lung Cancer of the International Association for the Study of Lung Cancer (IASLC) Staging Committee in 2009. Lungcancer is classified into two categories: small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC). NSCLC is a group of primary lung neoplasms with the same staging system and therapy and can be cured with resection if it is in an early stage. In this review we will discuss: T1 Tumor T2 Tumor > 3 cm but T3 Tumor > 7 cm or any of the following: T4 Tumor of any size that invades the mediastinum, heart, great vessels, trachea, recurrent laryngeal nerve, esophagus, vertebral body, carina, or with separate tumor nodules in a different ipsilateral lobe. In 2009 a new Lung cancer lymph node map was proposed by the International Association for the Study of Lung Cancer (IASLC) in order to reconcile the differences between the Naruke and the MD-ATS maps and refine the definitions of the anatomic boundaries of each of the lymph node stations. Stages NSCLC includes adenocarcinoma, squamous cell carcinoma and large cell carcinoma and is staged according to the TNM-staging system. TNM subsets are grouped into certain stages, because these patients share similar prognostic and therapeutic options. For instance all stage IIIA patients have a 5 year-survival of 10%. In the table on the left resectable stages are indicated in green and unresectable stages are indicated in red. Stage IIIA is possibly resectable, usually after combined-modality therapy consisting of platinum-based chemotherapy and radiation. Stage IIIB, i.e. any patient who has T4 or N3 disease is virtually unresectable, but in some countries there are subgroups of patients that will get a resection. Evidently all patients with distant metastases (stage IV) are inoperable.  In the table on the left the changes for the new stage grouping in the revised TNM staging system . The goal of imaging is to decide whether the tumor is resectable and whether it should be a lobectomy or a pneumonectomy. On the left a patient with a tumor near the fissure. On coronal reconstructions it was demonstrated that there was no transfissural growth. Lobectomy therefore is a possibilty. Lobectomy is not possible if there is: Thin collimation and MPR are necessary in order to clearly demonstrate the relation of a tumor with the fissure. On the left a case with transfissural growth on both coronal and sagittal reconstructions. Lobectomy is not possible. In the Table on the left the new T-staging according to the 7th edition of the TNM-staging of lungcancer. T-staging is best done with CT to determine the local extent and to look for satellite nodules. There are advantages if CT precedes bronchoscopy and the information from CT is used by the bronchoscopist. CT however has important limitations in overall staging. Preoperative predictions with CT differ from operative staging in 45% of cases. Patients are being both over- and understaged. CT staging remains unsatisfactory for detecting hilar (N1) and mediastinal (N2 and N3) lymph node metastases, and for chest wall involvement (T3) or mediastinal invasion (T4), in which sensitivity and specificity can be less than 65%. MR is more useful than CT in the following cases: PET has a limited role in T-staging because of its lack of resolution. PET however is of great value in N- and M-staging.  On the left a typical T1 tumor.  On the left a typical T2 tumor with obstructive infiltrate of the left lower lobe. The tumor is located in the main bronchus, but the distance is more than 2 cm from the carina.  Chest pain usually indicates chest wall invasion (i.e.T3). A Pancoast tumor is a tumor that involves the superior sulcus and the chest wall is almost always involved in these patients (i.e. T3). Local chest wall invasion can be treated with en-bloc resection. On the left a typical T3 tumor.  On the left an endobronchial tumor of the left main bronchus within 2 cm of the carina. This means that it is at least a T3 tumor. There is also invasion of the mediastinum (blue arrow) and invasion of the pulmonary artery (small yellow arrow), indicating that this is a T4-tumor. In many cases T4-tumors do not pose a diagnostic dilemma. On the left we see a large mass that invades the mediastinum. There is complete obliteration of the superior vena cava with collaterals in the para-aortic and paraspinal regions. The aortic arch is partially surrounded by tumor. T4 - tumor (2) On the left another straight forward case. The tumor invades the mediastinum and surrounds and narrows the right pulmonary artery. T4 - tumor (3)- mediastinal invasion In many cases it is not certain whether there is mediastinal invasion. These are patients with a mass that is not clearly invading the mediastinum but that do not have an intervening fat-plane (Figure). In the case on the left there is an intimate relationship of the tumor with the right brachiocephalic vein. This should be dagnosed as 'indeterminate mediastinal invasion'. This patient should be given the benefit of the doubt and get an operation, since that is the only chance for definitive cure. At surgery the tumor fell away from the mediastinum and was subsequently succesfully resected. On the left an odd case, that was recently published in the NEJM, to illustrate the difficulty of determining mediastinal invasion (11). A CT showed a mass in the right upper lobe, closely associated with the paratracheal soft tissues, indicating possible mediastinal invasion. A needle biopsy of the mass resulted in a pneumothorax. Repeat CT imaging with the patient in a right decubitus position revealed that the mass had moved with the lung and had separated completely from the trachea and mediastinum. Evidently this is not a T4-tumor, but a T2-tumor. The patient underwent resection of the right upper lobe. T3/T4 - tumor (4)- Satellite nodules In patients with NSCLC the presence of satellite metastatic nodules may be considered a contraindication to surgical treatment. However the use of multidetector CT has led to the detection of a considerable number of indeterminate satellite lesions. Obtaining a differential diagnosis of these lesions is extremely important in defining the therapeutic strategy (12). In many institutions T3/T4-tumors due to satellite metastatic nodules in the ipsilateral lung will be resected, since these patients have a slightly better prognosis than other T4-patients and have a chance of completeness of resection. PET scan is of limited value in the detection and differentiation of satellite nodules. Many nodules are not detected by PET because of their small size (figure). On the left a patient with a lungcancer in the left upper lobe. In the right upper lobe a second nodule is seen. Both lesions show FDG-uptake. A malignant node in another lobe can either be a metastasis or a synchronous primary tumor. In this case, if it is a metastasis, this means stage IV disease, which is not resectable, because it is not in the ipsilateral lung. If it is a synchronous primary tumor, it is possibly resectable. T4 - tumor (6) - resectability There is some controversy concerning the resectability of certain T4-tumors. As mentioned above in some institutions T4-tumors due to ipsilateral satellite nodules will be resected. In some institutions even limited invasion of the vertebral body or the heart is no contraindication for surgery. On the left a T4 tumor with invason of the left atrium. The invasion is usually through the pulmonary veins. A Pancoast tumor is a tumor of the superior pulmonary sulcus characterized by pain due to invasion of the brachial plexus, Horner's syndrome and destruction of bone due to chest wall invasion. MR is superior to CT for local staging due to its superior soft tissue contrast. Sagittal T1WI will best demonstrate ingrowth into thoracic and cervical nerves. Axial MR will best demonstrate ingrowth into mediastinum and vertebral canal. Pancoast tumors are staged at least as T3, because there is almost always chest wall invasion. When there is ingrowth into a vertebral body or vital mediastinal structures, the tumor is staged as T4. Nodes in pancoast tumors are treated differently than in other tumors. Ipsilateral supraclavicular nodes (N3) are potentially resectable with en bloc resection, while mediastinal nodes (N2) are not. In 20% of patients there will be N2 nodes, so all patients with a pancoast tumor should undergo mediastinoscopy with sampling of mediastinal nodes before surgery. On the left a detail of an AP-film of the cervical spine of a patient with pain in the neck and shoulder. A mass is seen in the apex of the left lung. This proved to be a Pancoast with ingrowth in the brachial plexus. On the left the chest film of the same patient. Notice how difficult it is to depict the tumor. We will continue with the MR-images. On the left sagittal T1-weighted images after the administration of Gadolinium. Scroll through the images by clicking on the arrows. Notice how the tumor grows through the chest wall and invades the structures of the neck. Only a small part of the tumor is actually within the lung. Pancoast tumors are potentially resectable if only one of the following occurs: Patients are nowadays treated with a trimodality approach. First they undergo preoperative radiotherapy and chemotherapy (chemoradiation), followed by restaging and finally en-bloc resection of the upper lobe and the chest wall if there is no evidence of progressive disease. On the left a Pancoast tumor. The tumor abuts the root T1 (long white arrow), but the other roots of the brachial plexus are free in the interscalene triangle (green arrow). This patient is operable. Pancoast tumors are unresectable if one of the following occurs: On the left another patient with a Pancoast tumor. The tumor is seen as an enhancing mass and invades the interscalene triangle, where the roots and trunks of the brachial plexus are located. There is encasement of the subclavian artery (A). Lymph node staging is done according to the American Thoracic Society mapping scheme. Supraclavicular zone (1) Superior Mediastinal Nodes (2-4) Aortic Nodes (5-6) Inferior Mediastinal Nodes (7-9) Hilar, Interlobar, Lobar, Segmental and Subsegmental Nodes (10-14) N1-nodes are ipsilateral nodes within the lung up to hilar nodes. N1 alters the prognosis but not the management. A T1-tumor without positive nodes within the lung has a 5-y survival of 61%. The same T1-tumor with N1-nodes has a 5-y survival of only 34%. On the left a T2 tumor (> 3cm) in the right lower lobe with ipsilateral hilar node (N1). Although we all have learned, that N2-nodes are resectable, there is only a subset of patients with N2 disease that benefits from resection. Those are the patients who, after a negative mediastinoscopy, are found to have microscopic metastatic disease at the time of thoracotomy. Those patients have a better prognosis. Patients however with bulky N2-nodes on CT and FDG-PET will not undergo surgery. They are treated with neo-adjuvant therapy followed by definitive locoregional treatment, which may consist of either radiotherapy or surgery. On the left a tumor in the right upper lobe with progression into the mediastinum (T4) with ipsilateral mediastinal N2 nodes in station 4R. On the left a patient with a lungcancer and hilar nodes (N1), right paratracheal (station 4R = N2) and a prevascular node (station 3A = N2). N3-nodes are clearly unresectable. These are contralateral mediastinal or contralateral hilar nodes or any scalene or supraclavicular nodes. On the left a central tumor in the right lung. Lymphadenopathy all the way up to the lower paratracheal station on the left (i.e. station 4L). This is N3-stage due to contralateral mediastinal nodes. On the left two patients with lungcancer in the right lung. Both have contralateral nodes. If these lymph nodes contain tumor cells, this means inoperable stage IIIB-disease. On the left another patient with lungcancer. Scroll through the images. There is possible ingrowth into the mediastinum. Notice the extensive spread into the mediastinal lymph nodes up to the station 1 level, i.e. N3. The next step should be US-guided FNA. Regardless of the threshold size of lymph node chosen, CT findings in isolation can not be taken as clear evidence of malignant nodal involvement. 20% of all nodes deemed malignant on CT criteria will be benign. Size alone cannot be an exclusion criterion and proof is needed by biopsy or resection that a node is indeed malignant. It is now well established that PET is a much better technique than CT in the determining the lymph node status in patients with NSCLC. In the Table on the left a list of studies that clearly demonstrated the superiority of PET over CT. False-positive mediastinal nodal scans occur in sarcoid, tuberculosis and other infections. In normal-sized mediastinal lymph nodes PET has a sensitivity and specificity of 74 and 96%, respectively, for detecting metastasis. This means that if the PET is positive in these normal-sized nodes, there is almost always a lymph node metastasis. Only in 4% of cases PET is false negative. In enlarged mediastinal lymph nodes (short axis diameter of 10 mm or more) PET has a sensitivity and specificity of 95 and 76%, respectively. This means that PET depicts almost all the metastases, but is false positive due to reactive nodes in 24%. On the left a patient who during follow up for obstructive pulmonary disease presented with a mass in the left upper lobe. On the chest film enlarged nodes are seen in the AP-window (i.e. station 5 nodes). The next diagnostic step was a PET-CT which clearly depicted the tumor, many positive mediastinal nodes, but also nodes in the neck on the left side (i.e. N3-disease). This indicates stage IIIB, non-operable. The PET also showed activity in the right upper lung, which can be a metastasis or a synchronous tumor. There were no palpable nodes in the neck region, but ultrasound depicted the nodes that were PET-positive. Fine needle aspiration was performed and the diagnosis of NSCLC was made (i.e. N3-disease). On the left a patient with a solitary pulmonary nodule. There is FDG-uptake in the nodule, but not in the mediastinaum or elsewhere in the body. This is probably a T1N0M0 tumor (Stage I) Classification of disease as stage I on the basis of a clinical examination and negative results from CT and PET examinations appears sufficient to exclude mediastinal disease. In this case there is no need for mediastinoscopy, because the accuracy of PET-CT is as high as it is. Classification of stage II and III diseases is more controversial. The negative predictive value of PET decreases in relation to the size of the metastases, the presence of centrally located primary disease or N1 nodes, and the avidity of the primary tumor for 18F-FDG. In addition, the presence of hypermetabolic central tumors or hilar lymph nodes can decrease the detectability of mediastinal lymph nodes and thus the negative predictive value of mediastinal PET. For stage II and III diseases, the incidence of false-negative results is still greater with PET than with mediastinoscopy (respectively 11.7% and 3%). Mediastinoscopy likely will remain part of the standard protocol for mediastinal staging for stage II and III diseases.  Distant metastases are very common in patients with lung cancer. Almost every organ may be involved but the extra-thoracic sites posing common clinical problems are brain metastases, bone metastases, sometimes with cord compression, nodal spread, adrenal and liver involvement. On the left a CT- and corresponding PET image of a tumor in the right lung. PET-CT is extremely helpful in the detection of distant metastases. In a study of 100 patients PET was compared with conventional imaging, which consisted of CT of chest and upper abdomen, bone scintigraphy, brain-CT or MR (3). PET had superior sensitivity and specificity in the lung, liver, adrenals and bone. Not suprisingly PET was only less sensitive in the brain due to the high glucose uptake in the normal brain. In this study 9% of patients had metastases demonstrated by PET that were not found with conventional imaging, whereas 10% of patients suspected of having metastases because of conventional imaging findings were correctly shown with PET not to have metastases. Because of the high negative predictive value, PET scanning should be performed in all patients with no evidence of metastatic disease on CT who are considered candidates for surgery. PET-image demonstrates FDG-uptake in mediastinal nodes, but also in the liver (yellow arrows) and the spine (red arrow) indicating distant metastases. M1a-tumor and pleural effusion Malignant pleural effusion has a poor prognosis. These patients usually die within 3 months. In patients with malignant pleural effusion aspiration will yield a false negative results in about a third of the cases. If the cytology is negative, you can either do another thoracocenthesis or do a VATS-procedure (video assisted thoracoscopy) and obtain pleural biopsies. By Peter Goldstraw Journal of Thoracic Oncology: June 2009 - Volume 4 - Issue 6 - pp 671-673 by Stephen G. Spiro and Joanna C. Porter. American Journal of Respiratory and Critical Care Medicine Vol 166. pp. 1166-1196, (2002) H. van Tinteren et al. The Lancet, Volume 359, Issue 9315, Pages 1388-1392 Edith M. Marom et al Radiology. 1999;212:803-809 Didier Lardinois et al. NEJM, Volume 348:2500-2507, June 19, 2003, Number 25 Sung Shine Shim et al. Radiology 2005;236:1011-1019. Semin Chong et al RadioGraphics 2006;26:1811-1824 Gustav K. von Schulthess et al Radiology 2006;238:405-422 Liesbet Schrevens, Natalie Lorent, Christophe Dooms, Johan Vansteenkiste The Oncologist, Vol. 9, No. 6, 633-643, November 2004 Eric M. Rohren, MD, PhD, Timothy G. Turkington, PhD and R. Edward Coleman, MD Radiology 2004;231:305-332 by Tira Bunyaviroch, MD and R. Edward Coleman, MD Journal of Nuclear Medicine Vol. 47 No. 3 451-469, 2006 Images in Clinical Medicine by Robert C. Hoch, M.D., Minnesota Lung Center. NEJM Volume 356:2312 May 31, 2007 Number 22 Positron Emission Tomography to Evaluate Lung Lesions Smith-Bindman et al. JAMA.2001; 285: 2711-2712. by Ramon Rami-Porta et al Ann Thorac Cardiovasc Surg 2009; 15: 4 - 9IJsbrand Zijlstra, Otto van Delden, Cornelia Schaefer-Prokop and Robin Smithuis What is new in 7th edition of TNM T1 - tumor T2 - tumor T3 - tumor T4 - tumor Regional Lymph Node Classification System N1 - Nodes N2 - Nodes N3 - Nodes CT vs PET-CT in N-staging PET-CTLung - Cancer New TNMDepartment of Radiology of the Academical Medical Centre, Amsterdam and the Rijnland Hospital, Leiderdorp, the Netherlands
chest7
1
Lung - HRCT Basic Interpretation
by Robin Smithuis, Otto van Delden and Cornelia Schaefer-Prokop
In this article a practical approach is given for the interpretation of HRCT examinations We will discuss the following subjects: by Robin Smithuis, Otto van Delden and Cornelia Schaefer-Prokop Secondary lobule Knowledge of the lung anatomy is essential for understanding HRCT. The secondary lobule is the basic anatomic unit of pulmonary structure and function. Interpretation of interstitial lung diseases is based on the type of involvement of the secondary lobule. It is the smallest lung unit that is surrounded by connective tissue septa. It measures about 1-2 cm and is made up of 5-15 pulmonary acini, that contain the alveoli for gas exchange. The secondary lobule is supplied by a small bronchiole (terminal bronchiole) in the center, that is parallelled by the centrilobular artery. Pulmonary veins and lymphatics run in the periphery of the lobule within the interlobular septa. Under normal conditions only a few of these very thin septa will be seen. There are two lymphatic systems: a central network, that runs along the bronchovascular bundle towards the centre of the lobule and a peripheral network, that is located within the interlobular septa and along the pleural linings. Centrilobular area is the central part of the secundary lobule. It is usually the site of diseases, that enter the lung through the airways ( i.e. hypersensitivity pneumonitis, respiratory bronchiolitis, centrilobular emphysema ). Perilymphatic areais the peripheral part of the secundary lobule. It is usually the site of diseases, that are located in the lymphatics of in the interlobular septa ( i.e. sarcoid, lymphangitic carcinomatosis, pulmonary edema). These diseases are usually also located in the central network of lymphatics that surround the bronchovascular bundle. Basic Interpretation A structured approach to interpretation of HRCT involves the following questions: These morphologic findings have to be combined with the history of the patient and important clinical findings. When we study patients with HRCT, we have to realize that we are looking at a selected group of patients. Common diseases like pneumonias, pulmonary emboli, cardiogenic edema and lungcarcinoma are already ruled out. So uncommon diseases like Sarcoidosis, Hypersensitivity pneumonitis, Langerhans cell histiocytosis, Lymphangitic carcinomatosis, Usual Interstitial Pneumonitis (UIP) and many others become regular HRCT diagnoses and can be real Aunt Minnies. In the reticular pattern there are too many lines, either as a result of thickening of the interlobular septa or as a result of fibrosis as in honeycombing. Septal thickening Thickening of the lung interstitium by fluid, fibrous tissue, or infiltration by cells results in a pattern of reticular opacities due to thickening of the interlobular septa. Although thickening of the interlobular septa is relatively common in patients with interstitial lung disease, it is uncommon as a predominant finding and has a limited differential diagnosis (Table). Smooth septal thickening is usually seen in interstitial pulmonary edema (Kerley B lines on chest film); lymphangitic spread of carcinoma or lymphoma and alveolar proteinosis. Nodular or irregular septal thickening occurs in lymphangitic spread of carcinoma or lymphoma; sarcoidosis and silicosis. On the left we see focal irregular septal thickening in the right upper lobe in a patient with a known malignancy. This finding is typical for lymphangitic carcinomatosis. There are also additional findings, that support this diagnosis like mediastinal lymph nodes and a nodular lesion in the left lung, that probably represents a metastasis. Pulmonary lymphangitic carcinomatosis (PLC) In 50% of patients the septal thickening is focal or unilateral. This finding is helpful in distinguishing PLC from other causes of interlobular septal thickening like Sarcoidosis or cardiogenic pulmonary edema. Hilar lymphadenopathy is visible in 50% and usually there is a history of (adeno)carcinoma. Identical findings can be seen in patients with Lymphoma and in children with HIV infection, who develop Lymphocytic interstitial pneumonitis (LIP), a rare benign infiltrative lymphocytic disease. On the left a patient who had a CT to rule out pulmonary embolism. There is a combination of smooth septal thickening and ground-glass opacity with a gravitational distribution. The diagnosis based on this CT was cardiogenic pulmonary edema. Cardiogenic pulmonary edema generally results in a combination of septal thickening and ground-glass opacity. There is a tendency for hydrostatic edema to show a perihilar and gravitational distribution. Thickening of the peribronchovascular interstitium, which is called peribronchial cuffing, and fissural thickening are also common. Common additional findings are an enlarged heart and pleural fluid. Usually these patient are not imaged with HRCT as the diagnosis is readily made based on clinical and radiographic findings, but sometimes unsuspected hydrostatic pulmonary edema is found. On the left a patient with both septal thickening and ground glass opacity in a patchy distribution. Some lobules are affected and others are not. This combination of findings is called 'crazy paving'. Crazy paving was thought to be specific for alveolar proteinosis, but is also seen in many other diseases such as pneumocystis carinii pneumonia, bronchoalveolar carcinoma, sarcoidosis, nonspecific interstitial pneumonia (NSIP), organizing pneumonia (COP), adult respiratory distress syndrome and pulmonary hemorrhage. Alveolar proteinosis is a rare diffuse lung disease of unknown etiology characterized by alveolar and interstitial accumulation of a periodic acid-Schiff (PAS) stain-positive phospholipoprotein derived from surfactant. Honeycombing represents the second reticular pattern recognizable on HRCT. Because of the cystic appearance, honeycombing is also discussed in the chapter discussing the low attenuation pattern. Pathologically, honeycombing is defined by the presence of small cystic spaces lined by bronchiolar epithelium with thickened walls composed of dense fibrous tissue. Honeycombing is the typical feature of usual interstitial pneumonia (UIP). The distribution of nodules shown on HRCT is the most important factor in making an accurate diagnosis in the nodular pattern. In most cases small nodules can be placed into one of three categories: perilymphatic, centrilobular or random distribution. Random refers to no preference for a specific location in the secondary lobule. Perilymphatic distribution In patients with a perilymphatic distribution, nodules are seen in relation to pleural surfaces, interlobular septa and the peribronchovascular interstitium. Nodules are almost always visible in a subpleural location, particularly in relation to the fissures. Centrilobular distribution In certain diseases, nodules are limited to the centrilobular region. Unlike perilymphatic and random nodules, centrilobular nodules spare the pleural surfaces. The most peripheral nodules are centered 5-10mm from fissures or the pleural surface. Random distribution Nodules are randomly distributed relative to structures of the lung and secondary lobule. Nodules can usually be seen to involve the pleural surfaces and fissures, but lack the subpleural predominance often seen in patients with a perilymphatic distribution. The algorithm to distinguish perilymphatic, random and centrilobular nodules is the following: Perilymphatic distribution Perilymphatic nodules are most commonly seen in sarcoidosis. They also occur in silicosis, coal-worker's pneumoconiosis and lymphangitic spread of carcinoma. Notice the overlap in differential diagnosis of perilymphatic nodules and the nodular septal thickening in the reticular pattern. Sometimes the term reticulonodular is used. On the left a typical case of perilymphatic distribution of nodules in a patient with sarcoidosis. Notice the nodules along the fissures indicating a perilymphatic distribution (red arrows). Always look carefully for these nodules in the subpleural region and along the fissures, because this finding is very specific for sarcoidosis. Typically in sarcoidosis is an upper lobe and perihilar predominance and in this case we see the majority of nodules located along the bronchovascular bundle (yellow arrow). On the left another typical case of sarcoidosis. In addition to the perilymphatic nodules, there are multiple enlarged lymph nodes, which is also typical for sarcoidosis. In end stage sarcoidosis we will see fibrosis, which is also predominantly located in the upper lobes and perihilar. Centrilobular distribution Centrilobular nodules are seen in: In many cases centrilobular nodules are of ground glass density and ill defined (figure). They are called acinair nodules. In centrilobular nodules the recognition of 'tree-in-bud' is of value for narrowing the differential diagnosis. Tree-in-bud describes the appearance of an irregular and often nodular branching structure, most easily identified in the lung periphery. It represents dilated and impacted (mucus or pus-filled) centrilobular bronchioles. Tree-in-bud almost always indicates the presence of: On the left a tree-in-bud is seen. In the proper clinical setting suspect active endobronchial spread of TB. In most patients with active tuberculosis, the HRCT shows evidence of bronchogenic spread of disease even before bacteriologic results are available (6). Random distribution On the left a patient with random nodules as a result of miliary TB. The random distribution is a result of the hematogenous spread of the infection. Small random nodules are seen in: Sarcoidosis usually has a perilymphatic distribution, but when it is very extensive, it spreads along the bronchovascular bundle to the periphery of the lung and may reach the centrilobular area. Langerhans cell histiocytosis is an uncommon disease characterised by multiple cysts in patients with nicotine abuse. In a very early stage, these patients show only nodules, that later on cavitate and become cysts (figure). As in all smoking related diseases, there is an upper lobe predominance. Increased lung attenuation is called ground-glass-opacity (GGO) if there is a hazy increase in lung opacity without obscuration of underlying vessels and is called consolidation if the increase in lung opacity obscures the vessels. In both ground glass and consolidation the increase in lung density is the result of replacement of air in the alveoli by fluid, cells or fibrosis. In GGO the density of the intrabronchial air appears darker as the air in the surrounding alveoli. This is called the 'dark bronchus' sign In consolidation, there is exclusively air left intrabronchial. This is called the 'air bronchogram'. Ground-glass opacity (GGO) represents: So ground-glass opacification may either be the result of air space disease (filling of the alveoli) or interstitial lung disease (i.e. fibrosis). The location of the abnormalities in ground glass pattern can be helpfull: Thus ground glass in itself is very unspecific. Not suprisingly, there is a big overlap in the causes of ground-glass opacity and consolidation and some diseases may present with both areas of ground-glass and consolidation. On the left we see consolidation and ground-glass opacity in a patient with persistent chest abnormalities and weight loss without signs of infection. This suggested a chronic disease. There is no honeycombing or traction bronchiectasis, so we can rule out fibrosis. The weight loss is suggestive of a malignant disease. Histology revealed broncho-alveolar cell carcinoma Broncho-alveolar cell carcinoma (BAC) may present as: Treatable or not treatable? Ground-glass opacity is nonspecific, but highly significant finding since 60-80% of patients with ground-glass opacity on HRCT have an active and potentially treatable lung disease. In the other 20-40% of the cases the lung disease is not treatable and the ground-glass pattern is the result of fibrosis. In those cases there are usually associated HRCT findings of fibrosis, such as traction bronchiectasis and honeycombing. On the left two cases with GGO, one without fibrosis and potentially treatable and the other with traction bronchiectasis indicating fibrosis. On the left a patient with GGO as dominant pattern. In addition there is traction bronchiectasis indicating the presence of fibrosis. This case is one of the possible patterns of nonspecific interstitial pneumonia (NSIP). NSIP is characterized histologically by a relatively uniform pattern of cellular interstitial inflammation associated with variable degrees of fibrosis. As in UIP (usual interstitial pneumonia) it mainly involves the dependent regions of the lower lobes, but NSIP lacks the extensive fibrosis with honeycombing. NSIP may be idiopathic or associated with collagen vascular diseases or exposure to drugs or chemicals. NSIP has a relative good prognosis and the majority of patients respond to treatment with corticosteroids. This outcome is quite different from that seen in UIP, which has a poor prognosis. The term 'mosaic attenuation' is used to describe density differences between affected and non-affected lung areas. There are patchy areas of black and white lung. The role of the radiologist is to determine which part is abnormal: the black or the white lung. When ground glass opacity presents as mosaic attenuation consider: It can be difficult to distinguish these three entities. There are two diagnostic hints for further differentiation: If the vesses are difficult to see in the 'black' lung as compared to the 'white' lung, than it is likely that the 'black' lung is abnormal. Then there are two possibilities: obstructive bronchiolitis or chronic pulmonary embolism. Sometimes these can be differentiated with an expiratory scan. If the vessels are the same in the 'black' lung and 'white' lung, then you are looking at a patient with infiltrative lung disease, like the one on the right with the pulmonary hemmorrhage. Temporary bronchiolitis with air trapping is seen in: On the left a patient with ground glass pattern in a mosaic distribution. Some lobules are involved and others are not. The differential diagnosis is hypersensitivity pneumonitis, bronchiolitis or thromboembolic disease. The history was typical for hypersensitivity pneumonitis. Hypersensitivity pneumonitis usually presents with centrilobular nodules of ground glass density (acinar nodules). When they are confluent, HRCT shows diffuse ground glass. Hypersensitivity pneumonitis (HP) is an allergic lung disease caused by the inhalation of antigens contained in a variety of organic dusts. Farmer's lung is the best-known HP syndrome and results from the inhalation of fungal organisms that grow in moist hay or exposure to birds as pets (1). HP usually presents in two forms either as ground glass in a mosaic distribution as in this case or as centrilobular nodules of ground glass density (acinar nodules). On the left a patient with ground glass pattern in a mosaic distribution. The clue here is the enlargement of pulmonary arteries (arrow) in the areas of ground glass. The ground glass appearance is the result of hyperperfused lung adjacent to oligemic lung with reduced vessel caliber due to chronic thromboembolic disease. On the left another patient with ground glass pattern in a mosaic distribution. Again the ground glass appearance is the result of hyperperfused lung with large vessels adjacent to oligemic lung with small vessels due to chronic thromboembolic disease. Emboli adherent to the wall and intravascular septa are typical for chronic thromboemboli in which partial recanalization took place. Crazy Paving is a combination of ground glass opacity with superimposed septal thickening (5). It was first thought to be specific for alveolar proteinosis, but later was also seen in other diseases. Crazy Pavin can also be seen in: Consolidation is synonymous with airspace disease. When you think of the causes of consolidation, think of 'what is replacing the air in the alveoli'? Is it pus, edema, blood or tumor cells (Table on the left). Even fibrosis as in UIP, NSIP and long standing sarcoidosis can replace the air in the alveoli and cause consolidation. Acute consolidation is seen in: Chronic consolidation is seen in: Most patients who are evaluated with HRCT, will have chronic consolidation, which limits the differential diagnosis. On the left two cases with chronic consolidation. There are patchy non-segmental consolidations in a subpleural and peripheral distribution. The differential diagnosis is the same as the list above. The final diagnosis was cryptogenic organizing pneumonia (COP). In chronic eosinophilic pneumonia the HRCT findings will be the same, but there will be eosinophilia. In fibrosis there will be other signs of fibrosis like honeycombing or traction bronchiectasis. Bronchoalveolar carcinoma can also look like this. Organizing pneumonia (OP) Organizing pneumonia represents an inflammatory process in which the healing process is characterized by organization and cicatrization of the exudate rather than by resolution and resorption. It is also described as 'unresolved pneumonia'. If no cause can be identified it is called cryptogenic organizing pneumonia (COP). It was described in earlier years as Bronchiolitis-obliterans-organizing pneumonia (BOOP). Patients with COP typically present with a several-month history of nonproductive cough. Many cases are idiopathic, but OP may also be seen in patients with pulmonary infection, drug reactions, collagen vascular disease, Wegener's granulomatosis and after toxic-fume inhalation. On the left a case of chronic eosinophilic pneumonia. It was a patient with low-grade fever, progressive shortness of breath and an abnormal chest radiograph. There was a marked eosinophilia in the peripheral blood. Like in COP we see patchy non-segmental consolidations in a subpleural distribution. Chronic eosinophilic pneumonia is an idiopathic condition characterized by extensive filling of alveoli by an infiltrate consisting primarily of eosinophils. Chronic eosinophilic pneumonia is usually associated with an increased number of eosinophils in the peripheral blood and patients respond promptly to treatment with steroids. The fourth pattern includes abnormalities that result in decreased lung attenuation or air-filled lesions. These include: Most diseases with a low attenuation pattern can be readily distinguished on the basis of HRCT findings. Emphysema typically presents as areas of low attenuation without visible walls as a result of parenchymal destruction.  Paraseptal emphysema Paraseptal emphysema is localized near fissures and pleura and is frequently associated with bullae formation (area of emphysema larger than 1 cm in diameter). Apical bullae may lead to spontaneous pneumothorax. Giant bullae occasionally cause severe compression of adjacent lung tissue. Panlobular emphysema On the left a typical case of panlobular emphysema. There is uniform destruction of the underlying architecture of the secondary pulmonary lobules, leading to widespread areas of abnormally low attenuation. Pulmonary vessels in the affected lung appear fewer and smaller than normal. Panlobular emphysema is diffuse and is most severe in the lower lobes. In severe panlobular emphysema, the characteristic appearance of extensive lung destruction and the associated paucity of vascular markings are easily distinguishable from normal lung parenchyma. On the other hand, mild and even moderately severe panlobular emphysema can be very subtle and difficult to detect on HRCT(1). Lung cysts are defined as radiolucent areas with a wall thickness of less than 4mm. Cystic lung diseases as listed in the table on the left. Cavities are defined as radiolucent areas with a wall thickness of more than 4mm and are seen in infection (TB, Staph, fungal, hydatid), septic emboli, squamous cell carcinoma and Wegener's disease. On the left a case with multiple round and bizarre shaped cysts. There was an upper lobe predominance. The patient had a long history of smoking. This combination of findings is typical for Langerhans cell histiocytosis. Langerhans cell histiocytosis (LCH) is an idiopathic disease characterized in its early stages by granulomatous nodules containing Langerhans histiocytes and eosinophils. In its later stages, the granulomas are replaced by fibrosis and the formation of cysts. It is an uncommon condition. The majority of patients are young or middle-aged adults presenting with nonspecific symptoms of cough and dyspnea. Up to 20% of patients present with pneumothorax and over 90% of patients are smokers. Most cysts appear round, but can also have bizarre shapes (bilobed or clover-leaf shaped). An upper lobe predominance in the size and number of cysts is common. On the left a case with multiple cysts that are evenly distributed througout the lung ( in contrast to LCH). Notice the pneumothorax. There was no history of smoking and this was a 40 year old female. This combination of findings is typical for Lymphangiomyomatosis (LAM). Lymphangiomyomatosis is a rare disease characterized by progressive proliferation of spindle cells, resembling smooth muscle. Proliferation of these cells along the bronchioles leads to air trapping and the development of thin-walled lung cysts. Rupture of these cysts can result in pneumothorax. Other features of LAM include adenopathy and pleural effusion. Lymphangiomyomatosis occurs only in women, usually of child-bearing age, between 17 and 50 years. Identical clinical, radiologic, and pathologic pulmonary changes are seen in about 1% of patients with tuberous sclerosis. Most patients die within 10 years of the onset of symptoms. Bronchiectasis Bronchiectasis is defined as localized bronchial dilatation. The diagnosis of bronchiectasis is usually based on a combination of the following findings: A signet-ring sign represents an axial cut of a dilated bronchus (ring) with its accompanying small artery (signet). The most common cause of bronchiectasis is prior infection, usually viral, at an early age. It also occurs in patients with chronic bronchitis, COPD and cystic fibrosis. Bronchiectasis may mimic cystic lung disease and bullous emphysema. Bronchiectasis caused by primary airway disease should be differentiated from tracion bronchiectasis as a result of fibrosis. On the left we see a chest film with a typical finger-in-glove shadow. The HRCT shows focal bronchiectasis with extensive mucoid impaction, which is in the appropriate clinical setting (asthma and serum eosinophilia) typical for Allergic bronchopulmonary aspergillosis (ABPA). Allergic bronchopulmonary aspergillosis is a lung disease occurring in patients with asthma or cystic fibrosis, triggered by a hypersensitivity reaction to the presence of Aspergillus fumigatus in the airways. It characteristically presents with the findings of central bronchiectasis, mucoid impaction and atelectasis. Honeycombing is defined by the presence of small cystic spaces with irregularly thickened walls composed of fibrous tissue. Honeycomb cysts often predominate in the peripheral and subpleural lung regions regardless of their cause. Subpleural honeycomb cysts typically occur in several contiguous layers. This finding can allow honeycombing to be distinguished from paraseptal emphysema in which subpleural cysts usually occur in a single layer. The case on the left shows subpleural honeycomb cysts in several contiguous layers. There is also a lower lobe predominance and widespread traction bronchiectasis. These findings are typical for Usual Interstitial Pneumonia (UIP). UIP or 'end-stage lung' is a pathology diagnosis and usually shown at lungbiopsy, when honeycombing is visible. Idiopathic pulmonary fibrosis (IPF), accounts for more than 60% of the cases of UIP. UIP with lung fibrosis is also a common pattern of auto-immune disease and drug-related lung injury. A long list of drugs have been implicated, but this pattern is most commonly the result of cytotoxic chemotherapeutic agents such as bleomycin, busulfan, vincristine, methotrexate, adriamycin, and carmustine (BCNU). On the left another case of UIP. The lower zone predominance is demonstrated when you scroll through the images. Notice the ground glass opacity in the left lower lobe as a result of fibrous tissue replacing the air in the alveoli. Upper lung zone preference is seen in: Lower zone preference is seen in: Central distribution is seen in sarcoidosis and cardiogenic pulmonary edema. Peripheral distribution is mainly seen in cryptogenic organizing pneumonia (COP), chronic eosinophilic pneumonia and UIP. Pleural effusion is seen in: Hilar and mediastinal lymphadenopathy In sarcoidosis the common pattern is right paratracheal and bilateral hilar adenopathy ('1-2-3-sign'). In lung carcinoma and lymphangitic carcinomatosis adenopathy is usually unilateral. 'Eggshell calcification' in lymph nodes commonly occurs in patients with silicosis and coal-worker's pneumoconiosis and is sometimes seen in sarcoidosis, postirradiation Hodgkin disease, blastomycosis and scleroderma . Reticular pattern Nodular pattern Nodular pattern(2) High Attenuation pattern High Attenuation pattern (2) Low Attenuation pattern Low Attenuation pattern (2) For a printed version of this article, just push the print button on the upper right at the start of the article or use the crtl-P print modus. If you encounter printing problems with the margins of the document, simply adjust the margins or the scale of the document in the print settings. a spoken lecture given by Jud W. Gurney for www.chestx-ray Santiago E. Rossi, MD et al Radiographics. 2003;23:1509-1519 M. Bakhshayesh Karam MD et al.Robin Smithuis, Otto van Delden and Cornelia Schaefer-Prokop Algorithm for nodular pattern Tree-in-bud Ground-glass opacity Mosaic attenuation Crazy Paving Consolidation Emphysema Cystic lung disease HoneycombingLung - HRCT Basic InterpretationRadiology Department of the Rijnland Hospital, Leiderdorp and the Academical Medical Centre, Amsterdam, the Netherlands
chest8
1
Lung - HRCT Common diseases
by Robin Smithuis, Otto van Delden and Cornelia Schaefer-Prokop
In this review we present the key findings in the most common interstitial lung diseases. There are numerous interstitial lung diseases, but in clinical practice only about ten diseases account for approximately 90% of cases. Knowledge of both, the radiological and clinical appearance of these more common interstitial lung diseases, is therefore important for recognizing them in the daily practice and including them in the differential diagnosis. Some less common interstitial lung diseases will also be presented because their HRCT presentation may be very typical, allowing for a 'spot diagnosis' in selected cases. In 'HRCT part I : basic interpretation' the terminology is introduced and a practical approach is given for the interpretation of HRCT examinations. More than 100 entities manifest as diffuse lung disease. Fortunately only about 10 of these account for about 90% of all diffuse lung diseases, that are assessed by open lung biopsy. Knowing the common and also uncommon HRCT-presentations of these frequently encountered diffuse lung diseases is extremely important. On the left you find three different lists of diagnoses. Accounting for 80 - 90% of all diagnoses according to various literature references.
 In some of them the old names are used and in some the newer ones. The 'mnemonic' for the first list is 'SHIT FACED' (alternative shaded fit). Sarcoidosis is a systemic disorder of unknown origin. It is characterized by non-caseating granulomas in multiple organs, that may resolve spontaneously or progress to fibrosis. Pulmonary manifestations are present in 90% of patients. Systemic symptoms such as fatigue, night sweats and weight loss are common. L?fgren's syndrome, an acute presentation of sarcoidosis, consists of arthritis, erythema nodosum, bilateral hilar adenopathy and occurs in 9-34% of patients. Erythema nodosum is seen predominantly in women and arthritis is more common in men. Two third of patients have a remission within ten years. One third have continuing disease leading to clinically significant organ impairment. Less than 5% of patients die from sarcoidosis usually as a result of pulmonary fibrosis. Stages Chest films in sarcoidosis have been classified into four stages: These stages do not indicate disease chronicity or correlate with changes in pulmonary function. On the left a patient with stage I disease. There is hilar and paratracheal adenopathy and no sign of pulmonary involvement. HRCT findings in Sarcoidosis. On the left a typical presentation of sarcoidosis with hilar lymphadenopathy and small nodules along bronchovascular bundles (yellow arrow) and along fissures (red arrows). On the left a detailed view with the typical HRCT-presentation with nodules along bronchovascular bundle (red arrow) and fissures (yellow arrow). This is the typical perilymphatic distribution of the noduless. The HRCT appearance of pulmonary sarcoidosis varies greatly and is known to mimic many other diffuse infiltrative lung diseases. Approximately 60 to 70% of patients with sarcoidosis have characteristic radiologic findings. In 25 to 30% of cases the radiologic findings are atypical. In 5 to 10% of patients the chest radiograph is normal. On the left another typical presentation of sarcoidosis with mediastinal lymphadenopathy and small nodules in a perilymphatic distribution along bronchovascular bundles and along fissures (yellow arrows). Always look for small nodules along the fissures, because this is a very specific and typical sign of sarcoidosis. Fibrosis in Sarcoidosis. Progressive fibrosis in sarcoidosis may lead to peribronchovascular (perihilar) conglomerate masses of fibrous tissue. The typical location is posteriorly in the upper lobes, leading to volume loss of the upper lobes with displacement of the interlobar fissure. Other diseases that commonly result in this appearance are: On the left a typical chest film of long standing sarcoidosis (stage IV) with fibrosis in the upper zones and volume loss of the upper lobes resulting in hilar elevation. Fibrosis results in obliteration of pulmonary vessels, which can lead to pulmonary hypertension. On the left another case of stage IV sarcoidosis.
 Notice the distribution of the conglomerate masses of fibrosis in the posterior part of the lungs. In addition there are multiple small well-defined nodules. Some of these nodules have the typical subpleural distribution. Alveolar Sarcoidosis. On the left a case of alveolar sarcoidosis. Scroll through the images. In this case the appearance resembles a ground glass attenuation, but with a close look you may appreciate that the increased attenuation is the result of many tiny grouped nodules. Also notice the hilar lymphadenopathy. Alveolar Sarcoidosis (2) On the left a 47-year old female patient with a dry cough, slightly breathless and a normal blood analysis. A chest film was taken and she was treated with antibiotics. A follow up film was made, because she did not improve. The first chest film shows bilateral consolidations in the lower lobes (arrow), initially interpreted as infection. After two weeks of treatment with antibiotics, there is no improvement. The differential diagnosis now includes tumor (bronchoalveolar carcinoma or lymphoma), eosinophilic pneumonia , organizing pneumonia, Wegener's disease or an uncommon presentation of sarcoidosis. Now continue with the HRCT. Scroll through the images on the left . There are multiple areas of consolidation. Ancillary findings are hilar and mediastinal lymphadenopathy. The differental diagnosis of the CT-images is basically the same as of the chest film. Histology revealed alveolar sarcoid. There is only one clue to the diagnosis and that is the presence of small nodules that can be identified in image 3, but these are difficult to see. This case nicely demonstrates that sarcoidosis truely is 'the great mimicker'. Sarcoidosis should be therefore in our differential diagnostic list!.
 On the left a case of fibrosing sarcoidosis, showing fibrosis, traction bronchiectases and crowding of the involved bronchi, predominantly in the perihilar region and upper lobes. Nodular abnormalities are absent, but the appearance and the location of the fibrosis are very suggestive of the diagnosis of sarcoidosis. Differential diagnosis of Sarcoidosis. On the left some diseases with a nodular pattern. Silicosis and Coal worker pneumoconiosis (CWP) are pathologically distinct entities with differing histology, resulting from the inhalation of different inorganic dusts. The radiographic and HRCT appearances of these diseases, however, may not be distinguishable from each other and may be similar to sarcoidosis. It is important to realize that these diseases are rare compared to sarcoidosis.
 Silicosis and CWP occur in a specific patient group (construction workers, mining workers, workers exposed to sandblasting, glass blowing and pottery). HRCT findings in Silicosis/CWP On the left a case of silicosis showing nodules of varying sizes with a random and subpleural distribution. One nodule contains calcification (arrow). Note the absence of a lymphatic distribution pattern (peribronchovascular and along fissures), which would be suggestive of sarcoidosis. Differential diagnosis of Silicosis / Pneumoconiosis. Lymphangitic Carcinomatosis results from hematogenous spread to the lung, with subsequent invasion of interstitium and lymphatics.
 The presenting symptoms are dyspnea and cough and can predate the radiographic abnormalities. In many cases however the patients are asymptomatic. Lymphangitic Carcinomatosis is seen in carcinoma of the lung, breast, stomach, pancreas, prostate, cervix, thyroid and metastatic adenocarcinoma from an unknown primary. HRCT findings in Lymphangitic Carcinomatosis On the left a patient with Lymphangitic Carcinomatosis. Notice the focal distribution. This finding is helpful in distinguishing Lymphangitic Carcinomatosis from other causes of interlobular septal thickening like pulmonary edema or sarcoid. There is also lymphadenopathy. On the left another patient with Lymphangitic Carcinomatosis. In this case there is distribution in both lungs. Additional pleural fluid and lung metastases Differential diagnosis of Lymphangitic Carcinomatosis. On the left multiple diseases showing septal thickening: Patients with pulmonary edema are not imaged with HRCT as their diagnosis is usually based on a combination of clinical and chest radiographic findings. However sometimes the diagnosis is not that straightforward and knowledge of the HRCT appearance of pulmonary edema can be helpful in avoiding misdiagnosis. HRCT findings in cardiogenic pulmonary edema On the left typical features of cardiogenic pulmonary edema There is smooth septal thickening and some ground glass opacity in the dependent part of the lungs. In addition there is bilateral pleural fluid. In a patient with a known malignancy lymphangitic carcinomatosis would be high in the differential diagnostic list. Differential diagnosis of cardiogenic pulmonary edema. On the left another example of cardiogenic pulmonary edema. This patient had a CT to rule out pulmonary emboli. There is smooth septal thickening and ground glass opacity in a more patchy distribution. Note: edema can have a very unusual appearance and be distributed very patchy: some areas are filled with fluid as opposed to other areas in immediate vicinity which appear normal. Hypersensitivity pneumonitis (HP) is also known as extrinsic allergic alveolitis (EAA). HP is an allergic lung disease caused by the inhalation of a variety of antigens (farmer's lung, bird fancier's lung, 'hot tub' lung, humidifier lung). The radiographic and pathologic abnormalities in patients can be classified into acute, subacute, and chronic stages. Mostly HRCT is performed in the subacute stage of HP, weeks to months following the first exposure to the antigen or in the chronic phase. Subacute hypersensitivity pneumonitis The key findings in the subacute hypersensitivity pneumonitis are: On the left a case with subacute hypersensitivity pneumonitis with a typical mosaic pattern and without any signs of septal thickening or distorsion of the airways. On the left a case of subacute hypersensitivity pneumonitis. There is subtle opacity in the centre of the secondary lobules (arrows) with sparing of the subpleural region. The case on the left also demonstrates subtle centrilobular opacity in a patient with subacute HP. Notice how ill-defined these centrilobular nodules are. Sometimes the centrilobular opacities are more nodular in appearance like in the case on the left. On the left another case of hypersensitivity pneumonitis. Here we see the mosaic pattern. Some secundary lobules demonstrate ground-glass opacity due to lung infiltration, while others are more lucent due to bronchiolitis with air trapping. On the left a patient who presented with acute dyspnoe and a normal chest film (not shown). The HRCT at presentation (left) shows lobular areas of ground glass attenuation. A control HRCT ten days later demonstrated, that the findings had resolved without any treatment. The findings were thought to be due to hypersensitivity pneumonitis. Chronic hypersensitivity pneumonitis The key findings in chronic hypersensitivity pneumonitis are: On the left a patient with chronic hypersensitivity pneumonitis. The HRCT shows a mosaic pattern with hyperaerated secondary nodules and secondry nodules of increased attenuation. Additionally there is septal and intralobular reticular thickening, indicating already existing irreversible fibrosis. Differential diagnosis of Hypersensitivity Pneumonitis. Chronic hypersensitivity pneumonitis (2) The case on the left shows an inspiratory and expiratory scan: the mosaic pattern with areas of ground-glass attenuation and areas of low attenuation, that become more evident on the expiratory scan, indicating air trapping. Signs of fibrosis such as distorted vessels and bronchi as well as septal thickening are more pronounced in the mid and lower lung zones, but not limited to the subpleural area. The images on the left suggest the diagnosis hypersensitivity pneumonitis. Based on the imaging findigs alone, alveolar proteinosis and other diseases with a mozaic pattern should be included in the differential diagnosis. Primary TB: Initial infection with consolidation, adenopathy and pleural effusion. Secondary TB : Post-primary or reactivation TB. This is the reactivation of the original infection. Usually located in the apical segments of upper lobes with cavitation Endobronchial spread: May occur in both primary and secondary TB, when the infection is not contained. Hematogenous spread (miliary TB): May occur in both primary and secondary TB, when the infection is not contained. HRCT findings in TB On the left a patient with TB. There is a cavitating lesion and typical tree-in-bud appearance. The blue arrow indicates the biopsy needle. Miliary TB This represents a hematogenous dissemination of infection and may occur in association with either primary or postprimary disease. It is characterized by uniform small nodules with a random distribution. Cavitation in TB The chest film on the left shows diffuse areas with nodular air space opacifications. The HRCT demonstrates multiple nodules in peribronchial distribution, partially confluent, and a cavitation, strongly suggestive for tuberculosis.
 Other diseases in the differential are Wegener granulomatosis or malignancy (both show no tree-in-bud). Endobronchial spread of TB This can occur with primary or postprimary infection. In most subjects, the primary infection is localized and clinically inapparent. However, in 5 to 10% of patients with primary TB, the infection is poorly contained and dissemination occurs. This is termed progressive primary tuberculosis. Extensive cavitation of the tuberculous pneumonia lead to endobronchial spread of the infection. Rupture of necrotic lymph nodes into the bronchi can also result in endobronchial dissemination. Tree-in-bud appearance is typical for active endobronchial spread of infection. It occurs in acute tuberculosis but also in any other bacterial infection. Differential diagnosis of TB. In both miliary TB and metastases the nodules have a random distribution. In miliary TB the nodules are more uniform in size. Chronic eosinophilic pneumonia is an idiopathic condition characterized by filling of the alveoli with eosinophils. It is associated with an increased number of eosinophils in the peripheral blood and patients present with fever, cough, weight loss, malaise, and shortness of breath. The symptoms are often severe and last three months or more. Patients respond promptly to treatment with steroids. HRCT findings in Chronic eosinophilic pneumonia On the left a contrast enhanced CT in a patient with chronic eosinophilic pneumonia. Notice peripheral distribution of the consolidations. Differential diagnosis of Chronic eosinophilic pneumonia The images on the left show the similarities between chronic eosinophilic pneumonia and organizing pneumonia. Differentiation has to be made on the basis of clinical and laboratory findings. Pneumocystis carinii pneumonia (PCP) or pneumocystis jiroveci as it is currently named, is an opportumistic infection in immunocompromised patients. PCP used to affect most HIV-infected patients at some point during the course of their disease, but with the new anti-viral drugs it has become less common. Nowadays PCP is seen more in immunosuppressed patients, i.e. transplant recipients and patients on chemotherapy. HRCT findings in PCP On the left an immunocompromised patient with PCP. The CT findings are diffuse ground-glass opacification. The findings are not specific for PCP, but in this clinical setting PCP is the most likely diagnosis. On the left another patient with PCP. Scroll through the images.  Acute respiratory distress syndrome (ARDS) is a sudden, life-threatening lung failure requiring mechanical ventilation. ARDS represents the result of increased permeability often in combination with injury to the respiratory epithelium. A variety of underlying conditions, from infections to major trauma, can cause ARDS. Primary pulmonary risk factors include aspiration, pneumonia, toxic inhalation and pulmonary contusion. Extrapulmonary risk factors are sepsis, pancreatitis, multiple blood transfusions, trauma and the use of drugs such as heroin. Mild forms of ARDS may resolve completely, while severe forms result in irreverible fibrosis. Why some people develop ARDS and others do not is unknown. Extra-pulmonary ARDS On the left a patient who was involved in a traffic accident and within hours developed ARDS. The dominant pattern is ground glass opacity. In the dependent parts of the lung there is also some consolidation, so there is a gradient from front to back. An important finding in extra-pulmonary ARDS is the symmetry of the abnormalities. Pulmonary ARDS On the left a patient who developed ARSD as a result of pneumonia (i.e. pulmonary ARDS). Note the patchy distribution of lung disease and the almost complete distorsion more basal. Patient is ventilated with PEEP (positive end expiratory pressure ) leading to a barotrauma of the lung parenchyma: there are multiple subpleural cysts and a bilateral pneumothorax. Consolidations have a protecting effect on the lung parenchyma under PEEP ventilation, while the ventrally located areas of more normal lung are most prone to the effects of barotrauma. As a result we find cystic destruction ventrally and residual fibrosis mostly in the ventral lung areas. The idiopathic interstitial pneumonias (IIPs) comprise a heterogenous group of disorders. They represent fundamental responses of the lung to injury and do not represent 'diseases' per se. Idiopathic indicates unknown cause and interstitial pneumonia refers to involvement of the lung parenchyma by varying combinations of fibrosis and inflammation. IIPs include seven entities listed in the table on the left in order of relative frequency. These diseases have specific patterns of morphologic findings on HRCT and histology. Before we call these findings idiopathic or cryptogenic, we should realise, that these patterns are also common findings in collagen vascular diseases (e.g., sclerodermia, rheumatoid arthritis) and drug-related lung diseases. For instance in patients with rheumatoid arthritis findings of NSIP, UIP, OP and LIP have been reported. Usual Interstitial Pneumonitis (UIP) is a histologic diagnosis. UIP has distinctive HRCT findings and is usually shown at lungbiopsy, when honeycombing is visible. If the UIP pattern is of unknown cause (i.e. idiopathic), the disease is called Idiopathic pulmonary fibrosis (IPF). IPF accounts for more than 60% of the cases of UIP. In the presence of a surgical biopsy showing a UIP pattern the diagnosis of IPF requires exclusion of other known causes of UIP including drug toxicities, environmental exposures (asbest), and collagen vascular diseases like RA, SLE, polyarteritis nodosa and sclerodermia. A long list of drugs have been implicated, but this pattern is most commonly the result of cytotoxic chemotherapeutic agents such as bleomycin, busulfan, vincristine, methotrexate, adriamycin, and carmustine (BCNU). The differentiation between NSIP and UIP has tremendous prognostic implication for the patient. UIP is more progressive and more than 50% of patients with UIP die within 3 years. On the left a chest film of a patient with UIP due to IPF. The findings on the chest film comprise volume loss and fibrotic changes in the basal lung area. The radiographic appearance of honeycombing comprises reticular densities caused by the thick walls of the cysts. Whenever you see a chest film with long standing reticulation with a lower lobe and peripheral preference also think 'UIP'. HRCT findings in UIP Differential diagnosis of UIP. Chronic HP may be indistinguishable. It is suspected if there is a mosaic pattern with sparing of the lung bases or when there are centrilobular nodules. Sarcoidosis is a more likely diagnosis if the fibrosis is located in the posterior parts of the upper lobes or in the perihilar area and if there are also nodules in a perilymphatic distribution or if there is extensive mediastinal lymphadenopathy. The presence of pleural plaques helps for the differentiation between IPF and asbestosis. On the left a patient with UIP. Notice the honeycombing and the preference of the subpleural and basal lung areas. It is usually easy to recognize the pattern of UIP on HRCT. Nonspecific interstitial pneumonia (NSIP) is by some considered as a specific entity, with specific histologic characteristics, but by others as a 'wastebasket' diagnosis, representing cases of idiopathic interstitial pneumonia that cannot be classified as UIP, DIP, or OP. NSIP is histologically characterized by a homogeneous, uniform pattern of cellular interstitial inflammation associated with variable degrees of fibrosis. In contrast, UIP is associated with extensive fibrosis which is temporally inhomogeneous (i.e. various lesions are of different ages). NSIP is a very inhomogeneous group. NSIP ranges from type I which is a cellular pattern seen as ground glass opacity on HRCT to type IV with a fibrotic pattern, which may be indistinguishable from UIP. On the left a patient with a NSIP. This patient had a rash and muscle weakness. Scroll through the images. The predominant finding is ground glass opacity (GGO). There is very subtle traction bronchiectasis, indicating that the GGO is the result of fibrosis and therefore irreversible. It is important to note that we do not see the classic distribution of UIP, from which NSIP has to be differentiated. The history of this patient is suggestive for the diagnosis dermatomyositis. NSIP is by far the most common interstitial lung disease in patients with connective tissue disease. NSIP (2) NSIP is not a diagnosis on it's own. It is a pattern of lung damage. For the pathologist the key feature is the uniformity of the abnormality within the lung. The role of the radiologist is more to ‘exclude UIP pattern’ rather than to make the diagnosis of NSIP. The diagnosis of NSIP requires histological proof. In all patients with a NSIP pattern, the clinician should be advised to look for connective tissue diseases, hypersensitivity pneumonitis or drugs . On the left two cases of NSIP. Note the varying combination of GGO and fibrosis (traction bronchiectasis), but the lack of honeycombing. NSIP (3) NSIP is the prevalent lung pattern in systemic sclerosis and polymyosisits/dermatomyositis (more than 90%), but also may occur in RA, SLE, Sj?gren's and MCTD. In the images on your left you can appreciate again the spectrum of findings seen in NSIP. All three patients were suffering from connective tissue disease, all cases were biopsy proven. 
 The first (top left) shows a very subtle GGO. Note the difference in the density of the air within the bronchus and surrounding lungparenchyma (dark bronchus sign). The second (top right) is a more obvious example of GGO with a superimposed fine reticular densities as a result of thickening of the intralobular septa. The last image also shows GGO with a fine reticular pattern. Notice the lack of honeycombing in all three cases, excluding UIP as diagnosis. NSIP (4) The HRCT of this patient with scleroderma and NSIP shows a fine subpleural reticular pattern in the upper lobes and more extensive abnormalities in the lower lung zones. There are also areas of ground-glass and traction bronchiectases, but honeycombing is typically lacking. Note also the mildly dilated esophagus, which is consistent with scleroderma. Cryptogenic organizing pneumonia (COP) used to be described as bronchiolitis obliterans with organizing pneumonia (BOOP) in an earlier version of the classification of idiopathic interstitial pneumonias. It is a inflammatory process in which the healing process is characterized by organization of the exudate rather than by resorption ('unresolved pneumonia'). Organizing pneumonia is mostly idiopathic and then called cryptogenic, but is also seen in patients with pulmonary infection, drug reactions, collagen vascular disease, Wegener's granulomatosis and after toxic-fume inhalation. OP presents with a several-month history of nonproductive cough, low-grade fever, malaise and shortness of breath. There is a good response to corticosteroid therapy and a good prognosis. OP is again a great mimicker and can show a broad variety of HRCT findings, which makes it a frequent differential diagnosis and actually represents a diagnosis of exclusion. Frequently biopsy is needed for final proof. HRCT findings in OP On the left a patient with the typical bilateral peripheral consolidations of OP. After exclusion of other diseases such as lymphoma, infection, bronchoalveolar carcinoma, the diagnosis of cryptogenic organizing pneumonia was made. On the left a patient who complained of arthritic pain.
 There are multiple small bilateral peripheral consolidations. The findings in this patient are not as specific as in the former case, but this was also organizing pneumonia, but now related to collagen vascular disease. On the left a patient with rheumatoid arthritis and bilateral peripheral consolidations as a result of organizing pneumonia. Patients with OP associated with collagen vascular diseases respond less well to therapy with steroids. Differential diagnosis of Organizing Pneumonia. The images on the left show the similarities between chronic eosinophilic pneumonia and organizing pneumonia. Differentiation has to be made on the basis of clinical and laboratory findings. Respiratory bronchiolitis (RB), respiratory bronchiolitis-associated interstitial lung disease (RB-ILD), and desquamative interstitial pneumonia (DIP) represent different degrees of severity of small airway and parenchymal reaction to cigarette smoke (8). All smokers have various degrees of respiratory bronchiolitis, but it is usually asymptomatic. However 5-10% of smokers have a clinically significant lung disease in association with RB, presenting with symptoms, lung function tests and auscultatory findings at clinical examination. The term RB-ILD was proposed to describe the bronchocentric (or centrilobular) lung disease in these patients and the term DIP was used to describe the more diffuse disorder. Radiologically however these diseases cannot be clearly separated because of the overlap of CT findings. HRCT findings in RB-ILD On the left a patient with RB-ILD. The dominant feature is ground glass opacification and there are some thickened interlobular septa (arrow). Usually these patients will also have smoking induced centrilobular emphysema and there is some evidence that respiratory bronchiolitis is the precursor of emphysema. RB-ILD (2) On the left a smoker with RB-ILD with subtle HRCT-findings. The dominant pattern is ground glass opacification. Additional findings in this patient are paraseptal emphysema in the upper lobes and some subtle septal thickening in the basal parts. Based on these non-specific CT findings there is a broad differential diagnosis and additional clinical information is mandatory for the interpretion of the HRCT. Since this patient is a smoker we first think RB-ILD. In a immunocompromised patient PCP would be on top of the list. If this patient was coughing up blood, this probably would be pulmonary hemorrhage (although we would expect more pulmonary densities in these patients). If this patient was a bird-fancier we would first think hypersensitivity pneumonitis, but mostly these patients do not smoke. On the left two different patients with similarl HRCT findings. The left one is a smoker with RB-ILD and the patient on the right has hypersensitivity pneumonits. Note the difference in severity of ground glass opacities and the well defined areas of airtrapping in HP. Somehow smoking seems to protect against HP. This is not a 100% specific criterium but is quite helpful for differential diagnosis. RB-ILD (3) On the left a patient with DIP. The HRCT shows diffuse areas of ground-glass density in the lower lobes and some mosaic pattern as the sole abnormality. Reticular abnormalities and signs of fibrosis are typically absent. These abnormalities are usually reversible and will disappear upon cessation of smoking. Acute interstitial pneumonia (AIP, earlier named Hamman Rich Pneumonitis) is a rare idiopathic lung disease characterized by diffuse alveolar damage with subsequent fibrosis. It has a fatal outcome in many cases. The histologic pattern aswell as the HRCT findings in AIP are indistinguishable from acute respiratory distress syndrome (ARDS). The HRCT characteristics are diffuse or patchy consolidation, often with a crazy paving appearance like in the case on the left. There are areas of consolidation and extensive areas of ground-glass density with a crazy-paving appearance. These abnormalities developed in several days and this rapid progression of disease combined with these imaging findings are very suggestive of the diagnosis AIP. Lymphocytic interstitial pneumonitis or LIP is uncommon, being seen mainly in patients with autoimmune disease, particularly Sj?gren's syndrome, and in patients with AIDS. Symptoms are nonspecific and often those of the patient's underlying disease HRCT findings are usually nonspecific. On the left a patient with Sjogren's syndrome with LIP. On the left three different patients with lung cysts. From left to right: Lymphangiomyomatosis, LIP and Langerhans cell histiocytosis. Drug-induced lung disease is a major source of iatrogenic lung injury. The major diagnostic problem is, that it may present with a large variety of radiologic patterns. It may present as organizing pneumonia, eosinophilic pneumonia, fibrosis, hypersensitivity pneumonitis or even as ARDS. The diagnosis of drug-induced pulmonary disease is usually one of exclusion. On the left a patient who is treated with cytotoxic drugs for a hematologic malignancy. The radiographic findings are areas of ground glass opacity, some traction bronchiectasis and subtle honeycombing in the left lower lobe. This could be the result of an idiopathic form of fibrosis like idiopathic pulmonary fibrosis and non-specific interstitial pneumonitis or fibrosis in chronic hypersensitivity pneumonitis and longstanding sarcoid. However it is not one of the typical forms of fibrosis, that we commonly encounter in patients with a UIP pattern or NSIP pattern seen in collagenvascular diseases. When there is fibrosis, that does not fit into any of the common diseases with fibrosis always consider drug-related lung disease in the differential. Drug-induced organizing pneumonia is commonly caused by Bleomycin and Cyclophosphamide and other drugs like Methotrexate, Amiodarone, Nitrofurantoin and Penicillamine (9). The HRCT findings are the same as in cryptogenic organizing pneumonia. Drug-induced non-specific interstitial pneumonita (NSIP) occurs most commonly as a manifestation of carmustine toxicity or of toxicity from noncytotoxic drugs such as amiodarone. The radiologic findings are the same as in other forms of NSIP.  Clinical findings: Key findings in Lymphangiomyomatosis: On the left a typical case of LAM with multiple evenly spread thin walled cysts complicated by a pneumothorax. Differential diagnosis of Lymphangiomyomatosis: On the left another typical case of LAM. Langerhans cell histiocytosis is also known as pulmonary histiocytosis X or eosinophilic granuloma. LCH is probably an allergic reaction to cigarette smoke since more than 90% of patients are active smokers. In the early nodular stage it is characterized by a centrilobular granulomatous reaction by Langerhans histiocytes. In the cystic stage bronchiolar obliteration causes alveolar wall fibrosis and cyst formation. HRCT findings in Langerhans cell histiocytosis: On the left early stage Langerhans cell histiocytosis with small nodules. There are no cysts visible. In a later stage the nodules start to cavitate and become cysts. These cysts start as round structures but finally coalesce to become the typical bizarre shaped cysts of LCH. In patients with LCH 95% have a smoking history. On the left radiological pathological correlation of Langerhans cell histiocytosis in respectively nodular stage and early and late cystic stage. On the left a chest film of a 19 year old patient with Langerhans cell histiocytosis. The dominant findig on the chest film is a reticular patern and that's about as far as you can go. There is also hyperinflation. No way you would have recognized that this pattern was caused by multiple cysts. This is late stage Langerhans cell histiocytosis. The most challenging differential diagnosis in this patient is centrilobular emphysema. Emphysema however is defined as airspaces without definable walls. Usually we can identify the central dot sign. The upper lobe predominance is not helpfull in the differential as we can appreciate this in many inhalational diseases and also in emphysema. On the left another case of Langerhans' cell histiocytosis. It started as small noduli, which progressed over time to cavitating nodules. In the end this will progress to bizarre shaped cysts, that replace normal lung tissue. Differential diagnosis of Langerhans cell histiocytosis. Emphysema, when it is severe, can mimick Langerhans cell histiosytosis. When it extends beyond the centrilobular area to the edge of the secondary lobule, it may look as if it is cystic with walls. In patients with LCH, the pathologist may find LCH, but also areas of emphysema, respiratory bronchiolitis and even fibrosis. So these smoking-related diseases do not represent discrete entities. Alveolar proteinosis is a rare disease characterized by filling of the alveolar spaces with PAS positive material due to an abnormality in surfactant metabolism. The diagnosis is based on the suggestive HRCT pattern (crazy paving) and the characteristic features of BAL fluid (Broncho Alveolar Lavage)  Key findings in alveolar proteinosis On the left a typical case of alveolar proteinosis with extensive thickening of interlobular and intra-lobular septa. This is caused by the fact that the proteinacious material, which is removed from the alveolar space by macrophages is transported to the interstitium and thus leads to thickening of septa. The crazy paving pattern is a rather non-specific finding. Many other diseases may present with this finding and are listed in the differential diagnosis. Differential diagnosis of alveolar proteinosis Special Thanks The authors like to thank Dr. Sujal Desai of the King's College Hospital in London for his inspiring lectures. Some of the images used in this overview were provide by him. We would also like to thank Dr. Richard Webb who produced such a fabulous educational CD (1). by Schaefer-Prokop, C, Prokop M, Fleischmann D, Herold CJ. European Radiology 2001;11: 373-392 This browser-based learning file is based on Dr. Webb's HRCT text. It offers a wide variety of cases dealing with common HRCT patterns of disease, diffuse lung diseases and their significance, and clinical characteristics. It is one of the best educational CD's ever made. by Hansell DM. Radiol Clin North Am 2001:39: 1115-35 by Webb, Mueller and Naidich. by Zampatori M, Sverzellati N, Poletti V et al. Semin Ultrasound CT MR 2005;20(3): 176-85 This Joint Statement of the American Thoracic Society (ATS), and the European Respiratory Society (ERS) was adopted by the ATS Board of Directors, June 2001 and by The ERS Executive Committee, June 2001 by Christina Mueller-Mang, MD, Claudia Grosse, MD, Katharina Schmid, MD, Leopold Stiebellehner, MD, and Alexander A. Bankier, MD RadioGraphics 2007;27:595-615 by LE Heyneman, S Ward, DA Lynch, M Remy-Jardin, T Johkoh and NL Muller American Journal of Roentgenology, Vol 173, 1617-1622 by Santiago E. Rossi, MD, Jeremy J. Erasmus, MD, H. Page McAdams, MD, Thomas A. Sporn, MD and Philip C. Goodman, MD. Radiographics. 2000;20:1245-1259Robin Smithuis, Otto van Delden and Cornelia Schaefer-Prokop UIP NSIP COP RB-ILD and DIP AIP LIP Lymphangiomyomatosis Langerhans cell histiocytosis Alveolar proteinosisLung - HRCT Common diseasesRadiology Department of the Rijnland Hospital, Leiderdorp and the Academical Medical Centre, Amsterdam, the Netherlands
chest9
1
Mediastinum - Lymph Node Map
by Robin Smithuis 
This is an update of the 2007 article, which used the Mountain-Dresler regional lymph node classification for lung cancer staging (MD-ATS maps)(1). In 2009 a new Lung cancer lymph node map was proposed by the International Association for the Study of Lung Cancer (IASLC) in order to reconcile the differences between the Naruke and the MD-ATS maps and refine the definitions of the anatomic boundaries of each of the lymph node stations (2). In this article we provide illustrations and CT-images for a better understanding of this IASLC lymph node map. Supraclavicular nodes 1. Low cervical, supraclavicular and sternal notch nodes From the lower margin of the cricoid to the clavicles and the upper border of the manubrium. The midline of the trachea serves as border between 1R and 1L. Superior Mediastinal Nodes 2-4 2R.Upper Paratracheal 2R nodes extend to the left lateral border of the trachea. From upper border of manubrium to the intersection of caudal margin of innominate (left brachiocephalic) vein with the trachea. 2L.Upper Paratracheal From the upper border of manubrium to the superior border of aortic arch. 2L nodes are located to the left of the left lateral border of the trachea. 3A. Pre-vascular These nodes are not adjacent to the trachea like the nodes in station 2, but they are anterior to the vessels. 3P.Pre-vertebral Nodes not adjacent to the trachea like the nodes in station 2, but behind the esophagus, which is prevertebral. 4R. Lower Paratracheal From the intersection of the caudal margin of innominate (left brachiocephalic) vein with the trachea to the lower border of the azygos vein. 4R nodes extend from the right to the left lateral border of the trachea. 4L. Lower Paratracheal From the upper margin of the aortic arch to the upper rim of the left main pulmonary artery. Aortic Nodes 5-6 5. Subaortic These nodes are located in the AP window lateral to the ligamentum arteriosum. These nodes are not located between the aorta and the pulmonary trunk but lateral to these vessels. 6. Para-aortic These are ascending aorta or phrenic nodes lying anterior and lateral to the ascending aorta and the aortic arch. Inferior Mediastinal Nodes 7-9 7.Subcarinal 8. Paraesophageal Nodes below carina. 9. Pulmonary Ligament Nodes lying within the pulmonary ligaments. Hilar, Lobar and (sub)segmental Nodes 10-14 These are all N1-nodes. 10. Hilar nodes These include nodes adjacent to the main stem bronchus and hilar vessels. On the right they extend from the lower rim of the azygos vein to the interlobar region. On the left from the upper rim of the pulmonary artery to the interlobar region. 1. Supraclavicular zone nodes These include low cervical, supraclavicular and sternal notch nodes. Upper border: lower margin of cricoid. Lower border: clavicles and upper border of manubrium. The midline of the trachea serves as border between 1R and 1L. 2R. Right Upper Paratracheal 2R nodes extend to the left lateral border of the trachea. Upper border: upper border of manubrium. Lower border: intersection of caudal margin of innominate (left brachiocephalic) vein with the trachea. 2L. Left Upper Paratracheal Upper border: upper border of manubrium. Lower border: superior border of aortic arch. On the left a station 2 node in front of the trachea, i.e. a 2R-node. There is also a small prevascular node, i.e. a station 3A node. 3. Prevascular and Prevertabral nodes Station 3 nodes are not adjacent to the trachea like station 2 nodes. They are either: 3A anterior to the vessels or 3B behind the esophagus, which lies prevertebrally. Station 3 nodes are not accessible with mediastinoscopy. 3P nodes can be accessible with endoscopic ultrasound (EUS). On the left a 3A node in the prevascular space. Notice also lower paratracheal nodes on the right, i.e. 4R nodes. 4R. Right Lower Paratracheal Upper border: intersection of caudal margin of innominate (left brachiocephalic) vein with the trachea. Lower border:lower border of azygos vein. 4R nodes extend to the left lateral border of the trachea. On the left we see 4R paratracheal nodes. In addition there is an aortic node lateral to the aortic arch, i.e. station 6 node.  4L. Left Lower Paratracheal 4L nodes are lower paratracheal nodes that are located to the left of the left tracheal border, between a horizontal line drawn tangentially to the upper margin of the aortic arch and a line extending across the left main bronchus at the level of the upper margin of the left upper lobe bronchus. These include paratracheal nodes that are located medially to the ligamentum arteriosum. Station 5 (AP-window) nodes are located laterally to the ligamentum arteriosum. On the left an image just above the level of the pulmonary trunk demonstrating lower paratracheal nodes on the left and on the right. In addition there are also station 3 and 5 nodes. On the left an image at the level of the lower trachea just above the carina. To the left of the trachea 4L nodes. Notice that these 4L nodes are between the pulmonary trunk and the aorta, but are not located in the AP-window, because they lie medially to the ligamentum arteriosum. The node lateral to the pulmonary trunk is a station 5 node. 5. Subaortic nodes Subaortic or aorto-pulmonary window nodes are lateral to the ligamentum arteriosum or the aorta or left pulmonary artery and proximal to the first branch of the left pulmonary artery and lie within the mediastinal pleural envelope. 6. Para-aortic nodes Para-aortic (ascending aorta or phrenic) nodes are located anteriorly and laterally to the ascending aorta and the aortic arch from the upper margin to the lower margin of the aortic arch. 7. Subcarinal nodes These nodes are located caudally to the carina of the trachea, but are not associated with the lower lobe bronchi or arteries within the lung. On the right they extend caudally to the lower border of the bronchus intermedius. On the left they extend caudally to the upper border of the lower lobe bronchus. On the left a station 7 subcarinal node to the right of the esophagus. 8 Paraesophageal nodes These nodes are below the carinal nodes and extend caudally to the diafragm. On the left an image below the carina. To the right of the esophagus a station 8 node. On the left a PET image demonstrating FDG uptake in a station 8 node. On the corresponding CT image the node is not enlarged (blue arrow). The probability that this is a lymph node metastasis is extremely high since the specificity of PET in unenlarged nodes is higher than in enlarged nodes. 9. Pulmonary ligament nodes Pulmonary ligament nodes are lying within the pulmonary ligament, including those in the posterior wall and lower part of the inferior pulmonary vein. The pulmonary ligament is the inferior extension of the mediastinal pleural reflections that surround the hila. 10 Hilar nodes Hilar nodes are proximal lobar nodes, distal to the mediastinal pleural reflection and nodes adjacent to the intermediate bronchus on the right. Nodes in station 10 - 14 are all N1-nodes, since they are not located in the mediastinum. Scroll through the images on the left.  The following nodal stations can be biopsied by cervical mediastinoscopy: the left and right upper paratracheal nodes (station 2L and 2R), left and right lower paratracheal nodes (station 4L and 4R) and the subcarinal nodes (station 7). Station 1 nodes are located above the suprasternal notch and are not routinely accessed by cervical mediastinoscopy. Left upper lobe tumors may metastasize to the subaortic lymph nodes (station 5) and paraaortic nodes (station 6). These nodes can not be biopsied through routine cervical mediastinoscopy. Extended mediastinoscopy is an alternative for the anterior-second interspace mediastinotomy which is more commonly used for exploration of mediastinal nodal stations. This procedure is far less easy and therefore less routinely performed than conventional mediastinoscopy. Endoscopic Ultrasound with Fine Needle Aspiration can be performed of all the mediastinal nodes that that can be assessed from the oesophagus. In addition the left adrenal gland and the left liver lobe can be visualized. EUS particularly provides access to nodes in the lower mediastinum (station 7,8 and 9) by CF Mountain and CM Dresler Chest, Vol 111, 1718-1723 by Valerie Rusch et al Journal of Thoracic Oncology: May 2009 - Volume 4 - Issue 5 - pp 568-577 by Paul De Leyn and Toni Lerut. in the Multimedia Manual of Cardiothoracic Surgery by Christian Lloyd, MD, and Gerard A.Silvestri, MD, FCCP Christian Lloyd, MD, and Gerard A.Silvestri, MD, FCCP Cancer Control, July/August 2001,Vol.8, No.4 Cancer Control 311 by J. T. Annema, and K. F. Rabe Endoscopy 2006; 38: 118-122 by Reginald F. Munden, MD, DMD, Stephen S. Swisher, MD, Craig W. Stevens, MD, PhD and David J. Stewart, MD Radiology 2005; 237:803-818Robin Smithuis Conventional mediastinoscopy Extended mediastinoscopy EUS-FNAMediastinum - Lymph Node MapRadiology department of the Rijnland Hospital in Leiderdorp, the Netherlands
chest10
1
Mediastinum - Masses
by Sanjeev Bhalla, Marieke Hazewinkel and Robin Smithuis
This review is based on a presentation given by Sanjeev Bhalla and was adapted for the Radiology Assistant by Marieke Hazewinkel and Robin Smithuis. Sanjeev Bhalla is section chief of the Cardiothoracic Imaging Section of the Mallinckrodt Institute of Radiology. This review will focus on how to narrow down the differential diagnosis of mediastinal lesions by localizing and characterizing them. Whenever you see a mass on a chest x-ray that is possibly located within the mediastinum, your goal is to determine the following: The table on the left is the overall table for mediastinal masses. In the next paragraphs we will discuss each compartment separately. Statistically, it is important to remember the following: The following characteristics indicate that a lesion originates within the mediastinum: A lung mass abutts the mediastinal surface and creates acute angles with the lung, while a mediastinal mass will sit under the surface creating obtuse angles with the lung (Figure). On the left you see two different patients. Describe the findings and continue. On the x-ray on the left there is a lesion that has an acute border with the mediastinum. This must be a lung mass. The chest radiograph on the right shows a lesion with an obtuse angle to the mediastinum. This must be a mediastinal mass. Since there is a silhouette-sign with the right heart border - which is located anteriorly - we can deduce that the mass must be located within the anterior mediastinum. The lesion on the left was a pancoast tumor. The lesion on the right was a thymoma, located within the anterior mediastinum. The mediastinum can be divided into anterior, middle and posterior compartments. It is important to remember that there is no tissue plane separating these compartments. On the lateral radiograph the anterior and middle compartments can be separated by drawing an imaginary line anterior to the trachea and posteriorly to the inferior vena cava. The middle and posterior compartments can be separated by an imaginary line passing 1 cm posteriorly to the anterior border of the vertebral bodies. This division allows us to make a more narrow differential diagnosis. In many hospitals a CT will be made to further analyze and characterize anterior and middle mediastinal masses. An MRI is usually made to analyze masses located in the posterior compartment because the majority of these masses turn out to be neurogenic in nature. An additional CT can be performed, when bone needs to be assessed. The anterior mediastinum contains the following structures: thymus, lymph nodes, ascending aorta, pulmonary artery, phrenic nerves and thyroid. The most common lesions that you will see in the anterior mediastinum will either be of thymic or lymph node origin. Even the germ cell tumors arise from the pluripotent cells of the thymus. Before you want to biopsy an anterior mediastinal mass, do not forget thta some of these lesions can be vascular in origin. The four T's make up the mnemonic for anterior mediastinal masses:: On conventional radiographs look for the signs listed in the table on the left. The finding of an obliterated retrosternal clear space is not so helpfull anymore, since nowadays many patients are obese. In these patients the retrosternal space can be filled with fat.  Describe the images on the left. Then continue. On the PA film there is a lobulated widening of the superior mediastinum. On the lateral chest film the retrosternal clear space is obliterated. This happened to be a patient with lymphoma. On the left FDG-PET images of the same patient. There are multiple lymphatic masses in the anterior, middle and even posterior mediastinum, spreading to the neck. When there is a mediastinal mass and you still can see the hilar vessels through this mass, then you know the mass does not arise from the hilum. This is known as the hilum overlay sign. Because of the geometry of the mediastinum most of these masses will be located in the anterior mediastinum. Describe the images on the left. Then continue. On the chest film there is a mass that has obtuse angles with the mediastinum, so it is a mediastinal mass. The hilar vessels are seen through this mass, so it does not arise from the hilum and probably will arise from the anterior mediastinum. The anterior location was confirmed on a CT. Most commonly this will be a mass of thymic or lymphatic origin. This proved to be a lymphoma in a HIV-positive patient. The anterior mediastinum is an important location for cystic masses. Masses can be entirely cystic (thymic cysts) or have solid components (lymphoma or cystic thymoma). Some masses are cystic with enhancing septations - in these cases you should think of a germ cell tumor. Describe the image on the left. Then continue. The CT shows an anterior mediastinal mass with water density attenuation. This is typical for a thymic cyst. Describe the image on the left. Then continue. The CT shows a mass located in the anterior mediastinum. The mass is cystic but has solid enhancing septa. This finding is very specific for a germ cell tumor. Now many think that germ cell tumors contain fat and if a lesion does not contain fat, it cannot be a germ cell tumor. You have to remember, that only about 60 % of germ cell tumors contain fat, so you must realize that the absence of fat does not exclude a germ cell tumor from the differential diagnosis. The more solid components a germ cell tumor has, the more likely the tumor is to be malignant. Describe the image on the left. Then continue. The CT shows a mass located in the anterior mediastinum. The mass is cystic but has solid enhancing components, so we are worried about lymphoma, germ cell tumor and cystic thymoma. This proved to be a cystic thymoma. The middle mediastinum contains the following structures: lymph nodes, trachea, esophagus, azygos vein, vena cavae, posterior heart and the aortic arch. The majority of middle mediastinal masses will consist of foregut duplication cysts (eg oesophageal duplication or bronchogenic cysts) or lymphadenopathy. Aortic arch anomalies can also present as middle mediastinal masses. Fluid containing lesions are usually duplication cysts or necrotic lymph nodes. A pancreatic fluid collection due to pancreatitis may also present as a mediastinal mass. A fibrovascular esophageal polyp is a mesenchymal lesion which almost always contains fat. Vascular lesions are arch anomalies, azygos continuation due to interrupted inferior vena cava or hyperenhancing lymph nodes. On conventional radiographs look for the signs listed in the table on the left. Displaced azygoesophageal recess wiil be seen on the right. On the left you may have a pseudoparavertebral line. This is a new interface that looks like a paravertebral line. Describe the image on the left. Then continue. On the AP chest radiograph of this patient there is widening of the azygoesophageal recess on the right. There is an apparent widening of the paravertebral line on the left. On the lateral film the mass is anterior to the spine and therefore is located in the middle mediastinal. On the CT the azygoesophageal recess is displaced to the right due to oesophageal varices (blue arrow) and there is also a new interface on the left. This is a patient with cirrhosis of the liver and varices as a result of portal hypertension. On the left a patient with a small cell lung carcinoma. Describe the images on the left. Then continue. On the PA film there is a lobulated paratracheal stripe on the right. On the lateral radiograph there is a density overlying the ascending aorta and filling the retrosternal space. These findings indicate a mass in the anterior aswell as in the middle mediastinum. The CT confirms the presence oof lymphomas in both the anterior and the middle mediastinum. On the left two different patients. One of these patients has pulmonary hypertension and the other has sarcoidosis. Describe the images on the left. Then continue. On the right image there is a lobulated mass surrounding the right bronchus creating a 'doughnut' with the bronchus as the hole in the doughnut. On the left image there is only density in the area from 9 o'clock to 3 o'clock and not in the 3 - 9 o'clock area. So the patient on the left has pulmonary hypertension with moderately enlarged vessels while the patient on the right has sarcoidosis with widespread lymphadenopathy. When there is a density in the 3 - 9 o'clock area, there should always be concern about mediastinal masses. The posterior mediastinum contains the following structures: sympathetic ganglia, nerve roots, lymph nodes, parasympathetic chain, thoracic duct, descending thoracic aorta, small vessels and the vertebrae. Most masses in the posterior mediastinum are neurogenic in nature. These can arise from the sympathetic ganglia (eg neuroblastoma) or from the nerve roots (eg schwannoma or neurofibroma). Don't forget lymphadenopathy, the vertebrae and the descending thoracic aorta as potential causes for posterior mediastinal masses. Cystic lesions will be either neuroenteric cysts, schwannomas or meningoceles. Fat containing lesions will be extramedullary hematopoiesis. When the anemia is resolved the extramedullary marrow will stop producing blood and become fatty. On conventional radiographs look for: The anterior mediastinum stops at the level of the superior clavicle. Therefore, when a mass extends above the superior clavicle, it is located either in the neck or in the posterior mediastinum. When lung tissue comes between the mass and the neck, the mass is probably in the posterior mediastinum. This is known as the Cervicothoracic Sign. If we study the image on the frontal view on the left, we see a mass extending above the level of the clavicle and there is lung tissue in front of it, so this must be a mass in the posterior mediastinum. On the left the MR of the same patient. It turned out to be a schwannoma. On the left images of a patient, who has a disease, that is the most commonly missed diagnosis in the emergency department resulting in the number one cause of law suits. Study the images and then continue. Notice the widening of the paravertebral stripes on both the left and the right on the PA radiograph. On the lateral radiograph there is a severely narrowed disc space. The diagnosis is discitis. On MR you will notice the edema of the soft tissues and the high signal intensity of the disc. Since there are no tissue planes separating the mediastinal compartments, there are lesions that do not respect our approach to the mediastinum. These lesions tend to occupy more than one compartment and include: mediastinitis, hematomas, vascular entities, bronchogenic cancer, metastases and lymphangiomas (fluid containing). Once you have localized a mediastinal mass, next try to charcterize it by assessing whether it has any of the following characteristics: This is a list of mediastinal msses that may contain fluid: If a mass contains fluid it could be a teratoma (on the left) or a thymic cyst (on the right). Note that this teratoma does not contain fat. Teratomas are the most common benign germ cell tumors. The most common malignant germ cell tumor is the seminoma. Describe the image on the left. Then continue. There is are multiple masses in both the anterior and middle mediastinum. The attenuation values are of water density. These findings favor the diagnosis of cystic lymphadenopathy in a patient with metastatic disease. Describe the image on the left. Then continue. There is a cystic lesion in the middle mediastinum. There is a fluid fluid level with milk of calcium. Foregut duplication cysts occasionally contain milk of calcium like in this example of an esophageal duplication cyst. The differential diagnosis of fat containing mediastinal masses is: On the left we see an fat-containing anterior mediastinal mass. This is the typical finding of a fat-containing teratoma. Describe the image on the left. Then continue. The axial CT and sagittal MR demonstrate a lipomatous lesion within the lumen of the esophagus. This is typical for a esophageal lipoma and its fibrovascular stalk. The differential diagnosis of enhancing mediastinal masses is: On the left multiple enhancing lesions. This is typical for hyperenhancing lymph nodes. Enhancing lymphomas can be seen in: Describe the image. Then continue. First notice the large thymus in this young child. There is also an enhancing mass in the posterior mediastinum extending into the vertebral canal. This is typical for a hemangioma. Describe the image. Then continue. Somewhat irregular enhancing mass in the anterior mediastinum. This proved to be a thyroid mass.Sanjeev Bhalla, Marieke Hazewinkel and Robin Smithuis Obliterated retrosternal clear space Hilum Overlay Sign Cystic masses Cervicothoracic sign Fluid containing masses Fat containing masses Enhancing massesMediastinum - MassesCardiothoracic Imaging Section of the Mallinckrodt Institute of Radiology, St. Louis, USA and the Radiology department the Medical Centre Alkmaar and the Rijnland Hospital, Leiderdorp, the Netherlands
chest11
1
Solitary pulmonary nodule: benign versus malignant
by Ann Leung and Robin Smithuis
A solitary pulmonary nodule is defined as a discrete, well-marginated, rounded opacity less than or equal to 3 cm in diameter that is completely surrounded by lung parenchyma, does not touch the hilum or mediastinum, and is not associated with adenopathy, atelectasis, or pleural effusion. Lesions larger than 3 cm are considered masses and are treated as malignancies until proven otherwise. The differential diagnosis of a solitary pulmonary nodule is broad and management depends on whether the lesion is benign or malignant. In this overview we will discuss some of the new features that can help to differentiate between benign and malignant nodules based upon CT and PET-CT findings. Diffuse, central, laminated or popcorn calcifications are benign patterns of calcification. These types of calcification are seen in granulomatous disease and hamartomas. All other patterns of calcification should not be regarded as a sign of benignity. The exception to the rule above is when patients are known to have a primary tumor. For instance the diffuse calcification pattern can be seen in patients with osteosarcoma or chondrosarcoma. Similarly the central and popcorn pattern can be seen in patients with GI-tumors and patients who previously had chemotherapy. A solitary pulmonary nodule (SPN) is defined as a single intraparenchymal lesion less than 3 cm in size and not associated with atelectasis or lymphadenopathy. A lesion greater than 3 cm in diameter is called a mass. This distinction is made, because lesions greater than 3 cm are usually malignant, while smaller lesions can be either benign or malignant. Swensen et al studied the relationship between the size of a SPN and the chance of malignancy in a cohort at high risk for lung cancer (1). Their findings are listed in the table on the left. They concluded that benign nodule detection rate is high, especially if lesions are small. Of the over 2000 nodules that were less than 4 mm in size, none was malignant. Comparison with prior imaging studies is often the most useful procedure to determine the importance of the finding of a SPN, since stability over 2 years is highly associated with benignity. Japanese screening studies showed that a polygonal shape and a three-dimensional ratio > 1.78 was a sign of benignity (2,3). A polygonal shape means that the lesion has multiple facets (multi-sided). A peripheral subpleural location was also a sign of benignity in this study. The three-dimensional ratio is measured by obtaining the maximal transverse dimension and dividing it by the maximal vertical dimension. A large three-dimensional ratio indicates that the lesion is relatively flat, which is a benign sign.  Recent studies have showed that an air bronchogram is more commonly seen in malignant pulmonary nodules. It is most commonly seen in BAC (bronchoalveolar cell carcinoma) and adenocarcinoma. The case on the left shows an airbronchogram seen as a linear lucency (broad arrow) and as a more cystic lucency (small arrow) due to the fact that the bronchus is seen en face. On the left two solitary pulmonary nodules. Based upon the morphology, which lesion has the most malignant features? The lesion on the far left has a spicuated margin and has lucencies within it. The lesion next to it is lobulated in contour and has some spicules radiating to the pleura. It is however homogeneous in attenuation. Based on these findings we should be most concerned that the lesion on the far left is malignant. It proved to be an adenocarninoma, while the other one was a fungal infection. The lucencies and frank air bronchograms should not mislead you in thinking that it probably is infection. Another result from screening studies is that nodules containing a ground-glass component are more likely to be malignant (5). On the far left a lesion that only has a ground-glass appearance and next to it a lesion that has both ground-glass and solid components. The likelihood of malignancy is 1:5 for the lesion on the far left and 2:3 for the lesion with both ground-glass and solid components. Contrast enhancement less than 15 HU has a very high predictive value for benignity (99%). After a baseline scan, 4 consecutive scans at 1 minute interval are performed. This applies only for nodules with the following selection criteria: PET-CT plays an increasingly important role in the evaluation of solitary nodules. When you perform PET-CT, you have to realize the following: With these specificity numbers, there will be false positives in about 20%, depending on the background prevalence of granulomatous disease. On the left a patient with an adenocarcinoma, that was not hypermetabolic on the PET, so it is a false negative PET. In the differentiation of benign versus malignant solitary pulmonary nodules nowadays new imaging features have to be added. We especially have to look for the presence of areas of ground-glass opacity, air bronchograms or cavities and the three-dimensional ratios of a lesion. With the increasingly important role of PET-CT, we have to be aware of the accuracy of PET-CT and we should have an idea about the prevalence of infectious and non-infectious granulomatous disease in the area that we practice. Stephen J. Swensen et al Radiology 2005;235:259-265. Shodayu Takashima et al. AJR 2003; 180:1255-1263 Shodayu Takashima et al. AJR 2003; 180:955-964 Claudia I. Henschke et al AJR 2002; 178:1053-1057Ann Leung and Robin Smithuis Calcification Size Growth Shape Margin Air Bronchogram sign Solid and Ground-glass components Contrast enhancementSolitary pulmonary nodule: benign versus malignantDepartment of Radiology, Stanford University Medical Center, Stanford, California and the Department of Radiology, Rijnland Hospital, Leiderdorp, the Netherlands
ped1
1
Acute Scrotum in Children
by Gael J. Lonergan
This article is based on a presentation given by Gael Lonergan at the 'Teaching in Holland' course and adapted for the Radiology Assistant by Robin Smithuis. In this overview we will discuss the following subjects: by Gael J. Lonergan Torsion occurs when an abnormally mobile testis twists on the spermatic cord, obstructing its blood supply. Patients present with acute onset of severe testicular pain. The ischemia can lead to testicular necrosis if not corrected within 5-6 hours of the onset of pain. Torsion can be intermittent and can undergo spontaneous detorsion. In a child with an acute scrotum, testicular torsion is not the most common condition (Table). Torsion of testicular appendices represents the more common cause of scrotal pain. Typically, it has a more gradual onset than testicular torsion and patients may endure pain for several days before seeking medical attention. Testicular appendage torsion appears as a lesion of low echogenicity with a central hypoechogenic area adjacent to the epididymis. Most of the time however, we don't see it and we do the US just to exclude a testicular torsion. We should see torsion of testicular appendices more as a diagnosis of exclusion. So although torsion of the testicular appendix and epididymitis are more common, our goal is mainly to detect or exclude a testicular torsion. We want to be able to tell the surgeon whether or not it is a surgical emergency. Complete absence of intratesticular blood flow and normal extratesticular blood flow on color Doppler images is diagnostic, if the flow is normal in the contralateral testis. Yet, the presence of flow within the testis does not exclude the presence of torsion, because incomplete vascular obstruction can sometimes occur or intermittent torsion. The case on the left shows a testicular torsion of the left testis. This case is very obvious because there is no flow on the affected side, but also a difference in echogenicity. With prolonged torsion, the testis is typically hypoechoic and inhomogeneous and is often accompanied by a surrounding hydrocele. By the time these sonographic findings occur, surgical salvage of the testicle is unlikely. Use at least a 10 MHz linear transducer. Always start with the examination of the normal side and optimize the settings for low flow, low resistance and low velocity. The background 'noise' should just be visible in the asymptomatic testis. Once you have a good image of the normal side, don't touch any of the settings' and go to the symptomatic side. In the very young child it can be difficult to examine the testes because they are very small and mobile. The prepubertal testis has a volume of about 1-2 cc, while the postpubertal testis has about 30cc. With age the testis increases in echogenicity, so in a very young child the small testis can be difficult to differentiate from the surrounding fat, especially if it is retracted into the inguinal canal (figure). At the start of the examination you can put your finger on the inguinal canal, so that the testis can not move around. Color Doppler imaging has limited sensitivity for detecting blood flow in pediatric patients with a testicular volume of less than 1cc. On the left two more cases. On the far left a child of 10 months old with torsio testis. There is more flow in the tissues around the testis than in the testis itself and that is abnormal, unless the child has cellulitis. The case next to it is an older child. The gray scale ultrasound shows an abnormal testis. So this could be torsio or orchitis, but the absent flow tells us, that this definitely is a torsio testis. The testis is usually elevated as a result of the torsion and the shortening of the cord itself and may be in a transverse lie. The affected side can be larger from the swollen testis itself, a hydrocele or skin thickening. Gray scale ultrasound is helpfull, not in making the diagnosis, but in predicting the outcome. For the first 4-6 hours there is a normal architecture. During this period the testis is salvagable, so a normal appearance on gray scale means good outcome. After this period the testis becomes heterogeneous and enlarged over the next 4-24 hours and the epididymis and scrotal wall may swell and become hypoechoic. A worsening appearance of the testis on gray scale US correlates with decreased viability (1). The way to look at differences in echogenicity, is to get a transverse image of both testes. The images on the left show the affected side to be bigger and more echogenic. This testis is probably not salvagable. The testis may appear more echogenic or less echogenic, it doesn't matter, as long as there is a difference, there is a poor outcome. Epididymitis is the most common inflammatory process involving the scrotum and more common in adults. Infections generally originate in the lower urinary tract from the bladder, urethra or prostate and are typically caused by urinary tract pathogens or sexually transmitted organisms (Chlamydia or gonorrrhhea) . Epididymitis also occurs in children, but is then due to infection with Streptococcus or Staphylococcus. In urinary tract abnormalities also infection with E.Coli is seen. A sterile chemical epididymitis can result from reflux of sterile urine through the ejaculatory ducts, for instance if the ureter inserts in the prostatic urethra, this may lead to increased pressure in the vas deferens. . On the left a child with a meningocele who had epididymitis. Due to increased bladder pressure and contractions against a closed sphincter, there was reflux not only into the left ureter and porstate, but also into the epididymis, which resulted in epididymitis. The case on the left shows the typical features of epididymitis. The epididymis is swollen and heterogeneous. There is a hydrocele and scrotal wall thickening. With color doppler there is increased flow. A normal epididymis has only limited colorflow. Orchitis is characterized by focal, peripheral, hypoechoic testicular lesions that are poorly defined, amorphous, or crescent-shaped. Orchitis also exhibits testicular hyperemia on color Doppler sonography images and is usually accompanied by epididymal hyperemia due to concomitant epididymitis. A reactive hydrocele is also frequently associated with epididymoorchitis. Focal testicular infarction can occur as a complication of epididymitis when swelling of the epididymis is severe enough to constrict the testicular blood supply. This appears as a hypoechoic intratesticular mass devoid of blood flow. The complications of orchitis are abscess formation and ischemia. The case on the left is a young patient, who came a week before with a hyperemic testis and epididymitis. Due to ongoing infection, the pressure within the testis increased leading to infarction. This is not due to torsion, but can be easily mistaken for torsion. On the left two cases with abnormal areas within the testis probably due to absces formation. In trauma there is either a hematocele or testicular hematoma. In the acute phase the hemorrhage is echogenic and in the chronic phase it is hypoechoic. A hematocele results from scrotal or intra-abdominal hemorrhage. It represents bleeding between the leaves of the tunica vaginalis and appears as a complex fluid collection. With time, this collection can develop loculations, which appear as thick septations. It is important to be able to tell if the testis is intact, because if there is a rupture, this can sometimes be treated surgically. On the left a patient with a typical hematocele. Testicular rupture is seen as focal alterations of testicular echogenicity correlating with areas of intratesticular hemorrhage or infarction in a patient with a hematocele. A discrete fracture plane is identified in fewer than 20% of cases, although visible alterations in the testicular contour are a common finding. On the left a complicated case. The ultrasound demonstrated a large hematocele. There was doubt whether the echogenic structure indeed was a testis. MR was performed and no testicular tissue was found, so we have to conclude that the echogenic structure is a result of fresh hematoma. On the left another patient with a rupture, that was seen on the abdominal CT. On US there was a large heterogeneous area with no identifiable testis due to rupture. Hernias in children are common especially in premature infants. Sometimes we can see them on plain films as we see in the case on the left. If they are filled with bowel, they are easy to detect, but sometimes these hernias are only filled with soft tissue The ultrasound examination starts with the child lying down and is then continued in the standing position. The bowel or omentum is visible separate from the testis (figure). The intestinal loop descends through the unclosed processus vaginalis. An incarcerated hernia is a cause of acute scrotal pain. Peristalsis suggests viability and absence of peristalsis is worrisome for incarceration. Idiopathic scrotal edema is seen in school-aged boys. They present with scrotal skin swelling. So the clinical question is, if there is torsion or infection. At examination the testes and epididymes are normal and all that we see on US is skin edema. If this is all we see and the child does not have fever or elevated white count, which can be seen in cellulitis, than we can make the diagnosis of Idiopathic scrotal edema. Although this is an idiopathic disease, so we don't know what it is, it is nevertheless important to make this diagnosis. It is far more reasuring for parents to be told, that their child has a specific diagnosis, that it is benign and will go away with time. That is better, than telling them, that you don't know what it is, but it is no torsion, so you don't really worry about it. Middleton WD, et al. J US Med 1997; 16: 23. Nussbaum Blask AR, Bulas D, Shalaby-Rana E, et al. Pediatr Emerg Care 2002;18 (2):67-71. Andrew C Peterson, MD, FACS et al in eMedicine Jason S Chang, MD et al in eMedicine Munden MM, Trautwein LM. in Curr Probl Diagn Radiol. 2000 Nov-Dec;29(6):185-205. Clinical and sonographic criteria of acute scrotum in children. Karmazyn B, Steinberg R, Kornreich L, et al. in Pediatr Radiol 2005;35:302-310.Gael J. Lonergan Color doppler Presentation Gray scale Ultrasound Hematocele Testicular ruptureAcute Scrotum in ChildrenChief of Radiology of the Children's Hospital of Austin, Washington, DC
neuro8
1
Multiple Sclerosis
by Frederik Barkhof, Robin Smithuis and Marieke Hazewinkel
This review is based on a presentation given by FrederikBarkhof at the Neuroradiology teaching course for the Dutch Radiology Society and was adapted for the Radiology Assistant by Robin Smithuis and Marieke Hazewinkel. This presentation will focus on the role of MRI in the diagnosis of Multiple Sclerosis. We will discuss the following subjects: One of the most common questions in daily radiology practice when we see an image like the one on the left is: In order to be able to answer that question, we have to realise that when we study white matter lesions (WMLs): Multiple sclerosis (MS) is the most common inflammatory demyelinating disease of the central nervous system in young and middle-age adults, but also affects older people. According to the McDonald criteria for MS, the diagnosis requires objective evidence of lesions disseminated in time and space. As a consequence there is an important role for MRI in the diagnosis of MS, since MRI can show multiple lesions (dissemination in space), some of which can be clinically occult, and MRI can show new lesions on follow up scans (dissemination in time). MS has a typical distribution of WMLs. This can be very helpful in differentiating them from vascular lesions (see Table). Typical for MS is involvement of corpus callosum, U-fibers, temporal lobes, brainstem, cerebellum and spinal cord. This pattern of involvement is uncommon in other diseases. In small vessel disease there may be involvement of the brainstem, but it is usually symmetrical and central, while in MS it is peripheral. Here we see typical differences in vascular brainstem lesions compared to MS. The image on the left is an axial T2 weighted image illustrating typical vascular brainstem involvement, with a central involvement of the transverse pontine fibers. The image on the right is an axial T2 weighted image of the brainstem of an MS-patient, showing typical peripherally located white matter lesions, often in or near the trigeminal tract, or bordering the 4th ventricle. Even when a patient is clinically suspected of MS, we still have to study the WMLs carefully to decide whether these lesions are indeed suggestive of MS, and not incidental age-related findings. We will discuss this more in detail when we look at the MRI criteria in the McDonald criteria for MS. Here we have a coronal PD image of a brain specimen with MS involvement. First look at the image and look for lesions that are specific for MS. Than continue. The lesions in the deep white matter (yellow arrow) are nonspecific and can be seen in many diseases. Typical for MS in this case is: Juxtacortical lesions are specific for MS. These are adjacent to the cortex and must touch the cortex. Do not use the word subcortical to describe this location, because that is a less specific term, indicating a larger area of white matter almost reaching the ventricles. In small vessel disease these juxtacortical U-fibers are not involved and on T2 and FLAIR there will be a dark band between the WML and the (also bright) cortex (yellow arrow). Temporal lobe involvement is also specific for MS. In hypertensive encephalopathy, the WMLs are located in the frontal and parietal lobes, uncommonly in the occipital lobes and not in the temporal lobes. Only in CADASIL there is early involvement of the temporal lobes. First look at the images on the left. Describe the lesions and decide which findings are typical for MS. Typical findings for MS as seen in this case are: These ovoid lesions are also called Dawson fingers. They represent areas of demyelination along the small cerebral veins that run perpendicular to the ventricles. First look at the spinal cord images on the left Describe the lesions and decide which findings are typical for MS. There are multiple lesions in the spinal cord. This is another typical feature of MS. Typical spinal cord lesions in MS are relatively small and peripherally located. They are most often found in the cervical cord and are usually less than 2 vertebral segments in length. By the way did you notice the lesion in the brainstem? A spinal cord lesion together with a lesion in the cerebellum or brainstem is very suggestive of MS. Spinal cord lesions are uncommon in most other CNS diseases, with the exception of ADEM, sarcoid, Lyme disease and SLE. Notice that this image is a proton density weighted image (PDWI). They are crucial for studying the spinal cord. On PDW-images the spinal cord has a uniformly low signal intensity (like CSF), which gives the MS lesions a good contrast against the surrounding CSF and normal cord tissue. Use a 512 matrix and cardiac gating for optimal results. First look at the images on the left Describe the lesions and decide which findings are typical for MS. Typical findings for MS as seen in this case are: Dawson fingers are typical for MS and are the result of inflammation around penetrating venules. These veins are perpendicular to the ventricular surface. Enhancement is another typical finding in MS. This enhancement will be present for about one month after the occurrence of a lesion. The simultaneous demonstration of enhancing and non-enhancing lesions in MS is the radiological counterpart of the clinical dissemination in time and space. The edema will regress and finally only the center will remain as a hyperintense lesion on T2WI. On the left a specimen showing the perivenular inflammation in MS. MS starts as inflammation around these veins. In the first four weeks of the inflammation there is enhancement with gadolinium due to local interruption of the blood brain barrier. First there is homogeneous enhancement but this can change to open ring enhancement. Juxtacortical lesions located in the U-fibers are also very specific for MS. You really have to look hard to notice them, because they are difficult to differentiate from the hyperintense cortex. This patien not only has multiple periventricular lesions of which some have the typical Dawson finger aspect (blue arrow), but there is also a juxtacortical lesion. The involvement of the U-fibers is best seen on the magnification view. On the left a patient who was re-examined 3 months after the first clinical attack. Describe the lesions and decide which findings are typical for MS. Typical findings for MS as seen in this case are: New lesions on T2W images also indicate dissemination in time. The patient on the left had a follow-up examination 3 months after the first clinical event. Notice how similar the positioning is. This allows good comparison of the images. Optimal positioning is discussed in the MRI protocol (see later). Tumefactive MS is a variant of Multiple Sclerosis. It on MRI presents as a large intra-parenchymal lesion with usually less mass effect than would be expected for its size. After the administration of gadolinium, there may be some peripheral enhancement, often with an incomplete ring. These lesions can be distinguished from gliomas or intraparenchymal abscesses, which typically have a closed-ring enhancement. These T2W and T1W post-gadolinium images are of a 39 year old male who presented with subacute onset of hemianopsia. He was referred for biopsy to differentiate between a glioma or demyelination. There is an intraparenchymal mass in the right temporal and occipital lobe with a hypointense rim on T2, which only partially enhances (open-ring sign) on the postcontrast images. There is surrounding edema, but relatively little mass effect. This was a biopsy-proven demyelinating lesion. The open-ring enhancement pattern with low signal T2 ring and low CBF are all indicative of demyelination. Balo's Concentric Sclerosis is an uncommon demyelinating disease. It is characterized by alternative bands of demyelination and myelin preservation, often in whorl-like configurations. Here T2 and postcontrast T1W images showing a large lesion in the left hemisphere with alternating T2-hyperintense and isointense bands. On the T1W images after gadolinium there is alternating linear enhancement. There is a smaller, similar lesion on the right. A very important differential to keep in mind, especially in patients with a bilateral optic neuritis, is Neuromyelitis Optica (NMO) or Devic's Disease. This is a demyelinating disease in which the optic nerves and spinal cord are usually involved. Often there are few T2-lesions in the brain. Think of NMO when there are extensive spinal cord lesions (more than 3 vertebral segments) with low T1-signalintensity and swelling of the cord. On axial images the lesions often involve most of the cord. This is unlike MS, in which the lesions are usually smaller and peripherally located. Here we see a sagittal T2-weighted image of the spinal cord in a patient with NMO showing a longitudinally extensive cord lesion with marked swelling. The clue to the diagnosis was the AQP4-AB titer 1:1024. Read more on NMO in the spinal cord in the article 'Spine - Myelopathy'. Acute Disseminated Encephalomyelitis (ADEM)is another important differential diagnosis of MS. This is a monophasic, immune-mediated demyelinating disease which often presents in children following an infection or vaccination. On MRI there are often diffuse and relatively symmetrical lesions in the supra-and infratentorial white matter which may enhance simultaneously. There almost always is preferential involvement of the cortical gray matter and the deep gray matter of the basal ganglia and thalami. Here we have axial FLAIR and T2W-images of a young patient with ADEM - notice the extensive involvement of the cortical and gray matter, including thalamus. Here another case of ADEM. Notice the involvement of the basal ganglia. Here another case of ADEM. Notice the similarity to the other two cases. The diagnosis of MS requires elimination of more likely diagnoses and demonstration of dissemination of lesions in space and time. Dissemination in Space (DIS) is: Dissemination in Time is: The McDonald criteria for MS were recommended in 2001 by an international panel and revised in 2005 and 2010. An Attack is: Neurological disturbance of kind seen in MS Subjective report or objective observation At least 24 hours duration in absence of fever or infection Excludes pseudoattacks, single paroxysmal symptoms (multiple episodes of paroxysmal symptoms occurring over 24 hours or more are acceptable as evidence) Time Between Attacks: Positive CSF is: The diagnosis is either: The McDonald criteria make use of the clinical presentation and the advances of MR imaging. When a patient presents with 2 or more attacks with clinical evidence of 2 or more neurological deficits, there is no need for additional requirements to make the diagnosis of MS, because there is dissemination in place and time. In all other cases (less than 2 attacks or less than 2 clinical lesions) there is a role for MRI to fulfill the diagnostic criteria by demonstrating dissemination in space, in time or both. The McDonald criteria are very specific, because if you want to use MRI for the diagnosis of MS, you have to make sure that the patient really has MS. You do not want a patient to start treatment daily if there is any doubt about the diagnosis. For dissemination in space (DIS) lesions in two out of four typical areas of the CNS are required: For dissemination in time (DIT) there are two possibilities: Indications for MRI of the brain are: Gadolinium is administered at the start of the examination because the longer you wait the more enhancement you will see on the T1W images (MS lesions are not spontaneously bright on T1-weighted images without contrast administration). A scout with additional mid-sagittal T1WI is made for optimal and constant positioning. The sagittal FLAIR is ideal for detection of lesions in the corpus callosum and the 3D sequence allows for better detection of smaller and juxtacortical lesions. The PD/T2W scan is preferably conventional SE or TSE/FSE. Finally the axial T1W-images are made after about 15 minutes to provide optimal contrast enhancement. Coronal and midsagittal scout views are needed for reproducible positioning of the slices, so you are able to compare follow-up studies. Use the coronal scout to plan the true midsagittal image parallel to the falx and other midline structures. On a true midsagittal image a line is drawn through the pituitary gland and the roof of the fourth ventricle (fastigium). This is called the HYFA: hypophysis-fastigium line. Subsequently the slices are positioned with the middle slice at the lower border of the splenium of the corpus callosum. Indications for MRI of the spinal cord are: Gadolinium is not necessary when only the spinal cord is examined. Contrary to the brain there will only rarely be enhancement in the cord. Only when other diagnoses are considered (e.g. sarcoid) Gd is necessary. The most diagnostic sequence is the conventional SE or FSE (TSE) PDW, because this is the most sensitive technique. FLAIR should NOT be used in the spinal cord and will only demonstrate 10% of the lesions. When we look at the prevalence of the white matter diseases, you will notice that there are enormous differences. Hereditary diseases are extremely uncommon as individual diseases, but as a group they are not that uncommon, but still far more uncommon than MS. If we look at the prevalence of Lyme disease, which is a rather popular disease at the moment, then we will notice that it still is a very uncommon disease despite of all the serological tests that are being performed nowadays. When incidental WMLs are found, these are usually the result of small vessel disease, since up to 50% of patients that get an MR examination for whatever reason, will have WMLs of vascular origin. They are more common in older people and in patients with vascular risk factors like atherosclerosis, high blood pressure, high cholesterol, diabetes, amyloid angiopathy, hyperhomocysteinemia, atrial fibrillation etc. If a patient is clinically suspected of having MS and the MR-images support that diagnosis, than you should not consider the possibility of Lyme's disease and neuro-SLE in the differential diagnosis, because they have such a low prevalence. There must be other ways to impress your colleagues. These diagnoses are only worth mentioning if there are clinical findings that support these diagnoses. Consequently, it is not wise to put MS in the differential diagnosis if the clinician does not suspect the patient of having MS and on the MR incidental WMLs are found. The odds are against the diagnosis of MS, because vascular WMLs are 50-500 times more likely than MS plaques. On the other hand if a patient is clinically suspected of having MS and multiple WMLs are found, our major concern is the differential diagnosis MS versus vascular disease and we have to follow the McDonald criteria. The differential diagnosis of white matter lesions is extremely long. In normal ageing WMLs are seen, but most WMLs are acquired and of hypoxic-ischemic origin. The most common inflammatory disease is Multiple Sclerosis. The most common viral infections are PML and HIV. Inherited diseases usually will have symmetrical abnormalities, so they have to be differentiated from intoxications. On the left a collection of images with multiple punctate and patchy lesions in the WM. Some will be discussed in more detail. There is no complete overlap between the images on the left and the text on the right. Borderzone infarction Key finding: typically these lesions are located in only one hemisphere, either in deep watershed area or peripheral watershed area. In the case on the left the infarction is in the deep watershed area. ADEM Key findings: Multifocal lesions in WM and basal ganglia 10-14 days following infection or vaccination. As in MS, ADEM can involve the spinal cord, U-fibers and corpus callosum and sometimes show enhancement. Different from MS is that the lesions are often large and in a younger age group. The disease is monophasic. Lyme 2-3mm lesions simulating MS in a patient with skin rash and influenza-like illness. Other findings are high signal in spinal cord and enhancement of CN7 (root entry zone). Sarcoid Sarcoid is the great mimicker. The distribution of lesions is quite similar to MS. PML Progressive Multifocal Leukoencephalopathy (PML) is a demyelinating disease caused by JC virus in immunosuppressed patients. Key finding: space-occupying, nonenhancing WMLs in the U-fibers (unlike HIV or CMV). PML may be unilateral, but more often it is asymmetrical and bilateral. Virchow Robin spaces Key finding: Bright on T2WI and dark on FLAIR. Small vessel disease WMLs in the deep white matter. Not located in corpus callosum, juxtaventricular or juxtacortical. In many cases there are also On the left a collection of images with multiple enhancing lesions in the WM. Some will be discussed in more detail. There is no complete overlap between the images on the left and the text on the right. Vasculitis Most diseases with vasculitis are characterized by punctiform enhancement. Vasculitis in the brain is seen in SLE, PAN, Behcet, syphilis, Wegener, Sjogren and Primary angiitis of CNS Behcet Behcet is more commonly seen in Turkish patients. Typical findings are brainstem lesions with nodular enhancement in the acute phase Metastases Metastases are mostly surrounded by a lot of edema. Borderzone infarction Peripheral border zone infarctions may enhance in the early phase. First look at the images on the left and describe the lesions. Then continue. On the T2W image there are multiple high intensity lesions in the basal ganglia. On the FLAIR image these lesions are dark, so they follow the intensity of CSF on all sequences (they were hypointense ion the T1WI). This signal intensity in combination with the location is typical for VR spaces. Virchow Robin spaces are CSF spaces around penetrating leptomeningeal vessels. They are typically located in basal ganglia, around atria, near the anterior commissure and in the middle of the brainstem. On MR they follow the signal intensity of CSF on all sequences. They are dark on FLAIR and PD unlike other WMLs. Usually they are small except around the anterior commissure, where perivascular spaces can be larger. On this image we see both very wide VR spaces as well as confluent hyperintense lesions in the WM. This case nicely illustrates the difference between VR spaces and WMLs. This is an extreme case and this condition is known as etat crible. VR spaces enlarge with age and hypertension as a result of atrophy of surrounding structures. In normal ageing we can see: Periventricular caps are hyperintense regions around the anterior and posterior pole of the lateral ventricles and are associated with myelin pallor and dilated perivascular spaces. Periventricular bands or 'rims' are thin linear lesions along the body of the lateral ventricles and are associated with subependymal gliosis. Normal Aging (2) The clinical significance of white matter changes in aging has not been fully elucidated. There is a relationship between several cerebrovascular risk factors and the presence of white matter changes. One of the strongest risk factors however, apart from hypertension, is that of age. What is still considered normal depends on the age of the patient. These white matter changes are classified according to Fazekas: First look at the images on the left and describe the lesions. Then continue. The location of these white matter lesions is in the deep white matter and it is important to notice that these lesions are not juxtaventricular, not juxtacortical and not located in the corpus callosum. Unlike in MS, they do not touch the ventricles or the cortex. Given the a priori greater chance of hypoxic-ischemic WM lesions, we must conclude that these WMLs probably have a vascular origin. Only if the clinical findings strongly direct us towards inflammatory, infectious, toxic or other diseases, should we consider these diagnoses. Suggesting the diagnosis of MS in a patient with these MR findings and with no clinical suspicion for MS would be unwise. The spinal cord in this patient was normal. In a patient with vasculitis or ischemia the spinal cord is usually normal, while in a MS patient in more than 90% of the cases it will be abnormal (2). If the differentiation between a vascular origin of WMLs and MS is difficult for instance in an older patient who is suspected of MS, than an MR of the spinal cord can be helpful (2). Vascular disease (2) When we go back to the first case that was shown, it is now very obvious that this is vascular disease. There is widespread disease in the deep white matter, but the U-fibers and corpus callosum are not involved. Ischemic WMLs present as lacunar infarcts, watershed infarcts or diffuse hyperintense lesions within the deep white matter. Lacunar infarcts are due to arteriolar sclerosis of small penetrating medullary arteries. Watershed infarctions are the result of atherosclerosis of larger vessels, for instance carotid obstruction or the result of hypoperfusion. Atherosclerotic brain changes are seen in 50% of patients older than 50 years. They are found in normotensive patients, but more common in hypertensives. First look at the images and describe the lesions. Then continue. The distribution of lesions is quite similar to MS. Besides lesions in the deep WM, there are some juxtaiventricular lesions and even Dawson finger-like lesions. The final diagnosis was sarcoid. Sarcoid has surpassed neurosyphilis as the great mimicker. Sarcoid (2) On the left we see the coronal Gd-enhanced T1W images of this patient. First study these images, than continue. There is punctate enhancement in the basal nuclei. This is seen in sarcoid and can also be seen in SLE or other vasculitis. Typical for sarcoid in this case is the leptomeningeal enhancement (yellow arrow). This is the result of granulomatous inflammation of the leptomeninges. Sarcoid (3) Another typical finding in this same case is the linear enhancement (yellow arrow). This is due to inflammation along the Virchow Robin spaces. This is also a form of leptomeningeal enhancement. This explains why sarcoid has a similar distribution to MS: the Virchow Robin spaces follow the small penetrating veins, which are involved in MS. Lyme disease is caused by a spirochete (borreliaBurgdorferi) that is transmitted by a tick. It first causes a skin rash. A few months later the spirochete can infect the CNS and MS-like WMLs are seen. Clinically Lyme presents with acute CNS symptoms (e.g.cranial nerve palsy) and sometimes transverse myelitis. Lyme disease (2) Key finding: 2-3mm lesions simulating MS in a patient with skin rash and influenza-like illness. Other findings are high signal in spinal cord and enhancement of CN7 (root entry zone). Key finding: Atrophy and symmetric periventricular or more diffuse WMLs in AIDS patient Cadasil is short for cerebral autosomal dominant arteriopathy with subcortical infarcts and leukencephalopathy. It is an inherited small vessel disease. Clinical clues: migraine, dementia and family history. Key finding: subcortical lacunar infarcts with small cystic lesions and leukoencephalopathy in young adults. Localizations of the WMLs in the anterior temporal pole and external capsule have a high specificity. F Barkhof, M Filippi, DH Miller, P Scheltens, A Campi, CH Polman, G Comi, HJ Ader, N Losseff and J Valk Department of Diagnostic Radiology, Vrije Universiteit Hospital, Amsterdam, The Netherlands. Brain, Vol 120, Issue 11 2059-2069 Joost C. J. Bot, MD, Frederik Barkhof, MD, PhD, Geert Lycklama ? Nijeholt, MD, PhD et al. Radiology 2002;223:46-56. Tintore M, Rovira A, Martinez MJ, et al. AJNR Am J Neuroradiol 2000;21:702-706 McDonald WI, Compston A, Edan G, et al. Ann Neurol 2001;50:121-127 Bot JC, Barkhof F, Polman CH, et al. Neurology 2004;62:226-233 Polman CH, Reingold SC, Edan G et al. Ann Neurol. 2005 Dec;58(6):840-6. M. Inglese American Journal of Neuroradiology 27:954-957, May 2006 Diagnostic Imaging: Brain by Anne Osborn, Susan Blaser, Karen Salzman by Robert I. Grossman and David M. Yousem. 2nd ed. St. Louis, MO: Mosby; 2003. Magnetic Resonance Imaging of CNS disease, A Teaching File by Douglas H. Yock second edition. Mosby Polman et al, Ann Neurol 2011; 69:292-302Frederik Barkhof, Robin Smithuis and Marieke Hazewinkel Typical MRI findings in MS Dawson fingers Tumefactive MS Balo’s Concentric Sclerosis Neuromyelitis Optica ADEM MS Brain Protocol MS Spinal cord Protocol DD multiple patchy lesions DD multiple enhancing lesions Virchow Robin spaces Normal Aging Vascular disease Sarcoid Lyme disease HIV CadasilMultiple SclerosisFrom the MR Center for MS Research, Radiology Department of the 'Vrije Universiteit' Medical Center, Amsterdam and the Rijnland Hospital, Leiderdorp, the Netherlands
ped5
1
Pediatric Chest CT 1 - Nonvascular Mediastinal Masses
by Marilyn J. Siegel and Valerie Niehe
This review is based on a presentation by Marilyn Siegel and was adapted for the Radiology Assistant by Valerie Niehe. Marilyn Siegel is specialized in pediatric and chest radiology. In this review we will discuss the most common non-vascular mediastinal masses in the chest. In Pediatric Chest CT part II we will discuss the most common vascular anomalies of the aorta, pulmonary vessels and systemic veins in the chest.  The differential diagnosis of a mediastinal mass is based on identifying its location in anterior, middle or posterior mediastinum and attenuation: soft tissue, fat, fluid and enhancement. In infants and young children ( In older children, the thymus gradually assumes a triangular or arrowhead configuration with straight or concave margins. By 15 years of age it is triangular in nearly all individuals. Marked lobularity of the thymus is always abnormal. In prepubertal children, the thymus is homogeneous. The attenuation value is equal to that of skeletal muscle. In adolescents it may be heterogeneous, containing areas of fat. Anatomic variations include extension into the posterior mediastinum or upper neck. Clues to the diagnosis are: The figure shows a thymus that extends cranially to the brachiocephalic vessels. It is contiguous with the normal thymus and extend between the superior vena cava and the trachea. There is no mass effect. There are no well-established data concerning size of normal lymph nodes in infants and young children. Mediastinal lymph nodes are generally not seen on CT prior to puberty. The nodes should then not exceed 1 cm in the widest dimension. The azygoesophageal recess is dextroconvex in children younger than 6 years of age, straight in children between 5 and 12 years of age, and concave in adolescents and adults. Recognizing the normal dextroconvex appearance is important so that it is not mistaken for lymphadenopathy. Anterior mediastinal masses are usually of thymic origin.  Lymphoma is the most common anterior mediastinal mass in children, with Hodgkin lymphoma occurring three to four times more frequently than non-Hodgkin lymphoma. Calcifications or cystic areas, due to ischemic necrosis consequent to rapid tumor growth, can be seen. Lymphadenopathy from lymphoma has varied appearances, ranging from mildly enlarged nodes in a single area to large conglomerate soft tissue masses in multiple regions. Thymic enlargement and lymphadenopathy show minimal if any enhancement after intravenous contrast. Additional findings include airway narrowing and compression of vascular structures. Hodgkin lymphoma in children is more common in the second decade of life. The lymphomatous mass is most common located in the anterior mediastinum and reflects lymphadenopathy or infiltration and enlargement of the thymus. The enlarged thymus has a quadrilateral shape with convex, lobular lateral borders. The chest film shows the typical features of Hodgkin lymphoma, e.g., an anterior mediastinal mass. The CT-images of the same patient show a large soft tissue mass in the anterior mediastinum, which arises in the thymus. There is associated paratracheal adenopathy (arrow). Two more cases of Hodgkin lymphoma. Again these cases show an anterior mediastinal mass and paratracheal adenopathy. Non-Hodgkin disease in children occurs in the first and second decade of life. The disease usually involves the nodes in the chest (paratracheal, subcarinal and hilar). The extension of the disease is not contiguous, it can skip a location. Non-Hodgkin disease, in contrast to Hodgkin disease, often spares the thymus. In this case, enlarged lymph nodes are seen in the right paratracheal , hilar and subcarinal areas. Thymic hyperplasia is another cause of thymic enlargement. In childhood, thymic hyperplasia is most often 'rebound' hyperplasia associated with chemotherapy, particularly therapy with corticosteroids. Rebound hyperplasia may be observed during the course of chemotherapy or after therapy completion and occurs 3 to 10 months after the start of chemotherapy. The mechanism of hyperplasia is believed to be initial depletion of lymphocytes from the cortical portion of the gland due to high serum levels of glucocorticoids, followed by repopulation of the cortical lymphocytes when the cortisone levels return to normal. On CT, hyperplasia appears as diffuse enlargement of the thymus, with preservation of the normal triangular shape. The definition of thymic hyperplasia is a > 50 % increase in volume of the thymus. CT, MRI of PET cannot differentiate rebound hyperplasia from infiltration of the thymus by tumor. The absence of other active disease and a gradual decrease in thymus size on serial CT's supports the diagnosis of rebound hyperplasia. The thymus usually returns to its normal size in 3 to 6 months. Thymomas are common and account for 20% of mediastinal neoplasms. Thymic carcinomas are extremely rare and acount for less than 1% of all thymic tumors. The images show a thymoma on the left and a carcinoma on the right. The thymic carcinoma has invaded the superior vena cava (arrow). Germ-cell tumors are the most common cause of a fat containing lesions in the anterior mediastinum and the second most common cause of an anterior mediastinal mass in children. Approximately 90 % are benign germ-cell tumors. Most arise in the thymus. On CT, a benign teratoma is a well-defined, thick-wall cystic mass containing a variable mixture of water, calcium, fat and soft tissue. The soft tissue component in benign teratoma is minimal. Size is not an indicator of malignancy. Mature teratomas can be very large and still be benign. Malignant teratomas make up 10 % of all teratomas. They tend to have irregular or nodular walls and a predominance of soft tissue components. They also may show pulmonary or liver metastases and chest wall invasion. The most common nonteratomatous germ-cell tumors in the pediatric population are choriocarcinoma, embryonal cell cancer and yolk-sac cancer. Thymolipoma is an infrequent fat-containing thymic tumor. At CT, it appears as a heterogeneous mass containing fat and soft tissue elements. Calcifications are absent. Thymolipoma does not have a capsule and does not have any mass effect. Lymphangiomas are developmental tumors of the lymphatic system. In the mediastinum they are almost always an inferior extension of a cervical lymphangioma. Isolated mediastinal involvement is rare. Lymphangioma is a benign, but aggressive tumor that shows mass effect and may encase vessels and other structures. It typically affects infants younger than 6 months of age. At CT it appears as nonenhancing, thin-walled, multiloculated mass with near water attenuation. MRI may sometimes be used to better delineate the extension of the lesion. The MRI in this patient shows a cystic mass in the neck extending into the right axilla and mediastinum. The tumor encases vessels.
 The presence of contrast enhancement of the wall or internal septations suggests superimposed infection or a hemangiomatous component. Thymic cysts are usually congenital lesions resulting from persistence of the thymopharyngeal duct. They can also occur after thoracotomy. Typically, they are thinwalled, homogeneous masses of near water attenuation on CT. The attenuation value may be higher than that of simple cysts when the contents are proteinaceous or hemorrhagic rather than serous. In children thymoma, thymic carcinoma and goiter are so uncommon, that you should put them very low in your differential diagnosis. In the middle mediastinum we will find foregut duplication cysts or lymph nodes. Foregut cysts in the middle mediastinum are classified as bronchogenic or enteric. Bronchogenic cysts are lined by respiratory epithelium and most are located in the subcarinal or right paratracheal area in close proximity to the trachea or bronchus. Enteric cysts are lined by gastrointestinal mucosa and are located in a paraspinal position in the middle to posterior mediastinum near the esophagus. The images show a well defined lesion of water attenuation in close proximity to the trachea or bronchus, which is typical for a bronchogenic foregut cyst. The images show more examples of bronchogenic cysts and their close proximity to the airway. The images show a well defined lesion of water attenuation in the lower mediastinum in close proximity to the esophagus, which is typical for an enteric foregut cyst. Mediastinal lymphadenopathy is usually caused by lymphoma or granulomatous disease. Metastatic disease from rhabdomyosarcoma, osteosarcoma or a Wilms tumor, is also a possibility. On CT, adenopathy can appear as discrete, round, soft tissue masses or as a single soft tissue mass with poorly defined margins. Calcification within lymph nodes suggests old healed granulomatous disease, fungal infection or metastatic disease from osteosarcoma. Areas of low attenuation suggest tuberculosis or fungal infection. Posterior mediastinal masses are of neural origin in approximately 95 % of cases and may arise from sympathetic ganglion cells (neuroblastoma, ganglioneuroblastoma or ganglioneuroma) or from nerve sheaths (neurofibroma or schwannoma). In the first decade of life they are usually malignant, most commonly neuroblastoma. In the second decade or life they are usually benign (ganglioneuroma, neurofibroma, rarely schwanoma). Neuroblastoma typically is fusiform in shape, of soft tissue density; 50% of thoracic tumors have calcifications. Neuroblastoma grows over several interspaces and frequently invades the vertebral canal. The CT-images show a calcified mass in the posterior mediastinum extending over several vertebrae, which grows into the vertebral canal. On the MR-images the invasion of the vetebral canal is better seen (arrows). In the 2nd decade other neurogenic tumors are seen like ganglioneuroma, neurofibroma and rarely schwanoma. They are round or oval in shape, smaller in size than ganglion cell tumors and usually extend over only one or two vertebrae. 
Both types of tumor may cause pressure erosion of a rib and invade the spinal canal. Neurogenic cysts contain neural and gastrointestinal element. They are commonly associated with vertebral anomalies and scoliosis. The cyst does not communicate with CSF. The cyst is well demarcated and has a near water attenuation value on CT and water signal intensity on MRI, as shown in the case on the left. Extramedullary hematopoiesis accounts for less than 0.1 % of the lesions in the posterior mediastinum. It is characterized by formation of blood elements outside of the bone marrow. It occurs with severe anemia. On CT it is seen as a paravertebral mass and occurs with coarse bone trabeculation in the adjacent vertebra.Marilyn J. Siegel and Valerie Niehe Thymus Normal Lymph nodes Hodgkin lymphoma Non-Hodgkin lymphoma Thymic hyperplasia Thymoma Germ-cell tumors Thymolipoma Lymphangioma or Cystic hygroma Thymic cysts Bronchogenic Cysts Enteric foregut cyst Mediastinal lymphadenopathy Neuroblastoma Other Neurogenic Tumors Neurenteric Cyst Extramedullary hematopoiesisPediatric Chest CT 1 - Nonvascular Mediastinal MassesMallinckrodt Institute of Radiology Washington University School of Medicine St. Louis, MO and the Medical Centre Haaglanden in the Hague, the Netherlands
abdomen8
1
Biliary Ducts - pathology
by Angela D. Levy MD
This review is based on a presentation given by Angela Levy and adapted for the Radiology Assistant by Robin Smithuis. We will discuss: by Angela D. Levy Whenever there is bile duct dilatation, the first priority is to look for obstruction. Obstruction If there is an obstruction, we first look for gallstones in the bile duct. If there are no gallstones involved, we then look for strictures. The differential diagnosis for a stricture is based on the location. No Obstruction Once we have excluded obstruction, we have to think about nonobstructive biliary diseases like: Caroli disease is an autosomal recessive disease secondary to the ductal plate malformation. It is associated with polycystic kidney disease, medullary sponge kidney and medullary cystic disease. So looking at the kidneys can sometimes help you make this diagnosis. On the left we see images of a patient with Caroli disease. Notice the intrahepatic duct dilatation and the normal caliber of the choledochal duct (extrahepatic bile duct). The hallmark of Caroli disease is intrahepatic duct dilatation. The dilatation can be very large and saccular as seen in the case on the left or it can be very linear. Caroli Disease (2) The duct dilatation in Caroli disease is due to a congenital malformation of the ductal plate, which is the precursor of the intrahepatic bile ducts. On the left we see the normal development of the ductal plate. Embryologically each bile duct begins as a single layer of cells that surrounds a portal vein. This layer then duplicates. Portions of this double layer fuse and resorb leaving unfused portions that become the bile ducts. So in the normal situation each portal vein is surrounded by interconnecting bile ducts (left image). However if the patient has ductal plate malformation, the bile ducts are too numerous and they are ectatic (right image). Whether or not we see this on imaging depends on which portion of the bile ducts is affected. If the large ducts are involved, we see this as Caroli disease. However if only the very small ducts are involved, the result is congenital hepatic fibrosis. If all ducts are involved, then there is a combination of fibrosis and Caroli disease, which is also known as the Caroli syndrome. Most commonly the intrahepatic duct dilatation is segmental (83%) in distribution. The diffuse form is less common (17%). The shape of the dilatation is saccular in 76% or fusiform in 24% of the cases. A very important sign is the central dot sign. The central dot corresponds to the portal vein that is surrounded by dilated bile ducts. On the left we see a dot within the dilated ducts. When we put on the color doppler, we will notice that these structures contain blood flow and represent portal veins. Caroli Disease (3) On the left CT-images of the same patient. Notice the central dot sign and the segmental involvement. This patient has cirrhosis with splenomegaly due to portal hypertension. Extrahepatic duct dilatation is present in 53% of cases, secondary to cholangitis and stone or sludge passage. These are secondary findings, that are not part of the primary disease. When there is extensive fibrosis, these patients can develop cirrhosis over time. Caroli Disease (4) The cholangiogram is important in the work up of these patients, because obstruction has to be excluded. This can be done with MRCP or ERCP, as is shown on the left. There was no sign of obstruction. The mild dilatation of the choledochal duct was the result of cholangitis. Study the image on the left. Then continue reading. There is focal dilatation with intermixing strictures of the bile ducts in segment IV (arrow). The other bile ducts and the choledochal duct are normal. In some of the cases of Caroli disease the imaging findings may simulate a cystic neoplasm as is seen in the case on the left. This case was originally diagnosed as a biliary cystadenoma. However, the gross specimen demonstrates dilated bile ducts and ductal plate malformation was present microscopically. Caroli Disease (5): Complications Patient with Caroli disease are usually brought to our attention, when they develop complications. These complications are commonly the result of biliary stasis, which leads to stone formation and infection. Complications: On the left a patient with dilated bile ducts with intraductal stones (arrow) Study the image on the left. Then continue reading. The findings are: The mass in the right lobe of the liver turned out to be an abscess. Remember that liver abscesses in the early phase can look quite solid. In the differential diagnosis we would also have to include a neoplasm, because patients with Caroli disease have an increased risk of developing a cholangiocarcinoma. Ultimately if there is substantial fibrosis and the entire liver is involved, these patients develop cirrhosis. The patient on the left developed severe liver failure and a resection was performed. Notice the intrahepatic bile duct dilatation, splenomegaly and dilated venous collaterals. In the resected specimen there is a central dot sign (blue arrow) and a small pus collection (yellow arrow). Cholangiocarcinoma can take on many forms in patients with Caroli disease. The cholangiogram on the left demonstrates a focal stricture (blue arrow), which turned out to be a infiltrating cholangiocarcinoma. The patient on the right has a cholangiocarcinoma in which the tumor was filling up the dilated ducts (yellow arrow). A choledochal cyst is a congenital dilatation of the extrahepatic bile duct. These patients do not have a ductal plate abnormality. The most common theory for the development of a choledochal cyst is that the dilatation is due to an underlying anomalous pancreatico-biliary junction. In the anomalous junction the biliary and pancreatic duct join proximal to the sphincter of Oddi. In these patients there is a long common channel. The theory is that when the sphincter of Oddi contracts, pancreatic enzymes will flow into the bile duct and causing dilatation and in some cases, narrowing of the distal duct. This classification classifies the choledochal cysts into 5 cathegories. Type V, which is not shown on the left is Caroli disease. We now know, that Caroli is a different disease. Type I is a true choledichal cyst with focal dilatation of the extrahepatic duct. This is the most frequent type (90-95% of the cases). Type IV is also a true choledichal cyst with dilatation of the entire extrahepatic duct with involvement of portions of the intrahepatic ducts. The intrahepatic ducts taper normally to the periphery, indicating that there is no obstruction. Type II and III are extremely rare and it is debatable whether or not these are true choledochal cysts. Type II is a diverticulum of the extrahepatic duct and many believe that this entity is not related to an anomalous pancreatico-biliary junction. Type III is a choledochocele, where there is dilatation of the distal part of the bile duct. These patients also have a normal pancreaticobiliary junction. Choledochal cyst (2) On the left a patient with dilatation of both the extrahepatic duct and part of the intrahepatic ducts. So this is a type IV. Notice that the peripheral ducts are normal, so this is not an obstructive pattern. On the left another type IV choledochal cyst. There is dilatation of the extrahepatic duct, cystic duct and a small portion of the left hepatic duct. There is no intrahepatic dilatation. Choledochal cyst (3) There is an association of bile duct adenocarcinoma and choledochal cysts. These carcinomas can occur within the choledochal cyst, like the case on the left, or in the gallbladder or anywhere else in the biliary ducts. In the bile ducts they can present as classic peripheral cholangiocarcinoma, Klatskin tumor or distal cholangiocarcinoma. Recurrent pyogenic cholangitis is an uncommon disease in the western world. Most of these cases are seen in Asian countries. The etiology is unknown, although some of these patients have biliary parasites. The disease is characterized by the presence of intrahepatic pigmented stones and recurrent infection. These patients are also at risk of developing biliary cirrhosis and cholangiocarcinoma. The left lobe is the most common location of the disease due to the delayed drainage of the left system. On the left a typical case. There is focal dilatation of the bile ducts in the left lobe with stones. Recurrent pyogenic cholangitis (2) On the left another example of recurrent pyogenic cholangitis. There is intrahepatic lithiasis with focal diatation. A case like this is indistinguishable from focal Caroli disease with secundary stone formation. When we see intrahepatic bile duct dilatation with strictures and only mild dilatation, the first diagnosis we think of is primary sclerosing cholangitis (PSC). We know however that there is a long differential diagnistic list which includes: The case on the left nicely demonstrates the bile duct strictures. Notice that there is only mild dilatation, which is common in PSC. The hallmark of PSC is strictures, but early on in the disease the strictures can be difficult to appreciate. The underlying abnormality in PSC is fibrosis, which is of unknown etiology. PSC is strongly associated with ulcerative colitis in up to 70% of patients, but it can also be associated with Crohn's disease of the large intestine. The reason for the association with IBD is unknown, but it is thought to be the result of an immune response. Primary sclerosing cholangitis (2) On the far left a histologic specimen demonstrating chronic inflammation surrounding the bile duct. The gross specimen demonstrates the thickening of the wall of the bile duct (arrow) causing many strictures. The strictures in PSC are short in the order of 3-5 mm in length, which is helpful to remember because if you are looking for cholangiocarcinoma, the malignant strictures usually exceed 10 mm in length. One of the earliest features in PSC is on sonography where we see thickening of the wall of the bile duct as is seen in the image on the left. This patient came for an ultrasound examination to rule out gallstones. Notice that the intrahepatic ducts are normal. The differential diagnosis would include PSC, AIDS-cholangitis and cholangiocarcinoma. A cholangiocarcinoma would be rather unlikely, because there is no obstruction. Continue with the CT. On the CT the liver looks quite normal. However if you look at the common bile duct in the pancreatic head you will notice the soft tissue density. The differential diagnosis would include an impacted stone or cholangiocarcinoma, but since this patient had no obstruction, it was thought to be the result of bile wall thickening. The gallbladder wall is also thickened. Primary sclerosing cholangitis (3) On the left a patient with more severe disease. We can see mild bile duct dilatation with a discontinuous pattern. On the left a patient with more pronounced CT findings. Study the image and then continue reading. The findings are: Primary sclerosing cholangitis (4) Late CT features are seen on the images on the left. Study the images and then continue reading. The findings are: Cholangiography is used in the initial diagnosis of the disease, when there are only subtle strictures and in patients known with PSC to look for new strictures that are suspicious for carcinoma. On cholangiography we can see: On the left the typical findings in PSC. Notice the diverticula on the image on the right. Diverticula are very specific for the diagnosis PSC. So when you see these diverticula, you should immediately search for subtle strictures in the intrahepatic ducts. Primary sclerosing cholangitis (5) On the left a MRCP in a pateitnt with PSC. Notic the large stricture, which is quite worriesome for cholagiocarcinoma (arrow). The strictures in PSC show an abrupt transission, while here we see 'shouldering', which indicates mass-efect. In addition there is intrahepatic dilatation proximal to this stenotic area. On the left a MRCP in a patient demonstrating a stricture at the level of the hilum. On MRCP this stricture looked long and worriesome for cholangiocarcinoma. However, on the ERCP, the ducts have been distended with contrast and we can see that this is a short stricture compatible with the diagnosis of PSC. During follow up this proved to be just PSC. Cholangiocarcinoma (i.e., adenocarcinoma of the bile duct) arises from the columnar epithelium of the bile duct. It is characterized by malignant glands within a desmoplastic stroma. These tumors have an infiltrative growth pattern and do not have a capsule. There are four basic patterns of chlangiocarcinoma: (illustration on the left). Cholangiocarcinoma is an uncommon tumor, that is mostly seen in patients with underlying benign bilairy disease. The incidence in the U.S. is 2000 to 2500 cases per year (coloncancer 150.000 per year). In Asian countries the incidence is ten times greater due to more chronic biliary infection. High risk groups are patients with: Clinical presentation: These arise in the very small peripheral ducts. These tumors have abundant fibrous stroma that can cause retraction of the liver capsule. The tumor typically enhances in the equilibrium and delayed phases (5-10 minutes). Although these tumors are usually quite heterogeneous because the contrast uptake is delayed and can be irregular. The key findings to look for are: On the left a typical case. Notice the capsular retraction (blue arrow) and the late enhancement (yellow arrow). These are very rare tumors. They present as a intrabiliary mass with biliary dilatation peripheral to the mass. The most common site of biliary adenocarcinoma is at or near the confluence of the right and left hepatic ducts. These tumors are also known as Klatskin tumors. Klatskin tumors have an aggressive biologic behavior. Imaging features: On the left a nice correlation between an illustration and a sonographic image of a Klatskin tumor. Notice how ill-defined the tumor is. On CT these tumors can be very difficult to visualize. Many of these patients already have a stent in place when they first come to CT which makes it even harder (figure). Klatskin Tumor (2) In the case on the left we can identify the tumor at the confluens of the left and right hepatic duct. The margins of the tumor however are imperceptible because of the infiltrative growth. Based on the CT it is not possible to stage the tumor correctly. In these tumors it may be difficult to get a definitive diagnosis pre-operatively. Biopsy is almost impossible and results of endoscopic brushing are commonly negative. The staging is done with cholangiography and is based on the finding of mass effect (shouldering), irregular margins and abrupt tapering at the obstruction. The limitations of MRCP in staging are the spatial resolution and the inability in the evaluation of the secondary ducts. ERCP is superior to MRCP (figure) Klatskin Tumor (3) : Resectability These tumors are unresectable when: Bismuth-Corlette type I A type I tumor is a lesion limited of the common hepatic duct, i.e. below the confluence. These patients can undergo resection with bile duct recontruction because the confluence is normal. Bismuth-Corlette type II A type II tumor is a lesion that extends to the confluence. These tumors are potentially resectable Bismuth-Corlette type IIIa and IIIb A IIIa-tumor extends into the right and a IIIb-tumor extends into the left hepatic duct. On the left PTC-images of a type IIIa-tumor. The arrow indicates the extention into the right hepatic duct. The left duct is normal. This patient can undergo a resection of the right lobe of the liver. Bismuth-Corlette type IV On the left an illustration and ERCP of a type IV-tumor with extention into the right and left duct. A type IV tumor is unresectable.Angela D. Levy MD Differential Diagnosis of bile duct dilatation Central dot sign Todani Classification Ultrasound findings CT findings Intrahepatic cholangiocarcinoma Intraductal Cholangiocarcinoma Klatskin Tumor - Hilar Cholangiocarcinoma Klatskin Tumor - Bismuth-classificationBiliary Ducts - pathologyChief Gastrointestinal Radiology, University Department of Radiologic Pathology, Armed Forces Institute of Pathology, Washington DC
abdomen1
1
Closed Loop Obstruction in Small bowel obstruction
by by Jay P. Heiken and Robin Smithuis
This article is based on a presentation given by Jay Heiken and adapted for the Radiology Assistant by Robin Smithuis. Jay Heiken is professor of radiology at the Mallinckrodt Institute of Radiology of the Washington University School of Medicine in St. Louis. He has a special interest in abdominal imaging and is co-author of the well known book 'Computed Body Tomography With MRI Correlation'. Closed Loop Obstruction Closed loop obstruction is a specific type of obstruction in which two points along the course of a bowel are obstructed at a single location thus forming a closed loop. Usually this is due to adhesions, a twist of the mesentery or internal herniation. In the large bowel it is known as a volvulus. In the small bowel it is simply known as small bowel closed loop obstruction. Especially in the small bowel the risk of strangulation and bowel infarction is high with a mortality rate of 10-35%. When we have a patient in the ER with what appears to be a small bowel obstruction (SBO), the most important thing we can do, besides making the diagnosis, is to identify the presence or absence of strangulation. Strangulation is defined as obstruction associated with vascular compromise. The morbidity and mortality rate in the SBO-group is mainly due to bowel infarction and subsequent necrosis. This is most commonly caused by a closed loop obstruction. CT is the imaging procedure of choice in the evaluation of patients suspected of SBO. The CT-presentation of a closed loop obstruction in the small bowel depends on two things: If we have a short closed loop oriented within the plane of imaging, we will see a U- or C-shaped loop of bowel. Another important appearance of a closed loop obstruction is that of a radial array of dilated small bowel loops with the mesenteric vessels converging to a central point. This is almost always due to a small bowel volvulus. The findings of ischemia in closed loop obstruction are the same as in patients with other causes of mesenteric ischemia: The case on the left shows another patient with closed loop obstruction. Although there is good enhancement of the vessels there seems to be a lack of enhancement of the bowel wall. Other signs of ischemia in this case are mesenteric edema and bowel wall thickening. Infarcted bowel was found at operation. If the closed loop is longer and is oriented perpendicular to the plane of section, we will see a clump of bowel loops as shown in the case on the left. Sometimes this is difficult to appreciate on just the axial images and coronal or sagittal reconstructions can be helpful. In this case there is also mesenteric edema and localised ascites in combination with dilated loops with wall thickening indicating strangulation and risk of infarction. CT is the imaging procedure of choise in patients who are suspected for bowel obstruction. When we examine these patients, we should not give oral contrast for following reasons: In some of the patients with a closed loop obstruction a bowel obstruction is not suspected. In the case on the left positive oral contrast was given. Notice the constriction in the small bowel in figure B. Distal to the constriction in figure C we see a cluster of dilated small bowel loops not filled with oral contrast, indicating the closed loop. Only rarely contrast will pass the point of obstruction and enter the area of the closed loop. If we go back to figure B, you may already have noticed that there are two points of narrowing in the small bowel (arrows). Therefore we have two adjacent collapsed small bowel segments representing the point of the closed loop obstruction. The bowel wall thickening, ascites and mesenteric edema indicate the presence of bowel ischemia. Notice that you cannot appreciate the degree of bowel wall enhancement in the loops that are filled with oral contrast. In some of these patients with SBO the proximal small bowel proximal to the point of obstruction may not be dilated. On the left we see images of a patient in whom obstruction was not suspected. This patient also received positive oral contrast. Look for the major findings and then continue. First you will notice that the small bowel is not dilated. When you go down to the pelvis you see a dilated loop of bowel with inhomogeneous content and finally deep down in the pelvis there is a C-shaped dilated bowel indicating a closed loop obstruction. The other important finding in this patient is the 'Small Bowel Feces Sign' (SBFS: arrow). The SBFS is a very useful sign as it is seen at the zone of transition from normal to obstructed bowel and thus facilitating identification of the point and the cause of the bowel obstruction. The SBFS has been defined as gas and solid material within a dilated small-bowel loop that simulates the appearance of feces. The CT images are of a patient with mild left flank pain. At presentation the lab findings were normal. Based on this CT it was thought that this patient had a diverticulitis (red arrow). The mild dilatation of the small bowel adjacent to the descending colon was thought to be a reactive sentinel loop. Scroll through the axial images. Notice the locally dilated small bowel with the radiating pattern of the mesentery (image 7/11). Three days later the CT was repeated with i.v. contrast to get a better impression of the small bowel. There is a progressive dilatation of the small bowel. First study the images, then continue with the next series. Notice the radial array of dilated small bowel loops on the left with the mesenteric vessels converging to a central point. These bowel loops are wider than other loops and show less enhancement. There are dilated mesenteric veins (yellow arrow). At the point of strangulation the afferent loop is dilated (blue arrow) and efferent loop is collapsed (red arrow). The distal small bowel is collapsed (red arrows). The proximal small bowel is dilated (blue arrow). There is a large amount of ascites in Douglas cave, which also indicates the possibility of ischemia (blue arrow). Sometimes multiplanar reconstructions can be helpful in making the diagnosis of closed loop obstruction. Scroll through the sagittal images. Notice how the afferent loop enters the strangulated bowel and mesentery (image 8-10/13). Notice reactive changes in the mesocolon simulating diverticulitis. The coronal reconstruction demonstrates the point of strangulation with the dilated afferent loop, the strangulated loop and the collapsed efferent loop. The yellow arrow marks the dilated veins. At operation an ischemic strangulated small bowel was found, which was herniated through a hole in the mesocolon. Here we see the resected part of the small bowel. Notice the areas of necrosis and the dilated veins, which were also seen on the CT-images. Coronal reconstructions of another patient with a closed loop obstruction. The obstructed afferent loops are indicated in red arrowheads. The collapsed efferent loop is indicated by a red arrow. Notice the closed loop cranially to the area of obstruction. At surgery the bowell was not ischemic. There are various types of internal herniation. The illustrations shows a left paraduodenal hernia. This is an uncommon form of internal herniation. The CT-images show a left paraduodenal hernia. Notice the engorged veins (blue arrow). The duodenum is dilated (red arrow) and there is retention of fluid in the stomach. At operation the herniated small bowel was not ischemic. On the left a plain abdominal film is shown of a 57 year old man with a two day history of increasing abdominal pain and distension. First look at the image and then continue. Besides diffuse dilatation of the bowel, the major finding on this film is a large air containing structure in the pelvis. An important diagnosis to consider would be a volvulus of the colon and many would diagnose this as a sigmoid volvulus because it is located in the pelvis. However this actually is a cecal volvulus as will be explained below. A volvulus always extends away from the area of bowel twist. So a sigmoid volvulus can only move upwards and usually goes to the right upper quadrant. Cecal volvulus however can go almost anywhere and can even be located in the pelvis (figure). On the left there are additional CT-images of the same patient as above. First look at these images and look for the major findings and then continue. First we see a collapsed descending colon and a non-dilated ascending colon, so this cannot be a sigmoid volvulus. Secondly, we see a beak-like structure in the right lower quadrant which is where the bowel is twisted. In the left lower quadrant we see the dilated cecum. Coronal recontructions can be very helpfull in demonstrating what is going on. On the left we see the non-dilated ascending and descending colon (straight arrows) and the transition point of the volvulus (curved arrow). Cecal volvulus is due to the cecum twisting around the ascending colon thus leading to small bowel obstruction. A long narrow based mesentery predisposes to volvulus. An incomplete midgut rotation is a predisposing factor. Infarction is usually the result of venous congestion, while the arterial supply is rarely compromised. Cecal volvulus accounts for about 25% of cases of colonic volvulus. On the left a typical cecal volvulus is seen. We can see the beak-like transition zone located in the right lower quadrant indicating that this is a cecal volvulus. The dilated cecum is located in the left upper quadrant. Also notice the collapsed descending colon posterior to the dilated cecum (curved arrow). The x-rays show a typical cecal volvulus. Notice that the dilated bowel points toward the area of twist, which is the area where you expect the cecum to be located. Continue with the CT-images. Scroll through the images. The areas of twist and obstruction are marked. On the left a patient with a sigmoid volvulus. We can see the distended sigmoid extending from the pelvis way up into the right upper quadrant. Look at the image and decide for yourself why this cannot be a cecal volvulus. Then continue. The key finding is the dilatation of the proximal colon. The dilated loop seen on the left side is the dilated transverse colon. At CT we can nicely appreciate the area of the twist with the sigmoid extending up to the diafragm. The sigmoid is the commonest site of colonic volvulus. It accounts for 75% of large bowel obstruction. AP supine and erect radiograph of the abdomen demonstrates the characteristic coffee bean sign in sigmoid volvulus. Notice that the dilated loops point towards the sigmoid area. Continue with the CT-images. Scroll through the images. Notice the transition point (red arrows). Here another sigmoid volvulus. On the abdominal x-rays it is difficult to recognize what is going on, since so many bowel loops are dilated. Continue with the CT-images. CT is very helpful in this case and demonstrates the twist at the transition point (arrow). The last image shows the collapsed rectum posterior to the dilated small bowel loops. In the pelvis dilated small bowel loops were seen and a collapsed distal sigmoid (arrow). by Emil J. Balthazar Department of Radiology, New York University-Tisch-Bellevue Medical Center, 550 First Ave., New York, NY 10016. AJR 1994;162:255-261 by Carolyn J. Moore, Frank M. Corl and Elliot K. Fishman Department of Radiology, Johns Hopkins Hospital, 601 N. Caroline St., Baltimore, MD 21287. AJR 2001; 177:95-98by Jay P. Heiken and Robin Smithuis Imaging technique in SBO Small Bowel Feces Sign Case of small bowel strangulation Paraduodenal herniation Cecal Volvulus Sigmoid VolvulusClosed Loop Obstruction in Small bowel obstructionMallinckrodt Institute of Radiology of the Washington University School of Medicine, St. Louis, Missouri and the Rijnland Hospital, Leiderdorp, the Netherlands
cardio1
1
Cardiac Anatomy
by Tineke Willems and Marieke Hazewinkel
This review is based on a presentation given by Tineke Willems and was adapted for the Radiology Assistant by Marieke Hazewinkel. We will discuss: Cardiac CT with ECG triggering offers good image quality of the heart when compared to CT performed for other purposes. As in any other field of radiology, analysis of the acquired images requires a systematic approach. First of all, it is important to understand that the orientation of the heart in the human body differs from that of other anatomic structures: the right ventricle, for example, does not lie completely on the right, but more anterior. The left ventricle does not lie on the left, but more posterior. Also, the heart does not always maintain the same position within the mediastinum - in young people it tends to have a vertical orientation, whereas in older people it tends to rest on the diaphragm, a more horizontal orientation. Cardiologists analyze the heart using cardiac axes. These axes are also used in cardiac CT/MR and must be reconstructed in order to assess the heart properly. Axial slices, such as those imaged on the left, are useful for a global assessment of the morphology of the heart and its relation to the pericardium. On the left is a 4-chamber view, achieved by rotating upwards from the apex of the heart on the axial slices. In this axis, the right ventricle is projected next to the right atrium, and the left ventricle next to the left atrium. The mitral valve comes into view and - depending on the contrast protocol - the tricuspid valve may also be visible. Another feature of this cardiac axis is that the apex of the heart is well demarcated. Note that the apex is formed by the left ventricle. When the border between the mitral and aortic valves is localized on the axial slices and the images are rotated from this point, a 3-chamber view like the image on the left can be reconstructed. On this image, the left atrium, left ventricle, mitral-, aortic valve and proximal aorta ascendens are visible. On the left is an image of a 5-chamber view, which is similar to the 4-chamber view, but additionally displays the aortic valve and left ventricular outflow tract. This view is achieved by rotating the 4-chamber view a little more cranially. The 2-chamber view in the image on the left is achieved by rotating the images perpendicularly to the mitral valve and parallel to the cardiac septum. This axis gives an overview of the left atrium ventricle and mitral valve. It is a good view for analyzing ventricular function, especially that of the inferior and anterior walls. For attaining functional data, consecutive short axes must be reconstructed making use of the 3- and 4-chamber views. The cardiac anatomy will be discussed in the order of normal blood flow: from the right to the left. In the normal situation, contrast will be injected intravenously - usually in the arm - reaching the right atrium via the superior vena cava. The right atrium has an anterolateral position in the heart, and lies inferior to the left atrium. The superior vena cava enters through the roof of the right atrium. The inferior vena cava enters the right atrium from below near the cardiac septum. Another structure that carries blood towards the right atrium is the coronary sinus, (venous return of the coronaries) which enters anterior to, and just to the left of the inferior vena cava. In the right atrium lies the crista terminalis, a muscular ridge that runs from the entrance of the superior- to that of the inferior vena cava. This structure separates the smooth part of the right atrium - the sinus venosus - from the trabecularized right atrial appendage. On the images on the left it is visible as a smooth linear structure (blue arrows). This is not always the case, however, it may simulate a mass. The coronary sinus is the main draining vein of the myocardium. It runs in the atrioventricular groove on the posterior surface of the heart and enters the right atrium in the vicinity of the tricuspid valve. On the left is a reconstruction illustrating the course run by the coronary sinus in the atrioventricular groove on the posterior surface of the heart. The right atrial appendage is the trabecularized part of the right atrium. It partially covers the atrioventricular groove and the right coronary artery that runs in it. Characteristically, it is flat and triangular in shape and contains small muscular bundles which run parallel to the atrium itself. Blood leaves the right atrium and enters the right ventricle via the tricuspid valve. This valve has three leaflets and three papillary muscles, which partially insert on the septum (in contrast to the papillary muscles of the mitral valve, which do not). The right ventricle is shaped differently to the left ventricle: the left ventricle is cylindrical in shape and the cavity of the right ventricle is effectively wrapped around it. The right ventricle also has a thinner wall which is more trabecularized, especially towards the apex. The moderator band is another distinguishing feature of the right ventricle. It runs from the septum to the lateral wall of the right ventricle, and plays a key role in the electrophysiological conduction of the right ventricle's free wall (blue arrows). Next, blood runs towards the pulmonary valve - first entering the smooth, muscular infundibulum of the right ventricle. The pulmonary valve has three cusps, and is separated from the tricuspid valve by a thick muscle known as the crista supraventricularis (blue arrow in the image on the left). This differs from the left ventricular outflow tract, where the mitral and aortic valves lie side by side. On the left is a summary of the characteristics which are specific for the right ventricle and are useful in distinguishing the left from the right ventricle in cases with complex congenital cardiac anatomy. Oxygen-rich blood enters the left atrium via the pulmonary veins. In most cases, there are two pulmonary veins on the left and two on the right. The middle pulmonary veins usually drain into the superior pulmonary vein. There are frequent variations in pulmonary vein anatomy however, especially on the right, where an anomalous insertion is associated with atrial fibrillation. The left atrial appendage is a finger like, trabecularized structure which originates supralaterally in the left atrium. It lies over the left atrioventricular groove, and partially covers the left coronary artery in it. Its small, parallel-running muscles should not be mistaken for thrombus. When assessing the coronary arteries, the left atrial appendage must be removed, so that the LCX and proximal LAD may be visualized. Blood enters the left ventricle via the mitral valve. This is a complex valve, consisting of an annulus and posterior and anterior leaflets. The leaflets are connected to the papillary muscles via cord-like tendons called chordae tendinae. The papillary muscles insert into the lateral and posterior walls as well as the apex of the left ventricle. In normal situations the left ventricle has a uniform thickness, varying end-diastolically from 0.6 to 1.0 cm. Blood enters the aortic valve via the left ventricular outflow tract. Note that there appears to be a fibrous connection between the mitral and aortic valve. Like the pulmonary valve, the aortic valve has three cusps. Just cranially to it there is a slight dilatation of the aortic root. This is the sinus of Valsalva. It fills with blood during diastole, supplying the coronary arteries with oxygen-rich blood. The image on the right shows that the coronary arteries originate fairly cranially, on the border of the ascending aorta. The cusps of the aortic valves are named according to their relationship with the coronary arteries, namely the right coronary, left coronary and non-coronary cusp (R, L and N). by Benoit Desjardins and Ella A. Kazerooni AJR 2004; 182:993-1010Tineke Willems and Marieke Hazewinkel 4-chamber view 3-chamber view 5-chamber view 2-chamber view Right atrium Crista terminalis Coronary sinus Right atrial appendage Right ventricle Pulmonary valve Pulmonary veins Left atrial appendage Left ventricle Aortic valveCardiac AnatomyRadiology department of the University Medical Centre Groningen and the Medical Centre Alkmaar, the Netherlands
abdomen9
1
Acute Abdomen - Practical approach
by Adriaan van Breda Vriesman and Robin Smithuis 
The 'acute abdomen' is a clinical condition characterized by severe abdominal pain, requiring the clinician to make an urgent therapeutic decision. This may be challenging, because the differential diagnosis of an acute abdomen includes a wide spectrum of disorders, ranging from life-threatening diseases to benign self-limiting conditions (Table 1). Indicated management may vary from emergency surgery to reassurance of the patient and misdiagnosis may easily result in delayed necessary treatment or unnecessary surgery. Sonography and CT enable an accurate and rapid triage of patients with an acute abdomen. We present practical guidelines on the radiological approach of these patients. Interactive cases are presented in the menubar to test your knowledge. Before you perform an examination, obtain relevant information from the referring clinician. Don't let the clinician simply 'order' a sonogram or CT, but discuss the patient's age and posture, laboratory results and the number one clinical diagnosis and differential diagnosis. Based on that information and your own degree of confidence with the modalities decide for yourself whether to perform sonography or CT. Sonography has the advantage of close patient contact, enabling assesment of the spot of maximum tenderness and the severity of illness without ionizing radiation. In general the diagnostic accuracy of CT is higher than sonography. In patients with inconclusive US-results, CT can serve as an adjunct to sonography, and vice versa. We advocate the following two-step radiological approach of an acute abdomen. 1. Confirm or exclude the most common disease 2. Screen for general signs of pathology You have to be familiar with all the diagnoses listed in Table 1 to be able to recognize them. The clinical presentation of patients with an acute abdomen is often nonspecific. Both surgical and nonsurgical diseases may present with a similar clinical history and symptoms. Laboratory findings (leucocyte count, erythrocyte sedimentation rate, CRP) are equally nonconclusive. Findings may be normal in patients who need emergency surgery (such as appendicitis) and may be abnormal in patients without a surgical disease (like salpingitis). A plain abdominal film has a limited value in the evaluation of abdominal pain. A normal film does not exclude an ileus or other pathology and may falsely reassure the clinician. An ileus may not be appreciated on a plain abdominal film if bowel loops are filled with fluid only without intraluminal air (figure). Alternatively if a plain abdominal film does indicate an ileus than sonography or CT are usually needed to identify its cause. Thus, a plain abdominal film is seldomly useful, with the exception of detection of kidney stones or a pneumoperitoneum. For all other indications use sonography or CT. Many disorders may cause an acute abdomen, but fortunately only a few of these are common and clinically important. Focus on confirming or excluding these frequent disorders: Pain in the RLQ, regardless of any other symptom or laboratory results, should be considered to be appendicitis until proven otherwise. If you are unable to find the appendix you cannot rule out the diagnosis of appendicitis unless a good alternative diagnosis is found. If you do not find the appendix and there is no altermative diagnosis call the results of the examination indeterminate. Do not call it:' no appendicitis'. Normal Appendix. Your first task is to identify the appendix. At sonography and CT the appendix is seen as a blind-ending nonperistaltic tubular structure arising from the base of the cecum. Do not mistake a small bowel loop for the appendix. Secondly determine if the appendix is normal or inflamed. The outer-to-outer diameter of the appendix is the most important imaging criterium. Although an overlap of appendiceal diameters in normal and inflamed appendices can incidentally be found, a threshold value of 6-7 mm is generally used. A normal appendix has a maximum diameter of 6 mm, is surrounded by homogeneous non-inflamed fat, is compressible and often contains intraluminal gas. Inflamed Appendix An inflamed appendix has a diameter larger than 6 mm, and is usually surrounded by inflamed fat. The presence of a fecolith or hypervascularity on power Doppler strongly supports inflammation. CT depicts an inflamed appendix as a fluid-filled blind-ending tubular structure surrounded by fat-stranding. In the case on the left a hyper-attenuating wall is seen on the enhanced CT. In patients who lack intra-abdominal fat the use of iv. contrast can be helpfull in depicting the inflamed appendix. If the pain is located in the LLQ your main concern is sigmoid diverticulitis. In diverticulitis sonography and CT show diverticulosis with segmental colonic wall thickening and inflammatory changes in the fat surrounding a diverticulum. Complications of diverticulitis such as abscess formation or perforation, can best be excluded with CT. An important pitfall is colon cancer, which may present with similar imaging features, especially when the colon cancer is surrounded by fat stranding due to invasive groth, desmoplastic reaction or inflammation. Frequently it is not possible to reliably distinguish diverticulitis from colon cancer and therefore we routinely include colon cancer in the differential diagnosis of sigmoid diverticulitis. Cholecystitis occurs when a calculus obstructs the cystic duct. The trapped bile causes inflammation of the gallbladder wall. As gallstones are often occult on CT, sonography is the preferred imaging method for the evaluation of cholecystitis, also allowing assesment of the compressiblity of the gallbladder. The diagnosis of a hydropic galbladder is solely made on the non-compressability of the galbladder. Do not rely on measurements. Some galbladders happen to be small and others are large. The imaging appearance of cholecystis consists of an enlarged hydropic (meaning non-compressible) gallbladder with a thickened wall in the region of maximum tenderness (the so-called 'Murphy sign') The inflamed gallbladder usually contains stones or sludge, whereas the obstructing calculus itself may or may not be identified because it is located deep within the galbladder neck or cystic duct. The gallbladder may be surrounded by inflamed fat, but on sonography this frequently is not seen, while CT sometimes does show fat-stranding. Potential pitfalls are pancreatitis, hepatitis or right-sided heart failure, which all may lead to thickening of the gallbladder wall without cholecystitis. Therefore be certain that hydropic obstruction of the gallbladder is present before assigning the diagnosis of cholecystitis. Pain in LUQ An acute abdomen with LUQ pain is rare. Its most common cause is gastric pathology in which radiological imaging plays a minor role. After excluding these frequent disorders, search for signs of any other pathology, by systematically screening the whole abdomen. Look for inflamed fat, bowel wall thickening, ileus, ascites and free air. Inflamed fat is hyperechoic, space occupying and noncompressible at sonography. Inflamed fat is shown as fat-stranding at CT. Inflamed fat usefully points out where and what the problem is. As a rule, the organ or structure in the centre or nearest to the inflamed fat is the cause of the inflammation. Thickening of bowel wall indicates inflammation or tumor, and has an extensive differential diagnosis. Thickening of small bowel loops usually indicates regional inflammation, as small bowel tumors (carcinoid, lymphoma, GIST) are relatively infrequent. In patients with local colonic wall thickening a carcinoma is a prime concern. Pathologic distention of bowel loops may be caused by obstruction or paralysis. Firstly determine which parts of the gut are affected: small bowel, large bowel, or both. Look for normal nondistended bowel loops, which, if present, strongly suggest an obstructive cause for the ileus. Small bowel obstruction (SBO) accounts for approximately 4% of all patients presenting with an acute abdomen. The diagnosis of SBO is made when you see dilated small bowel and collapsed small bowel loops. If obstruction is present, try to identify its cause and location (adhesion, tumor, volvulus, intussusception, inguinal hernia). Adhesions account for 60-80% of all cases and are the likely cause when a smooth transition from dilated to collapsed small-bowel loops is noted. The 'Small Bowel Feces Sign' (SBFS) is a very useful sign as it is seen at the zone of transition thus facilitating identification of the cause of the obstruction. The SBFS has been defined as gas and particulate material within a dilated small-bowel loop that simulates the appearance of feces. Scroll through the images on the left to see the small bowel feces sign indicating the site of obstruction. Alternatively, an ileus without any normal bowel loops strongly suggests a paralytic cause. This is usually a response to general peritonitis, wich may have many possible causes of the inflammation. Asymptomatic volunteers do not have a detectable amount of free intraperitoneal fluid, with the exception of an incidental drop of fluid in Douglas in fertile women. The presence of ascites is a nonspecific sign of abdominal pathology, indicating that 'something is wrong'. You may want to perform a US-guided diagnostic puncture of the ascites, in order to investigate whether it is sterile reactive fluid, pus, blood, urine, or bile. The presence of free intraperitoneal air is proof of bowel perforation, and indicates a surgical emergency. A pneumoperitoneum has only two frequent causes: - Perforation of a gastric ulcer - Perforation of colonic diverticulitis Free air is usually not seen in perforated appendicitis). Always examine the images in lungsetting for better detection of free intraabdominal air (figure). A complete list of all possible causes of an acute abdomen is of little use in daily practice, therefore we just provide some imaging examples of several frequent causes of acute abdominal pain Mesenteric lymphadenitis is a common mimicker of appendicitis. It is the second most common cause of right lower quadrant pain after appendicitis. It is defined as a benign self-limiting inflammation of right-sided mesenteric lymph nodes without an identifiable underlying inflammatory process, occurring more often in children than in adults.. This diagnosis can only be made confidently when a normal appendix is found, because adenopathy also frequently occurs with appendicitis. Key finding: Lymphadenopathy with a normal appendix and normal mesenteric fat. On the left a CT of mesenteric lymphadenitis in a child suspected of appendicitis. Infectious enterocolitis may cause mild symptoms resembling a common viral gastroenteritis, but it may also clinically present with features indistinguishable from appendicitis especially in bacterial ileocecitis, caused by Yersinia, Campylobacter, or Salmonella. Key finding: ileocecal wall thickening without inflamed fat, adenopathy, normal appendix Right-sided colonic diverticulitis may clinically mimic appendicitis or cholecystitis, though the patient's history is generally more protracted. In contrast to sigmoid diverticula, right-sided colonic diverticula are usually true diverticula, that is, outpouchings of the colonic wall containing all layers of the wall. This may possibly explain the essentially benign self- limiting character of right-sided diverticulitis. Salphingitis is a common mimicker of both of appendicitis and diverticulitis. Transvaginal sonography depicts an inhomogeneous enlarged inflamed ovary. Epiploic appendages are small adipose protrusions from the serosal surface of the colon. An epiploic appendage may undergo torsion and secondary inflammation causing focal abdominal pain that simulates appendicitis when located in the right lower quadrant or diverticulitis when located in the left lower quadrant. The characteristic ring-sign corresponds to inflamed visceral peritoneal lining surrounding an infarcted fatty epiploic appendage. Epiploic appendagitis has been reported in approximately 1% of patients clinically suspected of having appendicitis. It is very important to make a positive diagnosis of this characteristic entity since epiploic appendagitis is a self-limiting disease. Both US and CT will depict an inflamed fatty mass adjacent to the colon. Key finding: inflamed fatty mass adjacent to the colon with characteristic ring sign. Urolithiasis often causes flank pain, but an ureteral stone (arrowhead) may occasionally present with clinical signs simulating appendicitis, cholecystitis or diverticulitis. Appendicitis on the other hand may cause hematuria, pyuria and albuminuria in up to 25% of patients because of ureteral inflammation from an adjacent inflamed appendix. Most abdominal aortic aneurysms rupture into the left retroperitoneum (4). Clinically this may simulate sigmoid diverticulitis or renal colic due to impingement of the hematoma on adjacent structures. However most patient will present with the classic triad of hypotension, a pulsating mass and back pain. Continuous leakage will lead to rupture into the peritoneal cavity and eventually death. Sonography is a quick and convenient modality, but it is much less sensitive and specific for the diagnosis of aneurysmal rupture than CT. The absence of sonographic evidence of rupture does not rule out this entity if clinical suspicion is high. CT depicts fat-stranding (arrowheads) surrounding the primary focus of the inflammation: the pancreas. Conclusion In patients with an acute abdomen 'the stakes are high'. A misdiagnosis may have serious consequences. We advocate a systematic approach: 1. First focus on the most common diseases and make a firm diagnosis or exclude them. 2. Always screen the whole abdomen for general signs of pathology. JB Puylaert et al; NEJM Volume 317:666-669 Adriaan C. van Breda Vriesman et al ; Radiology 2003;226:556-557 Dawn E. Lazarus et al, AJR 2004; 183:1361-1366 by Walter A Tan, MD, MS and Michel S Makaroun, MDAdriaan van Breda Vriesman and Robin Smithuis RLQ : Appendicitis LLQ : Diverticulitis RUQ : Cholecystitis Inflamed fat Bowel wall thickening Ileus Ascites Free air Mesenteric lymphadenitis. Bacterial ileocecitis Right-sided diverticulitis Salphingitis Epiploic appendagitis. Urolithiasis Ruptured Aneurysm PancreatitisAcute Abdomen - Practical approachRadiology department of the Rijnland Hospital, Leiderdorp, the Netherlands
abdomen2
1
Acute Abdomen - Role of CT in Trauma
by Stephen Ledbetter and Robin Smithuis
This review is based on a presentation given by Stephen Ledbetter and was adapted for the Radiology Assistant by Robin Smithuis. Stephen Ledbetter is director of the emergency radiology department of the Brigham and Women's Hospital in Boston, which is a teaching affiliate of the Harvard Medical School. This review will focus on the role of CT in the evaluation of patients with traumatic abdominal injuries. Some of the cases will be presented in an interactive way. Trauma is the leading cause of death under the age of forty. Of all traumatic deaths, abdominal trauma is responsible for 10%. The findings to look for in abdominal trauma are the following: Nowadays there is a trend towards non-operative management of blunt abdominal trauma. More than 50% of splenic injury, 80% of liver injury and virtually all renal injurys are managed non-operatively, because patients proved to have better outcomes on the long term related to visceral salvage. CT is used to evaluate patients with blunt trauma not only initially, but also for follow up, when patients are treated non-operatively. CT is also used to clear patients before they are dismissed from the ER, because CT has a very high negative predictive value and can rule out injury in patients who have had a significant mechanism of injury. These patients do not have to be admitted for observation. CT is also increasingly used for penetrating trauma, which traditionally was evaluated operatively. Blunt injury A relatively simple protocol can be used for patients with blunt trauma based on scanning the entire abdomen in the portal venous phase and a subsequent delayed excretory scan 3-5 minutes later if injury is detected on the initial scan. No oral contrast is administered. Penetrating injury Most patients with penetrating trauma are injured in the flank, so there is great risk for bowel perforation. If there is no reason for immediate surgery on the initial scan, these patients get an additional scan after the administration of rectal contrast (50 ml contrast in 1000 ml saline). 500 ml can be administered if there is isolated left flank injury, but in all other cases 1000 ml is administered. The spleen is the most commonly injured solid organ (25%). The standard CT grade of splenic injury of the American Association for the Surgery of Trauma (AAST) is of limited value since it does not predict the succes rate of a non-operative management. The finding of contrast extravasation on the other hand, which is not part of the grading system, has great impact on the patients management, because when there is active bleeding, there will be failure of a non-operative management in 80% of the cases. In these patients the need for intervention is almost ten times as high compared to patients without extravasation. In a recent article a new CT grading system is proposed, which is better than the AAST system (3). On the left a case of splenic injury. Scroll through the images and determine the degree of splenic injury. Then continue. The findings are the following: Because of the absence of hemoperitoneum or active bleeding, this patient has a good prognosis and will be managed non-operatively. On the left another patient with splenic injury. Scroll through the images and describe the lesions. Then continue. The findings are the following: Depending on the clinical condition this patient will be managed non-operatively, because there is no active bleeding. On the left the most commonly used Splenic CT Injury Grading Scale. A way to remember this system is: The shortecommings of this grading scale are: On the left images of a 22-year old male who presented 3 hours after a snowboarding accident with LUQ and left shoulder pain. Scroll through the images and describe the lesions. Then continue. The findings are the following: So in this case there is a great chance of failure of non-operative management. A contrast blush is defined as an area of high density with density measurements within ten HU (Houndsfield Units) compared to the nearby vessel (or aorta). The differential diagnosis is: How can these entities be differentiated? On the left a different case of splenic injury with lacerations. There is also active bleeding with a contrast blush with the density within the range of the density of the aorta. There also is hemoperitoneum, so this patient will probably need surgery. In trauma the liver is the second most commonly involved solid organ in the abdomen after the spleen. However liver injury is the most common cause of death. This is due to the fact that there are many major vessels in the liver, like the IVC, hepatic veins, hepatic artery and portal vein. It is important to remember, especially if you are doing ultrasound, that the posterior segment of the right liver lobe is the most frequently injured part. This part also involves the bare area and this can lead to retroperitoneal bleeding rather than bleeding into the peritoneal cavity. First look at the images on the left of a patient with liver injury. Describe the findings. Then continue. The findings are: CT grading system for liver injury On the left the CT grading system for liver injury, which is almost the same as the grading system for splenic injury. The only difference with the spleen is that the liver has two lobes. So before you come to grade 5, which is devascularization or maceration of both lobes, you have grade 4, which is devascularization or maceration of only one lobe or laceration greater than 10 cm. Now regarding the consequences of the CT grading system the following somewhat conflicting remarks can be made: First look at the images on the left of a patient with liver injury. What are the CT findings in this case? What is the CT grade of injury? The findings are the following: So the next question is: does the presence of a contrast blush alter the CT grade of injury? The answer is: it does not, because active bleeding is not part of the grading system. However there is increased likelihood of failure of non-operative management. Whenever there is a contrast blush, it is important to note if the contrast blush is associated with a hemoperitoneum and if it extends beyond the parenchyma, as in this case. First look at the images on the left of a patient with liver injury. What are the CT findings in this case? What is the CT grade of injury? The findings are the following: So despite the fact that there is a grade 4 injury and contrast extravasation, this patient will be treated non-operatively and probably will do fine, because there is no bleeding into the peritoneal cavity. So the important thing to remember it that, the grading system is of limited help in the management of the patient. Contrast extravasation on the other hand is of great importance especially if it is associated with hemoperitoneum. On the left two more examples of laceration. Lacerations can be stellate, like the example on the left or branching like the one on the right. First look at the images on the left of a patient with liver injury. Ask yourself the following questions: There is i.v. contrast and images were taken in the portal phase. There is also oral contrast filling of the stomach. The contrast surrounding the liver could be a result of stomach or bowel perforation, but since there was no pneumoperitoneum, this was thought to be unlikely. So the extravasation was thought to be a result of active bleeding and since there is a great amount of contrast surrounding the liver, this was thought to be a huge leak. At the OR an avulsed right hepatic vein was found. This diagnosis has a 90-100% mortality and this patient died in the OR. Some final remarks conceirning liver injury: Penetrating injury Look at the images on the left and try to answer the following questions: Answers: The next question is, whether the protocol is correct or do we need to give rectal contrast to see if there is bowelperforation, because there is a penetrating trauma? In this case the answer is no, do not give this patient rectal contrast. The reason is that we already have reached the treshold for this patient to go to the OR. There are 3 reasons for this patient to go to surgery: If rectal contrast was given at the start of the examination, this might pose the problem that it would have been unclear, whether the contrast deposition was due to active bleeding or bowel perforation. So the bleeding could have been missed. Rectal contrast should only have been given if there were no other findings in need of surgery. Although this patient had severe renal injury, there was no hematuria. This is often the case in penetrating trauma and does not rule out renal injury. In blunt trauma however the abcense of hematuria does rule out renal injury. On the left another patient with a penetrating injury due to a knife stab in the flank. The CT demonstrates nicely, that the injury is limited to the retroperitoneal space with a small renal hematoma. There is no sign of peritoneal violation and on delayed images (not shown) there was no extravasation of the collecting system. This patient will be treated non-operatively Blunt injury In 90% of cases there will be renal injury due to blunt trauma. Unlike in injury to the spleen and the liver, in renal trauma we also need to evaluate the collecting system. The grading system on the left has proven to be of value in the management of the patient. However unlike the grading for spleen and liver injury it is not that simple to remember. In grade I there is nothing wrong with the parenchyma, just contusion or subcapsular hematoma. Grade II and III injuries are either less or greater than 1 cm lacerations, but with no injury to the collecting system. Grade IV is injury to the collecting system or large lacerations> Grade V is a shattered or devascularized kidney. First look at the images on the left of a patient with renal injury after a blunt trauma. What is the CT grade of injury? The answer is, that like all grading systems, this system also has its limitations. What we see on the left is not a laceration, because it is not linear. It is not a contusion, because it is sharply demarcated. This is an post traumatic segmental infarction. On the left a typical subcapsular hematoma, which is also a grade I renal injury. Some final remarks on renal injury: Michael Federle placed renal injuries into four categories: On the left a 65-year old male struck by a car traveling at moderate speed. Loss of consciousness for 2 minutes. A foley catheter was passed and there was gross hematuria. The x-ray shows a moderately displaced fracture of the pubic bone with bony spicules in the bladder region. So the question is: For what other pelvic injuries is this patient at risk and how will it affect our protocol? First this patient is at risk for arterial injury with pelvic hematoma, rectal, vaginal injury and bladder injury. Secondly, a CT-cystogram is indicated after the routine CT. On the left the images of the routine trauma-CT. What are the findings? There is a displaced pelvic fracture with a spicule pointing towards the bladder. There is fluid in the prevesicle space (space of Rezius). If there is a pelvic fracture the chance of a bladder rupture is 10%. If there is a bladder rupture, there is almost always a pelvic fracture. First it was thought that the rupture was caused by the pelvic fracture itself, but now we know that only in one third of cases the bladder rupture is at the site of the bone spicule. Two third of rupture occur at the opposite site, meaning that shearing forces play a significant role in bladder injuries. On the left the pre- and post-cystogram images. There is contrast in the bladder surrounding the foley catheter and there is extravasation of contrast in the prevesicle space or space of Rezius. This has been referred to as the 'molar tooth sign' indicating extraperitoneal bladder rupture. On the left a sagittal and coronal reconstruction. Notice that there is no contrast extending into the pericolic gutter, so there is no intraperitoneal extention. The sensitivity and specificity of CT Cystography is very high. For extraperitoneal rupture it is respectively 100% and 99% and for intraperitoneal rupture it is 92% and 100%. The most important factor is that you have to have good distention of the bladder. First we drain the bladder, because we want to get rid of the urine and contrast that was excreted by the kidneys. The contrast solution that we use is the same as we use for oral or rectal contrast (i.e. 50 cc contrast in 1L saline). We instill the contrast retrograde through the foley catheter until one of three things happen: On the left another case to illustrate why you do not administer contrast in the bladder at the same time as the administration of iv. contrast. The next question that comes up, is whether we should perform an additional CT-cystogram? The answer to the first question is that if you would have administered contrast to the bladder at the start of the examination, you would have been puzzled whether the contrast that is seen is due to a bladder rupture or to active bleeding. Since no contrast was instilled in the bladder, it is obvious, that this is arterial bleeding. Secondly because of the enormous extravasation, this patient is in need of immediate embolisation without further delay. Concerning pancreatic injury the following remarks can be made: On the left an unrestrained driver who had a car accident. Vital signs were stable and there was only a mildly tender abdomen. First look at the images on the left and then continue. All the intraperitoneal organs were normal and there was no intraperitoneal fluid. The only findings were a vague hypodense area in the pancreatic tail and some fluid behind the pancreas, best seen anteriorly to the left kidney. So the radiologist said that there was concern about pancreatic injury. The reason that he was not more definitive was that, an isolated pancreatic injury is exceptionally rare, since the pancreas is protected by the liver and spleen and the bony thorax. During follow up this patient experienced more pain and on a follow up scan (not shown) there was impressive accumulation of fluid around the pancreas. So this patient had an isolated pancreatic injury. The case above is an exceptional case. When the pancreas is involved in a trauma, it is almost always part of a package injury (Table). The more common presentation of pancreatic injury is what is seen on the left. Scroll through the images and describe the findings. Then continue. This is a typical left sided package injury. There is pancreatic tail injury and also splenic injury, renal injury and pneumoperitoneum. On the left another common presentation of pancreatic injury. Look at the images and describe the findings. Then continue. There is a right sided package injury. There is a liver laceration crossing the major vessels associated with a transsection of the pancreas at the junction of the head and the body. The force must have come from the right anterior side squeezing the liver and the pancreas against the spine. Sometimes this kind of injury also involves the duodenum. On the left a chest film of a 79-year old restrained driver who had a car accident. Initially unresponsive at the scene. He was transferred from an outside hospital after placement of tubes. Look at the image on the left and describe the findings. Then continue. The first thing you'll notice is that the tube is in the right main bronchus. Chest tube looks okay. Nasogastric tube comes down and coils in the stomach. The superior mediastinum looks widened and indistinct, so this certainly has to be evaluated. In the left lower zone we have an indistinct diafragmatic border and an opacity. This could be a lot of things like hematothorax, lung contusion, diafragmatic rupture or splenic injury. So based on the chest film we are conceirned about possible aortic injury, pulmonary contusion and injury to the diaphragm, spleenic and left kidney. Continue with the CT images. Scroll through the images on the left. What contrast is on board adn wh at are the findings? There is i.v. contrast in the late arterial phase and when we follow the nasogastric tube we will notice that there is no contrast in the stomach. The most important finding in this case is the area of soft tissue density next to the atelectatic lower lobe of the lung and lateral to it an amount of fat. This is very suggestive of diafragmatic rupture. What can we do to get more certainty about this structure? Since the nasogastric tube is in place, we can administer contrast to the stomach. The images on the left prove that the structure is the stomach, which is in a high position. Secondly there is a waist in the stomach compatible with the 'collar sign'. These findings are specific for diafragmatic rupture. On the left the coronal reconstruction of hte same patient demonstrating the 'collar sign', where the stomach passes through the diafragmatic rupture. So in the case above specific signs of diafragmatic injury are present. Non-specific signs are discontinuity or thickening of the diafragm or the 'dependent viscera' sign. On the left a demonstration of the 'dependent viscera' sign. On the left side there clearly is a diafragmatic rupture with herniation of the stomach. Notice that the stomach and the spleen lie against the posterior thoracic wall, which is abnormal. This is unlike on the right side where the liver is away from the chest wall due to the presence of the diafragm. On the left a patient with a right-sided injury. On the chest film it looks as if there is just elevation of the hemidiafragm or maybe there is a subpulmonic pleural fluid collection. There also could be a baseline diafragmatic paralysis. Now continue with the CT images. Describe the findings on the left and then continue. The axial image demonstrates that the opacity on the chest film is actually the liver. As we follow the livercontour, there is this unusual shape (yellow arrow). There is discontinuity of the crus which is a non-specific sign (small blue arrow). On the axial image there is indentation of the liver on the posterior side due to blood in the thorax. On the sagittal MPR there is indentation of the liver and the 'collar' sign is nicely demonstrated. On the left some final remarks conceirning diafragmatic rupture. On the left an unrestrained 22 y.o. male involved in a high-speed motor vehicle accident. He was ejected from the vehicle. At presentation he was unconscious and intubated with diminished femoral pulses. Scroll through the images on the left and describe the findings. The findings are: So the questions are: A unilateral renal infarct can be the result of a localized injury. However when there are multiple bilateral infarcts, we have to think of an embolic source. The most common location after injury for these emboli is in the thoracic aorta at the isthmus, because the aorta is fixated there. In this patient however the source was a traumatic dissection of the aorta at the level of the diaphragm. This is the second most common location for injury to the aorta due to the relative fixation. . On the left images of a 44 y.o. male who jumped 40 feet from building onto concrete surface in suicide attempt. History of treatment for depression BP 90/54. Pale, diaphoretic, confused. No head injury. Ecchymoses around chest and abdomen. Distended abdomen. Pelvis grossly unstable. Gross hematuria. Scroll through the images on the left and describe the findings. The findings are: The questions in this patient are: Concerning pneumoperitoneum some important remarks have to be made: In fact the most common findings in small bowel injury are non-specific findings like thickening of the bowel wall and unexplained intraperitoneal fluid. In the patient that we discussed the diffuse wall thickening was only a result of hypoperfusion or 'shock' bowel due to the active bleeding. Direct injury to the bowel wall usually results in focal thickening and is mostly a non-transmural injury. It is very uncommon to identify findings that are specific for bowel injury like extravasation of oral contrast or bowel content. More commonly you will find a combination of intraperitoneal fluid and mesenteric stranding, focal bowel thickening or interloop fluid, that is very suggestive for bowel injury. by Akira Kawashima, MD, Carl M. Sandler, MD, Frank M. Corl, MS, O. Clark West, MD, Eric P. Tamm, MD, Elliot K. Fishman, MD and Stanford M. Goldman, MD Radiographics. 2001;21:557-574 This review considers the issue of blunt abdominal trauma in adults. A continued trend is noted for detection of specific findings that do predict the need for therapeutic surgery or for angiographic embolization or that predict a period of close observation is needed for an injured patient. This trend in imaging parallels a strong trend in trauma therapy toward nonoperative management of injuries of the spleen, liver, and kidney even when hemoperitoneum is present. by Helen Marmery et al. AJR 2007; 189:1421-1427Stephen Ledbetter and Robin Smithuis Contrast blush Categories of Renal Injuries CT Cystography CT 'collar' sign 'Dependent viscera' signAcute Abdomen - Role of CT in TraumaDepartment of Radiology of the Brigham and Women's Hospital, Boston and the Rijnland Hospital in Leiderdorp, the Netherlands
abdomen3
1
Acute Abdomen - Role of Ultrasound
by Julien Puylaert
Multi-slice CT is increasingly replacing ultrasonography for the evaluation of patients with acute abdominal pain. Ultrasound however has specific advantages. This review will focus on: Multi-slice CT is increasingly replacing ultrasonography (US) for the evaluation of patients with acute abdominal pain CT has major advantages over US: it is extremely fast and its time burden is often less than that of a US examination (1-4). CT is not disturbed by gas and bone, while obesity is even an advantage. Most of all, CT is not operator-dependent and can be reviewed by others, even at a distance. With all these advantages, it is not surprising that US is losing field in the evaluation of the acute abdomen. US however has some advantages.  Specific advantages of US (2)  Specific advantages of US (3) Worldwide, there is a large variation of who performs the US examination of the acute abdomen. US is done by technicians, general radiologists, radiologists specialized in US, abdominal radiologists, and all sorts of clinicians, urologists, gynecologists and even family doctors. The US examination performed as described above, requires a person with a thorough medical background, knowledge of all possible causative conditions (urological, gynecological, gastrointestinal, vascular, etc.), and with a large expertise in US as well as in CT, imaging guided puncture and other radiological imaging. There is no doubt that the person who meets these conditions best is the radiologist, and preferably a radiologist with special interest in abdominal US and CT. Additional advantages of concentrating all primary, diagnostic abdominal US examinations within the Radiology Department are obvious: It guarantees integrated imaging, constant quality, round-the-clock coverage, continuity, central archiving and accurate and early triage of patients with abdominal symptoms (6). US examination in patients with acute abdominal pain requires a specific technique of graded compression. In this way fat and bowel are displaced or compressed. This eliminates the disturbing influence of bowel gas and reduces the distance from the transducer to the appendix, allowing the use of a high frequency probe with better image quality (Figure). This technique also allows assessing the rigidity of a structure by evaluating its reaction upon compression In order to avoid pain, the compression should be applied slowly and gently, similar to the classic palpation of the abdomen. The entire abdomen is examined to exclude disease of gallbladder, pancreas, kidney, aorta, stomach, small and large bowel, appendix, uterus and ovaries. A moderately filled bladder allows better survey of the distal ureters, and of uterus and ovaries in women; however, a full bladder does not allow proper graded compression. Transvaginal US may be used for gynecological conditions but also for pelvic appendicitis, diverticulitis and Douglas abscesses. The peritoneal cavity is screened for bowel pathology with five to six vertically oriented, overlapping lanes using a broad based, high frequency probe. We refer to this as 'mowing the lawn' (Figure). This form of screening is facilitated by the use of thin-liquid US-gel. Bowel pathology is usually conspicuous, because the diseased and empty bowel has a thickened and hypoechoic wall, which contrasts with the surrounding hyperechoic fatty tissue. On the left a patient with segmental colitis caused by Crohn's disease. The pathologic bowel segment is easily picked up using the 'mowing-the-lawn' technique. Acute appendicitis is the most common abdominal surgical emergency in the Western World. The diagnosis may be easy but may also be very difficult. The clinical diagnosis of appendicitis is as often wrongly made as it is initially overlooked, leading to unnecessary surgry, respectively to ill advised delay. Using US it is possible to confirm appendicitis by visualizing the inflamed appendix (successful in 90 %) or to exclude appendicitis, either by visualization of the normal appendix (successful in 50%) or by demonstrating an alternative condition (possible in 20 %). This means that there will always be a rather large group of patients in whom the US result is equivocal making further studies necessary. A fortunate circumstance is that most of the patients in the latter group are obese, and therefore suitable for CT. The normal appendix presents as a small, easily compressible, concentrically layered, mobile, blind-ending, sausage-like structure. A diameter up to 7mm is regarded as normal. The normal appendix is mobile, may have a collapsed lumen, but also may conntain air or some fecal material, and rarely a little fluid (6). Power Doppler reveals scarce or no vascular signal and there is no hyperechoic, non-compressible inflamed fat around the appendix. The typical appearance of an inflamed appendix is that of a concentrically layered, non-compressible sausage-like structure demonstrated in a fixed position at the site of maximum tenderness (Figure). The average maximum diameter is 9 mms with a variation from 7 to 17 mms. In 30% intraluminal fecoliths are found actually obstructing the lumen. Six to 12 hours after the onset of symptoms, the inflammation progresses to the adjacent fat of the meso-appendix, which becomes larger, more hyperechoic and less compressible. Later on, this fatty tissue will tend to increase in volume around the appendix: this represents mesentery and omentum, which have migrated towards the appendix in an attempt to wall-off the imminent perforation. Slowly applied intermittent compression is the best way to identify the non-compressible inflamed fat (Figure). An irregular, asymmetrical contour and loss of the layer structure of the appendix indicate perforation or imminent perforation. Vascularization of the appendiceal wall is either markedly increased or absent due to high intraluminal pressure with concomitant ischemic necrosis, however there is always increased vascularization in the directly surrounding fatty tissue . The presence of a generalized, adynamic ileus is suspect for perforated appendicitis, even if the inflamed appendix cannot be visualized. A small quantity of free intraperitoneal fluid is aspecific. It may be present in both non-perforated and perforated appendicitis as well as in many other conditions, both surgical and non-surgical. A large quantity of fluid in the presence of an inflamed appendix may represent pus from perforated appendicitis and then is usually accompanied by paralytic ileus. Larger quantities of free fluid are also found in perforated peptic ulcer (note air and food particles) and gynecological conditions (puncture usually reveals blood). In most patients with appendicitis inflamed mesenteric lymph nodes can be demonstrated higher up in the mesenterial root. In case of an abnormal position of the inflamed appendix far from where the usual gridiron-incision is made, it is useful to indicate the location of the appendix on the skin of the patient with a waterproof marker. This may influence site, size and orientation of the incision. If the clinical symptoms rapidly subside despite the presence of an unequivocally inflamed appendix on US, one should consider the diagnosis of spontaneously resolving appendicitis. These patients initially have the typical clinical signs of appendicitis, but within 12-48 hours after the onset of pain the clinical symptoms relatively abruptly subside, probably due to relief of obstruction. On US follow up, the appendix usually decreases in size in the course of days. If the patient recalls similar previous attacks, immediate appendectomy is advisable, even if the patient is again completely free of symptoms at that time. Histology in such cases will nevertheless- confirm acute inflammation. In case conservative management is opted for, keep in mind that there is a recurrence rate of approximately 40 % (8). Patients, who are admitted with considerable delay may present with a palpable mass and relatively mild peritonitis. In these patients, who usually have a high ESR, US shows a large mass of non-compressible fat around the appendix, interspersed with echolucent streaks . These patients are diagnosed as 'appendiceal phlegmon' and are usually managed conservatively because the surgeon knows that appendectomy in such cases is technically difficult or even impossible (9). On the left a 35-year old man with a 10-day history of abdominal pain in the right lower quadrant. At examination a palpable mass was found. There was no evidence of peritonitis. On the top image the US reveales a large noncompressable inflammatory mass consisting of the inflamed appendix, mesentery and omentum. Patient was treated conservatively. Follow up examination at 8 days and another 6 weeks later showed still some residual abnormalities. The patient was completely symptom free. There were no recurrent symptoms and the patient did not undergo operation. If next to the inflamed appendix, a fluidcollection is found, this is suggestive for an appendiceal abscess. The collection often contains air and is surrounded by inflamed non-compressible hyperechoic tissue representing omentum and mesentery as well as secondarily thickened neighboring bowel loops, attempting to seal-off the abscess from the peritoneal cavity. If an appendiceal abscess is demonstrated and there is no frank peritonitis, percutaneous drainage is the treatment of choice. In stable patients who have no fever and only mild pain, it can be even wise to await spontaneous drainage of the abscess to neighboring bowel. On the left an abscess cavity that contains a fecolith. Note the inflamed appendix (arrows) lying next to the abscess. Finally, there are some patients with an appendiceal abscess who are better off with immediate surgery: this goes in general for children and for those patients with severe peritonitis, which indicates that the walling-off process is failing. Immediate surgery is also indicated for patients who have a small abscess with a history of only a few days of symptoms, in whom appendectomy with evacuation of the abscess is usually technically easy (Figure). On the left a patient with a 4-day history of right lower quadrant pain. There was peritonitis at physical examination. The sedimentation rate was 48mm/hour. Palpation was unreliable. Subsequent appendectomy with evacuation of the abscess was performed without technical difficulties. Prior to percutaneous drainage, CT is necessary to delineate the extent of the abscess and to determine the safest access route. If expertise is available in US guided puncture, the combination US plus fluoroscopy has several advantages over CT guided drainage. It is rapid, allows continuous control, any angulation, and can be performed as a bed-side procedure. A false positive diagnosis can be made if the normal appendix is mistaken for an inflamed one. Not infrequently the normal appendix is larger than 7 mms, especially in children due to lymphoid hyperplasia and in adults due to fecal impaction. Appendiceal compressibility, the absence of a Doppler signal and the absence of inflamed fat are the most important features in deciding if it is normal or inflamed. Mistaking a normal appendix for an inflamed one may also occur if there is secondary thickening of the appendix associated with cecal carcinoma. In the latter case, the appendiceal lumen is obstructed giving rise to sterile accumulation of mucus in the lumen. The patient often has remarkably mild symptoms and is managed conservatively under the erroneous diagnosis of an appendiceal phlegmon. If the underlying tumour is small and is not recognized, this may lead to considerable delay in surgical treatment. The combination of a relatively large appendix with paradoxically mild and atypical symptoms should raise suspicion of underlying malignancy. Other conditions with secondary thickening of the appendix are perforated peptic ulcer, Crohn's disease and sigmoid diverticulitis. Pitfalls in the US diagnosis of appendicitis (2) A false negative ultrasound examination is mostly the result of overlooking the inflamed appendix. In experienced hands the inflamed appendix can be visualized in 90% of patients with acute appendicitis. Generalized peritonitis hampers graded compression which may account for a lower score in patients with free appendiceal perforation. Also air-filled dilated bowel loops from adynamic ileus may hide the appendix from view. Air in the lumen can make it difficult to identify the inflamed appendix. Pitfalls in the US diagnosis of appendicitis (3) Another pitfall is demonstration of the normal proximal part of the appendix while the distal inflamed tip is overlooked, because it is obscured by bowelgas. Rarely, the inflamed appendix has a maximal diameter of less than 7 mms. In those cases rigidity, hypervascularity and the presence of inflamed fat must give the clue. On thr left a patient with acute pain in the right lower quadrant. The appendix has a diameter of only 6.5mm. However there is inflamed fat and an increased Doppler signal indicating that it is acutely inflamed. Pitfalls in the US diagnosis of appendicitis (4) Another pitfall is advanced appendicitis where there is secondary wall thickening of the ileum. Often the ileal thickening is more prominent and conspicuous on US than the underlying inflamed appendix (Figure). If only the ileum is appreciated and the appendix is overlooked, an erroneous diagnosis of infectious ileocecitis or Crohn's disease can be made, leading to ill-advised surgical delay. Similarly, if in an adult patient enlarged mesenteric lymph nodes are the sole US finding, one should be cautious to diagnose mesenteric lymphadenitis because these nodes could be secondarily enlarged due to acute appendicitis, while the inflamed appendix is overlooked. If in a patient with appendicitis only the fecolith in the appendiceal base is visualized and the rest of the appendix is overlooked, this may lead to an erroneous diagnosis of cecal diverticulitis. Pitfalls in the US diagnosis of appendicitis (5) Also, if in a woman a relatively large rightsided ovarian cyst is found, this is not necessarily the cause of her symptoms and appendicitis should still be searched for. Finally, if in advanced appendicitis only the hyperechoic non-compressible inflamed fat of omentum and mesentery is visualized, and the inflamed appendix is overlooked, this may lead to an erroneous diagnosis of omental infarction or epiploic appendagitis (10,11). In patients with equivocal US findings, CT scan is indicated. A fortunate circumstance is that these are often obese patients. Patients with ileocecal Crohn's disease often have protracted and atypical symptoms causing marked diagnostic delay. On the other hand, Crohn's disease may also present with acute, appendicitis-like symptoms and lead to an ill-advized operation. In both scenarios US may play an important role in establishing the initial diagnosis (12,13). The sensitivity of US for detecting ileocecal Crohn's disease of over 95%. On the left a patient with Crohn's ileitis. US reveals marked wall thickening of the terminal ileum with focal disruption of the wall and a small abscess, walled of by hypoechoic, inflamed fat. Sonographically, there is marked mural thickening of the ileum, which shows decreased or no peristalsis and is not compressible. Classically, all layers are involved and layerstructure is often locally disturbed , the earliest sign being echolucent changes in the submucosa. There is inflammation of the fatty mesentery and omentum, recognizable as hyperechoic, non-compressible tissue adjacent to the ileum. In the echolucent wall bright eccentric foci may indicate deep ulceration. Echolucent streaks within the hyperechoic tissue indicate liponecrotic tracts, which may herald fistulaformation (Figure). Cecum and appendix may also show mural thickening. Mesenteric lymph nodes are often markedly enlarged, but hypovascular. In longstanding Crohn's disease, 'creeping fat' is found which is recognized as a large, moderately well-compressible fatty mass encompassing most of the circumference of the ileum and iso-echoic to normal fat. Eventually, there are often US signs of prestenotic dilatation, abscess formation, or fistulaformation. Infectious ileocolitis is a bacterial infection of terminal ileum and colon which is characterized by diarrhea and abdominal pain. The most frequently cultured bacteria are Campylobacter and Salmonella, and Yersinia. The infection is generally limited to the mucosa, is self-limiting and rarely poses diagnostic problems. There is an interesting variant of infectious ileocolitis in which the infection is mainly limited to the ileocecal area and is therefore has been coined infectious ileocecitis (14). It is usually caused by the same bacteria and the importance of this variant is that its clinial symptoms are dominated by acute right lower abdominal pain, while diarrhea is absent or only mild. These symptoms masquerade as the clinical signs of appendicitis and explain why infectious ileocecitis often leads to an unnecessary laparotomy. The symptoms of Yersinia are often more protracted and then both the clinical symptoms and the US features may mimick those of Crohn's disease. The absence of a transmural component, the self-limiting course and of course positive stool cultures or serology will yield the correct diagnosis. The frequency of infectious ileocecitis is fairly high and has a ratio of 1 to 8 compared to appendicitis (14). On the left a 26-year old woman with clinical symptoms of appendicitis. US shows prominent ileocecal valve and to marked mucosal and submucosal wallthickening of ileum and cecum. Enlarged lymph nodes were found in the radix of the mesentery. The appendix was normal. Appendectomy was cancelled. The next day the patient developed diarrhea and stool cultures eventually revealed Campylobacter jejuni. In infectious ileocecitis US shows fairly characteristic features. There is diffuse thickening of mucosa and submucosa of the terminal ileum and the cecum. The appendix has to be sonographically normal (Figure). In contrast to ileocecal Crohn's disease, in infectious ileocecitis, the wall layers are always intact and the muscularis and serosa, are never affected. Also omentum and mesentery are never involved and there are never signs of bowel obstruction, abscess- or fistula-formation. The various micro-organisms have a slightly different pattern of affecting the ileocecal area (Figure). On the left a schematic representation of relative involvement of ileum, cecum and mesenteric nodes in infectious ileocolitis caused by Yersinia, Campylobacter, and . This is an ill-defined entity, probably of viral origin, in which the mesenteric lymph nodes become inflamed and enlarged. It is a typical disease of childhood and is only rarely seen in young adults. It mimicks the clinical signs of appendicitis and may therefore lead to an unnecessary appen?dectomy. The US findings are solely enlarged, hypervascular mesenteric lymph nodes. However if these are the only US findings in a symptomatic young adult, it is well possible that these nodes are in fact secondarily enlarged due to acute appendicitis and the inflamed appendix is overlooked. Patients with cecal carcinoma can present with acute or subacute abdominal symptoms, in several ways. The tumour may cause acute small bowel obstruction, the appendix may be involved, the tumour may perforate and the tumour itself may cause direct pain. The often bulky nature of the tumour and the close proximity of the right colon to the abdominal wall makes cecal carcinoma in most cases fairly conspicuous on US. The majority presents as a hypoechoic, solid, well-vascularized irregular and asymmetrical thickening of the cecal wall (Figure). In the proximity enlarged mesenteric lymph nodes can be found, and in most cases there is also some inflamed fat around the tumour. In a minority of cases the tumour is of the scirrhous type, which is less easy to detect. The finding of livermetastases of course strongly supports the diagnosis of malignancy. Ingrowth of the tumour in the appendiceal base only rarely causes a full-blown appendicitis, but rather will lead to mucinous dilatation of the appendiceal lumen. On US the enlarged appendix is often more conspicuous than the underlying tumour and, since there is often a palpable mass and protracted symptoms, these patients are often misdiagnosed as to have an appendiceal phlegmon, leading to significant surgical delay. A clue to the correct diagnosis is the discrepancy between the relatively mild and protracted symptoms of intermittent, nagging pain and the impressive size of the appendix and the surrounding tissue. Another helpful sign are markedly enlarged mesenteric lymph nodes (short axial diameter > 12 mms). If CT is not helpful and both clinical symptoms and US abnormalities do not resolve within weeks, colonoscopy is indicated. The diagnosis of sigmoid diverticulitis is often made on clinical grounds. In the classical case the patient presents with localized pain and guarding in the left lower abdomen, fever, leukocytosis and, later on, elevation of the sedimentation rate. However, the diagnosis is not always clear. On one hand the clinical signs may be so atypical that initially another diagnosis is considered, as urinary tract infection, renal colic, perforated peptic ulcer, adnexitis or, -in case of diverticulitis in a rightsided loop of sigmoid- appendicitis. On the other hand, the clinician may think of sigmoid diverticulitis while in fact another condition is present, as sigmoid carcinoma, epiploic appendagitis, a gynecological or urological condition or even a ruptured aortic aneurysm. In all of these cases, US may play a role by making the correct diagnosis at an early point in time. On US, the normal descending colon and upper part of the sigmoid can reliably be identified in virtually all patients due to its consistent location laterally in the left paracolic gutter. The US appearance of the normal sigmoid is variable. The lumen can be empty or filled with feces, and the sigmoid can be contracted or relaxed (Figure). A third factor influencing the aspect is compression by the transducer, which flattens the colon. The muscularis in diverticulosis is often markedly thickened and fecolith-containing diverticula can easily be recognized, as large (4-12 mms), strongly reflective, round-ovoid structures casting an acoustic shadow and localized on the outside of the contour of the contracted colon. If the sigmoid is filled with feces, the diverticula are hardly recognizable. On the left sigmoid diverticulosis in two asymptomatic patients. The US appearance of diverticulitis depends on the stage of the disease. In the earliest stage there is usually local wall thickening of the colon, at first without but later with local blurring of the layer structure. Around the fecolith there is hyperechoic, non-compressible tissue, which represents the inflamed mesentery and omentum trying to seal off the imminent perforation. This inflamed fat, which is best identified during gentle, intermittent compression with the transducer, is obligatory for the diagnosis of diverticulitis (15). On the left a schematic presentation of the benign, natural course of sigmoid diverticulitis as it is observed in 80 % of patients. In stage 0 the neck of the diverticulum becomes obstructed, followed by high intra-diverticular pressure and an impaired defense system against the bacteria lodging within the fecolith. Surrounding inflamed fat represents mesentery and omentum attempting to wall-off the imminent perforation. In stage 1 there is development of a small paracolic abscess, successfully walled-off by mesentery and omentum. The fecolith usually disintegrates and the sigmoid wall is locally weakened. In over 80 % of patients, after one or two days, the pus and the fecolith evacuate towards the colonic lumen via local weakening of the colonic wall at the level of the original diverticular neck. (Figure) . Correspondingly, the patient's symptoms resolve. Note that the residual inflammatory changes (stage R) may remain present for a long time after the evacuation, so the patient can be completely symptomfree when there are still considerable US visible abnormalities present. On the left the natural, benign course of sigmoid diverticulitis. TOP: US reveals mural thickening of the sigmoid at the level of an inflamed diverticulum (arrow) containing a fecolith (stage 0). Note the surrounding hyperechoic, non-compressible tissue representing the omentum and mesentery effectively walling-off the imminent perforation. Within the fat, echolucent linear streaks (arrowheads) are visible. MIDDLE: One day later the patient feels slightly better. The fecolith cannot be recognized as such and the contents of the diverticulum are bulging towards the sigmoid lumen, sign of impending evacuation. BOTTOM: Another two days later, the patient was almost symptomfree. Pus and fecal material have completely evacuated to the sigmoid lumen, leaving an empty diverticulum (curved arrow). In about 20 % of patients, diverticulitis takes a complicated course. Free perforation without any sealing-off by mesentery or omentum, is relatively rare. Spill of fecal material and/or pus to the peritoneal cavity quickly leads to severe peritonitis rendering laparotomy inevitable. Even in case of a larger diverticular abscess (> 2.5 cms) spontaneous evacuation to the colonic lumen remains the rule (Figure). On the left a paracolic abscess due to diverticulitis, effectively walled-off by large masses of inflamed fat, representing mesentery and omentum. The abscess eventually evacuated completely, and the patient recovered without surgery. In some patients however the abscess may evacuate in a less favourable direction (Figure). In the first place, it may find its way to neighbouring diverticula, thus giving rise to more longitudinally oriented abscesses undermining the colonic wall. These abscesses tend to heal badly and often lead to recurrent inflammation with stenosis eventually requiring elective surgery. In rare cases the abscess breaks through to the peritoneal cavity which may lead to diffuse peritonitis or to secondary abscessformation. If a diverticular abscess evacuates into bladder or vagina, a fistula may result. On the left a schematic presentation of the natural evolution of sigmoid diverticulitis, once a paracolic abscess has developed. The most frequent and most favourable pathway is evacuation to the sigmoid lumen. Less favourable is breakthrough to neighboring diverticula (D), giving rise to persistent, longitudinal, cuff-like abscesses. Even worse is the formation of secondary abscesses (A), and eventual perforation to the peritoneal cavity (P). Finally, evacuation to bladder (B), vagina (V) and through the skin will lead to fistulaformation. On the left a colovesical fistula from sigmoid diverticulitis. LEFT: From the lumen of the sigmoid an air-track (arrow) can be followed all the way to the bladder. RIGHT: In the dome of the bladder gas (arrowheads) is seen. From the orificium of the fistula, from time to time the passage of air-bubbles (arrows) could be witnessed. Finally, US has an important role in the diagnosis of alternative conditions: ureterolithiasis, sigmoid carcinoma, ruptured aortic aneurysm, perforated peptic ulcer, appendicitis, epiploic appendagitis. On the left a sigmoid carcinoma in 39-year old patient with clinical signs of diverticulitis. LEFT: Transverse image of the sigmoid 5 cms cranial to the tumour: the colon is thin-walled and well-compressible. RIGHT: Axial US image of the tumour shows asymmetrical, moderately echolucent, wallthickening of the sigmoid. There is also non-compressible fat around the tumour, representing a desmoplastic reaction. On the left a 48-year old man with clinical signs of diverticulitis. br> US reveals an ovoid, non-compressible, avascular fatty mass (arrowheads) while the adjacent sigmoid has a normal aspect. br> The neighboring fat shows hyperemia (arrows). br> During respiration the mass was seen to be adherent to the parietal peritoneum. br> The patient's symptoms disappeared within a week without treatment. These findings are typical for epiploic appendagitis.br> The mass represents the infarcted epiploic appendagebr> The patient's symptoms disappeared within a week without treatment.br> Although in lean patients and in women using the transvaginal probe, differentiation of diverticulitis from colonic carcinoma is often well-possible, it is good practice that in every patient with diverticulitis, colonoscopy is performed when the inflammatory changes have subsided. Percutaneous drainage of a large diverticular abscess is indicated in case of persistent spiking fever, however it is only rarely needed. The presence of a persistent, large paracolic abscess should always raise the suspicion of underlying malignancy. Rightsided colonic diverticulitis in many respects differs from sigmoid diverticulitis. Diverticula of the right colon are usually congenital, solitary, true diverticula containing all bowel layers. The fecoliths within these diverticula are larger, the diverticular neck is wider and there is no hypertrophy of the muscularis of the right colonic wall. Understandably, right colonic diverticulitis, which can occur at any age, almost invariably has a favourable course and never leads to free perforation with peritonitis or large abscesses. Although relatively rare, it is crucial to make a correct diagnosis, since the clinical symptoms of acute RLQ pain may lead to an unnecessary operation for suspected appendicitis. In 40% of patients it even leads to a right hemicolectomy because the surgeon during the operation assumes he is dealing with a colonic malignancy. Although much more frequent in Asians, the diagnosis in the Western world is not rare: in a recent study one case of right colonic diverticulitis is seen for every 15 cases of sigmoid diverticulitis, and for every 30 cases of appendicitis [Oudenhoven]. US, if necessary complemented by CT, has characteristic features and prevents unnecessary surgery for this benign and self-limiting condition. For proper understan?ding of the US images, it is vital to realize the dynamic sequence of the inflammatory process, where each stage of the disease has its own US image [Oudenhoven]. A dangerous pitfall is to mistake a fecolith in the base of an inflamed appendix for a case of cecal diverticu?li?tis. The finding of pneumoperitoneum on a standing chest X-ray in combination with severe acute upper abdominal pain, is strongly suggestive for perforated peptic ulcer. A laparotomy will usually follow without additional imaging. In some cases however, symptoms of a perforated ulcer may be atypical and mimic those of appendicitis, in which case no chest X-ray is made. In other cases of perforated ulcer free air is not present or not detectable. In all those cases, US and CT may be of help. Free air is easier detected by CT than by US, but US better defines the ulcer, demonstrates the free fluid, and can guide puncture of this fluid. On the left a patient with a perforated duodenal ulcer. In the right upper quadrant wall thickening of the duodenal bulb is seen. There are both transmural and extramural (arrow) gasconfigurations. The inflamed fat represents mesenterial and omental fat (fat) attempting -in vain- to wall off the perforation. In the right lower quadrant a large amount of debris-like peritoneal fluid is found (right image). (Continued) On the left another image of the patient with the perforated duodenal ulcer. In the left decubitus position free air can be seen to collect between liver and the lateral abdominal wall. In peptic ulcer US visualizes asymmetrical thickening of the duodenal wall which contains a constant air configuration reaching from the duodenal lumen to the periphery of the wall or even penetrating into the adjacent inflamed fat. Right decubitus position will allow gastric fluid- which is usually present in peptic ulcer disease- to proceed to the duodenum, enabling a better visualization of the ulcer. In case of perforation, an air-track can be found from the ulcer to the peritoneal cavity usually in ventral or cranial direction. Free air is best demonstrated in the left decubitus position between liver and right abdominal wall. A lot of free fluid is usually present which contains airbubbles and foodparticles. Puncture reveals turbid or purulent fluid Birnbaum BA, Jeffrey RB, Jr. CT and sonographic evaluation of acute right lower quadrant abdominal pain. AJR Am J Roentgenol 1998;170(2):361-71. Mindelzun RE, Jeffrey RB Jr (1997) Unenhanced helical CT for evaluating acute abdominal pain: a little more cost, a lot more information. Radiology 205 :43-45. Mindelzun RE, Jeffrey RB Jr. The acute abdomen: current CT imaging techniques. Semin Ultrasound CT MR 1999;20 :63-67. Nisenbaum HL, Birnbaum BA, Myers MM, Grossman RI, Gefter WB, Langlotz CP. The costs of CT procedures in an academic radiology department determined by an activity-based costing (ABC) method. J Comput Assist Tomogr 2000; 24:813-823. Jeffrey RB Jr. In patients with right lower quadrant pain, is sonography or CT the preferred imaging technique for initial evaluation? AJR 1995;164: 1547-1548. Rioux M. Sonographic detection of the normal and abnormal appendix. AJR 1992;158:773-778 Mindel S. The full potential of ultrasound.Lancet 1988; 1: 244. Cobben LPJ, Mol van Otterloo A, Puylaert JBCM. Spontaneously resolving appendicitis: frequency and natural history in 60 patients. Radiology 2000;215:349-352. Jeffrey RB. CT and sonography of the acute abdomen. Raven Press, 1996. Rioux M, Langis P. Primary epiploic appendagitis: clinical, US and CT findings in 14 cases. Radiology 1994; 191: 523-6. Van Breda Vriesman AC, Lohle PNM, Coerkamp EG, Puylaert JBCM. Infarction of omentum and epiploic appendage: diagnosis, epidemiology and natural history. Eur Radiol 1999; 9:1886-92 Puylaert JBCM. US of acute GI tract conditions. Eur Radiol 2001; 11:1867-77. Sarrazin J, Wilson SR. Manifestations of Crohn disease at US. Radiographics 1996;16(3):499-520. Puylaert JBCM, Van der Zant FM, Mutsaers JAEM. Infectious ileocecitis caused by Yersinia, Campylobacter and Salmo?nella: clinical, radiological and US findings. Eur Radiology 1997;7:3-9 Wilson SR (1996) Gastrointestinal tract sonography. Abdom Imaging 21:1-8. Oudenhoven LFIJ, Puylaert JBCM, Koumans RKJ. Right colonic diverticulitis: US and CT findings- new insights about frequency and natural history. Radiology 1998;208:611-618).Julien Puylaert Why perform ultrasonography when you have CT ? Specific advantages of US Who does the US examination ? Normal appendix Appendicitis Spontaneous resolving appendicitis Appendiceal mass Better off with immediate surgery Pitfalls in the US diagnosis of appendicitis Differential diagnosis of diverticulitisAcute Abdomen - Role of UltrasoundDepartment of Radiology, MCH Westeinde Hospital, The Hague, The Netherlands
abdomen4
1
Adrenals
by Theo Falke and Robin Smithuis
Update: 19-5-06 Adrenal masses are seen in 1% of CT-examinations. Most of these masses are benign. Even in patients with a known malignancy these masses are usually non-functioning adenomas. The issue is how to differentiate these benign adenomas from malignant adrenal masses. The most common tumor in the adrenal gland is the adenoma. Adenomas are reported to occur in from 1.4% to 8.7% of postmortem examinations. Adenomas large enough to be recognized at abdominal CT examination are found in 1% of patients. Adrenal adenomas have two properties that differentiate them from non-adenomas. (1) 1. 70% of adenomas contain high intracellular fat (lipid-rich adenomas) and will be of low attenuation on unenhanced CT. 2. Adenomas rapidly wash out contrast. Unenhanced CT. Using a safe threshold value of 10HU on a native CT scan results in a sensitivity of 70% and a high specificity of 98% for the diagnosis of an adenoma. A density equal to or below 10 HU is considered diagnostic of adenoma. 30% of adrenal adenomas do not contain enough intracellular lipid to have a density of less than 10 HU and cannot be differentiated from malignant masses on an unenhanced CT. These adenomas are called lipid-poor (3). Enhanced and Delayed scan. Although on the initial enhanced CT (at 60 sec) most adenomas show mild enhancement, while malignant tumors and pheochromocytomas show strong enhancement, there is too much overlap in attenuation values to allow differentiation between malignant and benign. A number of these adenomas however can be differentiated from malignant masses on the basis of their fast wash-out of contrast. The wash-out can be calculated by comparing the attenuation value at 60 sec with the attenuation value on a delayed scan at 15 minutes. The most commonly used formula is the 'enhancement wash out' formula presented on the left (sometimes called absolute wash out). Attenuation values are measured on unenhanced, initial enhanced (at 60 sec) and delayed CT (at 15 min) . A calculator for the enhancement washout formula aswell as another formula for the 'relative wash out' (only based on the enhanced and delayed scan) is given in reference 1. You only need to fill in the attenuation values and an answer is given whether the mass is probably an adenoma or not. (1) Mostly an adrenal mass will be found on an enhanced CT that is performed in patients with abdominal complaints or patients that are referred for lungcarcinoma staging. As differentiation between benign and malignant is usually not possible on the initial enhanced CT (at 60sec), ordering the patient back for a dedicated adrenal-CT is the best strategy ( although some prefer MRI). If on the unenhanced-CT the density is equal to or below 10 HU the lesion is considered to be an adenoma and no further workup is neccessary If the density is more than 10 HU the wash out should be calculated. If the washout is not compatible with an adenoma, a biopsy can be performed if a definitive diagnosis is crucial to the patients management. On the left an adrenal mass identified during staging for lungcarcinoma. On an enhanced CT at 60 sec the attenuation value was 22HU. The next day patient was ordered back for dedicated adrenal CT. On the unenhanced CT the attenuation value was -19HU indicating the presence of a lipid-rich adenoma. No further work up was needed. On the left a dedicated adrenal protocol in a patient with an adrenal mass. On the unenhanced CT there is a small homogeneous mass that is well defined. The density is 9 HU, which is characteristic of a lipid-rich adenoma. Although the protocol should have stopped at that moment, i.v. contrast was given to determine the washout. The enhancement washout = (43 - 22) : (43 - 9) = 62% indicating a fast washout characteristic of an adenoma. The lower the density on the unenhanced CT and the faster the washout the more confident you can be in making the diagnosis of an adenoma.. The discriminating parameters on CT based on attenuation values only apply to homogenous lesions. Metastases may have a relative low HU due to central necrosis. Although chemical shift MRI is commonly performed, it is believed by some not to provide additional information beyond that which is already available on unenhanced CT (4). The characterization of a lesion as an adenoma relies on the ratio of a decreased relative signal intensity from in phase to opposed phase images and the ratio of adrenal mass and various organs on T2-weighted and chemical shift images. There are no reported studies yet that compare unenhanced CT, delayed enhanced CT, and chemical shift for the discrimination between adenomas and nonadenomas. Adenomas are generally small, homogeneous and well-defined lesions with clear margins. Although the presence of these features are non-specific the absence strongly suggests a nonadenoma. In a retrospective study Gufler et al (5) combined morphologic criteria with the density measurements on unenhanced CT and found a high accuracy in differentiating adrenal adenomas from metastases in patients with a known malignancy. They proposed a scoring system based on density (10% of HU), contour (plus 2 if blurred), homogeneity (plus 1 if inhomogeneous) and size (in cm). By setting a threshold at 7 points all but one lesion in 56 patients were classified correctly. With the new imaging algorithms there is a decreasing need to perform percutaneous Fine needle aspiration (FNA) for definitive characterization. Because a benign cytological diagnosis does not exclude malignancy, FNA cannot be recommended as a standard procedure in the diagnostic work-up. Major complications (2.8-3.6%) include pneumothorax that requires treatment, hemorrhage, abscess, pancreatitis, and seeding along the track. Prior to FNA a clinical and laboratory assessment should be done to exclude the possibility of a pheochromocytoma as FNA may precipitate a hypertensive crisis. Adrenal biopsies can be performed via a posterior approach with the patient in the prone position. The risk of a pneumothorax can be reduced by caudal angulation of the gantry. The lateral decubitus approach is also safe and well tolerated. The patient is placed 'downside' for whichever adrenal gland that is being biopsied. This position elevates the diaphragm on the lesion side and decreases the volume of the lung, thereby reducing the risk of the needle traversing the lung en route to the adrenal gland. The reported accuracy of FNA is 90-96%. Findings from FNA are most likely to be conclusive if the mass is a metastatic tumor. FNA should only be performed when the diagnosis is crucial to patient management (figure). Adrenocortical carcinomas are rare and often diagnosed at an advanced stage. They tend to be large at diagnosis. Patients present with abdominal pain, palpable mass or Cushing's syndrome (50%). The combination of Cushing's syndrome and virilization is frequently found. CT demonstrates a large inhomogeneous mass with heterogeneous enhancement. An adrenal carcinoma is not likely to be less than 5 cm in diameter. Central necrosis is common. Calcification is seen in 20-30% of cases. Most tumours spread by both the haematogenous (lung, liver and bone) and the lymphogenous route. Metastases to the contralateral adrenal, or simultaneous bilateral involvement may occasionally be found. As in renal cell carcinoma tumour tends to spread early by direct invasion of surrounding structures. Extension of the tumour into the renal vein or inferior vena cava is not unusual. MR can be helpful in defining the cephalic extent of the tumour . This is important to the surgeon to gain vascular control. On the left a patient with a small right adrenal carcinoma on CT with high SI on T2-weighted MRI, indistinguishable from lipid-poor adenoma except for invasion into the inferior vena cava. Adrenal metastases are found in 27% of postmortem studies in patients with malignant neoplasms. Lung and breastcarcinoma and melanoma are the most common primary tumors. A diagnosis of adrenal metastasis is important in examining patients with cancer because the metastasis indicates inoperable stage IV disease (except in ipsilateral renal cancer). Adrenal metastases have no specific imaging features. Statistically most non-adenomas are metastases. On the left a patient with partial liver resection for metastasis of a colon carcinoma. Left adrenal metastasis in follow up with no specific imaging findings. The lesion is indistinguishable from a true lipid poor adenoma or non-adenoma such as a neuroendocrine tumour, primary adrenocortical carcinoma, sarcoma or lymphoma and infection. As mentioned above adenomas can be divided into lipid-rich adenomas ( 10HU on unenhanced CT). Lipid-poor adenomas do contain intracellular lipid but not enough to be of an attenuation value The NIH state-of-the-science conference has proposed a minimal standard evaluation for adenomas to rule out endocrine function (4). Myelolipomas are benign tumors composed of bone marrow elements. Usually they are easy to recognize on CT or MR because they contain areas of fat. Calcifications are seen in 20% of cases. On the left another adrenal mass mainly composed of fat. Diagnosis myelolipoma. Pheochromocytomas are paragangliomas arising from the adrenal medulla. They are hormonally active in 90% of cases. Morphologic findings on CT and MRI include large variation in size, homogeneity, and margination of the tumours and significant enhancement in most cases. On MRI tumours have a low SI on T1-weighted images and a very high SI on T2-weighted images. Pheochromocytomas are sometimes called the 10% tumor because they are associated with a 10% risk of malignancy, 10% of the tumors are bilateral, 10% are hormonally inactive and 10% are extra-adrenal (figure). Usually, tumors are larger than 3 cm when seen. They are highly vascular, and larger tumors are prone to hemorrhage and necrosis, even when they are benign. Extensive adrenal hemorrhage may occur at any age and under various circumstances such as severe stress as in surgery, sepsis, burns, hypotension, trauma, hemorrhagic diathesis and underlying conditions such as adenoma, cyst and tumour. Cysts may be of any size and in most instances are unilateral. Large cysts may be complicated by hemorrhage and consequent onset of acute symptoms. Pathological substrates include epithelial , endothelial, parasitic, and pseudocysts. Most importantly lesions show a thin wall and no enhancement after intravenous contast material. This site provides a calculator to measure the wash-out of adrenal masses for differentiation of benign masses (usually adenomas) from malignant lesions (usually metastases). N. Reed Dunnick and Melvyn Korobkin Am. J. Roentgenol., Sep 2002; 179: 559 - 568. Elaine M. Caoili et al. Radiology 2002;222:629-633. State of the Science Statement (html and pdf) and 3 day video conference Gufler H, Eichner G, Grossmann A, Krentz H, Schulze CG, Sauer S, Grau G. J Comput Assist Tomogr. 2004 Nov-Dec;28(6):818-22. Ian C. Mitchell, Fiemu E. Nwariaku The Oncologist, Vol. 12, No. 2, 168-174 JJ Erasmus, EF Patz Jr, HP McAdams, JG Murray, J Herndon, RE Coleman and PC Goodman American Journal of Roentgenology, Vol 168, 1357-1360 Semin Chong et al RadioGraphics 2006;26:1811-1824 Michael A. Blake et al RadioGraphics 2006;26:1335-1353 Georg Mansmann, Joseph Lau, Ethan Balk, Michael Rothberg, Yukitaka Miyachi and Stefan R. Bornstein Endocrine Reviews 25 (2): 309-340Theo Falke and Robin Smithuis CT-examination CT-Algorithm benign versus malignant MRI Morphologic features Percutaneous biopsy Primary Adrenocortical carcinoma Metastases Adenoma Myelolipoma Pheochromocytoma Hemorrhage CystAdrenalsRadiology Department of the Groene hart hospital in Gouda and the Rijnland hospital in Leiderdorp, the Netherlands
abdomen5
1
Anatomy of the liver segments
by Robin Smithuis
Liver anatomy can be described using two different aspects: morphological anatomy and functional anatomy. The traditional morphological anatomy is based on the external appearance of the liver and does not show the internal features of vessels and biliary ducts branching, which are of obvious importance in hepatic surgery. C. Couinaud (1957) divided the liver into eight functionally indepedent segments. This classification will be presented here with several illustrations. The Couinaud classification of liver anatomy divides the liver into eight functionally indepedent segments. Each segment has its own vascular inflow, outflow and biliary drainage. In the centre of each segment there is a branch of the portal vein, hepatic artery and bile duct. In the periphery of each segment there is vascular outflow through the hepatic veins. Right hepatic vein divides the right lobe into anterior and posterior segments. Middle hepatic vein divides the liver into right and left lobes (or right and left hemiliver). This plane runs from the inferior vena cava to the gallbladder fossa. Left hepatic vein divides the left lobe into a medial and lateral part. Portal vein divides the liver into upper and lower segments. The left and right portal veins branch superiorly and inferiorly to project into the center of each segment. Because of this division into self-contained units, each segment can be resected without damaging those remaining. For the liver to remain viable, resections must proceed along the vessels that define the peripheries of these segments. This means, that resection-lines parallel the hepatic veins, The centrally located portal veins, bile ducts, and hepatic arteries are preserved. There are eight liver segments. Segment 4 is sometimes divided into segment 4a and 4b according to Bismuth. The numbering of the segments is in a clockwise manner (figure). Segment 1 (caudate lobe) is located posteriorly. It is not visible on a frontal view. The illustrations above are schematic presentations of the liversegments. In reality however the proportions are different. On a normal frontal view the segments 6 and 7 are not visible because they are located more posteriorly. The right border of the liver is formed by segment 5 and 8. Although segment 4 is part of the left hemiliver, it is situated more to the right. Couinaud divided the liver into a functional left and right liver (in French 'gauche et droite foie') by a main portal scissurae containing the middle hepatic vein. This is known as Cantlie's line. Cantlie's line runs from the middle of the gallbladder fossa anteriorly to the inferior vena cava posteriorly. On this illustration it looks as if the medial part of the left lobe is separated from the lateral part by the falciform ligament. However it actually is the left hepatic vein, that separates the medial part (segment 4) from the lateral part (segments 2 and 3). The left hepatic vein is located slightly to the left of the falciform ligament. The far left figure is a transverse image through the superior liver segments, that are divided by the hepatic veins. The right figure shows a transverse image at the level of the left portal vein. At this level the left portal vein divides the left lobe of the liver into the superior segments (2 and 4A) and the inferior segments (3 and 4B). The left portal vein is at a higher level than the right portal vein. The image on the far left is at the level of the right portal vein. At this level the right portal vein divides the right lobe of the liver into superior segments (7 and 8) and the inferior segments (5 and 6). The level of the right portal vein is inferior to the level of the left portal vein. At the level of the splenic vein, which is below the level of the right portal vein, only the inferior segments are seen (right image). The caudate lobe or segment 1 is located posteriorly. The caudate lobe is anatomically different from other lobes in that it often has direct connections to the IVC through hepatic veins, that are separate from the main hepatic veins. The caudate lobe may be supplied by both right and left branches of the portal vein. On the left a patient with cirrhosis with extreme atrophy of the right lobe, normal volume of the left lobe and hypertrophy of the caudate lobe. Due to a different blood supply the caudate lobe is spared from the disease process and hypertrophied to compensate for the loss of normal liverparenchyma. There are many other anatomical and functional descriptions of the liver anatomy. In the classical description the external appearance of the liver is used to describe the anatomy. However there are many differences between this classical model and the fuctional models, as popularized by Couinaud and Bismuth. A more detailed discussion of the various models is given in reference 4. The classical description of the liver anatomy is based on the external appearance. On the diaphragmatic surface, the ligamentum falciforme divides the liver into the right and left anatomic lobes, which are very different from the functional right and left lobes (or right and left hemiliver). In this classical description, the quadrate lobe belongs to the right lobe of the liver, but functionally it is part of left lobe. This classification is very similar to the Couinaud classification, although there are small differences. It is popular in the United States, while Couinaud's classification is more popular in Asia and Europe. According to Bismuth three hepatic veins divide the liver into four sectors, further divided into segments. These sectors are termed portal sectors as each is supplied by a portal pedicle in the centre. The separation line between sectors contain a hepatic vein. The hepatic veins and portal pedicels are intertwined, as are the fingers of two hands. The left portal scissura divides the left liver into two sectors: anterior and posterior. Left anterior sector consists of two segments: segment IV, which is the quadrate lobe and segment III, which is anterior part of anatomical left lobe. These two segments are separated by the left hepatic fissure or umbilical fissure. Left posterior sector consists of only one segment II. It is the posterior part of left lobe. In the Couinaud classification little attention is given to the high prevalence of anatomical variations which occur, especially in the right hemiliver. Using volumetric acquisition techniques, such as magnetic resonance imaging or spiral computed tomography scanning, detailed insight into the individual segmental anatomy can now be obtained in a non-invasive manner (2,3). The significance of this anatomical insight lies in the planning of anatomical resections, whereby the relationship between tumour and individual segmental anatomy can be depicted in a three-dimensional format. Three dimensional liver imaging is of most practical value if a resection of one or more segments or sectors is considered, especially in the right hemiliver. In these cases, 3D liver imaging can demonstrate the precise location of the scissuras to the surgeon pre-operatively. 3-D tutorials of the Division of Physiologic Imaging, Dept. of Radiology, Univ. of Iowa MS van Leeuwen, J Noordzij, MA Fernandez, A Hennipman, MA Feldberg and EH Dillon Department of Radiology, University Hospital Utrecht, The Netherlands van Leeuwen MS, Noordzij J, Hennipman A, Feldberg MA. Department of Radiology and Surgery, University Hospital Utrecht, The Netherlands. Saulius Rutkauskas et al. Clinic of Radiology, Institute of Anatomy, Clinic of Surgery, Kaunas University of Medicine, LithuaniaRobin Smithuis Couinaud classification Segments numbering Transverse anatomy Caudate lobe Classical Anatomy Bismuth's classification VariationsAnatomy of the liver segmentsRadiology Department of the Rijnland Hospital, Leiderdorp, the Netherlands
abdomen6
1
Aorta - Aneurysm rupture
by Jay P. Heiken, M.D.
This article is based on a presentation given by Jay Heiken and adapted for the Radiology Assistant by Robin Smithuis. Jay Heiken is professor of radiology with special interest in abdominal imaging and co-author of the well known book 'Computed Body Tomography With Mri Correlation'. The classical findings in aortic aneurysm rupture are well known. In this article we will present the more subtle findings of contained leak and pending rupture of aortic aneurysm. Aortic aneurysm rupture is the most important diagnosis you want to be able to exclude in patients with acute abdominal pain especially when they present with back or flank pain. The primary signs of AAA rupture are periaortic stranding, retroperitoneal hematoma and extravasation of iv. contrast. On the left we see three patients with aortic aneurysm rupture. In the image on the far left we only see a little bit of peripheral soft tissue density adjacent to the aneurysm and indeed this is a sign that this patient is at risk for frank rupture. The other two cases show retroperitoneal hematoma and contrast leakage outside the aorta, which makes it easier to diagnose. On the left a classical case in a patient with an aneurysm of the abdominal aorta and a large hyperdense retroperitoneal hematoma due to rupture. The majority of these cases show posterior periaortic hemorrhage and in cases of massive hemorrhage, the posterior pararenal and perirenal compartments are the most frequently involved sites. The CT features of contained leak or pending rupture of an aortic aneurysm may be subtle and easily overlooked. We have to look for the high-attenuating crescent sign, focal discontinuity of intimal calcification or tangential calcium or the draped aorta sign (table on the left). The high attenuating crescent represents an acute hematoma within either the mural thrombus or the aneurysmal wall. This sign is strongly associated with AAA rupture. Sensitivity of the high-attenuating crescent sign as an indication of complicated aneurysm is 77%; specificity, 93%; and positive predictive value of 53%. So even if there are no primary signs of rupture, we need to indicate to the referring physician, that this patient is at very high risk for aneurysm rupture within the next few days. On the left two more cases of the high-attenuating crescent sign. In the case on the right we can also identify a retroperitoneal hematoma, so in this case there is a frank AAA rupture. Another sign of impending rupture or contained leakage is focal discontinuity of intimal calcification. In most of these cases we can also identify the tangential calcium sign. In these cases it looks as if the calcium is pointing out away from the expected circumference of the aneurysm. On the left we see another example of the tangential calcium sign. The intimal calcification points away from the aneurysm and there is retroperitoneal leakage. On the left a patient who presented with backpain. The image on the far left shows a bulge of the aorta. This either represents a focal weakening of the aortic wall or a localized leak. This patient unfortunately was reported as having no leak and discharged from the emergency departement. Two weeks later there was a frank rupture. A positive aortic drape sign is considered to be present when the following features are seen: On the left another patient who presented with backpain. There was no evidence of aneurysm leakage, but we see a draped aorta. The posterior contour of the aorta follows the contour of the spine as if the aorta is draped over the vertebral body. There is no imaging follow up in this patient, but three hours after this image was taken, this patient exsanguinated from a ruptured AAA. The leakage was probably where the bulge was (arrow). Radiology 1994; 192:359-362 WB Mehard, JP Heiken and GA Sicard Mallinckrodt Institute of Radiology, Washington University School of Medicine, St Louis, MO 63110. CL Siegel, RH Cohan, M Korobkin, MB Alpern, DL Courneya and RA Leder Department of Radiology, University of Michigan Medical Center, Ann Arbor AJR 1994;163:1123-1129 KE Halliday and A al-Kutoubi Department of Radiology, St Mary's Hospital, London, England. Radiology 1996; 199:41-43Jay P. Heiken, M.D. High-attenuating crescent Focal discontinuity of intimal calcification Tangential calcium sign Draped AortaAorta - Aneurysm ruptureMallinckrodt Institute of Radiology of the Washington University School of Medicine, St. Louis, Missouri
abdomen7
1
Appendicitis - Mimics
by Adriaan van Breda Vriesman M.D. and Julien Puylaert M.D.
Introduction In this overview we focus on nonsurgical appendicitis-mimicking diseases. A correct imaging diagnosis prevents an unnecessary operation or costful in-hospital observation. If you encounter printing problems with the margins of the document, try to adjust the margins or the scale of the document in the print settings. Sonography and CT allow direct visualization of the normal or inflamed appendix. The normal appendix can be identified in 67-100% of patients without appendicitis who undergo CT [1]. At sonography the normal appendix is less frequently visualized, with results varying between 0-82% [1], reflecting the operator dependency of sonography. One of the most important imaging criteria in the evaluation of appendicitis is the outer diameter of the appendix. Although an overlap of appendiceal diameters in normal and inflamed appendices has been reported, a threshold value of 6-7 mm is most commonly used [1]. (Fig. 1). A normal appendix has a maximum outer diameter of 6 mm, is surrounded by homogeneous non-inflamed fat, and often contains intraluminal gas [2] (Fig. 2). An inflamed appendix has a diameter larger than 6 mm, and is usually surrounded by hyperechoic inflamed fat at sonography (Fig. 3a). Other strongly supportive signs of inflammation include the presence of an appendicolith, cecal apical thickening Another supportive sign for appendicitis is hypervascularity of the appendix wall on color Doppler sonography [1] (Fig. 3b). At CT the inflamed appendix is surrounded by fat-stranding (Fig. 4). Mesenteric adenitis has been reported to be the second most common cause of right lower quadrant pain after appendicitis, accounting for 2-14% of the discharge diagnoses in patients with a clinical suspicion of appendicitis [3]. It is defined as a benign self-limiting inflammation of right-sided mesenteric lymph nodes without an identifiable underlying inflammatory process, occurring more often in children than in adults. Sonography and CT show clustered adenopathy (Fig. 5). Because adenopathy also frequently occurs with appendicitis, the normal appendix must be confidently visualized on imaging studies before assigning a diagnosis of mesenteric adenitis. Infectious enterocolitis may cause mild symptoms resembling a common viral gastroenteritis, but it may also clinically present with features indistinguishable from appendicitis [4]. This latter presentation may occur in bacterial ileocecitis, caused by Yersinia, Campylobacter, or Salmonella. Imaging studies show mural thickening of the terminal ileum and cecum without inflammation of the surrounding fat (Fig. 6), and moderate mesenteric adenopathy. Epiploic appendages are small adipose protrusions from the serosal surface of the colon. An epiploic appendage may undergo torsion and secondary inflammation, causing focal abdominal pain that simulates appendicitis when located in the right lower quadrant. Epiploic appendagitis is a self-limiting disease that has been reported in approximately 1% of patients clinically suspected of having appendicitis [5]. Sonography and CT depict an inflamed fatty mass adjacent to the colon (Fig. 7), containing a characteristic hyperattenuating ring of thickenend visceral peritoneal lining on CT [5]. Omental infarction has a pathophysiology and clinical presentation similar to that of epiploic appendagitis, with the infarcted fatty tissue being a right-sided segment of the omentum. Imaging shows a cakelike inflamed fatty mass (Fig. 8), larger than in epiploic appendagitis and lacking a hyperattenuating ring on CT. In some cases it may be difficult to distinguish epiploic appendagitis from omental infarction (Fig. 9), however, this distinction has no clinical importance as both have a similar benign natural history [5]. Right-sided colonic diverticulitis may clinically mimic appendicitis or cholecystitis, though the patient's history is generally more protracted. In contrast to sigmoid diverticula, right-sided colonic diverticula are usually true diverticula, that is, outpouchings of the colonic wall containing all layers of the wall. This may possibly explain the essentially benign self- limiting character of right-sided diverticulitis [6]. Sonography and CT findings consist of inflammatory changes in the pericolic fat with segmental thickening of the colonic wall, at the level of an inflamed diverticulum (Fig. 10). Crohn disease often causes long-standing symptoms, but up to one third of patients with ileocecal Crohn disease present with initial symptoms so acute that they are misdiagnosed as appendicitis [7]. In the acute active phase of ileocecal Crohn disease, imaging shows transmural bowel wall thickening, often predominantly of the submucosal layer, with frequent inflammatory changes of the surrounding fat (Fig. 11). Uncomplicated Crohn disease can initially be treated with anti-inflammatory drugs. Gynecologic conditions such as pelvic inflammatory disease or a hemorrhagic functional ovarian cyst can cause acute pelvic pain that may simulate appendicitis. In pelvic inflammatory disease the imaging findings vary according to the severity of the disease, and may be normal in early conditions. In more advanced stages, findings may include enlargement of the internal genital organs with indistinct contours, and free pelvic fluid (Fig. 12). In absence of a drainable tubo-ovarian abscess, treatment is medically with antibiotics. An hemorrhagic ovarian cysts appears as a complicated cyst at sonography and a high- attenuation adnexal mass at unenhanced CT, and does not require any treatment. Urolithiasis may present with right lower quadrant pain when obstruction is caused by a distal ureteral stone. Unenhanced CT (Fig. 13) is more accurate in detecting ureteral stones than sonography, Ultrasound may show both hydronephrosis and hydroureter as signs of obstruction (Fig. 14). A rectus sheath hematoma may be easy to diagnose in patients presenting with a painful palpable mass under anticoagulant therapy, however, small nonpalpable hematomas may clinically masquerade as appendicitis and also occur in patients without anticoagulantia [8]. Sonography and CT show a hemorrhagic mass within the sheath of the rectus abdominis muscle (Fig. 15). No treatment is required other than adjusting any anticoagulant therapy. A broad spectrum of nonsurgical diseases may clinically present as appendicitis in patients without appendicitis. The radiologist should be aware of the sonographic and CT features of these alternative disorders, as a correct imaging diagnosis prevents an unwarranted operation and unnecessary hospital resource use. Birnbaum BA, Wilson SR Appendicitis at the millennium. Radiology 2000;215:337-348 Rettenbacher T, Hollerweger A, Macheiner P, et al. Presence or absence of gas in the appendix: additional criterion to rule out or confirm acute appendicitis. Radiology 2000; 214:183-187 Macari M, Hines J, Balthazar E, Megibow A. Mesenteric adenitis: CT diagnosis of primary versus secondary causes, incidence, and clinical significance in pediatric and adult patients. Am J Roentgenol 2002;178:853-858 Puylaert JBCM, van der Zant FM, Mutsaers JA. Infectious ileocecitis caused by Yersinia, Campylobacter, and Salmonella: clinical, radiological and US findings. Eur Radiol 1997; 7:3-9 Breda Vriesman AC, Puylaert JBCM. Epiploic appendagitis and omental infarction: pitfalls and look-alikes. Abdom Imaging 2002;27:20-28 Oudenhoven LFIJ, Koumans RKJ, Puylaert JBCM. Right colonic diverticulitis: US and CT findings - new insights about frequency and natural history. Radiology 1998;208:611-618 Sturm EJC, Cobben LPJ, Meijssen MAC, Werf SDJ, Puylaert JBCM. Detection of ileocecal Crohn's disease using ultrasound as the primary imaging modality. Eur Radiol 2004;14:778-782 Lohle PN, Puylaert JB, Coerkamp EG, Hermans ET. Nonpalpable rectus sheath hematoma clinically masquerading as appendicitis: US and CT diagnosis. Abd Imaging 1995;20:152- 154Adriaan van Breda Vriesman M.D. and Julien Puylaert M.D. Normal Appendix Appendicitis Mesenteric adenitis Bacterial ileocecitis Epiploic appendagitis Omental infarction Right-sided colonic diverticulitis Crohn disease Gynecologic conditions Urolithiasis Rectus sheath hematomaAppendicitis - MimicsRadiology Department, Rijnland Hospital, Leiderdorp and Medical Centre Haaglanden, the Hague, the Netherlands
abdomen10
1
Diagnostic Work up of Ovarian Cysts
by Wouter Veldhuis, Robin Smithuis, Oguz Akin and Hedvig Hricak
Ovarian cancer is the second most common of all gynecologic malignancies. It is the leading cause of death in this category of diseases, frequently presenting as a complex cystic mass. The finding of an adnexal cyst causes considerable anxiety in women due to the fear of malignancy. However, the vast majority of adnexal cysts - even in postmenopausal women - are benign. In this article we will focus on specific features of ovarian cysts that are helpful in making a differential diagnosis. We will present a roadmap for the diagnostic work-up and management of ovarian cystic masses, based on ultrasound and MRI findings. In the article 'Ovarian Cysts: Common Lesions' the imaging features of normal ovaries and the most common ovarian cystic masses will be presented, as well as several less common cystic lesions. Based on these steps we can determine further management: ignore, follow-up with US, further evaluation with MRI or excision. Role of Ultrasound For characterization of ovarian masses, ultrasound is often the first-line method of choice, especially for distinguishing cystic from complex cystic-solid and solid lesions. Role of CT CT is useful for the N- and M-staging of proven malignant lesions. Role of MRI For complex lesions, primary evaluation with ultrasound is often followed by further evaluation with MRI. Even with MRI it is often not possible to make an accurate diagnosis of neoplastic subtype. By using MRI as an adjunct to sonography a delay in the treatment of potentially malignant ovarian lesions is prevented. This is not only beneficial to the small number of women who do have ovarian cancer, but also a proven cost-effective approach to the management of sonographically indeterminate adnexal lesions. If a cystic adnexal mass is present and you suspect an ovarian origin, the first thing to do is try to identify the ovaries. If the gonadal vessels lead to the lesion with no separately identifiable normal ovaries, then most likely you are dealing with an ovarian lesion. If both ovaries are separately identifiable from the lesion, you are dealing with a non-ovarian cystic lesion, or a lesion that mimics a cystic mass. The next step would be to check if there is uni- or bilateral disease and to look for any solid components that may indicate malignancy. Also look for secondary findings like ascites, enlarged lymph nodes and peritoneal deposits. The table shows a differential diagnosis for possible cystic ovarian masses. A helpful tool to identify the ovaries is to follow the ovarian veins caudally. Scroll through the CT-images and follow the right ovarian vein from where it joins the inferior vena cava, and the left ovarian vein where it joins the left renal vein, until you identify the ovaries. Pattern recognition on ultrasound often allows a fairly confident diagnosis of common cystic ovarian masses. This means that in many cases the diagnostic work-up is based on determining the probability that we are dealing with a lesion which falls into the category of a simple cyst, hemorrhagic cyst, endometrioma or a mature cystic teratoma (commonly referred to as a dermoid cyst). Most other cystic lesions are indeterminate and therefore possibly malignant. These therefore require further evaluation, either with MRI or surgical excision. US findings that allow a confident diagnosis of a simple ovarian cyst are: The US-image shows two simple cysts in the right ovary with ovarian stroma in between. The surrounding vessels are normal and there are no vascularized septations. These were simple follicular cysts in a premenopausal woman. Differential diagnosis Most simple cysts are functional cysts, usually follicular cysts. They are commonly seen in premenopausal women, but functional cysts also still do occur in postmenopausal women. Some simple cysts may turn out to be paraovarian or paratubal cysts. A hydrosalpinx may also mimic an ovarian cyst. Cystadenomas can also present as simple cysts, but they usually present as a large cyst in a postmenopausal woman. In a large cancer screening study from 1987 to 2002 including 15,106 women of 50 years or older, 2763 women (18%) were diagnosed with a unilocular ovarian cyst. None of these isolated unilocular cysts turned out to be ovarian cancer (4). In women of reproductive age, cysts up to 3 cm are a normal physiologic finding. These simple physiologic cysts do not need to be described in the imaging report and do not require follow-up (1). Cysts up to 7 cm in both pre- and postmenopausal woman are almost certainly benign. Cysts larger than 7 cm may be difficult to assess completely with US and therefore further imaging with MR or surgical evaluation should be considered. When a Graafian follicle or follicular cyst bleeds, a complex hemorrhagic ovarian cyst (HOC) is formed. US findings that allow a confident diagnosis of a hemorrhagic ovarian cyst are: In premenopausal women short term follow-up is recommended in hemorrhagic cysts > 5 cm. The same follow-up is recommended in early postmenopausal women who have a cyst with all the characteristics of a HOC Larger hemorrhagic cysts in the early menopause and any hemorrhagic cyst in the late menopause should be considered possibly neoplastic and MRI or surgical evaluation should be considered. Differential diagnosis When hemorrhagic cysts present with diffuse low-level echoes, their appearance can be similar to that of endometriomas. In the acute phase a hemorrhagic cyst may be completely filled with low-level echoes, simulating a solid mass (5). Clot in a hemorrhagic cyst may occasionally mimic a solid nodule in a neoplasm. Clot, however, often has concave borders due to retraction, while a true mural nodule has outwardly convex borders. In both cases there will be no internal flow at Doppler US and there will be good through-transmission. Hemorrhagic cysts typically resolve within 8 weeks. The ultrasound image shows multiple simple and one complex right ovarian cyst, with diffuse low-level echos and absence of flow on Doppler US. Note that there is good through-transmission, also through the complex cyst (blue arrow). On the T1 with fatsat the lesion remains bright, ruling out a fatty lesion. After Gd administration there is no enhancement, confirming that this is a cystic hemorrhagic lesion, most likely a hemorrhagic ovarian cyst, although your differential may include an endometrioma. US findings that allow a confident diagnosis of an endometrioma are: In women of any age, probable endometriomas require initial 6-12 week follow-up to rule out a hemorrhagic cyst. Until surgically removed, endometriomas require follow-up with ultrasound, for example on a yearly basis. This image from a vaginal ultrasound shows a large hypoechoic, cystic lesion with diffuse low-level echoes and two small echogenic foci. These have been postulated to be cholesterol deposits, but may also constitute small blood clots or debris. It is important to differentiate these echogenic foci from true wall nodules. Finding these echogenic foci makes the diagnosis of an endometrioma very likely. US findings that are characteristic of a mature cystic teratoma are: Shown are transvaginal ultrasound images of two patients that demonstrate the 'tip-of-the-iceberg' sign: acoustic shadowing from the hyperechoic part of the dermoid cyst (arrow). When misinterpreted as bowel gas, the lesion may be overlooked. All other cystic lesions are regarded as possibly neoplastic and therefore possibly malignant. Surgical resection is needed by an oncologic gynaecologist, who may request prior imaging-based staging. Findings indicating possible neoplasm: Once we have determined a cystic ovarian lesion is either a probable simple cyst, hemorrhagic cyst, endometrioma or mature cystic teratoma, or is indeterminate, the next step is to place the patient in a low-risk or high-risk group (table). The final decision to ignore, follow or excise a cystic ovarian lesion is based on: That said, the great majority of cystic ovarian lesions is benign. While the risk of malignancy does increase with age, even in post-menopausal women the risk of malignancy in a simple ovarian cyst Although complex ovarian cysts in post-menopausal women are also most often benign, they do require further work-up, because of the chance of malignancy. The natural history of incidentally detected pelvic masses with benign US morpgology is not known and therefore the optimal management is also unknown. The roadmap is based on the 2010 Consensus Guidelines published in (1) and (2) and on the findings in (3) and (4). The mentioned size cut-offs and follow-up frequencies are accepted practices but not ironclad rules. Local guidelines may differ based on the clinical scenario and institutional practice preferences. Many of the imaging criteria described in this article are the same for ultrasound, CT and MRI, although of course not every feature is equally detectable on all modalities. Risk factors Age is the most important risk factor for all women. Lesions in pre-menopausal and post-menopausal women are managed differently. Several other factors (see table) may place a woman in a higher risk category. Concordantly, the roadmap shows two pathways, one for lower-risk and one for higher-risk patients. MRI protocol There are many possible 'Pelvic/Ovarian mass' protocols. The basic building blocks are simple and are the same for all protocols: A very short protocol may consist of only 1, 2 and 3 (e.g., when the request is to 'rule out an ovarian mass'). Many radiologists prefer a slightly more comprehensive protocol including 4, and often 5. When the clinical setting is characterization or staging of a known ovarian lesion, 4 (or CT) and 5 should always be included. The role of diffusion-weighted MRI is yet to be determined, but DWI is a useful aid in the detection of lymph nodes, tumors and peritoneal deposits. For the purpose of detection, the DW images are sometimes fused with (superimposed on) anatomical T2W images. DWI cannot discriminate benign from metastatic lymph nodes. Further differences in protocols all arise as variations on this simple theme. For example: MR imaging is a valuable adjunct to US, as it allows identification of blood products within hemorrhagic masses that may mimic solid tumor at US. Fat-suppressed T1-weighted MR images may reveal small amounts of fat, which allows the diagnosis of a mature teratoma ('dermoid'). Contrast-enhanced T1-weighted MR imaging depicts features of malignancy such as enhancing mural nodules and/or enhancing solid areas with or without necrosis (3). These MR images show a lesion with high signal on T1. This indicates either blood, other high protein content or fat. On the image with fat-saturation there is suppression of the signal. This means that we are dealing with a fat-containig lesion, i.e. a mature cystic teratoma. The US image shows an echogenic lesion. The corresponding lesion has a high signal on the T1-weighted MR image. This indicates either blood, high protein or fat. On the image with fat-saturation there is no suppression of the signal. This means that we are dealing with a blood-containig lesion, i.e. most likely a hemorrhagic cyst. by Deborah Levine et al September 2010 Radiology, 256, 943-954. by Spencer JA et al Eur Radiol. 2010 Jan;20(1):25-35. by John A. Spencer et al September 2010 Radiology, 256, 677-694. by Modesitt SC, Pavlik EJ, Ueland FR, DePriest PD, Kryscio RJ, van Nagell JR Jr. Obstet Gynecol. 2003 Sep;102(3):594-9. by Douglas L. Brown, MD, Kika M. Dudiak, MD and Faye C. Laing, MD February 2010 Radiology, 254, 342-354. by Penelope L. Moyle et al July 2010 RadioGraphics, 30, 921-938. by Paula J. Woodward et al RadioGraphics 2001; 21:193-216. by Rajkotia K, Veeramani M, Macura KJ Top Magn Reson Imaging 2006; 17:379-97 by Michael P. Stany et al AJR 2010; 194:337-342 by Maitray D. Patel, MD, Vickie A. Feldstein, MD and Roy A. Filly, MD 2005 J Ultrasound Med 24:607-614 by Stefano Guerriero et al 2004 J Ultrasound Med 23:1193-1200 by Kiran A. Jain J Ultrasound Med 21:879-886Wouter Veldhuis, Robin Smithuis, Oguz Akin and Hedvig Hricak Role of imaging Simple cyst Hemorrhagic ovarian cyst - HOC Endometrioma Mature cystic teratoma Any other cyst - possible neoplasmDiagnostic Work up of Ovarian CystsDepartment of Radiology of the University Medical Center of Utrecht, of the Rijnland hospital in Leiderdorp, the Netherlands and the Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, USA
abdomen11
1
Gallbladder - Wall Thickening
by Adriaan C. van Breda Vriesman, Robin Smithuis, Dries van Engelen and Julien B.C.M. Puylaert
Thickening of the gallbladder wall is a relatively frequent finding at diagnostic imaging studies. Historically, a thick-walled gallbladder has been regarded as proof of primary gallbladder disease, and it is a well-known hallmark feature of acute cholecystitis. The finding itself, however, is non-specific and can be found in a wide range of gallbladder diseases and extracholecystic pathological conditions. In this review we discuss and illustrate the various causes of a generalized thickened gallbladder wall. If you encounter printing problems with the margins of the document, try to adjust the margins or the scale of the document in the print settings. Sonography, CT and MRI all allow direct visualization of the normal and thickened gallbladder wall. Traditionally, sonography is used as the initial imaging technique for evaluating patients with suspected gallbladder disease, because of its high sensitivity in the detection of gallbladder stones, its real-time character, speed and portability [1]. However, CT has become popular for evaluating the acute abdomen and often is the first modality to detect gallbladder wall thickening [2], or it may be used as an adjunct to an inconclusive sonography or for staging of disease. The potential value of MRI in the evaluation of gallbladder pathology has been shown [3], but it still plays little role. The normal gallbladder wall appears as a pencil-thin echogenic line at sonography. The thickness of the gallbladder wall depends on the degree of gallbladder distention and pseudothickening can occur in the postprandial state. The normal gallbladder wall is usually perceptible at CT as a thin rim of soft-tissue density that enhances after contrast injection. Thickening of the gallbladder wall is a relatively frequent finding at diagnostic imaging studies. A thickened gallbladder wall measures more than 3 mm, typically has a layered appearance at sonography [1], and at CT frequently contains a hypodense layer of subserosal oedema that mimics pericholecystic fluid [2] The differential diagnosis of gallbladder wall thickening is listed on the left. Diffuse gallbladder wall thickening may produce a diagnostic problem, as it occurs in symptomatic and asymptomatic patients, and in patients with and without an indication for a cholecystectomy. Diffuse thickening of the gallbladder wall may occur in patients who do not have a primary gallbladder disease, but in whom the gallbladder is secondarily involved in an extrinsic pathological condition. In these patients a cholecystectomy is unwarranted, and gallbladder abnormalities will usually return to normal after correction of its extrinsic cause. Acute cholecystitis is the fourth most common cause of hospital admissions for patients presenting with an acute abdomen [4], and it is the prime diagnostic concern when a thick-walled gallbladder is found at imaging. This feature, however, is not pathognomonic and additional imaging signs should be present to support the diagnosis of acute calculous cholecystitis such as an obstructing gallstone, hydropical dilatation of the gallbladder, a positive sonographic Murphy's sign ( i.e., pain elicited by pressure over the sonographically located gallbladder), pericholecystic fat inflammation or fluid and hyperemia of the gallbladder wall at power Doppler. On the left images of a 62-year-old man with acute calculous cholecystitis. Transverse sonogram at the spot of maximum tenderness shows a non-compressible hydropically distended thick-walled gallbladder (arrowheads), with an intraluminal stone and sludge or debris. Contrast-enhanced CT depicts extensive fat inflammation (arrowheads) surrounding the gallbladder (arrow). Acute acalculous cholecystitis mainly occurs in critically ill patients, presumably due to increased bile viscosity from fasting and medication that causes cholestasis. The imaging features are those of acute cholecystitis, except for the absence of stones whereas gallbladder sludge is usually present (Fig). Because in critically ill patients gallbladder abnormalities are frequently found secondary to systemic disease (see below), acalculous cholecystitis can be difficult to diagnose [5]. In these patients a percutaneous cholecystostomy can be both diagnostic and therapeutic. Chronic cholecystitis is a term used clinically to refer to symptomatic gallbladder stones that cause transient obstruction, leading to a low-grade inflammation with fibrosis [1]. Correlation of the imaging finding of a stone-containing slightly thick-walled gallbladder with the clinical history is critical. On the left images of a 49-year-old woman with chronic cholecystitis. This patient had fasted overnight, so the wall-thickening does not represent physiologic contraction. Correlation of these findings with her clinical history of recurrent colic-like right upper quadrant pain, due to transient gallbladder obstruction, is essential for the diagnosis. Xanthogranulomatous cholecystitis is an unusual variant of chronic cholecystitis, characterized by a lipid-laden inflammatory process comparable to xanthogranulomatous pyelonephritis. Imaging studies show marked gallbladder wall thickening, often containing intramural nodules that are hypoechoic at sonography and hypoattenuating at CT, representing abscesses or foci of xanthogranulomatous inflammation. These features overlap with those of gallbladder carcinoma, making preoperative distinction between these entities often impossible [6]. On the left a 71-year-old man with xanthogranulomatous cholecystitis. Contrast-enhanced CT shows a deformed and thickened gallbladder wall containing hypoattenuating nodules . These represent abscesses or foci of inflammation. A porcelain gallbladder is a rare disorder in which chronic cholecystitis produces mural calcification. In these patients a prophylactic cholecystectomy has been advocated because of its association with gallbladder carcinoma [4]. However, this association appears to be weak. Gallbladder carcinoma is the fifth most common malignancy of the gastrointestinal tract, and is found incidentally in 1% to 3% of cholecystectomy specimens [4]. It is often detected at a late stage of the disease, due to lack of early or specific symptoms. Gallbladder carcinoma has various imaging appearances, ranging from a polypoid intra-luminal lesion to an infiltrating mass replacing the gallbladder, and it may also present as diffuse mural thickening. Associated findings such as invasion of adjacent structures, secondary bile duct dilatation, and liver or nodal metastases may help in differentiating a carcinoma from acute or xanthogranulomatous cholecystitis [2, 4]. In absence of these associated findings, it may not be possible to differentiate a carcinoma from xanthogranulomatous cholecystitis. Adenomyomatosis of the gallbladder is characterized by epithelial proliferation, muscular hypertrophia and intramural diverticula (Rokitansky-Aschoff sinuses), which may segmentally or diffusely involve of the gallbladder. It is a benign condition that requires no specific treatment, occurring as an incidental finding in up to 9% of cholecystectomy specimens [6]. The sonographic finding of cholesterol crystals, shown as 'comet-tail' reverberation artifacts (Fig), within a thickened wall of the gallbladder strongly suggests this diagnosis. Air may produce a similar artifact, however, patients with emphysematous cholecystitis are usually ill in contrast to those with adenomyomatosis. MR imaging may be able to differentiate adenomyomatosis from gallbladder carcinoma by depicting Rokitansky-Aschoff sinuses [7]. Systemic diseases such as hepatic dysfunction, heart failure, or renal failure may lead to diffuse gallbladder thickening [1, 2]. The exact pathophysiologic mechanism leading to oedema of the gallbladder wall in these diverse conditions is uncertain, but it is likely due to elevated portal venous pressure, elevated systemic venous pressure, decreased intravascular osmotic pressure, or a combination of these factors. Hypoproteinemia has also been reported as a cause of extrinsic gallbladder disease, but this has been disputed [8]. Liver cirrhosis, hepatitis and congestive right heart failure are relatively frequent causes. The case on the left is a patient with liver cirrhosis. The secondary gallbladder wall thickening is presumably due to elevated portal venous pressure and decreased intravascular osmotic pressure. On the left a 75-year-old man with drug-induced hepatitis. Longitudinal sonogram of a non-distended gallbladder shows diffuse wall thickening (arrow), and incidental cholelithiasis which may be confusing. In the same patient with the drug-induced hepatitis MR images were obtained to evaluate the bile ducts because of abnormal liver function tests. On the far left Axial SPIR T2-weighted image (A) shows a small amount of ascites (arrowhead) which indicates that the thickened gallbladder wall (arrow) probably has an extrinsic systemic cause. Next to it an oblique HASTE image for MR cholangiography that excludes choledocholithiasis. Mural thickening of the gallbladder (arrowhead) is also shown. On the left a 74-year-old man with congestive right heart failure. Ultrasound depicts diffuse wall thickening of a stone-free painless gallbladder and large-caliber hepatic veins (arrowheads) and inferior vena cava, as supporting evidence of right heart failure. Extracholecystic inflammation may secondarily involve the gallbladder causing wall thickening, due to direct spread of the primary inflammation, or less frequently due to an immunologic reaction [8]. Theoretically, it may be caused by any inflammation that extends to the region of the gallbladder, but only few are regularly encountered including hepatitis, pancreatitis (Figure), and pyelonephritis. Gallbladder wall thickening has also been reported in patients with infectious mononucleosis [9], and in patients with AIDS due to opportunistic infections or secondary neoplastic infiltration [2]. Diffuse gallbladder wall thickening can result from a broad spectrum of pathological conditions, including surgical and non-surgical diseases. At times a definite imaging diagnosis may be impossible. In most cases however, the cause can be determined by correlation of the associated imaging findings with the clinical presentation. Rumack CM, Wilson SR, Charboneau JW. Diagnostic Ultrasound, 2nd ed. St.Louis: Mosby, 1998:175-200 Zissin R, Osadchy A, Shapiro M, Gayer G. CT of a thickened-wall gallbladder. Br J Radiol 2003; 76:137-143 Jung SE, Lee JM, Lee K, et al. Gallbladder wall thickening: MR imaging and pathologic correlation with emphasis on layered pattern. Eur Radiol 2005; 15:694-701 Gore RM, Yaghmai V, Newmark GM, Berlin JW, Miller FH. Imaging of benign and malignant disease of the gallbladder. Radiol Clin N Am 2002; 40:1307-1323 Boland GWL, Slater G, Lu DSK, Eisenberg P, Lee MJ, Mueller PR. Prevalence and significance of gallbladder abnormalities seen on sonography in intensive care unit patients. AJR 2000; 174:973-977 Levy AD, Murakat LA, Abbott RM, Rohrmann CA. Benign tumors and tumorlike lesions of the gallbladder and extrahepatic bile ducts: radiologic-pathologic correlation. RadioGraphics 2002; 22:387-413 Yoshimitsu K, Honda H, Jimi M, et al. MR diagnosis of adenomyomatosis of the gallbladder and differentiation from gallbladder carcinoma: importance of showing Rokitansky-Aschoff sinuses. AJR 1999;172:1535-1540 Kaftori JK, Pery M, Green J, Gaitini D. Thickness of the gallbladder wall in patients with hypoalbuminemia: a sonographic study of patients on peritoneal dialysis. AJR 1987; 148:1117-1118 Yamada K, Yamada H. Gallbladder wall thickening in mononucleosis syndromes. J Clin Ultrasound 2001; 29:322-325Adriaan C. van Breda Vriesman, Robin Smithuis, Dries van Engelen and Julien B.C.M. Puylaert Normal gallbladder Thickened gallbladder wall Differential diagnosis of gallbladder wall thickening Acute cholecystitis Acalculous cholecystitis Chronic cholecystitis Xanthogranulomatous cholecystitis Porcelain gallbladder Gallbladder carcinoma Adenomyomatosis Liver cirrhosis Hepatitis Congestive right heart failure PancreatitisGallbladder - Wall ThickeningRadiology Department of the Rijnland Hospital, Leiderdorp; the Groene Hart Hospital, Gouda and the Medical Centre Haaglanden, the Hague, the Netherlands
abdomen12
1
Kidney - Cystic masses
by David S. Hartman, MD and Ileana Chesaru, MD.
This article is based on a presentation given by David Hartman and adapted for the Radiology Assistant by Ileana Chesaru. In the menubar in the upper left, you will find interactive cases. Renal cysts are commonly encountered lesions in daily radiological practice. Usually these are simple benign cysts, but they can become complicated in case of hemorrhage, infection and ischemia. When this occurs it can be difficult to differentiate these complicated cysts from cystic renal cell carcinomas (10% of all renal cell carcinomas) Since the only treatment for renal cell carcinoma is surgery or ablation, we need to recognize these cystic renal cell carcinomas. Imaging is a reliable means for differentiating benign from malignant cystic lesions. Renal cysts can be classified according to the Bosniak classification depending on their features. Type I cysts are simple cysts. Type II are the minimally complicated cysts. Type I and II can be ignored. Type II F are probably benign, but need to be followed. Type III and IV both are surgical lesions. Type IV is inevitably malignant and in the type III group about 80-90% turn out to be malignant as well. In our communication with the clinicians it is important, that we explain the significance of our findings and the meaning of the classification in terms of: Ignore (type I and II), Follow (type IIF) or Excise (type III and IV). So in this lecture we will only talk about Ignore, Follow or Excise. For those who want to see the original Bosniak classification, look at the table which is presented at the end of the lecture. Although the final differentiation of cystic renal masses is based upon histologic diagnosis, there are imaging findings that tell you that a cyst is not a simple cyst and whether it is probably benign or malignant. The following imaging features indicate that a cyst is NOT simple: - Calcification - Hyperdense / high signal - Septations - Multiple locules - Enhancement - Nodularity / wall thickening The table on the left summarises these imaging features together with the management consequences: Ignore, Follow or Excise. When we look at these imaging features, we have to realise, that the most worrisome portion of a cystic mass should be used in deciding appropriate management. So when the findings are discordant either within one examination or using different radiological examinations, the lesion should be managed based upon the most aggressive imaging findings. When we look at the table on the left, we can say that we are pretty good with the first 3 parameters (calcification, hyperdens and septations), because we are correct in about 95% of the cases. The other four are even more easy, because when you have any of these (enhancement, multiloculated, nodularity or wall thickening), the lesion is almost always a surgical lesion. Regarding follow up, there are no rules at the moment. One could do a follow up at 6 months and if the lesion is stable then double the follow up time. We will now discuss all these imaging features in detail. The most important thing is a good description of the type of calcifications. We can ignore small amounts of calcification that are smooth, septal or if it is milk of calcium, which moves to the lowest point with positional changes. We have to make sure, that no enhancement (= All lesions that show enhancement and lesions with wall thickening or nodularity of the wall outside the calcifications should be excised. We can follow lesions with thick or nodular calcification without any enhancement. On the left we see a cystic lesion. There is a small punctate calcification that we can ignore. On the bottom of the cyst there is a layer of calcium typical for milk of calcium. This is also a benign calcification that we can ignore. On the left a patient with nefrolithiasis. There is also a cystic lesion with linear and nodular calcification . If there were only these linear calcifications we could ignore the lesion. In case of nodular calcification we can follow it, if there is no enhancement. In this case however we see enhancement, so this lesion has to be excised. On CT hyperdense means: > 20 HU on a NECT On MRI hyperintense means all that has higher signal intensity than water on a T1 weighted image. Hyperdensity or hyperintensity usually indicates hemorrhage or high protein content of the cyst. Ignore all lesion with sharp margins; lesions On US they have to be clearly cystic Follow all lesions that are totally intrarenal, because you can not appreciate the wall and follow all lesions > 3 cm, because there is at the moment not much experience with these lesions. All these lesions must show no enhancement. Excise all lesions that are poorly defined or heterogenous or show enhancement. Also when ultrasound shows that the lesion is solid, the lesion should be excised. On the left we see a hyperdens cystic lesion on CT and a hyperintense lesion on a T1-weighted MR. Both lesions have a sharp margin and are homogeneous, although there is some noise in the CT image. On the enhanced scans (not shown) the lesions didn't show any enhancement. We therefore can ignore the lesion on the left and we have to follow the lesion on the right, because it is totally intrarenal. Ignore thin septations ( Follow all septations that are only slightly 'greater than hairline'. They still have to show no enhancement. Excise all septations that are thick, irregular or nodular and all septations that show enhancement. On the left we see 2 cases There is a cystic lesion with a thin smooth non enhancing septation that we can ignore. The other case is a thick enhancing septation that has to be excised. The ultrasound image on the left shows a thick septation. There is also a nodule in the wall of the cyst. So we have two reasons to excise this cystic lesion. Enhancement is our best predictable sign of malignancy. So we have to excise all lesions, that clearly show enhancement. The only exception is infection. Enhancement is defined as follows: Increase in Hounsfield Units of the mass after contrast injection: On MR: >15% = enhancement = surgical The case on the left doesn't shows much on the NECT. However when we give contrast we can appreciate a thick wall and we see enhancement both of the wall and of a central area in the medial part of the cystic lesion. We should never see this in an benign cyst, so this is a surgical lesion. Masses with three or more septa are not called multiseptated but multiloculated. All multiloculated lesions should be excised, unless there is clear evidence of infection. In the adult, the two most common multiloculated masses are MLCN (multilocular cystic nephroma) which is usually benign, but sometimes malignant and MLRCC (multilocular renal cell carcinoma) which is always malignant. On imaging there is no way that we can separate these two and therefore, all multiloculated masses are surgical (unless infection). Even the pathologist can not separate the usually benign multilocular cystic nephroma from the malignant multilocular renal cell carcinoma on gross specimen . The differentiation is based on the histology of the epithelial cells lining these locules. So all multilocular renal masses have to be excised. Ignore: none Follow: only very small nonenhancing nodules, and follow carefully Excise: all other nodular lesions The case on the left shows very small nodules on a CECT and a T2WI. From all the other images we could tell that they were not enhancing. So we can probably follow this lesion. If they start to grow or show any enhancement, then we have to excise the lesion. On the left an easy case. There is a big nodule with enhancement, so this lesion has to be excised. Even if there was no enhancement, the lesion still had to be excised. The case on the left is a T2WI with Fatsat. We see multiple nodules so this mass has to be excised. It doesn't matter whether or not these nodules enhance. All lesions with a thickened wall, with or without enhancement, should be excised, unless there is clear evidence of infection. In these latter cases the lesions should be followed. On the left we see two renal lesions that are cystic. The lesion on the far left clearly shows a thickened wall. This is easy to appreciate because part of the lesion is extrarenal. The cystic lesion next to it is totally intrarenal, which makes it harder to appreciate, but there is wall tickening. So both lesions have to be excised, whether there is enhancement of the wall or not. The only exception would be if there were evident signs of infection. The cystic lesion on the left is clearly a surgical lesion. It has a thick irregular wall, it is exophytic and shows enhancement. In the literature there are different opinions concerning the role of biopsy in renal cystic lesions. Here we provide Dr Bosniak's opinion. A different opinion is given by Dr Lang in the artice on CT-guided biopsy of indeterminate renal cystic masses (Bosniak 3 and 2F): accuracy and impact on clinical management (see references) Biopsy plays a limited role. Only in case of clinical suspicion that the mass is inflammatory (pyuria, etc.) or there is radiological evidence suggesting inflammation (such as perinephric fat stranding), puncture is acceptable. When the outcome is infection, the lesion can be treated and followed. Although considered rare, needle track spread of tumor is an 'underestimated risk'. A core biopsy of the wall of a cystic lesion (benign or malignant) can cause it to rupture and spill its contents into the surrounding tissues. A negative biopsy result does not rule out malignancy, particularly in cystic lesions that have less bulk of tissue to sample. The Bosniak classification was described in 1986. After the original description, it became obvious that some revision was needed because there were some category II cysts that were slightly more complicated than most category II lesions, but not complicated enough to place them in category III. For that reason, a category IIF (F for follow-up) was introduced in 1993. Category I (simple cysts), Category II (mildly complicated benign cysts), and Category IV (cystic neoplasms) are easy to diagnose. Their clinical management is straightforward, with surgery indicated for category IV masses while category I and II masses can be dismissed as benign. Category IIF ('F' for follow-up; moderately complicated cystic masses that require follow-up imaging to demonstrate stability and therefore benignity) and Category III masses (indeterminate masses that require surgery in most cases) can be difficult to differentiate and are subject to interobserver variability. This distinction is essential because their treatment is different. From the RSNA Refresher Courses David S. Hartman, MD, Peter L. Choyke, MD and Matthew S. Hartman, MD From the Department of Radiology, Milton S. Hershey Medical Center, Penn State University School of Medicine Gary M. Israel, MD and Morton A. Bosniak, MD From the Department of Radiology, New York University Medical Center, New York, NY. Morton A. Bosniak. Erich K. Lang, Richard J. Macchia, Brian Gayle et al Eur. Radiol, (2002) 12:2518-2524 by Michael Suh, MD et al. Radiology 2003;228:330-334 by Chrystia M. Slywotzky and Morton A. Bosniak AJR 2001; 176:843-849 by Keith A et al. Radiology 2002;225:451-456David S. Hartman, MD and Ileana Chesaru, MD. Ignore, Follow or Excise Radiological Interpretation Dr Bosniak's opinionKidney - Cystic massesFrom the Department of Radiology, Milton S. Hershey Medical Center, Penn State University School of Medicine, U.S.A and the Westeinde hospital the Hague, the Netherlands.
abdomen13
1
Liver - Incidentalomas
by Maarten van Leeuwen, Joost Nederend and Robin Smithuis
This review is based on a presentation given by Maarten van Leeuwen for the Dutch Radiology Society and was adapted for the Radiology Assistant by Joost Nederend and Robin Smithuis. With the increasing use of multidetector CT small hepatic lesions are frequently depicted. In many cases the pathological nature of these incidentally found liver lesions or incidentalomas is not known. This results in a diagnostic problem, which is initiated by radiology so radiologists should take responsibility in correctly categorizing these lesions as to their clinical significance. In this article we will discuss the management of two different type of incidentally found liver lesions: First study the images on the left. Then continue reading. We see multiple hypodense lesions. We cannot diagnose them with certainty as: For this type of lesions which, due to their small size and atypical imaging features, cannot be confidently categorized, the term TSTC (to small to characterize) lesions has been coined. Jones (1992) studied 1500 patients who had an abdominal CT examination (1). He found: Schwartz (1999) studied 2978 patient with a known malignancy (2). He found TSTCs in 12% of patients with a known malignancy. In 88% of patients the lesions were benign and in 12% they proved to be metastases (1.4% of all patients). The percentage of malignancy depended much on the known primary tumor. Most metastases were found in patients with breast cancer. This is in accordance with the observation that breast metastases usually present as multiple small lesions, while liver metastases of colorectal cancer and lymphoma usually present as a solitary or a few larger masses. Robinson (2003) studied various characteristics of TSTCs and their correlation with malignancy (3). The lesions where classified by their behavior on follow up CT, as either stable or unstable. Image features of stable (benign) lesions where small size and sharp edge. Heterogeneity and soft tissue attenuation were associated with unstable behavior, but only seen in a small minority of cases. Krakora (2004) studied the prognostic importance of small hypoattenuating hepatic lesions seen at initial CT in patients with breast cancer, who did not have definite hepatic metastases at initial examination (4). One or more small hypoattenuating hepatic lesions (TSTCs ) were seen in 54 of 153 patients (35%). During a median follow-up of 584 days definite hepatic metastases developed in 43 of 153 patients (28%). No difference was found in the chance for development of liver metastases in patients with or without TSTCs at initial CT. Krakora concluded that in patients with breast cancer, who do not have definite hepatic metastases at presentation, there is no evidence that small hypoattenuating hepatic lesions seen at initial CT contribute to an increased risk of subsequently developing hepatic metastases. In a patient without a known malignancy these small hypodense lesions, as a rule, should be considered as benign. In a patient with a known malignancy a single TSTC lesion can also be assumed to be benign. Even multiple TSTCs in these patients are mostly benign, especially when they are small, sharply defined and hypodens. In these latter cases you should not be too defensive! Don't dictate 'we can't rule out metastases'. In patients with breastcancer and no known livermetastases at presentation, these TSTC lesions have no positive predictive value for the development of livermetastases in the long term. Incidental hypervascular lesions are also very common findings in liver imaging. It is important to differentiate between 'touch' and 'don't touch' lesions. Benign 'don't touch' hypervascular tumors include hemangioma, FNH and small adenomas. 'Touch' lesions include large adenomas (more then 5 cm) and malignant tumors like Hepatocelular carcinoma (HCC), Fibrolamellar carcinoma (FLHCC) and metastases. These enhancing, solid lesions should be differentiated from vascular lesions , like hepatic aneurysm, aortaportal shunt or pseudoaneurysm. Karhunen (1986) found at autopsy an incidence of 20 % hemangioma, 3% FNH and 1% adenoma (5). A study in 1989 by the AFIP showed a FNH : adenoma ratio of 8:1 in a series of 9000 autopsies (6). Enhancement in Hemangioma A hemangioma is a slowly perfused vascular space. So the timing and amount of enhancement will follow, but lag behind the arterial system. Hemangiomas less than 1 cm frequently demonstrate immediate homogenous enhancement, isodense to the aorta. Hemangiomas larger than 1cm generally show slow centripetal spread of nodular enhancement, slowly decreasing in density. On the left a typical hemangioma. Enhancement in arterial phase is almost isodense to the aorta, and, as contrast diffuses toward the center of the lesion, the level of enhancement lowers slowly, and in the late phase is still hyperdense compared to the vascular spaces. Enhancement in 'capillary blush' The typical, slowly perfused vascular space enhancement of a hemangioma has to be differentiated from the 'capillary blush' due to an abundant capillary network which characterizes FNH, adenoma, HCC and hypervascular metastases. As capillaries are surrounded by tissue the overall enhancement will be less dense compared to the enhancement of the vascular spaces in hemangioma. Hence, in capillary blush, the enhancement occurs slightly later compared to the aorta and is less dense than the aorta. Hemangiomas on dynamic MR will show the same enhancement characteristics as on contrast-enhanced CT. The advantage of MR over CT is its higher sensitivity to contrast as will be shown in the next case. On the left an atypical, apparently hypovascular lesion on CT, possibly metastasis. Same case on dynamic MR. Notice how MR depicts the nodular, peripheral, slowly progressing enhancement (blue curved arrow) which CT failed to depict. On the left an atypical hypoechoic lesion, surrounded by a small but definite halo. In the arterial phase there is homogeneous enhancement of arterial intensity, frequently seen in small hemangiomas. In the portal venous phase and in the equilibrium phase it has the same enhancement as the aorta. So all appearances are consistent with a hemangioma, a benign, non-solid vascular lesion. Once we have excluded hemangiomas, our main goal is to determine whether a hypervascular lesion is a FNH, which is the most prevalent hypervascular solid lesion, or whether it is a lesion which needs further management like adenoma, HCC, FLHCC or hypervascular metastases. For this purpose we have to look for morphologic features like inhomogeneity and presence of capsule, scar, calcification or fat. On the left two adjacent hypervascular lesions with homogeneous enhancement in arterial phase and hypodense central scars in arterial and venous phase, which enhance in the equilibrium phase. This is characteristic of FNH. Notice that the small FNH, which is anterior and right to the bigger one, has the same enhancement pattern. FNH is considered a non-neoplastic, hyperplastic response to a congenital vascular malformation. Histologically, FNH is not a tumor and consists of benign-appearing hepatocytes occurring in a liver that is otherwise normal (i.e. no cirrhosis). At late arterial phase, FNH typically presents with a bright homogeneous enhancement, but less intense than the aorta with a hypodense central scar. Smaller ( The radiating hypodense fibrous bands or septa, arising from the scar, are not infrequent and quite characteristic. At portal phase, FNH is often iso-attenuating to the normal liver and may be difficult to deliniate. Delayed phase often shows hyperattenuation of the central scar and septa due to late opacification of the fibrotic components. No calcifications, inhomogeneity or capsule should be seen in FNH. Focal Nodular Hyperplasia (2) On the left a typical FNH on MR. Slightly hypointense on T1WI and slightly hyperintense on T2WI. The scar is somewhat hyperintense on T2. The enhancement is as we expect with 'capillary blush' with a scar that enhances late in the equilibrium phase. Focal Nodular Hyperplasia (3) On the left a lesion, that has all the characteristics of FNH except for lack of late enhancement of the central scar. In addition, it is slightly hypodense to normal parenchyma in the portal and equilibrium phase. However, all other characteristics are present like lobular enhancement, central scar and no capsule, and therefore we characterize this lesion as FNH. Focal Nodular Hyperplasia (4) The ultrasound image on the left shows two lesions. The small one (blue arrow) is characteristic of a hemangioma, while the larger one (green arrow) is non specific on US. On T2WI the hemangioma shows the typical homogeneous hyperintensity . The larger lesion is somewhat hypointense on T1 and somewhat hyperintense on T2. The enhancement is almost homogeneous with small septae that do not enhance in the arterial phase and do show late enhancement (yellow arrows). On T2WI the hemangioma shows the typical homogeneous hyperintensity . The larger lesion is somewhat hypointense on T1 and somewhat hyperintense on T2. The enhancement is almost homogeneous with small septae that do not enhance in the arterial phase, and do show late enhancement (yellow arrows). We also characterize this lesion as FNH. Focal Nodular Hyperplasia (5) On the left another FNH on MR. The lesion is almost isointense to liver on T1WI and T2WI, but shows more contrast to the liver on a T1W-MPRGRE (gradient-echo). The enhancement in the arterial phase is lobulated with nonenhancing septation and in the equilibrium phase the lesion is not different from normal liver parenchyma. Notice that the lesion has a small scar. Small FNHs often do not have a central scar on imaging and even not on pathologic examination. Focal Nodular Hyperplasia (6) Another FNH on the left, in order to get really familiar with these common lesions. On a CTA for pulmonary emboli a small hypervascular lesion is seen in the liver. Further evaluation was done with MR. On T1WI the lesion is not seen and on T2WI it is only slightly hyperintense. In the arterial phase there is homogeneous enhancement and in the venous phase the lesion is not seen. Provided that this patient does not have liver cirrhosis, this is probably a benign lesion, probably FNH. As the appearance was not pathognomonic for FNH, a follow up examination was done and the lesion had not changed, making the diagnosis FNH most likely. When does it stop, this comfortable feeling, that something is a FNH? It stops when there are too many features that do not belong to a FNH. Like the case on the left. Decide for yourself which findings are compatible with the diagnosis typical FNH and which are not. The lesion on the left does have a central scar like FNH , but on the T1WI the lesion is inhomogeneous and not sharply defined. On T2WI the scar has a low signal intensity. For typical FNH the signal intensity however should be high and the lesion is again inhomogeneous. In the arterial phase the lesion does enhance like FNH, but in the portal and equilibrium phase the enhancement persists and is inhomogeneous. In addition, the central scar does not enhance in the late phase. So there are many findings that are not compatible with the diagnosis FNH. Since the specificity for diagnosing a lesion as benign should be very high, we cannot stop here and we have to get a histological diagnosis. When we encounter lobulated hypervascular masses in the liver, an important diagnosis that you don't want to miss is a fibrolamellar hepatocellular carcinoma (FLHCC). This particular form of HCC may mimick FNH on imaging. In contrast to HCC, the prognosis is reasonable. Like FNH, FLHCC also is a hypervascular, lobulated mass with a central scar . Both FNH and FLHCC appear in normal liver, unlike HCC that is most frequently seen in a cirrhotic liver. In distinction to FNH, FLHCC is inhomogeneous, large (> 5 cm), frequently has calcifications (>70%), a blunt central scar and usually there is lymphadenopathy. Calcifications in FNH are so uncommon that it should make you consider another diagnosis like FLHCC. Fibrolamellar HCC (2) On the left a pathologic specimen of FLHCC and FNH. At first glance they look very similar. However when you look carefully you will notice the more lamellar and heterogeneous structure of FLHCC compared to the homogeneous appearance of FNH. Fibrolamellar HCC (3) On the left CT- and MR-images of a left-lobe fibrolamellar HCC in a 19-year-old man. A. Non-enhanced transverse CT scan shows calcification (curved arrow) within the hypoattenuating tumor (straight arrows). B. Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). C. Ten-minute delayed transverse CT scan demonstrates subtle areas of hyperattenuation that represent fibrous tissue within the central scar, radiating septa, and capsule (open arrows). Curved arrow = calcification. D. Transverse T2-weighted MR image (5,000/105) also demonstrates the central scar and septa (open arrow). The tumor itself (straight arrows) is nearly isointense to liver (the only such case in our series). On the left a photograph of the cut surface of the gross pathologic specimen shows a large tumor with eccentric and central scars (open arrows) and radiating septa. The mass has an irregular lobulated pushing margin (solid arrows) and a variegated appearance with areas of bile staining. In a series of 31 cases of FLHCC, Ichikawa et al (7) found the following: An adenoma is regularly characterized by bleeding, fat or peliosis. Although we cannot see peliosis itself, it can result in a hyperintense lesion on T1WI. On the left two incidentalomas. Decide for yourself why these are not FNH lesions. In the arterial phase there are two hypervascular lesions, somewhat less dense than we would expect in FNH. Both lesions demonstrate a halo of a capsule, which should not be apparent in FNH. Unlike in FNH, the enhancement is inhomogeneous and in the portovenous and equilibrium phase the lesions are not isodens to the liver. Adenoma (2) Regularly adenomas present with bleeding. On the left images of a woman who presented with acute abdominal pain. On US a livermass was seen and free fluid surrounding the liver. This is a typical presentation of an adenoma. On portal phase CT, the lesion is hypointense with haemorrhage adjacent to the lesion, extending subcapsularly. Adenoma (3) On the left an US image of an incidentally found lesion in a 50 y old female. Work up was done with CT, but only non-specific features were found without signs of hypervascularity. . Continue with next images. In contrast to the CT, there clearly is enhancement in the arterial phase on MR, again demonstrating that MR depicts enhancement better than CT. The enhancement is due to a capillary blush, most intense in the arterial phase with apparent wash-out in portal and equilibrium phase, due to greater enhancement of the surrounding parenchyma. In the 'out of phase' image there is signal loss indicating that the lesion contains fat, which is very suggestive for adenoma. A HCC may also contain fat, but in this case there is no cirrhosis and the entire lesion shows signal loss, which we would not expect in HCC. Concerning the diagnosis of HCC, there is one thing to remember: 'Every hypervascular lesion in a cirrhotic liver is HCC until proven otherwise' '. On the left we see a cirrhotic liver with irregular margins (arrows), suggesting that the hypervascular lesion is a HCC. The inhomogeneous enhancement and the partial capsule are helpful for the diagnosis HCC, but even if these features were not present, our diagnosis still would be HCC. Characteristics of hypervascular metastases are: On the left hypervascular metastases. Notice that the larger ones show central necrosis, as they outgrow their blood supply. In the workup of incidentally found hypervascular lesions, we first have to decide whether the lesion is a hemangioma, because these are the most common lesions and usually have specific imaging findings. If not, we have to find out whether it is an FNH. For this differentiation we have to look at differences in enhancement pattern and differences in morphology like presence of a capsule, scar, calcification and inhomogeneity. Hypervascular lesions most often can be characterized, even when small. FNH and hemangiomas need no further investigation or treatment. The preferred modality to characterize incidentalomas is MR, as it is better for lesion characterization and incidentalomas often occur in young females, where radiation burden should be minimized. If HCC or FLHCC is considered further investigation is always needed. In the table on the left we have summarized the typical findings in FNH, Adenoma and HCC. Since FNH is so common, we have to get a clear mental picture of the many ways that these lesions present. As radiologists we have a great responsibility here. In FNH not all features have to be present, but there should be no calcification or high signal intensity on T1WI and the lesion should not be inhomogeneous or have a capsule. Sometimes the term 'stealth lesion' is used to describe the phenomenon that some of these small FNH lesions are only seen in the arterial phase. EC Jones, JL Chezmar, RC Nelson and ME Bernardino Department of Radiology, Emory University School of Medicine, Atlanta, GA 30322. American Journal of Roentgenology, Vol 158, 535-539, Lawrence H. Schwartz, MD, Eric J. Gandras, MD, Sandra M. Colangelo, MD, Matthew C. Ercolani, BS and David M. Panicek, MD Radiology. 1999;210:71-74. P J Robinson, MB, FRCP, FRCR, P Arnold, BSc and D Wilson, MSc Clinical Radiology Research Unit and Medical Physics Department, St James's University Hospital, Beckett Street, Leeds LS9 7TF, UK British Journal of Radiology (2003) 76, 866-874 George A. Krakora, MD et al Radiology 2004; 233:667-673 by Karhunen PJ. J Clin Pathol. 1986 Feb;39(2):183-8. by Robert L. Craig Tomoaki Ichikawa, MD, Michael P. Federle, MD, Luigi Grazioli, MD, Juan Madariaga, MD, Michael Nalesnik, MD and Wallis Marsh, MD Radiology. 1999;213:352-361.Maarten van Leeuwen, Joost Nederend and Robin Smithuis TSTCs in patients without a known malignancy TSTCs in patients with a primary malignancy TSTCs in breastcarcinoma Conclusion Incidence of hypervascular lesions Hemangioma Focal Nodular Hyperplasia (FNH) Fibrolamellar HCC Adenoma Hepatocellular carcinoma (HCC) Hypervascular metastases Work up Differential diagnosisLiver - IncidentalomasRadiology department of the University Medical Centre of Utrecht, the Leiden University and the Rijnland hospital, Leiderdorp, the Netherlands
abdomen14
1
Liver - Masses I - Characterization
by Richard Baron
This article is based on a presentation given by Richard Baron and adapted for the Radiology Assistant by Robin Smithuis. Richard Baron is Chair of Radiology at the University of Chicago and well known for his work on hepatobiliary diseases. He has been president of the Society of Computed Body Tomography and Magnetic Resonance. In Part I a basic concept is given on how to detect and characterize livermasses with CT. In Part II the imaging features of the most common hepatic tumors are presented. Interactive cases are presented in the menubar to test your knowledge (Liver mass 1 and 2). The conspicuity of a liver lesion depends on the attenuation difference between the lesion and the normal liver. On a non enhanced CT-scan (NECT) liver tumors usually are not visible, because the inherent contrast between tumor tissue and the surrounding liver parenchyma is too low. Only a minority of tumors contain calcifications, cystic components, fat or hemorrage and will be detected on a NECT. So i.v. contrast is needed to increase the conspicuity of lesions. When we give i.v. contrast, it is important to understand, that there is a dual blood supply to the liver. Normal parenchyma is supplied for 80% by the portal vein and only for 20% by the hepatic artery, so it will enhance in the portal venous phase. All liver tumors however get 100% of their blood supply from the hepatic artery, so when they enhance it will be in the arterial phase. This difference in bloodsupply results in different enhancement patterns between liver tumors and normal liver parenchyma in the various phases of contrast enhancement (figure). In the arterial phase hypervascular tumors will enhance via the hepatic artery, when normal liver parenchyma does not yet enhances, because contrast is not yet in the portal venous system. These hypervascular tumors will be visible as hyperdense lesions in a relatively hypodense liver. However when the surrounding liver parenchyma starts to enhance in the portal venous phase, these hypervascular lesion may become obscured. In the portal venous phase hypovascular tumors are detected, when the normal liver parenchyma enhances maximally. These hypovascular tumors will be visible as hypodense lesions in a relatively hyperdense liver. In the equilibrium phase at about 10 minutes after contrast injection, tumors become visible, that either loose their contrast slower than normal liver, or wash out their contrast faster than normal liver parenchyma. These lesions will become either relatively hyperdense or hypodense to the normal liver. Optimal timing and speed of contrast injection are very important for good arterial phase imaging. Hypervascular tumors will enhance optimally at 35 sec after contrast injection (late arterial phase). This time is needed for the contrast to get from the peripheral vein to the hepatic artery and to diffuse into the liver tumor. On the left a patient who underwent two phases of arterial imaging at 18 and 35 seconds. In the early arterial phase we nicely see the arteries, but we only see some irregular enhancement within the liver. In the late arterial phase we can clearly identify multiple tumor masses. Notice that in the late arterial phase there has to be some enhancement of the portal vein. The only time that an early arterial phase is needed is when you need an arteriogram, for instance as a roadmap for chemoembolization of a liver tumor. Timing of scanning is important, but almost as important is speed of contrast injection. For arterial phase imaging the best results are with an injection rate of 5ml/sec. There are two reasons for this better enhancement: at 5ml/sec there will be more contrast delivered to the liver when you start scanning and this contrast arrives in a higher concentration. On the left a patient with cirrhosis examined after contrast injection at 2.5ml/sec and at 5ml/sec. At 5ml/sec there is far better contrast enhancement and better tumor detection. Portal venous phase imaging works on the opposite idea. We image the liver when it is loaded with contrast through the portal vein to detect hypovascular tumors (figure). The best moment to start scanning is at about 75 seconds, so this is a late portal venous phase, because enhancement of the portal vein already starts at 35 sec in the late arterial phase. This late portal venous phase is also called the hepatic phase because there already must be enhancement of the hepatic veins. If you do not seen enhancement of the hepatic veins, you are too early. If you only do portal venous imaging, for instance if you are only looking for hypovascular metastases in colorectal cancer, fast contrast injection is not needed, because in this phase the total amount of contrast is more important and 3ml/sec will be sufficient. The equilibrium phase is when contrast is moving away from the liver and the liver starts to decrease in density. This phase begins at about 3-4 minutes after contrast injection and imaging is best done at 10 minutes after contrast injection. This phase can be valuable if you're looking for: fast tumor washout in hypervascular tumors like HCC or retention of contrast in the blood pool as in hemangiomas or the retention of contrast in fibrous tissue in capsules (HCC) or scar tissue (FNH, Cholangioca). Relative hyperdense lesions in the delayed phase Fibrous tissue that's well organized and dense is very slow to let iodine or gadolineum in. Once contrast gets in however, it is equally slow to get back out in the equilibrium phase. So when the normal liver parenchyma washes out, the fibrous components of a tumor will look brighter than the background liver tissue. Cholangiocarcinoma may have a fibrous stroma and in the delayed phase it may be the only time when you see the tumor (figure). Relative hypodense lesions in the delayed phase On the left the importance of the delayed phase in a cirrhotic patient with an HCC is demonstrated. Notice that you do not see the tumor on the nonenhanced scan and also not in the portal venous phase. This is often the case and demonstrates the importance of the arterial phase. Now the issue at hand is in small enhancing lesions in a cirrhotic liver whether it is a benign lesion like a regenerating nodule or a HCC. In the delayed phase we see that the tumor is washed out more than the surrounding liver parenchyma. Benign lesions typically will not show this kind of wash out. For instance a FNH or adenoma will show fast enhancement in the arterial phase, become isodense in the portal venous phase, but it will stay isodense with liver in the equilibrium phase. These benign tumors do not have enough neoplastic neovascularity to have a fast wash out. Especially in cirrhotic patients you have to rely heavily on this delayed phase to differentiate benign little enhancing lesions from small HCC's. Normally when we look at lesions filling with contrast, the density of these lesions is always compared to the density of the liver parenchyma. In hemangiomas however you should not compare the density of the lesion to the liver, but to the blood pool. This means that the areas of enhancement in a hemangioma should match the attenuation of the appropriate vessels (bloodpool) at all times. So in the arterial phase the enhancing parts of the lesion must have almost the same attenuation value as the enhancing aorta , while in the portal venous phase it must match the enhancement of the portal vein. If it does not match the bloodpool in every single phase of contrast enhancement forget the diagnosis of a hemangioma. On the left a characteristic hemangioma. Notice that on the NECT the density of the tumor is the same as the density of the vessels. In the arterial phase it is matching the bloodpool and the attenuation is almost the same as the aorta. In the portal venous phase it matches the density of the portal vein. In the equlibrium phase it has the same enhancement as the vessels. Eventually the lesion will become iso-attenuating to the liver, but only because the vessels become iso-attenuating with the liver. It has nothing to do with the density of the liver parenchyma itself. So think of bloodpool rather than liver if you're thinking of a hemangioma. You have to adapt your protocol to the type of scanner, the speed of contrast injection and to the kind of patient that you are examining. If you have a single slice scanner, it will take about 20 seconds to scan the liver. For late arterial phase imaging 35 sec is the optimal time, so you start at about 25 seconds and end at about 45 seconds. However if you have a 64-slice scanner, you will be able to examine the whole liver in 4 seconds. So you start scanning at about 33 seconds, which is much later. In aterial phase imaging the time window is narrow, since you have only limited time before the surrounding liver will start to enhance and obscure a hypervascular lesion. For portal venous phase imaging it is different. Here you don't want to be too early, because you want to load the liver with contrast and it takes time for contrast to get from the portal vein into the liver parenchyma. Besides you have more time, because the delayed or equilibrium phase starts at about 3-4 minutes. So you start at 75 seconds with whatever scanner you have. Only when you inject with high speed at 5ml/sec you may start earlier at about 65-70 seconds. Use arterial phase imaging in the following situations: From a practical point of view, the approach to characterizing a focal liver lesion seen on CT begins with the determination of its density. If the lesion is of near water density, homogeneous, has sharp margins and shows no enhancement, then it is a cyst. If the lesion does enhance, then the next step is to determine whether the lesion could be a hemangioma, since this is by far the most common liver tumor. The enhancement should be peripheral and nodular, with the same density as the bloodpool in all phases. If it is not a cyst nor a hemangioma, then we further have to study the lesion. Based on the enhancement pattern, we divide masses into hypervascular and hypovascular lesions. Usually a combination of the enhancement pattern and gross pathologic features, like the presence of fat, blood, calcifications, cystic or fibrotic components, in combination with the clinical history will limit the differential diagnosis (figure). Arterially enhancing lesions are mostly benign lesions and include primary liver tumors as FNH, adenoma and small hemangiomas that fill rapidly with contrast. These benign tumors have to be differentiated from the most common hypervascular malignant liver tumor, which is HCC and metastases from hypervascular tumors like melanoma, renal cell carcinoma, breast, sarcoma and neuroendocrine tumors (islet cell tumors, carcinoid, pheochromocytoma). Hypervascular lesions may look very similar in the arterial phase (figure). Differentiation is done by looking at the enhancement pattern in the other phases and additional gross pathologic features together with clinical findings. Hypervascular metastases will be considered in patients with a known primary tumor. In general HCC is considered when there is a setting of cirrhosis, while FNH is considered in young women and hepatic adenoma in patients on oral contraceptives, anabolic steroids or with a history of glycogen storage disease. Hypovascular liver tumors are more common than hypervascular tumors. Most hypovascular lesions are malignant and metastases are by far the most common. Although primary liver tumors are mostly hypervascular, there are exceptions. 10% of HCC is hypovascular. Cholangioca is hypovascular, but may show delayed enhancement (figure). On the left a hypovascular mass with irregular enhancement in the late arterial and late portal venous phase. This is a sign of malignancy. On the delayed images a relative dense structure is seen centrally, which looses its contrast slower compared to normal liver. This means that this tumor is mainly composed of fibrous tissue. The fibrous tissue has also retracted the liver capsule. These imaging findings are very suggestive of a cholangiocarcinoma. Liver lesions which may have a central scar are FNH, fibrolamellar carcinoma, cholangiocarcinoma, hemangioma and hepatocellular carcinoma. On CT a scar is sometimes visible as a hypodense structure. On MR scar tissue is hypointense on both T1WI and T2WI due to intense fibrotic changes. An example is the central scar of fibrolamellar carcinoma (FLC) An exception to this rule is the central scar in FNH which is hyperintense on T2WI due to edema. T2WI can be very helpfull if there is a problem in differentiating FNH from FLC. Both on CT and MRI scar tissue will enhance in the delayed phase. On the left a lesion with a typical central scar. It has a hypodense centre on the NECT. In the portal venous phase there is homogeneus enhancement of the lesion except for the scar. Enhancement of the fibrous tissue of the central scar is seen only on the delayed phase images. The combination of homogeneous enhancement and central scar is typical for the diagnosis of FNH. Liver lesions which may have a capsule are Adenoma, HCC and cystadenoma or cystadenocarcinoma. The most common tumor with a capsule is HCC. The capsule will not enhance in the arterial phase and even in the portal venous phase it will be hypodense, because the fibrous tissue enhances very slowly. A capsule is usually best seen in the delayed phase as a relative hyperdense structure. Adenoma frequently has a thin fibrous capsule seen in 30% of cases. It has a well defined contour and subcapsular feeding arteries. On the left a different patient with HCC. Only in the equilibrium phase a relatively bright capsule was seen. The image on the left was taken 8 minutes after contrast injection. Notice that the tumor itself is relative hypodense in the equilibrium phase. So it has a fast wash out. Central calcifications are seen in: These calcifications are hyperdense on CT and hypointense on T1 and T2 MR images. In FLC these calcifications are located within the central scar as seen on the left. Fat within liver tumors is seen in: The case on the left shows an adenoma with fat depositions within the tumor. Hemorrhage in liver tumors is seen in: Hemorrhage is most commonly seen in adenomas. The case on the left shows a well circumscribed lesion with hemorrhage. At resection the lesion proved to be an adenoma. If a lesion has a near water density in the centre and does not show enhancement in the centre, we usually will call it a cystic lesion. You have to realize, that it still can be a tumor as in cystic metastases or metastases with central necrosis. Secondly you always have to add absces to the differential diagnosis. On the left a patient with hypovascular lesions with a low density, so it may be cystic i.e fluid containing. These lesions are multiple, but not spread out through the liver, so we describe them as clustered or satelite lesions. This is a typical finding which makes the lesions suspective for liver abcesses. This was a case of diverticulitis. The common route is through the portal vein as a result of abdominal infection. The bacteria enter the slow flow portal system, where they layer within the vessel and finally these bacteria 'fall down' into the dependent portion of the right lobe. On the left a typical case of a echinococcus cyst with 'daughter cysts' within the large cyst. Most cases of echinococcus cysts however are not that typical. If you look at the CT image on the left, the first impression might be that there are only simple cysts within the liver. However, if you look more carefully, you will notice that some of the hypodense lesions show vague rim enhancement. And although you might think that these could be cystic metastases, the US-findings clearly show, that these lesions are hyperechoic solid masses. So you have to be very carefull in calling a lesion cystic, because you might end up missing metastases or looking in the wrong file for a differential diagnosis. Most liver tumors will present as a mass. Some tumors however have an infiltrative growth pattern with a lot of fibrous tissue and do not cause mass effect. As the fibrous stroma matures, the tissue will contract and cause retraction of the liver capsule (figure). Breast cancer metastases can be infiltrative. When they shrink they can cause multiple retractions. This will give a pseudo-cirrhosis appearance. The most common tumor however to cause retraction is cholangiocarcinoma. The delayed image on the left shows a large cholangiocarcinoma with dense enhancing fibrous tissue and retraction of the liver capsule. Notice the resemblance with the case above. Another cause of local retraction is atrophy due to biliary obstruction or chronic portal venous obstruction. On the left another case of cholangiocarcinoma with multifocal lesions. Notice the retraction and the delayed enhancement of the fibrotic component of the tumor. Many will regard 'peripheral enhancement and progressive fill in' as a typical feature of hemangioma, but it is not. Peripheral rim enhancement is a typical feature of malignant lesions and only discontinuous nodular peripheral enhancement that matches bloodpool is a typical feature of hemangioma. Many lesions will show progressive fill in. In hemangiomas this progressive fill in must have the same density as the bloodpool. Many hypovascular metastases will show contrast diffusion into a lesion starting on the outside. Usually the center does not fill in. Cholangiocarcinomas will show progressive fill in because the fibrous centre will enhance slowly. You will see it enhance in the delayed phase (see part II) So if you want to make the diagnosis of a hemangioma you have to look at all the other phases to see if the enhancement matches the bloodpool. Oliver JH, Baron RL: State of the art, helical biphasic contrast enhanced CT of the liver: Technique, indications, interpretation, and pitfalls. Radiology 1996; 201:1-14. Brancatelli G., Baron RL, Peterson MS, Marsh W. Helical CT screening for HCC in patients with Cirrhosis: Frequency and causes of False-Positive interpretation. AJR 2003; ISO: 1007-1014.Richard Baron Arterial phase imaging Portal Venous phase Equilibrium Phase Blood pool and Hemangioma Tailored CT protocol Hypervascular lesions Hypovascular lesions Scar Capsule Calcifications Fat Hemorrhage Cystic components Retraction of liver capsule Peripheral enhancement and progressive fill inLiver - Masses I - CharacterizationRadiology department of the University of Chicago
abdomen15
1
Liver - Masses II - Common Tumors
by Richard Baron
This article is based on a presentation given by Richard Baron and adapted for the Radiology Assistant by Robin Smithuis. Richard Baron is Chair of Radiology at the University of Chicago and well known for his work on hepatobiliary diseases. He has been president of the Society of Computed Body Tomography and Magnetic Resonance. In Part I a basic concept is given on how to detect and characterize livermasses with CT. In Part II the imaging features of the most common hepatic tumors are presented. Hemangioma is the most common benign liver tumor. It is composed of multiple vascular channels lined by endothelial cells. In 60% of cases more than one hemangioma is present. The size varies from a few millimeters to more than 10 cm (giant hemangiomas). Calcification is rare and seen in less than 10%, usually in the central scar of giant hemangioma. CT will show hemangiomas as sharply defined masses with the same density as the vessels on NECT and CECT. The enhancement pattern is characterized by sequential contrast opacification beginning at the periphery as one or more nodular areas of enhancement. All these areas of enhancement must have the same density as the bloodpool. This means that in the arterial phase the areas of enhancement must have almost the density of the aorta, while in the portal venous phase the enhancement must be of the same density as the portal vein. Even on delayed images the density of a hemangioma must be of the same density as the vessels. Finally most hemangiomas show complete fill in with contrast. Small hemangiomas may show fast homogeneous enhancement ('flash filling'). Small HCC and hypervascular metastases may mimic small hemangiomas because they all show homogeneous enhancement in the arterial phase. By looking at the other phases to see if the enhancing areas match the bloodpool, it is usually possible to differentiate these lesions. Large hemangiomas can have an atypical appearance. Complete fill in is sometimes prevented by central fibrous scarring. These lesions need to be differentiated from other lesions with a scar like FLC, FNH and Cholangiocarcinoma. Again looking at the bloodpool will help you. On the left two large hemangiomas. Notice that the enhancing parts of the lesion follow the bloodpool in every phase, but centrally there is scar tissue that does not enhance. Peripheral enhancement The enhancement of a hemangioma starts peripheral . It is nodular or globular and discontinuous. Rim enhancement is continuous peripheral enhancement and is never hemangioma. Rim enhancement is a feature of malignant lesions, especially metastases. Progressive fill in First look at the images on the left and describe what you see. Then continue. The lesion definitely has some features of a hemangioma like nodular enhancement in the arterial phase and progressive fill in in the portal venous and equilibrium phase. In the portal venous phase however, the enhancement is not as bright as the enhancement of the portal vein. The conclusion must be, that this lesion does not match bloodpool in all phases, so it cannot be a hemangioma. So progressive fill in is a non-specific feature, that can be seen in many other lesions like metastases or primary liver tumors like cholangiocarcinoma. The delayed enhancement in this lesion is due to fibrotic tissue in a cholangiocarcinoma and is a specific feature of these tumors. Ultrasound Most hemangiomas are detected with US. If you had to pick one word to characterize a hemangioma on US, you would probably say 'hyperechoic'. You have to realize however, that this simply means that the lesion is hyperechoic to normal liver. If the liver is hyperechoic due to steatosis, the hemangioma can appear hypoechoic (figure). Another important feature of hemangiomas is the increased sound transmission. This is because the lesion is made of these channels containing blood. Differential diagnosis Hemangiomas must be differentiated from other lesions that are hypervascular or lesions that show peripheral enhancement and progressive fill in. HCC is the most frequent abdominal malignancy worldwide and is especially common in Asia and mediterrean countries. HCC may be solitary, multifocal or diffusely infiltrating. HCC consists of abnormal hepatocytes arranged in a typical trabecular pattern. Larger HCC lesions typically have a mosaic appearance due to hemorrhage and fibrosis. In patients with cirrhosis or with hepatitis B/C our major concern is HCC, since 85% of HCC occur in these patients. If you take a cohort of patients with hepatitis C and you follow them for 10 years, 50% of them will have end stage liver disease and 25% will have HCC. It is important to separate the early appearance from the late appearance of HCC. Nowadays we encounter very small HCC's in patients, that we screen for HCC (figure). These are small lesions that transiently enhance homogeneously. You will only see them in the arterial phase. Sometimes there is rim enhancement and you might mistake them for a hemangioma. Always look how they present in the other phases and compare with the bloodpool and remember that rim enhancement is never hemangioma. These early HCC's are very different from the large ones that we see in the non-cirrhotic patients. HCC is a silent tumor, so if patients do not have cirrhosis or hepatitis C, you will discover them in a late stage. They tend to be very large with a mozaic pattern, a capsule, hemorrhage, necrosis and fat evolution. HCC becomes isodense or hypodense to liver in the portal venous phase due to fast wash-out. On delayed images the capsule and sometimes septa demonstrate prolonged enhancement. HCC and Portal Vein thrombosis Many patients with cirrhosis have portal venous thrombosis and many patients with HCC have thrombosis. These are two common findings and they can be coincidental. It is very important to make the distinction between just thrombus and tumor thrombus. First, if you have a malignant thrombus in the portal vein, it will always enhance and you'll see it best in arterial phase. Secondly, if you have a malignant thrombus in the portal vein, it will increase the diameter of the vessel. Sometimes a tumor thrombus may present with neovascularity within the thrombus (figure). Early HCC needs to be differentiated from other hypervascular lesions, that will be hyperdense in the arterial phase. First look at the images on the left and look at the enhancement patterns. Then continue. In the arterial phase we see two hypervascular lesions. Now do not just concentrate on the images, where you see the lesions best. You have to look at all the other images, because they give you the clue to the diagnosis. The upper images show a lesion that is isodens to the liver on the NECT. In the arterial phase there is enhancement, but not as dense as the bloodpool. In the portal venous phase the lesion is again isodense to the surrounding liver parenchyma and you can't see it. If you only had the portal venous phase you surely would miss this lesion. The lower images show a lesion that is visible on all images. You see it on the NECT and you could say it is hypodens compared to the liver. Does this help you? No, not in the least. However if you look at the bloodpool, you will notice that on all phases it is as dense as the bloodpool. So we have a HCC in the right lobe on the upper images and a hemangioma in the left lobe on the lower images. The key is to look at all the phases. The importance of a non enhanced scan is demonstrated in the case on the left. In the arterial phase we see a hyperdense structure in the lateral segment of the left lobe of the liver. This looks like an enhancing nodule very suspective of early HCC. However if we look at the NECT on the right, we'll notice, that it is not enhancement that we're looking at. It is just a siderotic iron containing hyperdense nodule. They are very common and are seen in up to 50% of patients with cirrhosis. 10% of HCC are hypodense compared to liver. The imaging findings will be non-specific. The case on the left proved to be HCC. Hepatocellular adenomas are large, well circumscribed encapsulated tumors. They consist of sheets of hepatocytes without bile ducts or portal areas. 80% of adenomas are solitary and 20% are multiple. Adenomas typically measure 8-15 cm and consist of sheets of well-differentiated hepatocytes. Adenomas are prone to central necrosis and hemorrhage because the vascular supply is limited to the surface of the tumor. The pathogenesis is believed to be related to a generalized vascular ectasia that develops due to exposure of the liver to oral contraceptives and related synthetic steroids. In young woman using contraceptives an adenoma is the most frequent hepatic tumor. CT will show most adenomas as a lesion with homogeneous enhancement in the late arterial phase, that will stay isodense to the liver in later phases. Unfortunately, this homogeneous enhancement in the late arterial phase is not specific to adenomas, since small HCC's and hemangiomas as well as hypervascular metastases and FNH can demonstrate similar enhancement in the arterial phase. Malignant lesions however have a tendency to loose their contrast faster than the surrounding liver, so they may become relatively hypodense in later phases. The finding of hemorrhage as an area of high attenuation can be seen in as many as 40% of adenomas. This is however also a feature of HCC and large hemangiomas. Fat deposition within adenomas is identified on CT in only approximately 7% of patients and is better depicted on MRI. Typically adenomas have well-defined borders and do not have lobulated contours. A low-attenuation pseudocapsule can be seen in as many as 30% of patients. This capsule will only show enhancement on delayed scans. Coarse calcifications are seen in only 5% of patients. On the left an adenoma with fat deposition and a capsule. MRI usually is more sensitive in detecting fat and hemorrhage. Chemical-shift imaging showing loss of signal on out-of-phase images can confirm the presence of fat. HCC is known to contain fat in as many as 40% of lesions, therefore the presence of fat does not help differentiate the lesions. Adenomas may rupture and bleed, causing right upper quadrant pain. The two most common liver lesions causing hepatic hemorrhage are HA and HCC. Although adenomas are benign lesions, they can undergo malignant transformation to hepatocellular carcinoma (HCC). Although malignant transformation is rare, for this reason, surgical resection is advocated in most patients with presumed adenomas. Significant overlap is noted between the CT appearances of adenoma, HCC, FNH, and hypervascular metastases, making a definitive diagnosis based on CT imaging criteria alone difficult and often not possible. Clinical correlation in such cases is most helpful. In otherwise healthy young women using oral contraceptives, adenoma is favored. Patients with glycogen storage disease, hemochromatosis, acromegaly, or males on anabolic steroids also are more prone to developing hepatic adenomas. A history of cirrhosis and high AFP levels favor HCC. A history of a primary hypervascular tumor favors metastases. As a result of the risk of intraperitoneal hemorrhage and the rare occurrence of malignant transformation to HCC, surgical resection has been advocated in most patients with presumed HA. The risk of significant bleeding from the tumor is as high as 30%. The exact risk of malignant transformation is unknown. Some advocate surgical resection only when tumors are larger than 5 cm or when AFP levels are elevated, since these two findings are associated with higher risk of malignancy. The value of percutaneous fine needle biopsy for the diagnosis of HA is controversial for two reasons. First, histologic studies may lead to misdiagnosis when differentiating HA from FNH. In addition, a considerable risk of hemorrhage exists when biopsy is performed on these hypervascular tumors. Adenomas may diminish after oral contraceptives are discontinued, but this does not lower the risk of malignant transformation. When a definitive diagnosis of FNH can be made using imaging studies, surgery can be avoided and lesions can be observed safely using radiologic studies. However, if HA or HCC remains in the differential diagnosis, surgery usually is indicated. FNH is the second most common tumor of the liver. FNH is not a true neoplasm. It is believed to represent a hyperplastic response to increased blood flow in an intrahepatic arteriovenous malformation. All the normal constituents of the liver are present but in an abnormally organized pattern. US will show a FNH as a non specific ill-defined lesion. The central scar may be detected as a hyperechoic area, but often cannot be differentiated. With color doppler sometimes the vessels can be seen within the scar. CT will show FNH as a vascular tumor, that will be hyperdens in the arterial phase, except for the central scar. On the left a typical FNH with a central scar that is hypodens in the portal venous phase and hyperdens in the equilibrium phase. MRI will show a hypointense central scar on T1-weighted images. On T2-weighted images the scar appears as hyperintense in 80% of patients, which is very typical. However in 20% of patients the scar is hypointense. Gadolineum enhanced MRI will reveal similar enhancement patterns as on CECT. The diagnosis of FNH is based on the demonstration of a central scar and a homogeneous enhancement. However, a typical central scar may not be visible in as many as 20% of patients (figure). Moreover a central scar may be found in some patients with fibrolamellar hepatocellular carcinoma, hepatic adenoma and intrahepatic cholangiocarcinoma. The key to the diagnosis in the lesion on the left is the fact that it is isoattenuating to normal liver in the portal venous phase and stays that way without a wash out on the delayed phase (not shown). This could also be an adenoma, but HCC would be unlikely because they show a fast wash out. If you look at the images on the left and just would consider the T2W-images, what could be the cause of the central area of high signal? The most common cause would be central necrosis in a tumor. However if you look at the delayed phase, you will notice that this area enhances. So this is fibrotic tissue and the diagnosis is FNH. Fibrolamellar carcinoma (FLC) has a dark scar on T2WI and FNH has a brigth scar on T2WI in 80% of the cases. This means that at times the differential between FNH and FLC will not be possible. FLC is an uncommon malignant hepatocellular tumor, but less aggressive than HCC. FLC characteristically manifests as a 10-20 cm large hepatic mass in adolescents or young adults. The typical risk factors for HCC such as cirrhosis, elevated alphafetoprotein, viral hepatitis, alcohol abuse are absent. FLC characteristically appears as a lobulated heterogeneous mass with a central scar in an otherwise normal liver. Calcifications occur in 30-60% of fibrolamellar tumors. Imaging features of FLC overlap with those of other scar-producing lesions including FNH, HCC, Hemangioma and Cholangiocarcinoma. FNH, in particular, may simulate FLC, since both have similar demographic and clinical characteristics. In contrast to FNH the central scar in FLC will usually be hypointense on T2WI and will less often show delayed enhancement. While FNH is always very homogeneous, FLC is usually heterogeneous following contrast administration. On the left pathologic specimens of FLC and FNH. At first glance they look very similar. However when you look carefully you will notice the lamellar and heterogenous structure of FLC compared to the homogeneous appearance of FNH. On non enhanced images a FLC usually presents as a big mass with central calcifications. Cholangiocarcinoma usually presents as a mass of 5-20cm. In 65% there are satellite nodules and in some cases punctate calcifications are seen. The diagnosis of a cholangiocarcinoma is often difficult to make for a radiologist and even a pathologist. That is because cholangiocarcinoma has a varied morphology and histology. It can be a constricting or an expanding lesion, because it can have a fibrous or a glandular stroma. It can be located anywhere in the intrahepatic bile ducts or common bile duct. First look at the images on the left and try to find good descriptive terms for what you see. Then continue. The lesion on the left has the folowing characteristics: The finding of an infiltrating mass with capsular retraction and delayed persistent enhancement is very typical for a cholangiocarcinoma. Infiltrative cholangiocarcinoma does not cause mass effect, because when the stroma matures, the fibrous tissue will contract and cause retraction of the liver capsule. There are not many tumors that cause retraction of the liver capsule, since most tumors will bulge. The most common tumor that causes retraction besides cholangiocarcinoma is metastatic breast cancer. This will give a pseudo-cirrhosis appearance. Another cause of local retraction is atrophy due to biliary obstruction or chronic portal venous obstruction. The case on the left demonstrates how difficult the detection of ta cholangiocarcinoma can be. Only on the delayed images at 8-10 minutes after contrast injection a relative hyperdense lesion is seen. This is the fibrous component of the tumor. Some cholangiocarcinomas have a glandular stroma. The liver is the most common site of metastases. The most common organs of origin are: colon, stomach, pancreas, breast and lung. Most liver metastases are multiple, involving both lobes in 77% of patients and only in 10% of cases there is a solitary metastasis. Hypovascular metastases are the most common and occur in GI tract, lung, breast and head/neck tumors. They are detected as hypodense lesions in the late portal venous phase. In this phase the attenuation of the normal liver parenchyma increases, revealing the relatively hypoattenuating metastases, sometimes with peripheral enhancement. The rim enhancement that occurs represents viable tumor peripherally, which appears against a less viable or necrotic center (figure). Hypervascular metastases are less common and are seen in renal cell carcinoma, insulinomas, carcinoid, sarcomas, melanoma and breast cancer. They are best seen in the late arterial phase at 35 sec after contrast injection. Although breast cancer metastases can be hypervascular, it was shown that routine use of adding arterial phase imaging, did not show any advantage. Calcified liver metastases are uncommon. Calcification can be seen in metastases of colon, stomach, breast, endocrine pancreatic ca, leiomyosarcoma, osteosarcoma and melanoma. When calcified liver metastases are revealed by CT in a patient with unknown primary tumor, colon cancer will be the most likely cause. Cystic liver metastases are seen in mucinous ovarian ca, colon ca, sarcoma, melanoma, lung ca and carcinoid tumor. On MRI metastases are usually hypointense on T1WI and hyperintense on T2WI. Peritumoral edema makes lesions appear larger on T2WI and is very suggestive of a malignant mass. On dynamic contrast-enhanced MRi the characteristics of metastases are the same as for CECT. Ultrasound findings At US, metastases may appear cystic,hypoechoic, isoechoic or hyperechoic. Bull's eye or target lesions is a common presentation of metastases. In these metastases the halo is most probably related to a combination of compressed normal hepatic parenchyma around the mass and a zone of cancer cell proliferation. This pattern suggests aggressive behavior and is seen in bronchogenic, breast and colon carcinoma, . However, this pattern is not specific for metastases as it can also be seen in primary malignant liver neoplasms (eg, HCC) and benign liver neoplasms (eg, adenoma in glycogen storage disease). A similar appearance has been described with liver abscesses. Calcified metastases may shadow when they are densely echogenic (figure). This pattern is commonly seen in colorectal cancer. Metastases can look like almost any lesion that occurs in the liver. Hypervascular metastases have to be differentiated from other hypervascular tumors that can be multifocal like hemangiomas, FNH, adenoma and HCC. Hypovascular metastases have to be differentiated from focal fatty infiltration, abscesses, atypical hypovascular HCC and cholangiocarcinoma. Metastases in fatty liver Focal fatty sparing in a diffusely fatty liver or foci of focal fatty infiltration can simulate metastases. However on nonenhanced scans these regions of fat variation tend to be nonspherical and geographic, with no mass effect or distortion of the local vessels. On the other hand a fatty liver can also obscure metastases. On a contrast enhanced CT hypovascular lesions can be obscured if the liver itself is lower in density due to fat deposition. On a NECT these lesions usually are better depicted (figure). If a patient is known to have a fatty liver, it is better to do an MRI or ultrasound for the detection of livermetastases. On the left a patient with fatty infiltration of large parts of the liver. No metastases were seen, but on an ultrasound of the same region multiple metastases were detected. The presentation of liver abcesses is very much dependend on the way the bacteria have entered the liver. There are four routes for bacteria to get into the liver. The common route is through the portal vein as a result of abdominal infection. The bacteria enter through the slow flow portal system and they are layered within the vessel. The bacteria will fall down into the dependent portion of the right lobe. In sepsis the spread will be via the arterial system as in patients with endocarditis and there will be multiple abscesses spread out through the periphery of the liver. The biliary route is often the result of biliary manipulation as in ERCP. It is usually central in location and then spreads out. Finally there is a direct route as in penetrating injury or direct spread of cholecystitis into the liver. First look at the images on the left and try to find good descriptive terms for what you see. Then continue. If you would describe the image on the left, you would use terms as: So these findings suggest liverabscesses especially because it's clustered. Only when you have a population with livertransplants, bilomas in an infarcted area would look the same. If it wasn't clustered than any cystic tumor could look like this. It is very important to make the diagnosis of liver absces because it is a benign disease that kills and the radiologist may be the first to raise the suspicion. Whenever you see a small cyst-like lesion in a patient who recently underwent an ERCP, be very carefull to assume it is just a simple cyst. Biliary abscesses start small but can progress rapidly. The figure on the left shows such a case. Within 3 weeks the small lesion in the left liver lobe progressed to this huge abces. So any cystic structure near the biliary tract in a patient, who recently has undergone a biliary procedure, is suspicious of a liver abces. Oliver JH, Baron RL: State of the art, helical biphasic contrast enhanced CT of the liver: Technique, indications, interpretation, and pitfalls. Radiology 1996; 201:1-14. Brancatelli G., Baron RL, Peterson MS, Marsh W. Helical CT screening for HCC in patients with Cirrhosis: Frequency and causes of False-Positive interpretation. AJR 2003; ISO: 1007-1014.Richard Baron Early appearance of HCC Late appearance of HCC Differential diagnosis Differential diagnosisLiver - Masses II - Common TumorsRadiology department of the University of Chicago
abdomen16
1
MRI detection of Endometriosis
by Jan Hein van Waesberghe, Marieke Hazewinkel and Milou Busard
Laparoscopy is the gold standard for the diagnosis of pelvic endometriosis. MRI is helpful in determining the extent of deep infiltrating endometriosis, especially when laparoscopic inspection is limited by adhesions. In this article we will focus on the diagnosis and preoperative assessment of endometriosis using MR imaging. You can enlarge images by clicking on them. This item is not available on the iPhone application.  Endometriosis is defined as the presence of endometrial tissue outside the uterine cavity. It is mainly found in the abdominal cavity, most commonly on the surface of the ovaries. It is an estrogen-dependent disease and is estimated to occur in 10% of the female population, almost exclusively in women of reproductive age. The most common symptoms are dysmenorrhea, dyspareunia, pelvic pain, and infertility - although it may also be asymptomatic. The symptoms depend on the localization of the endometriosis, the depth of the infiltration and whether the endometriosis is complicated by adhesions. The illustration shows the typical localizations of endometriosis: If the only reason for performing MRI is to determine the presence or extent of endometriosis, the sequences listed in the table on the left are sufficient. Lesions usually demonstrate low to intermediate signal intensity on T2- and T1-weighted images. In some cases punctate foci of high signal intensity are seen on T2-weighted imaging, indicating dilated endometrial glands. Foci of high signal intensity may be seen on T1-weighted images with fat saturation, indicating the presence of hemorrhage. If these foci have a high signal intensity on T1-weighted images with fat saturation, it indicates the presence of hemorrhage. T1-weighted images with fat saturation are necessary to differentiate blood in endometriomas from fat in mature cystic teratomas, since both show high signal intensity on T1-weighted images without fatsat. If the questions that need answering are more diverse, for example in cases of suspected malignancy, T1- and T1-fatsat sequences before and after the administration of intravenous gadolinium may supplement this protocol. Diffusion-weighted imaging may also be added. In superficial endometriosis – also known as Sampson's syndrome - superficial plaques are scattered across the peritoneum, ovaries and uterine ligaments. These patients tend to have minor symptoms and usually also less structural changes in the pelvis. At laparoscopy, these implants may be seen as superficial powder-burn or gunshot lesions. On MRI these lesions are most often not visible because they are tiny and flat, and therefore undetectable. Only when they exceed 5mm or when they appear as hemorrhagic cysts, showing high signal intensity on T1 and low signal intensity on T2-weigthed images, they may be detected (figure). Neither transvaginal ultrasound nor MRI are sufficiently sensitive to screen for these endometriotic plaques. In deep pelvic endometriosis - also called Cullen's syndrome - there is subperitoneal infiltration of endometrial deposits. The symptoms are more severe and related to the localization and depth of invasion. MRI is of use for the diagnosis of deep infiltrating endometriotic lesions and for the assessment of disease extension. Preoperative mapping of disease extension is important to decide whether surgical intervention is indicated, and if so, for planning complete surgical excision. The cul-de-sac is the most common site of pelvic involvement. Presence of deep infiltrating endometriosis in the cul-de-sac can be easily overlooked at laparoscopy due to the creation of a false peritoneal floor by endometriosis in the pouch of Douglas, partly caused by anterior rectal wall adhesions. This phenomenon gives an erroneous impression of extraperitoneal orgin. Consequently, the location of the deep infiltrating endometriosis in the rectovaginal septum may also be a misnomer as the rectovaginal septum is located caudal to the posterior vaginal fornix and, on the basis of normal anatomy, may therefore not be a primary site for endometriosis to develop. This differentiation between normal anatomy and the presence of endometriosis in the cul-de-sac is readily made using MRI. This sagittal T2-image shows deep infiltrating endometriosis in the posterior cul-de-sac with infiltration of the rectal wall. The torus uterinus - where the sacrouterine ligaments attach - and posterior fornix are common localizations of endometriosis. Clinically these patients often present with dyspareunia. T2-images of endometriosis involving the torus uterinus. T2-images showing deep infiltrating endometriosis in the posterior fornix and torus uterinus. There is no infiltration of the bowel wall. T2-weighted images demonstrating involvement of the left sacrouterine ligament. Bowel endometriosis affects between 4% and 37% of women with endometriosis. Transvaginal ultrasonography is the first line of investigation in patients with suspected bowel endometriosis. Additionally, MRI can determine the depth of bowel wall infiltration, the length of the affected area and the distance of the lesion from the anus. The T2-images demonstrate two fan-shaped hypointense lesions (red arrows). These findings are typical for endometriotic lesions infiltrating the muscular layer of the bowel wall. There is also some submucosal swelling, seen as hyperintensity on the luminal side of the bowel wall. In case of circular involvement, extensive deep infiltrating endometriosis of the bowel wall can lead to stenosis of the bowel lumen. Patients may clinically present with pencil-like stool or constipation. The T2-images show focal stenosis of the rectum as a result of circular endometriotic involvement. The urinary tract is involved in only 4% of women with endometriosis of which around 90% involve the bladder. The T2-images show endometriosis infiltrating the bladder wall. The sagittal T2-image shows full-thickness bladder endometriosis with isointense signal compared to muscle and foci of high signal intensity, indicating dilated endometrial glands. The fatsat T1-image shows small cysts with hyperintense signal within the lesion caused by hemorrhage. Endometriosis is frequently complicated by adhesion formation. On MRI adhesions can be seen as spiculated, low- to intermediate signal intensity strandings on T1 and T2. Adhesions can fixate the pelvic organs, leading to posterior displacement of uterus and ovaries, elevation of the posterior vaginal fornix and angulation of bowel loops. They may also lead to hydronephrosis, although in most cases hydronephrosis is caused by fibrosis secondary to the endometriosis. The T2- and fatsat T1-images on the left show a patient with endometriosis in whom the ovaries are stuck together ('kissing ovaries'), as a result of extensive adhesion formation. In this patient a small hemorrhagic cyst of the left ovary and a hemorrhagic superficial plaque are also shown (high signal on T1 red arrows). These T2-images show dilatation of the left distal ureter caused by extensive deep infiltrating endometriosis involving the left sacrouterine ligament extending to the sigmoid colon. Endometriomas - also known as chocolate cysts - develop when superficial endometriotic lesions on the surface of the ovary invaginate. Blood produced by such an implant during each menstrual cycle cannot escape and will accumulate within the ovary, forming a cyst known as an endometrioma. Endometriomas present as complex cystic masses, often thick-walled with a homogeneous content. On transvaginal ultrasound, endometriomas may be seen as thick-walled cysts with low level echoes. On the left a transvaginal ultrasound image and the corresponding laparoscopic image during cystectomy. On MRI, endometriomas present as solitary or multiple masses with a homogeneous hyperintense signal intensity on T1- and T1-fatsat sequences. The T1-fatsat helps differentiate endometriomas from mature cystic teratomas, which usually contain fat. On T2WI, endometriomas may range from having a low signal intensity (also known as shading) to an intermediate or high signal intensity. The low signal intensity reflects the hemoconcentration of a cyst. Endometriomas generally have a thick, fibrous capsule with low signal intensity on T2, caused by hemosiderin-laden macrophages (figure). On the left T2WI and fatsat T1WI of a patient with an endometrioma of the right ovary that demonstrates high signal intensity on T1-fatsat and intermediate signal on T2. Moreover, a hydrosalpinx with low signal on T1-fatsat (blue arrow) and high on T2WI (red arrow) and a leiomyoma with intermediate signal on T1-fatsat and low signal intensity on T2WI. On the left another example of an endometrial cyst. The T2- and fatsat T1-images show a cyst with a bloodclot (hypointense on T2, intermediate on T1). Sometimes these clots are accompanied by fibrotic tissue at histopathology. They may be recognized as irregularly shaped, hypointense lesions (on T2) found in the dependent portion of the endometrial cysts. In this case there is also a hematosalpinx (curved arrow). The T2- and fatsat T1-images on the left show an endometrial cyst of the left ovary. The wall of the cyst is hypointense on T2WI and T1WI caused by hemosiderin. The differential diagnosis of endometrial cysts includes: hemorrhagic functional cysts, fibrothecoma, cystic mature teratoma, cystic ovarian neoplasm and ovarian abscess. For more information about the differential diagnosis see the artices 'Diagnostic Work up for Ovarian Cysts' and 'Ovarian Cysts: Common Lesions'. Endometrial implants have been reported in many unusual sites outside the pelvis including the chest. Abdominal wall endometriosis is the most common location of extrapelvic endometriosis and usually occurs after cesarean section. Sonography shows a solid hypoechoic lesions in the abdominal wall , frequently containing internal vascularity on power Doppler examination. These sonographic findings are nonspecific, and a wide spectrum of disorders should be considered in the differential diagnosis including neoplasms such as a sarcoma, desmoid tumor, or metastasis and nonneoplastic causes such as a suture granuloma, hernia, hematoma, or abscess. However, abdominal wall endometriosis should always be your prime concern in patients with an abdominal wall mass nearby a cesarean section scar. The CT and MR characteristics of abdominal wall endometriosis are nonspecific, both showing a solid enhancing mass in the abdominal wall. On the left MR-images of a patient with abdominal wall endometriosis. On T2WI, the lesions have an isointense signal to muscle with small foci of high signal intensity, indicating dilated endometrial glands. They have a slightly higher signal intensity to muscle on the fatsat T1-image (arrow). A characteristic clinical symptom of abdominal wall endometriosis is cyclic pain associated with the menses, but patients may also present with continuous pain or no pain at all. The axial T2-weighted image on the left demonstrates another case of abdominal wall endometriosis. by Ray Garry Hum. Reprod. (2004) 19 (4): 760-768. by Paula J. Woodward, Roya Sohaey and Thomas P. Mezzetti Jr. Januari 2001 RadioGraphics 21, 193-216. by Jan-Hein Hensen and Julien Puylaert AJR 2009; 192:1618-1624 by Jan-Hein J. Hensen, Adriaan C. Van Breda Vriesman and Julien B. C. M. Puylaert. AJR March 2006 vol. 186 no. 3 616-620 by Milliam L. Kataoka et al. March 2005 Radiology, 234, 815-823. by Busard MP, van der Houwen LE, Bleeker MC, Pieters-van den Bos IC, Cuesta MA, van Kuijk C, Mijatovic V, Hompes PG, van Waesberghe JH. Abdominal Imaging 2011;Doi: 10.1007/s00261-011-9790-1 by Busard MP, Mijatovic V, Luchinger AB, Bleeker MC, Pieters-van den Bos IC, Schats R, van Kuijk C, Hompes PG, van Waesberghe JH. Eur J Radiol. 2011 Sep 9Jan Hein van Waesberghe, Mariek Hazewinkel and Milou Busard MRI-protocol Cul-de-sac localization Uterus Bowel involvement Bladder involvement Adhesions Differential diagnosisMRI detection of EndometriosisRadiology department of the VU University Medical Center Amsterdam, the Netherlands
abdomen17
1
Ovarian Cysts - Common lesions
by Wouter Veldhuis, Robin Smithuis, Oguz Akin and Hedvig Hricak
In this review the imaging features of normal ovaries and the most common ovarian cystic masses are presented. In Ovarian Cystic Masses Part I a roadmap for the diagnostic workup and management of ovarian cystic masses is presented based on the findings of ultrasound and MRI. Images can be enlarged by clicking on them. On the iPhone application this results in hi-res images at full retina resolution. The normal ovary contains over two million primary oocytes at birth, about 10 of which mature each menstrual cycle. Of the 10 Graafian follicles that begin to mature, only one becomes the dominant follicle and grows to a size of 18-20 mm by mid-cycle, when it ruptures to release the oocyte. The other nine follicles become atretic and fibrous. After release of the oocyte, the dominant follicle collapses, and the granulosa cells in the inner lining proliferate and swell to form the corpus luteum of menstruation. Over the course of 14 days the corpus luteum degenerates, leaving the small scarred corpus albicans. Graafian follicles The normal ovary in pre-menopausal women contains small cysts. The images show two normal ovaries with several anechoic, simple cysts consistent with Graafian follicles. On T2-weighted MR-images the Graafian follicles are seen as bright cysts surrounded by darker solid ovarian stroma. FDG-PET pitfall - normal premenopausal ovaries In some pre-menopausal women the normal ovaries may be avidly PET positive, depending on the date in the menstrual cycle. Because in pre-menopausal women a PET-positive ovary may be either an adnexal neoplasm or completely normal, it is important to be aware of the possibility of physiologic mid-cycle FDG uptake and to correlate this finding with the clinical history. FDG-PET in pre-menopausal women should therefore preferably be scheduled in the first week of the menstrual cycle. In post-menopausal women, the normal ovaries show only minimal uptake of FDG. Any increased ovarian FDG uptake in this age group is suspicious for a possible neoplasm. Post-menopause is defined as 1 year or more of amenorrhea. In Western countries the average age of menopause is 51-53 years. In post-menopausal women the ovaries are generally smaller and gradually stop forming Graafian follicles. Note, however, that follicular cysts may persist several years after menopause. In the coronal T2-weighted image of a postmenopausal woman the ovary is no more than a dark tissue clump near the proximal end of the round ligament. The axial T2-weighted image also shows a dark left ovary, devoid of follicles. Although a bit prominent, this is likely to be completely normal. Only if, by chance, there happened to be prior imaging showing that the lesion was growing, your differential diagnosis would start with a benign solid lesion such as ovarian fibroma or fibrothecoma. By far the most common cystic ovarian lesions are benign functional ovarian cysts. Functional cysts are Graafian follicles or corpora lutea that have grown too large or have bled, but are otherwise benign. In the early post-menopause phase, 1-5 years after the final menstrual period, sporadic ovulatory cycles still may occur and ovarian cysts may be seen. Even in late menopause, which is defined as more than 5 years since the final menstrual period, when ovulation is unlikely to occur, small simple cysts may be seen in up to 20% of women. A dominant Graafian follicle sometimes fails to ovulate and does not involute. When it becomes larger than 3 cm, it is called a follicular cyst. Follicular cysts are usually 3-8 cm, but may become much larger. On ultrasound follicular cysts present as simple unilocular, anechoic cysts with a thin, smooth wall. There should be no enhancing nodules or other solid components, no enhancing septations, and no more than physiologic ascites. Follicular cysts will usually resolve spontaneously on follow-up. A corpus luteum may seal and fill with fluid or blood, forming a corpus luteum cyst. The transvaginal ultrasound images show a small complex ovarian cyst with wall vascularity on power Doppler analysis. The characteristic circular Doppler appearance is called the 'ring of fire'. Note, there is good through-transmission and no internal vascularity, consistent with a, partially involuted, corpus luteum cyst. Remember that women who are on birth control pills usually won't form a corpus luteum, as birth control pills prevent ovulation. On the other hand, use of fertility drugs that induce ovulation, increases the chance of developing corpus luteum cysts. Another case with the typical the 'ring of fire' on ultrasound. At pathologic examination the collapsed bloody cyst can be clearly seen. Corpus luteum cyst at MRI: an axial T2-weighted image demonstrating an involuting corpus luteum cyst (arrow). This is a normal finding. The right ovary is also normal. When a Graafian follicle or follicular cyst bleeds, a complex hemorrhagic ovarian cyst (HOC) is formed. On ultrasound hemorrhagic ovarian cyst presents as an unilocular thin-walled cyst with fibrin-strands or low-level echoes and good through transmission. On MRI hemorrhagic cysts are bright on pre-contrast T1-FatSat, and dark on T2. There should be no internal vascularity on Doppler ultrasound or internal enhancement on CT or MRI. Hemorrhagic ovarian cysts have variable wall thickness, and often some circumferential vascularity can be seen. Clinically the classic presentation is with acute pain. However HOC can also be an incidental finding in an asymptomatic patient. The ultrasound images show multiple simple and one complex right ovarian lesion (red arrow). The latter demonstrates diffuse low-level echos and no flow on Doppler. Note that there is a good through transmission (blue arrow). These findings indicate the presence of a hemorrhagic cyst. Continue with the MR-images. Axial and sagittal T2W images from the same patient. The right ovary contains multiple simple T2 bright cysts with thin borders and no solid components. On the axial image there is one lesion, that is dark on T2, i.e. a complex cyst (arrow). There is a small amount of ascites around the right ovary, but not enough to raise concern of a possible neoplasm. On the T1-weighted image without fatsat the complex cyst is bright, indicating either fat or blood content. On the T1-weighted image with fatsat the lesion remains bright, ruling out a fatty lesion. After the administration of Gd there is no enhancement, confirming that this is a hemorrhagic ovarian cyst. An endometrioma would be in your differential. Note that subtraction images are best to demonstrate the lack of enhancement in a lesion, that is bright on the pre-contrast T1-weighted image. The ultrasound images show the right and left ovary: on both sides there is what appears to be a solid lesion. There is however good through transmission, which indicates that we are probably dealing with hemorrhagic cysts. On Doppler US (not shown) there was no vascularity. Continue with the MR examination. On an axial T1-weighted image both lesions are bright indicating fat, blood or high protein fluid. Fat saturation does not suppress the signal in these lesions. In an image with overall reasonably good fat suppression this rules out a fat-containing teratoma and confirms the suggestion of hemorrhagic fluid. On the axial T2-weighted image both lesions show typical 'shading'. The gradual drop in T2 is thought to be caused by a combination of increasing viscosity and increasing concentration of protein and iron towards the dependent portion of the lesion. There is no enhancement on the subtraction image (Post-Gd minus Pre-Gd). Again, subtraction is useful in cases like this: Gd-induced signal increase over the already very bright pre-contrast image would be very hard to appreciate otherwise. Cystic endometriosis or endometrioma is a type of cyst formed when endometrial tissue grows in the ovaries. It affects women during the reproductive years and may cause chronic pelvic pain associated with menstruation. The ovaries are involved in approximately 75% of patients with endometriosis. On ultrasound endometrioma can be variable but the great majority (about 95%) of patients present with a classic homogeneous, hypoechoic cyst with diffuse low level echoes. Rarely it is anechoic, mimicking a functional ovarian cyst. Endometriomas can be multilocular and have thin or even thick septations. In about one third of patients, on careful examination, small echogenic foci can be seen adhering to the wall. These have been postulated to be cholesterol deposits, but may also constitute small blood clots or debris. It is important to differentiate these foci from true wall nodules. In the presence of these foci, the diagnosis of an endometrioma is very likely. The transvaginal ultrasound shows a typical endometrioma, with hyperechoic wall foci. At Doppler US no vascularity was seen in these foci (not shown). The next case is a transvaginal US-image that shows a cystic lesion with a hyperechoic structure. There is a wide differential diagnosis including ovarian cystic neoplasm with solid component, mature cystic teratoma with hyperechoic Rokitansky nodule, hemorrhagic cyst with clot and endometrioma with clot or debris. Continue with the CT and MR. A CT was requested that turned out to show the same, predominantly cystic lesion. If additional imaging is needed for cysts that are indeterminate at ultrasound, it is better to perform MRI. The T2-weighted image on the right correlates nicely with the ultrasound image. On T2-weighted images endometriomas typically show 'shading'. MRI confirms the absence of any enhancement, confirming that it is most likely debris within the cyst. On MRI the hemorrhagic content will make endometrioma appear bright on T1-weighted images. On T1-fatsat images an endometrioma will remain bright. This in contrast to teratomas, that are also bright on T1 but dark on T1-fatsat images. Always include a T1 fat suppressed sequence, because this makes small T1 bright lesions more conspicuous. The next case is a unilocular, mildly hypoechoic ovarian lesion with through transmission. There is no internal or wall vascularity on Doppler. On ultrasound this can again either be a hemorrhagic cyst or an endometrioma. Continue with the MR images 6 months later a follow-up MRI was performed. The lesions are bright on T1-weighted images. The bright signal persists on fat saturation indicating the presence of blood. There is T2 shading consistent with a hemorrhagic lesion. There is no enhancement. The fluid-fluid level in the right ovarian lesion also confirms its cystic nature. The fact that the lesions persist after 6 months makes bilateral endometrioma much more likely than hemorrhagic cysts. The Poly-Cystic Ovary Syndrome (PCOS) is also known as Stein-Leventhal syndrome. Imaging can confirm or suggest the diagnosis. Imaging criteria: These patients usually have menstrual cycle irregularities and either typical clinical signs of hirsutism, obesity, infertility, acne, male balding pattern or biochemically show increased androgen levels. On the left a sagittal T2-weighted image in a patient with increased serum androgen levels. The ovary is enlarged and shows multiple small peripherally located simple cysts The obesity associated with this syndrome is evident from the abundance of fat, showing bright on these FSE T2-weighted images. In this patient, MRI confirmed the diagnosis of PCOS. Ovarian hyperstimulation syndrome is a relatively rare condition. It is caused by hormonal overstimulation by hCG, and is therefore usually bilateral. Hormonal overstimulation can occur in gestational throphoblastic disease, PCOS or in patients receiving hormonal therapy. It can also be seen in pregnancy, but seldom in a normal single pregnancy. If it does occur in normal pregnancies, the reported natural course is spontaneous resolution after birth. In normal pregnancies the reported natural course is spontaneous resolution after birth. Hormonal overstimulation more often occurs in molar pregnancy, erythroblastosis fetalis or in plural pregnancies. On imaging there is - usually bilateral - ovarian enlargement with multiloculated cyst that can totally replace the ovary. The clinical history is the distinguishing feature to make the diagnosis of ovarian hyperstimulation syndrome. The US images are of a young pregnant woman, who had multiple ovarian cysts. The other ovary is not shown but showed a similar appearance. The features needed to make the diagnosis of ovarian hyperstimulation syndrome are in the clinical history - a young pregnant woman - and in the last image of the uterus, which shows an invasive uterine mass, consistent with invasive molar pregnancy. Tubo-ovarian abscess (TOA) usually arises as a complication of Chlamydia or Gonorrhoeae infection that rises from the vagina or cervix to the fallopian tubes. On imaging a thick-walled complex cystic ovarian lesion is seen with abundant flow. The presence of a thickened endometrium or hydrosalphinx makes the diagnosis of a PID more likely. The axial CECT image shows a left complex cystic lesion with thick enhancing walls and internal gas. It looks like an abscess. Note the relatively unremarkable aspect of the overlying mesentery: this is unlikely to be a peri-diverticular abscess. Continue with the reconstructed images. On the sagittal image notice, that the lesion is connected to the ovarian vein confirming that this is an ovarian lesion (arrow). The coronal image shows the anatomic connection to the uterus. There is a gasbubble in the uterine cavity, which confirms the suggestion of an infection rising from the uterine cavity via the salphinx to involve the ovary (click or tap the image to enlarge). A very common benign ovarian lesion that may appear cystic is a mature cystic teratoma, also called dermoid cyst. Mature in this context means benign, as opposed to the immature, malignant teratoma. Benign cystic teratomas typically occur in young women of child-bearing age. At imaging they are usually unilocular (up to 90%) but can be multilocular, and are bilateral in ~15%. Up to 60% may contain calcifications. The cystic component is fluid fat, produced by sebaceous glands in cyst lining. The presence of fat is diagnostic. The characteristic ultrasound appearance is that of a cystic mass, with a hyperechoic solid mural nodule, which is called a Rokitansky nodule or dermoid plug (figure). In another case the transvaginal ultrasound shows the 'tip-of-the-iceberg' sign: acoustic shadowing from the hyperechoic part of the dermoid cyst. This may be misinterpreted as bowel gas and the lesion may be overlooked. A fat-fluid level may be present, caused by fat floating on more aqueous fluid. Multiple thin, echogenic lines or stripes may be seen, caused by hair floating in the cyst cavity. Mature cystic teratomas, even though benign, are often resected because of increased risk of ovarian torsion, the most commonly associated complication. Other complications associated with teratoma are infection, rupture (spontaneous or trauma) and, rarely, hemolytic anemia (resolves with resection). Malignant transformation can occur but is also rare ( Axial T1-weighted image in the same patient shows a bright lesion with an internal septation. A septation is seen in about 10% of thse lesions. On the T1-weighted image with fat suppression there is suppression of the signal. This confirms the fatty content and is diagnostic of a teratoma. Classic low attenuation consistent with fat in a right sided cystic teratoma at CT. Cystadenoma and cystadenofibroma are also common benign ovarian tumors. They can be either serous or mucinous. At imaging a serous cystadenoma is most often unilocular and anechoic, and may look like a simple cyst. Mucinous cystadenomas are most often multilocular with thin ( The locules may contain complex fluid, due to proteinaceous debris or hemorrhage, or both. The finding of papillary projections should raise the suspicion of a possible borderline malignancy or a cystadenocarcinoma. Transvaginal ultrasound shows a 5.1x5.2-cm dominant left ovarian cyst. The cyst is anechoic and no septations are seen. Also there is no ascites. There is, however, a nodule on the posterior wall that shows no flow on Doppler. This may be a follicular cyst with some debris, but a cystic neoplasm cannot be excluded. Work-up with MRI is recommended. T2-weighted image of the same patient shows thin enhancing septations (as well as motion artifacts that should not be mistaken for septations). There are no tumor nodules and no adenopathy or peritoneal deposits. There is only a small amount of ascites. This proved to be a cystadenoma. The next case is another cystic lesion On the posterior wall a solid mural nodule is found, which is avascular. No secondary signs of malignancy. Continue with the MRI. Five years later the lesion has grown. Axial T2 shows a complex cystic left ovarian lesion, with a solid nodule on the posterior wall. At post-contrast axial T1W-FatSat the thin septa and the mural nodule show slight enhancement. On the basis of these findings the distinction between a benign ovarian lesion such as a cystadenofibroma and a malignant lesion cannot be made. The lesion was resected and was found to be a cystadenofibroma. The next case is a transabdominal ultrasound that shows a left-sided multiloculated cystic mass. This looks like a cystic ovarian neoplasm but no ovary could be identified. CT of the same patient shows a multi-loculated cystic mass adjacent to the bladder, connected to the left ovarian vein (arrow). There are thick septations and irregular wall thickening. On the basis of this CT the distinction between a benign ovarian lesion such as as cystadenofibroma and a malignant ovarian lesion cannot be made. The lesion was resected and found to be a cystadenofibroma. Remember, the role of imaging is not to determine the histological nature of a lesion, but to distinguish benign from malignant lesions and guide decisions on further management. The examples given here serve as a demonstration of suspicious imaging features, not as a guide for determination of histologic lesion type. Ultrasound shows a complex solid-cystic mass in the left ovary, and another, very large complex solid-cystic mass in the right hemi-pelvis. CT of the same patient shows a complex solid-cystic mass with thick, enhancing septations in the right ovary. These findings are very suspicious for a malignant cystic neoplams. There is also bilateral lymphadenopathy (arrows). Pathology showed a serous ovarian cystadenocarcinoma. This is the most common type of ovarian cancer. Ultrasound shows a very large multi-loculated cystic lesion in the region of the right adnex. Some locules are anechoic. Others contain uniform low-level echoes, consistent with proteineous content, such as hemorrhage or, in this case, mucin. The septations are thin, except for the dorsal septations that appear somewhat thicker, partially caused by the lower scanresolution at great depth. The septations are avascular. There are no solid components. There was no ascites. Despite the absence of solid components and despite the absence of vascularity on color Doppler, the size and the multi-loculated aspect of this lesion are suspicious for a cystic neoplasm and warant further evaluation. The CECT shows similar findings. The locules are of different attenuation, consistent with varying protein content. There is no ascites orperitoneal deposits and no lymphadenopathy. At pathology this was a mucinous cystadenocarcinoma of low malignant potential. The thin, relatively avascular septae, the absence of frank solid components, the absence of ascites and peritoneal carcinomatosis and the absence of invasion, suggest a lesion of low malignant potential (LMP). Note however, that this diagnosis can not be made on imaging findings alone. Especially the absence of invasion in ovarian stroma cannot be judged reliably on imaging. On ultrasound both ovaries are markedly enlarged and contain cystic components with intracystic solid components (arrows). The complex solid-cystic lesions, in addition to being bilateral, are suspicious for a cystic ovarian neoplasm and warrant further evaluation. Again, the role of imaging is to confirm a lesion is present and to decide that this is not a lesion that can be classified as definitely benign nor a lesion that can be safely followed-up: action is required. CT of the same patient confirms large bilateral complex solid-cystic lesions, bulging into the abdomen. The purpose of the CT is not to confirm what was already known from the ultrasound, but to stage disease. On the basis of CT (or of MRI) it is not possible to determine the histologic type of the tumor. This is not relevant. This patient will undergo surgery. For epithelial tumors - by far the most common group of malignant ovarian tumors - even after surgery, the exact tumor subtype is much less important for the prognosis than factors such as FIGO-stage, tumor differentation grade, and how succesful surgery was in removing all of the disease. For this patient the relevant findings are on the image on the left. There is a peritoneal implant. The tumor was resected and pathology showed this was an endometrioid ovarian carcinoma. While metastases to the ovary are most commonly solid - such as for example Krukenbergs metastases - cystic ovarian metastases do occur. The CT image shows complex cystic masses in both ovaries. While a serous cystadenocarcinoma may very well be bilateral, they are more often unilocular than multilocular. Barely visible is part of a circumferential colorectal cancer (blue arrow). Clearly visible are cystic implants on the peritoneal reflection (red arrow). These were cystic ovarian metastases of a colorectal cancer. This is an uncommon finding. by Deborah Levine et al September 2010 Radiology, 256, 943-954. by Spencer JA et al Eur Radiol. 2010 Jan;20(1):25-35. by John A. Spencer et al September 2010 Radiology, 256, 677-694. by Modesitt SC, Pavlik EJ, Ueland FR, DePriest PD, Kryscio RJ, van Nagell JR Jr. Obstet Gynecol. 2003 Sep;102(3):594-9. by Penelope L. Moyle et al July 2010 RadioGraphics, 30, 921-938. by Paula J. Woodward et al RadioGraphics 2001; 21:193-216. by Rajkotia K, Veeramani M, Macura KJ Top Magn Reson Imaging 2006; 17:379-97 by Michael P. Stany et al AJR 2010; 194:337-342 by Maitray D. Patel, MD, Vickie A. Feldstein, MD and Roy A. Filly, MD 2005 J Ultrasound Med 24:607-614Wouter Veldhuis, Robin Smithuis, Oguz Akin and Hedvig Hricak premenopausal Post-menopausal Follicular cyst Corpus luteum cyst Hemorrhagic ovarian cyst Endometrioma Polycystic ovary syndrome Ovarian hyperstimulation syndrome - Theca lutein cysts PID with tubo-ovarian abscess Mature cystic teratoma Cystadenoma and cystadenofibroma Serous ovarian cystadenocarcinoma Mucinous ovarian cystadenocarcinoma Endometrioid ovarian carcinoma Cystic metastases to the ovariesOvarian Cysts - Common lesionsDepartment of Radiology of the University Medical Center of Utrecht, of the Rijnland hospital in Leiderdorp, the Netherlands and the Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, USA
abdomen18
1
Pancreas - Acute Pancreatitis
by Thomas Bollen, Marieke Hazewinkel and Robin Smithuis
This review is based on a presentation given by Thomas Bollen at the Abdominal teaching course for the Dutch Radiology Society and was adapted for the Radiology Assistant by Marieke Hazewinkel and Robin Smithuis. We will discuss the following subjects:  Acute pancreatitis varies from a mild uneventful disease to a severe life-threatening illness with multisystemic organ failure (MOF) with shock, renal failure, respiratory failure and death. Incidence The frequency of acute pancreatitis varies among different countries (1). In the US the rate is 270 cases per 100,000, which accounts for more than 200.000 hospital admissions each year. In Finland the rate is 70 cases per 100,000 and in the Netherlands the rate is 18 per 100,000. Etiology Gallstones and alcohol abuse are the most common causes of acute pancreatitis, accounting for 80% of cases. Post-ERCP pancreatitis is the third most common cause of pancreatitis, but usually has a mild course. In the Netherlands there are approximately 3500 admissions for acute pancreatitis a year. In the last 10 years the incidence of acute pancreatitis has increased by 75%, due to increased alcohol consumption and increasing obesity. Clinical outcome The 1992 Atlanta Symposium on Acute Pancreatitis has classified this entity into a mild acute pancreatitis and a severe acute pancreatitis. 80-85% of cases of acute pancreatitis run a mild course without the development of multiple organ failure. This group has a mortality of 15-20% of cases of acute pancreatitis run a serious clinical course with pancreatic necrosis and the development of multiple organ failure. Of these, pancreatic necrosis remains sterile in 60% of patients, whereas in 40% of these patients the necrosis becomes infected. This last category of patients has the highest mortality rate of 25-70%. It is important to realize that severe acute pancreatitis runs a biphasic course (2). During the first 1-2 weeks there is a pro-inflammatory response, which results in a systemic inflammatory response syndrome (SIRS). It is a sterile response in which sepsis or infection hardly ever occurs. If the SIRS is severe it will lead to early multiple organ failure. After the first 1-2 weeks there is a transition from a pro-inflammatory to an anti-inflammatory response. It is during this anti-inflammatory response that the patient is at risk for the translocation of intestinal flora and the development of infection of necrotic tissue and fluid collections. The subsequent sepsis will result in late multiple organ failure. Mortality Early mortality in acute pancreatitis is the result of the systemic inflammatory response with multiple organ failure. Late mortality is the result of infection of pancreatic necrosis and peripancreatic fluid collections which results in sepsis and is seen in more than 50% of deaths. In the diagnosis and staging of acute pancreatitis and its complications CT is the imaging modality of choice. Ultrasound is important in determining whether gallstones are the cause of the acute pancreatitis (i.e. biliary pancreatitis). ERCP with sphincterotomy and stone extraction should only be used if a patient has biliary pancreatitis and signs of biliary obstruction. MRI is as sensitive as CT, but not as practical or accessible. In certain cases it can be of additional value. There is no additional value of an early CT (within 72 hours) in patients with acute pancreatitis. The diagnosis is usually made on clinical and laboratory findings. An early CT may be misleading concerning the severity of the pancreatitis, since it can underestimate the presence and amount of necrosis. Early CT is only recommended when the diagnosis is uncertain, or in case of suspected early complications such as perforation or ischemia. The case on the left shows a normally enhancing pancreas with enhancement comparable to that of the spleen on day 1. As the patient's condition worsened, a second CT was performed on day 3. Notice how the majority of the pancreatic body and tail no longer enhance - in fact only a small part of the pancreatic head enhances. The patient died on day 5 due to severe SIRS and multiple organ failure. It is critical to identify patients who are at high risk for severe disease, since they require close monitoring and possible intervention. Early staging is based on the presence and degree of systemic organ failure (cardiovascular, pulmonary, renal) and on the presence and extend of pancreatic necrosis (3,4). Balthazar et al constructed a CT severity index (CTSI) for acute pancreatitis that combines the grade of pancreatitis (A-E) with the extent of pancreatic necrosis. The CTSI assigns points to patients according to their grade of acute pancreatitis - which can be determined on a non-contrast CT as well as a contrast-enhanced CT - as well as the degree of pancreatic necrosis - which requires the use of intravenous contrast material. More points are given for a higher grade of pancreatitis and for more extensive necrosis. Mild pancreatitis Patients with pancreatitis but no collections or necrosis (i.e. Balthazar grade A-C) have a mild pancreatitis. This is also called 'edematous or interstitial pancreatitis' (no pancreatic necrosis). It is a self-limiting disease with an uneventfull recovery occurring in 80% of patients with acute pancreatitis. There is an intermediate form of pancreatitis without pancreatic necrosis with an intermediate clinical course. This is called extrapancreatic necrosis (EXPN) (5,6). Sometimes the term exudative pancreatitis is used. These patients have Balthazar grade D or E. These patients have a relatively mild course because there is no pancreatic necrosis, but there is higher morbidity than in interstitial pancreatitis, because they have peripancreatic collections, that can become infected (7). Severe pancreatitis or necrotizing pancreatitis Severe pancreatitis, also called 'necrotizing pancreatitis' occurs in 20% of patients. It is characterized by a protacted clinical course, a high incidence of local complications and a high mortality rate. On the left there is normal enhancement of the entire pancreatic gland with only mild surrounding fatty infiltration. There are no fluid collections or necrosis (Balthazar grade C, CTSI: 2). In exudative pancreatitis, or better called EXPN, there is normal enhancement of the entire pancreas associated with extensive peripancreatic collections. These are often heterogeneous in appearance and may be progressive. EXPN consists of necrosis of peripancreatic fat, extravasated pancreatic fluid and inflammatory and hemorrhagic components. When peripancreatic collections persist or increase, it is usually due to the presence of fat necrosis (i.e. EXPN). Since fat does not enhance on CT, we cannot diagnose fat necrosis. In the case on the left on day 18 there is expansion of the peripancreatic collections. There are two or more collections, but no pancreatic necrosis. (Balthazar grade E, CTSI: 4) On the left a patient with acute necrotizing pancreatitis. There are 2 or more fluid collections and more than 50% of the gland does not enhance (Balthazar grade E, CTSI :10). This patient died two days later due to severe systemic inflammatory response syndrome (SIRS) and multi organ failure. The detection of pancreatic necrosis in clinical practice is important because most life-threatening complications occur in patients with pancreatic necrosis (1,2). Necrotizing Pancreatitis (2) In the case on the left the body and tail of the pancreas do not enhance after i.v. contrast (blue arrows). There is however normal enhancement of the pancreatic head (yellow arrow). More than 50% of the pancreas is necrotic and there are at least two collections (CTSI : 10). Central gland necrosis is a subtype of necrotizing pancreatitis. It represents necrosis between the pancreatic head and tail and is nearly always associated with disruption of the pancreatic duct. This leads to persistent collections as the viable pancreatic tail continues to secrete pancreatic juices. These collections react poorly to endoscopic or percutaneous drainage. Definitive treatment often requires distal pancreatectomy. The images on the left illustrate a case of central gland necrosis. There is a fluid collection in the omental bursa, adjacent to the stomach. Notice the normal enhancement of the pancreatic head and tail, but the lack of enhancement of the majority of the pancreatic body. Two weeks later the collection in the omental bursa and pancreatic body has increased significantly. The pancreatic tail still enhances and so does the pancreatic head (arrows). This patient developed septicaemia and underwent surgery. Intraabdominal fluid collections and collections of necrotic tissue are common in acute pancreatitis. These collections develop early in the course of acute pancreatitis. In the early stage such a collection does not have a wall or capsule. Preferred locations are the omental bursa and the retroperitoneal space (anterior and posterior pararenal space). These collections are the result of the release of activated pancreatic enzymes (namely lipase, trypsin and amylase) which also causes necrosis of the surrounding tissues. This explains why a lot of these collections contain solid debris. 50% of these collections show spontaneous regression (image on the left). The other 50% either remain stable or increase and undergo organization and demarcation with liquefaction. They may remain sterile or develop infection. Based on imaging alone it is often not possible to determine whether these collections contain fluid or necrotic tissue and whether they are infected or not. Consequently, instead of naming them as 'pseudocysts', 'abscesses' or 'necrosis', it is better to describe them as 'peripancreatic collections'. On the left CT scans of a patient with acute pancreatitis. There is a collection in the area of the pancreatic head in the right anterior pararenal space. On a follow up scan the collection in the right anterior pararenal space is larger. It has a fluid density and a thin enhancing wall. One day later the patient developed septicaemia and percutaneous drainage was performed. After drainage the collection has barely diminished in size and consequently there was suspicion of necrotic tissue. The patient therefore underwent surgery and the collection was found to consist of necrotic debris, which was not appreciated on CT. The necrotic debris was too thick for successful percutaneous drainage.  The case on the left is a typical example of infected pancreatic necrosis. On day 3 there is no enhancement of the pancreas, consistent with necrosis (compare to enhancing spleen). On follow up the peripancreatic collections increase in size and finally there are air bubbles in the heterogeneous collection, consistent with infected pancreatic necrosis. Infected necrosis (2) In the patient on the left there is a normal enhancement of the pancreas with surrounding septated heterogeneous peripancreatic collections with fluid- and fat densities (images on the left). Two weeks later (images on the right) there are air bubbles in the peripancreatic collection, consistent with infected necrosis. This patient underwent surgery. The surgeon had to remove a lot of necrotic tissue and estimated he had removed over 90% of the pancreas. Remarkably, a CT performed 6 months after surgery showed a normal pancreas. This indicates that during surgery the differentiation between pancreatic necrosis and necrosis of the peripancreatic tissue is sometimes impossible.  The patient on the left presented with a gastric outlet obstruction 2 months after an episode of acute pancreatitis. There is a large, homogeneous, well-demarcated peripancreatic collection which abuts the stomach and the pancreas. The patient did not have a fever. The collection underwent successful percutaneous drainage and subsequently resolved along with the patient's symptoms. This collection therefore proved to be a pancreatic pseudocyst. On the other hand on the left a CT of an ICU patient on day 40 with central gland necrosis with a spiking fever. The CT shows a similar collection to that of the previous patient, exept for its pancreatic location. The collection is homogeneous and well-demarcated with a thin wall abutting the stomach. During endoscopic debridement this collection contained fluid and necrotic tissue which was removed from the area of the pancreas (image on the right). Although the imaging characteristics in this case are similar to the previous case, this proved to be infected necrosis. The patient on the left also has a homogeneous pancreatic and peripancreatic collection, well-demarcated with an enhancing wall, on day 25 of an episode of acute pancreatitis. Since this patient had fever and multiple organ failure, this collection was suspected to be infected necrosis and not a pseudocyst. At surgery the collection contained a lot of necrotic debris, which was not recognizable on CT. These cases illustrate that CT cannot reliably differentiate between collections that consist of pure fluid and those that contain fluid and solid debris with or without infection. MRI is superior to CT in differentiating between fluid and solid debris. On the left a patient with several homogeneous peripancreatic collections on CT. Since these collections show homogeneous high signal intensity on a fat-suppressed T2 MRI sequence, they must be fluid-filled. On the left another patient 2 months after an episode of acute exudative pancreatitis with also a homogeneous peripancreatic collection in the transverse mesocolon (arrow). A T2-weighted MRI sequence however, shows that the collection has low signal intensity (arrow), and is therefore mainly solid. This patient had no fever or signs of sepsis. If endoscopic or percutaneous drainage were attempted in this patient, there would be little or no effect on its size. The only result would be an increased risk of infection when the collection is colonized following drain placement. The current management of acute pancreatitis is to be conservative for as long as possible. During the first two weeks patients with severe acute pancreatitis and multi organ failure should be stabilized in the ICU (8). Interventions should be delayed for as long as possible. Allow for demarcation of collections. Remember that many collections will resorb spontaneously. Once the clinical condition of the patient deteriorates and the patient is febrile, fine needle aspiration (FNA) can be used to differentiate between sterile and infected collections. Important remarks concerning FNA: Important remarks concerning Drainage: Collections can be approached through the transhepatic, transgastric or transabdominal route, but the preferred approach is to stay in the retroperitoneal compartment. This approach has some advantages over the others: Surgical intervention in infected necrotizing pancreatitis generally consists of necrosectomy via laparotomy (2,9). The morbidity and mortality after this procedure might be reduced by minimally invasive strategies. In the Netherlands there is a nationwide study into the optimal treatment of patients with infected necrotizing pancreatitis: the PANTER trial. (10). The 20 hospitals of the Dutch Acute Pancreatitis Study Group are currently enrolling patients in a randomised trial to compare:  by Whitcomb N. Engl. J. Med., May 18, 2006; 354(20): 2142 - 2150 Management of acute pancreatitis: from surgery to interventional intensive care. by J. Werner et al Gut 2005; 54: 426-36 by Emil J. Balthazar, MD Radiology 2002;223:603-613 by J. Dar?o Casas et al AJR 2004; 182:569-574 Extrapancreatic necrotizing pancreatitis with viable pancreas: a previously under-appreciated entity by G.H. Sakorafas et al J Am Coll Surg 1999;188: 643-8 Update on Acute Pancreatitis: Ultrasound, Computed Tomography, and Magnetic Resonance Imaging Features by Thomas Bollen et al Semin Ultrasound CT MR 2007. 28: 371-83 by Dipti K. Lenhart and Emil J. Balthazar AJR 2008; 190:643-649 Recommendations 2003 by Besselink MGH et al Br J Surg. 2006 May;93(5):593-9 by Besselink MGH et al BMC Surg 2006 Apr 11;6:6Thomas Bollen, Marieke Hazewinkel and Robin Smithuis CT Severity Index Interstitial pancreatitis Exudative Pancreatitis Necrotizing Pancreatitis Central gland necrosis Infected necrosis Pseudocyst FNA and Drainage Surgical intervention Take home messagesPancreas - Acute PancreatitisRadiology department of the St. Antonius hospital, Nieuwegein, the Medical Centre Alkmaar and the Rijnland hospital, Leiderdorp, the Netherlands
abdomen19
1
Pancreas - Carcinoma
by Otto van Delden and Robin Smithuis
Pancreatic adenocarcinoma has a poor prognosis. Complete resection of the tumor is the only curative treatment. About 10-15% of all patients with a pancreatic carcinoma will finally undergo resection and only in half of these cases the resection will prove to be radical. In this article we will focus on the CT-findings that are used to select patients with probable resectable tumors. As the clinical presentation, staging and treatment of other types of pancreatic neoplasms is distinctly different from adenocarcinomas, these are not discussed in this article. by Otto van Delden and Robin Smithuis Pancreatic carcinoma is a relatively common tumor with an incidence of 7,6 per 100.000 per year in Western-Europe. It comprises about 2,5 % of all newly diagnosed tumors and 5% of all cancer. The majority of pancreatic cancers (85%) are adenocarcinoma of ductal origin. It is more common in men (men:woman 1,5:1) between the age of 60 and 70 years [1-4]. In spite of the limited tumor size the majority of pancreatic head cancers (80%) are not eligible for resection at the time of diagnosis. This is due to advanced local tumor extension (40%) or the presence of distant metastatic disease (40%) mostly due to liver metastases of para-aortic lymphadenopathy. Operation The only curative treatment option is surgical resection. Out of every hundred patients with pancreatic carcinoma only 20 patients will be sceduled for explorative laparotomy. Out of these 20 only about 13-14 patients will undergo resection of the tumor, but only half of these resections will finally prove to be radical at pathologic examination of the resected specimen.. The resection consists of a partial pancreaticoduodenectomy according to Whipple or the modern variant, the so-called 'pylorus-preserving' pancreaticoduodenectomy. Palliation When the tumor proves to be unresectable during exploratory laparotomy, a so-called 'double bypass' (gastro-enterostomy en hepaticojejunostomy) is usually performed for palliative reasons. When curative resection is not considered an option, based on preoperative imaging and cytology or histology, palliation consists of endoscopic or percutaneous biliary stenting and celiac plexus block for relief of pain. Patients with a relatively short life expectancy (e.g. patients with extensive hepatic metastases), are probably best served by palliation by means of endoscopic bile duct stenting. In patients with a longer life expectancy (e.g. patients with a small, but locally unresectable tumor without distant metastatic disease), a double bypass is generally also considered acceptable palliation. Most pancreatic cancers occur in the head of the pancreas (75%). A minority is found in the body (15%) and tail (10%). At the time of diagnosis a pancreatic head carcinoma is usually a little larger than 3 cm. When tumors of the pancreatic body and tail are diagnosed, they are usually much larger, because they present late with aspecific symptoms. These tumors are usually irresectable. Tumors originating in the distal common bile duct or ampulla may also grow into the pancreatic head and together with pancreatic head carcinoma these tumors are often grouped together under the name periampullary tumors. This has some practical value as diagnostic imaging, staging and treatment of all these periampullary tumors is the same. The most striking clinical symptom leading to diagnostic imaging is painless obstructive jaundice, which is caused by compression or ingrowth of the distal common bile duct. US is the first line imaging test for the evaluation of these patients. US can determine the level of obstruction in most cases (sensitivity >90%). In patients with a pancreatic head tumor, typically dilatation of the common bile duct and pancreatic duct (double duct sign) is seen, which is very suggestive for a mass in the pancreatic head, even in the absence of a visible mass. The tumor itself usually presents as a hypoechoic mass (figure). In the detection of pancreatic cancers US has an overall sensitivity of 75% and a specificity of 75%. However in many cases US will suffice as the only imaging test for diagnosis and staging. This is particularely true in patients with tumors > 3 cm and liver metastases > 2 cm. The overall sensitivity and specificity for determining resectability of all pancreatic carcinomas however is only 63% and 83% respectively. If the cause of a distal bile duct obstruction is not revealed by US and there is a high suspicion for a pancreatic or periampullary tumor, the next diagnostic test is CT. ERCP (or MRCP) is only the next step when there is a high suspicion of bile duct stones. Whenever a pancreatic tumor is detected with US and no definite signs of unresectability are found, the next step is CT. CT should be done before ERCP and insertion of an endoprosthesis, because artifacts and post-ERCP pancreatitis may hamper the diagnostic accuracy of CT. As pancreatic carcinoma is a hypovascular tumor, it presents as a hypodens mass on a CECT. The mass is usually ill-defined. In 10 - 15% the tumor is isodens and therefore may be difficult to detect. Tumors smaller than 2 cm. may also be difficult to detect on CECT. In these cases  indirect signs may be helpful such as the presence of the double duct sign, atrophy of the pancreatic tail, or fullness of the pancreatic head (loss of the lobular appearance of the pancreatic parenchyma). CT and MRI both have a higher sensitivity than ultrasound for the detection of small ( MRI-sequences should involve at least T2W-images en dynamic T1W-images after intravenous administration of gadolinium. MRCP is also very sensitive for detecting a periampullary mass, but offers no significant additional staging information [9]. Many patients in whom a pancreatic head tumor is detected by ultrasound still undergo ERCP. Although ERCP has a high sensitivity for detecting pancreatic head tumors, it is nowadays no longer indicated because the diagnosis can usually be made with non-invasive tests. ERCP offers no usefull tumor staging information. It is doubtfull whether pre-operative bile duct drainage by ERCP is beneficial for the patient [12]. Pre-operative biliary drainage may potentially even increase the risk for post-operative infectious complications. Endoscopic ultrasound is generally accepted as the most sensitive imaging test for the detection of small pancreatic head tumors, particularly when smaller than 2 cm [10]. These pancreatic head tumors can be missed even on a technically excellent CT and therefore a 'negative' CT-scan in a patient with a strong suspicion for pancreatic head cancer requires additional imaging with endoscopic ultrasound. Unfortunately, there are only a few centers in The Netherlands with sufficient experience in this operator-dependent-technique. Endoscopic ultrasound has also been used for local tumor staging, but is currently not frequently used as such in the Netherlands. Diagnostic laparoscopy, sometimes complemented by laparoscopic ultrasound has been advocated by some as a staging tool. Laparoscopy is much more sensitive than any other technique for the detection of peritoneal implants and superficial liver metastases. Local staging is also feasible by laparoscopic ultrasound. However, a large series has shown, that the yield of laparoscopy after CT is not high enough to justify using this technique routinely [19,20]. It may be usefull in selected cases where there is doubt about resectability or when suspected metastatic disease cannot be proven otherwise. Water should be used as oral contrast material. A precontrast scan of the pancreas can be performed to look for calcifications within the pancreas, which may indicate the presence of a focal pancreatitis. On a contrast enhanced CT (CECT) these calcifications may not be appreciated as they will be confused with enhancement. Depending on the type of multidetector CT, 120 - 150 ml contrast is given at an injection rate of 3-5 ml/s. Slice thickness depends on the type of scanner that is used, but should be preferentially 2-3 mm or less. An early arterial phase-scan (delay 20 sec) does not add significant information on the staging of the pancreastumor, since there is not enough contrast in the pancreas [8]. Only if the surgeons want to get optimal pre-operative 3D-information on the anatomy of the mesenteric arteries this phase is included. The early-portal phase is also called the pancreatic phase. It has a scan-delay of 40-50 sec. This is the most important phase for detecting and staging a pancreatic tumor. At that moment the normal pancreatic parenchyma will enhance optimally, because it gets all of its bloodsupply through the arterial and capillary system. In this phase there is optimal attenuation difference between the hypodense tumor and the normal enhancing pancreatic parenchyma. This phase helps in the differentiation of liverlesions and usually the mesenteric arteries and veins are well opacified [7]. The 'late portal' or hepatic phase has a scan-delay of 70-80 sec. At that moment the normal liverparenchyma will enhance optimally, because normal livercells get 80% of their bloodsupply through the portal venous system. Liver metastases do not get their bloodsupply from the portal venous system and will be seen in this phase as hypovascular or hypodense lesions. This phase is performed for the overall assessment of the abdomen to look for liver metastases, lymphnodes and peritoneal implants. This phase is also helpfull for local staging of the tumor and detection of venous ingrowth. It is of the utmost importance to stage a pancreatic tumor correctly as the clinical consequences of this are enormous. Overstaging will lead to undertreatment if a laparotomy is not performed in a patient with a potentially resectable tumor. Understaging will lead to an unnecessary laparotomy with all the associated risks. To withhold the chance for curative resection from as few patients as possible, it is important to determine unresectability with a very high specificity, even if this means a lower sensitivity. Some patients will therefore get the benefit of the doubt and undergo a negative exploratory laparotomy. In some series the unresectability rate at laparotomy may be as high as 30%. Since the pancreas has no capsule, pancreatic tumor will easily spread into adjacent structures (figure). Because the confluens of the portal and superior mesenteric vein is in direct continuity with the pancreatic head, ingrowth into this vessel will often be the first sign of tumor extension outside the pancreas. Ingrowth into the celiac axis or superior mesenteric artery is always considered a criterium for unresectability. Although partial resection of the portal vein or superior mesenteric vein are technically possible and are being performed, ingrowth into these vessels is considered a criterium for unresectability by most oncologic surgeons in the Netherlands. Some centers in the US and Japan will resect part of the portal vein in case of tumor ingrowth. Although associated with a worse prognosis, the presence of peripancreatic lymphnode metastases does not constitute a definite contraindication for resection. Limited ingrowth into the peripancreatic fat, ingrowth into the duodenum or the gastroduodenal artery does not render a tumor unresectable as this vessel and the duodenum can be resected en-bloc with the tumor. When there is contiguity between the tumor and the portal or superior mesenteric vein, but the vessel is surrounded by tumor for less than half the circumference ( On the left two cases of pancreatic tumors with tumor-vessel contiguity These patients generally will be given the benefit of the doubt and will be sceduled for operation. Tumor ingrowth into stomach, colon, mesocolon, inferior vena cava or aorta constitute definite criteria for unresectability. Also the presence of hepatic metastases, peritoneal metastases or para-aortic lymfnode metastases is an absolute sign of unresectability. Mesenteric lymph node metastases, not immediately adjacent to the pancreas usually also indicate unresectability. Liver metastases and distant lymph node metastases should allways be proven by means of cytologic or histologic biopsy before refraining from exploratory laparotomy. Ingrowth into the celiac axis, hepatic artery or superior mesenteric artery also preclude resection. When a fatplane or normal pancreatic parenchyma is visible between the tumor and the vessel, the tumor is usually locally resectable. When there is tumor-vessel contiguity, but the vessel is surrounded by tumor for less than half the circumference ( This group of patients will usually get the benefit of the doubt and undergo exploratory laparotomy. On the left a pancreatic tumor in direct contiguity with the confluens of the portal and superior mesenteric vein. The tumor surrounds the confluens for more than half the cirumference (>180?). This tumor was regarded as unresectable. When the tumor surrounds the vessel for more than half the cirumference (>180?), the tumor will nearly allways be unresectable. Most surgeons will consider this a solid criterium for unresectability [13-16]. Flattening of the vessel or irregular vascular contours are also indicative of ingrowth. When the tumor surrounds the portal vein or superior mesenteric vein completely (360?) or occludes the vessel, the tumor is allways unresectable [13-16]. Other criteria for vascular ingrowth have been described, such as dilatation of the gastrocolic trunk (a sidebranch of the superior mesenteric vein) and the 'mesenteric teardrop sign'. These signs are not more sensitive or specific than the abovementioned criteria and therefore probably do not have much additional value as other criteria for unresectability are usually also present in these cases[17,18]. On the left a tumor thrombus is present in the lumen of the superior mesenteric vein. This is also a sign of unresectability. On the left a pancreatic carcinoma with encasement of the hepatic artery. The pancreatic duct is obstructed with subsequent atrophy of pancreatic tail. This tumor is not resectable. The differential diagnosis of a pancreatic head tumor includes carcinoma, focal pancreatitis, lymphoma and metastasic disease. Sometimes it is difficult to differentiate between a pancreatic head tumor and focal pancreatitis in the pancreatic head. There are no solid imaging criteria to decide this with certainty in all cases. Although a double duct lesion may be seen in cases of pancreatitis, this finding should always lead to a strong suspicion for pancreatic carcinoma. Imaging guided biopsy has a limited value in these cases because false-negative results frequently occur and can therefore not be used to rule out cancer. PET has currently not proven it?s use in the differentiation as tumor and inflammation may both show increased uptake of the radiofarmacon. In cases where differentiation is impossible, laparotomy or imaging follow-up may be performed, depending on the specific clinical circumstances. Most large surgical series show, that in about 5% of the patients, who undergo resection for suspicion of pancreatic head cancer only pancreatitis is eventually found in the resected specimen. [22]. Ahlgren JD. Epidemiology and risk factors in pancreatic cancer. Semin Oncol 1996;23:241-50. Rosewicz S, Wiedenmann B. Pancreatic carcinoma. Lancet 1997;349:485-9 Sener SF, Fremgen A, Menck HR, Winchester DP. Pancreatic cancer: a report of treatment and survival trends for 100,313 patients diagnosed from 1985-1995, using the National Cancer Database. J Am Coll Surg 1999;189:1-7. Di Magno EP, Reber HA, Tempero MA. AGA technical review on the epidemiology, diagnosis, and treatment of pancreatic ductal adenocarcinoma. American Gastroenterological Association. Gastroenterology 1999;117:1464-84 Karlson BM, Ekbom A, Lindgren PG, Kalskogg V, Rastad J. Abdominal US for diagnosis of pancreatic tumor: prospective cohort analysis. Radiology 1999;213:107-11. Sheridan MB, Ward J, Guthrie JA, Spencer JA, Craven CM, Wilson D, et al. Dynamic contrast-enhanced MR imaging and dual-phase helical CT in the preoperative assessment of suspected pancreatic cancer: a comparative study with receiver operating characteristic analysis. AJR Am J Roentgenol 1999;173:583-90 Boland GW, O?Malley ME, Saez M, Fernandez-del-Castillo C, Warshaw AL, Mueller PR. Pancreatic-phase versus portal vein-phase helical CT of the pancreas: optimal temporal window for evaluation of pancreatic adenocarcinoma. AJR Am J Roentgenol 1999;172:605-8. Graf O, Boland GW, Warshaw AL, Fernandez-del-Castillo C, Hahn PF, Mueller PR. Arterial versus portal venous helical CT for revealing pancreatic adenocarcinoma: conspicuity of tumor and critical vascular anatomy. AJR Am J Roengtenol 1997;169:119-23 Adamek HE, Albert J, Breer H, Weitz M, Schilling D, Riemann JF. Pancreatic cancer detection with magnetic resonance cholangiopancreatography and endoscopic retrograde cholangiopancreatography: a prospective controlled study. Lancet 2000;356:190-3. Wiersema MJ. Accuracy of endoscopic ultrasound in diagnosing and staging pancreatic carcinoma. Pancreatology 2001;1:625-35. Rose DM, Delbeke D, Beauchamp RD, Chapman WC, Sandler MP, Sharp KW, et al. 18F-fluorodeoxyglucose-positron emission tomography in the management of patients with suspected pancreatic cancer. Ann Surg 1999;229:729-38 Sewnath ME, Karsten TM, Prins MH, Rauws EJ, Obertop H, Gouma DJ. A meta-analysis on the efficacy of pre-operative biliary drainage for tumors causing obstructive jaundice. Ann Surg 2002;236:17-27. Lu DS, Reber HA, Krasny RM, Kadell BM, Sayre J. Local staging of pancreatic cancer: criteria for unresectability of major vessels as revealed by pancreatic-phase, thin-section helical CT. AJR Am J Roentgenol 1997;168:1439-43. O?Malley ME, Boland GW, Wood BJ, Fernandez-del-Castillo C, Warshaw AL, Mueller PR. Adenocarcinoma of the head of the pancreas: determination of surgical unresectability with thin-section pancreatic-phase helical CT. AJR Am J Roentgenol 1999;173:1513-8. Phoa SS, Reeders JW, Rauws EA, Wit LT de, Gouma DJ, Lam?ris JS. Spiral computed tomography for preoperative staging of potentially resectable carcinoma of the pancreatic head. Br J Surg 1999;86:789-94. Phoa SS, Reeders JW, Stoker J, Rauws EA, Gouma DJ, Lam?ris JS. CT criteria for venous invasion in patients with pancreatic head carcinoma. Br J Radiol 2000;73:1159-64. Hough TJ, Raptopoulos V, Siewert B, Mattews JB. Teardrop superior mesenteric vein: CT sign for unresectable carcinoma of the pancreas. AJR Am J Roentgenol 1999;173:1509-12. Hommeyer SC, Freeny PC, Crabso LG. Carcinoma of the head of the pancreas: evaluation of the pancreaticoduodenal veins with dynamic CT-potential for improved accuracy in staging. Radiology 1995;196:233-8. Nieveen van Dijkum EJ, Romijn MG, Terwee CB, Wit LT de, Meulen JH van der, Rauws EA, et al. A multicenter trial of laparoscopic staging and subsequent palliation in patients with periampullary carcinoma. Ann Surg 2002 Barreiro CJ, Lillemoe KD, Koniaris LG, Sohn TA, Yeo CJ, Coleman J, et al. Diagnostic laparoscopy for periampullary and pancreatic cancer: what is the true benefit? J Gastrointest Surg 2002;6:75-81. Tilleman EH, Phoa SS, Delden OM van, Rauws EA, Gulik TM van, Lam?ris JS, Gouma DJ. Reinterpretation of radiological imaging in patients referred to a teriary care referral center with a suspected hepatobiliary or pancreatic malignancy; impact on treatment strategy. Eur Radiol 2002 Gulik TM van, Moojen TM, Geenen R van, Rauws EA, Obertop H, Gouma DJ. Differential diagnosis of focal pancreatitis and pancreatic cancer. Ann Oncol 1999;10(Suppl 4):85-8. Diederichs CG, Staib L, Vogel J, Glasbrenner B, Glatting G, Bramts HJ, et al. Values and limitations of 18F-fluorodeoxyglucose-positron-emission tomography with preoperative evaluation of patients with pancreatic masses. Pancreas 2000;20:109-16. Fr?hlich A, Diederichs CG, Staib L, Vogel J, Beger HG, Reske SN. Detection of liver metastases from pancreatic cancer using FDG PET. J Nucl Med 1999;40:250-5. Gouma DJ, Tilleman EH, Benraadt J, Bossuyt PM. Diagnostiek en behandeling van het distale galweg- of pancreascarcinoom; richtlijn van de Integrale Kankercentra Amsterdam en Stedendriehoek Twente. Ned Tijdschr GeneeskdOtto van Delden and Robin Smithuis Treatment Ultrasound CT MRI ERCP Endoscopic ultrasound Diagnostic laparoscopy Early portal phase Late portal phase Local Tumorspread Resectable Not resectable Vessel ingrowthPancreas - CarcinomaFrom the Radiology Department of the Academical Medical Centre, Amsterdam and the Rijnland Hospital, Leiderdorp, the Netherlands
abdomen20
1
Pancreas - Cystic Lesions
by Marc Engelbrecht, Jennifer Bradshaw and Robin Smithuis
Cystic pancreatic lesions are increasingly identified due to the widespread use of CT and MR. Most of these cysts are incidental findings and are benign or low-grade neoplasms. The characterization and management of these cysts is a dilemma since there is a significant overlap in the morphology of benign and premalignant lesions. MR is the imaging study of choice to both characterize and follow these pancreatic cysts. In many cases these lesions remain undetermined and guidelines are needed for follow up and management. Pancreatic cysts can be categorized into the following groups: When a cystic pancreatic lesion is detected, the first step is to decide whether the lesion is most likely a pseudocyst or a cystic neoplasm. This scheme is a simplified roadmap for the differentiation of pancreatic cysts.  The left CT-image is of a patient with a history of pancreatitis. There are two unilocular or simple cysts. Notice also the retroperitoneal fat-stranding on the right. The most likely diagnosis is pseudocysts. The CT on the right shows a cyst in the pancreatic tail in a 36 year old woman, which was found incidentally with US. The cyst has a thick irregular rim and contains solid 'non-dependent' components. The most likely diagnosis is a cystic neoplasm. CT will depict most pancreatic lesions, but is sometimes unable to depict the cystic component. MR with heavily weighted T2WI and MRCP will better demonstrate the cystic nature and the internal structure of the cyst and has the advantage of demonstrating the relationship of the cyst to the pancreatic duct as is seen in IPMN. The images show a serous cystic neoplasm (SCN). MRI better shows the central scar (figure). There are cases when CT can be helpful, since it better depicts a central calcification in SCN or peripheral calcification in a mucinous cystic neoplasm (MCN). MRI is usually of more diagnostic than CT. MRI can show the cystic nature of a pancreatic lesion and it's internal structure. The MRI shows a large cyst with dependent internal debris (figure). Presence of internal dependent debris appears to be a highly specific MR finding for the diagnosis of pancreatic pseudocyst (6). MRI shows a lesion, which consists of multiple small cysts. This could be a serous cystic neoplasm or a branch-duct IPMN. The connection of the cystic lesion to the pancreatic duct indicates that this is a branch-duct IPMN. Small pancreatic cysts have been documented in approximately 2.3% of CT studies and up to 19% of MR studies (11). Most of these cysts are found in asymptomatic patients, who are studied for other reasons and represent benign or low-grade indolent neoplasms. The ability of imaging to enable a specific diagnosis of an individual pancreatic cyst is limited, but is easier in larger cystic lesions. In most small cysts we should not attempt to characterize the lesion and when we do, we should not be too confident. The management of cystic neoplasms has not yet been standardized and continues to evolve. According to the recent 2012 consensus guidelines by Tanaka et al the items mentioned in the Table should be addressed (8). Age, life-expectancy and comorbidity should be considered in the possible surveillance or treatment. Cysts smaller than 3 cm and no worrisome or high risk-features can be considered for follow-up with either MRI, CT or ultrasound. Cysts with obvious high risk stigmata should be considered for resection. The table shows the American College of Radiology flowchart for imaging of incidentally discovered pancreatic cysts in asymptomatic patients (11). Pancreatic cysts are regarded symptomatic when there is hyperamylasemia, recent-onset diabetes, severe epigastric pain, weight loss, steatorrhea, or jaundice. key findings: The CT demonstrates a large cyst in the upper abdomen in a patient who had an acute pancreatitis. Notice that there is also some ascites and pleural fluid. There wall enhances. Here an example of the value of MRI compared to CT. The MRI shows dependant debris (arrow) as a discriminator for walled off necrosis in a patient with a pseudocyst. CT demonstrates two large cysts in a 45 year old woman, who had a trauma. Notice some fat stranding in the retroperitoneal space (arrow). The imaging findings combined with the history make it very likely that these are traumatic pseudocysts. Most pseudocyst occur in the peripancreatic region, but rarely they may extend to the mediastinum. Scroll through the images. This patient has a chronic pancreatitis. Notice the calcifications in the pancreatic head (curved arrow). There are multiple pseudocysts extending all the way to the mediastinum compressing the heart. The diagnosis of a cystic neoplasm should be considered when there is no history of pancreatitis or trauma. Morphological characteristics of a cystic neoplasm are: thick irregular rim, septations, solid components, a dilated pancreatic duct > 3mm and calcifications. Fluid aspirated from a cyst with an amylase level In the table some discriminating features of cystic neoplasms. In many cases however it is not possible to make a definitive diagnosis. It is important to make the diagnosis of a serous cystic neoplasm, since this is the only tumor that has no malignant potential. In many cases differentiation from a branch-duct IPMN is difficult, since both have multiple small cysts. Some cystic neoplasm are seen almost exclusively in women, like mucinous cystic neoplasm (99%) and serous cystic neoplasm (75%). Solid pseudopapillary epithelial neoplasm is another pancreatic tumor which may have cystic components. It is uncommon, but is seen exclusively in young women. Hence the following rule:  key findings: The pathology specimen shows multiple microcysts, which gives the tumor a lobulated appearance. A macrocystic serous cystic neoplasm is rare and, although benign, can be similar in appearance to the potentially malignant macrocystic mucinous cystic neoplasm. A characteristic feature of a serous cystic neoplasm is a central scar, sometimes with calcifications. Sometimes the microcystic component of this tumor is difficult to identify on CT. MR will better identify the internal architecture. MRI is also useful in determining if the cysts communicate with the pancreatic duct or not to differentiate this lesion from a branch-duct IPMN (see below). The pathology specimen shows a cystic tumor with multiple small cysts and a central scar. There are no calcifications. CT-image of a 51 year old woman with a history of gallstones and abdominal pain. There is a hypodense lesion with central calcification in the head of the pancreas. The lesion has a lobulated contour. Continue with the MR. MRI better demonstrates the morphologic features of the lesion. On T2WI the lesion is multicystic. Note the central low signal due to the central scar with calcifications. Although some of the cysts are rather large, this is still a characteristic appearance of a serous cystic adenoma. Another example of a serous cystic neoplasm. The enhanced image on the right shows a hypodense lesion with central calcification in the body of the pancreas. On the right image subtle enhancement of septations are seen. Notice that on CT it is very difficult to appreciate the cystic nature of these lesions and you might think that you are dealing with a pancreatic adenocarcinoma. MRI will easily demonstrate the cystic nature of these lesions. The T2WI with fatsat nicely demonstrates a lobuated hyperintense lesion with central scar, which is characteristic of a SCN. It may be difficult to differentiate a serous microcystic adenoma from a branch-duct IPMN or intraductal papillary mucinous neoplasm. IPMN is always connected to the pancreatic duct. T2WI of a 71 year old man with a history of weight loss and nondescript upper abdominal complaints. This was initially thought to be a branch-duct IPMN, but turned out to be a SCN. Notice the central hypointensity. This is scar tissue in a SCN. Notice also the characteristic lobulated surface. Another example of a serous cystadenoma. Notice the central enhancement. Sometimes differentiation from a hypervascular cystic neuroendocrine tumor can be difficult, but in this case the central calcifications are helpful. Scroll through the images. In the pancreatic tail is a cystic lesion with a central scar with calcifications (arrow). Eventhough some of the cyst are larger than 2 cm, this presentation still is typical for a serous cystic neoplasm, because of the central scar, multilocular appearance and the lobulated contour. This patient had abdominal complaints which were attributed to the tumor, which was resected and proved to be a serous cystic neoplasm. This is the resected specimen. The tumor was attached to the spleen, which also had to be resected. Another case of a typical serous cystic neoplasm. There is a microcystic lesion with a central scar in the pancreatic head. This patient felt a mass in her abdomen. Otherwise there were no complaints. Because resection would mean extensive surgery, it was decided to follow the lesion. During 5 year follow up there was no growth and the patient has no symptoms otherwise. key findings: CT-images of a 32 year-old female with pain in the upper left quadrant radiating to the back. There is a large cyst in the pancreatic tail with peripheral calcification. There is subtle septation as seen on the left image and wall thickening. You may have to enlarge the image to see the septation. A specific diagnosis of a MCN can be made. key findings: Macroscopic specimen of a IPMN showing mucinous tumor, with extensive mucin producing papilary neoplasm (arrow). On imaging Main-duct IPMN is distinct from branch-duct IPMN, but sometimes there is a mixed type. Scroll through the images of a large main duct and branch-duct IPMN. There is obstruction of the common bile duct with dilatation of the intrahepatic bile ducts (blue arrows). Notice the extremely widened main pancreatic duct (red arrow). Normal T2WI and heavily T2WI with fatsat of a large main duct IPMN with extremely dilated pancreatic duct. This patient presented with pancreatitis. The MRCP shows both a main-duct aswell as a branch-duct IPMN (arrow). IPMN is a lesion with malignant potential. Signs of malignancy are: CT-images of an IPMN with a dilated pancreatic duct (blue arrows). Notice enhancing solid nodule in the pancreatic head (red arrow). Continue with US-image. The US-image shows a large branch-duct component within the pancreatic head. The CT-image shows a hypodense lesion in the pancreatic head. This could be an adenocarcinoma, but the low density makes you think of a cystic tumor. The microcystic appearance raises the possibility of a serous cystic neoplasm although there is no calcified scar. On MRCP the cystic nature is better appreciated and there is a connection to a widened duct (blue arrow). A detail nicely demonstrates that some of the mucus-filled branches are seen in cross-section and some longitudinally. In a 73 year old male a hypoechoic lesion was found in the pancreatic body, that looked like a cystic lesion. CT also identifies the lesion but isn't of much help. The heavily T2WI nicely demonstrates the multicystic lesion with the connection to the pancreatic duct, i.e. a branch-duct IPMN. CT-images of a patient with a branch-duct IPMN who choose not to have surgery. Over time growth of the tumor is seen with dilatation of the main duct indicating malignant transformation. Sometimes it takes 5-8 years before a transformation is seen. key findings: CT-images of a 26 year old woman with a large mass in the pancreatic head and metastases in the liver. In the center there is lack of enhancement due to cystic or necrotic degeneration. key findings: CT-images of a 61 year old woman with weight loss. There is a large mass in the body of the pancreas that is hypervascular, unlike an adenocarcinoma, with some cystic or necrotic parts. CT-image of a neuroendocrine tumor with central necrosis. Sometimes this can simulate a cystic component. Notice the peripheral enhancement. by Stephen J. Handrich; David M. Hough; Joel G. Fletcher; Michael G. Sarr Am J Roentgenol. 2005;184(1):20-23 by Dushyant V. Sahani, MD et al. November 2005 RadioGraphics, 25, 1471-1484. by Bobby Kalb, MD, Juan M. Sarmiento, MD, David A Kooby, MD, N. Volkan Adsay, MD and Diego R. Martin, MD, PhD October 2009 RadioGraphics, 29, 1749-1765. AJR June 2011 196:W668-W677 by Sahani DV, Saokar A, Hahn PF, Brugge WR, Fernandez-del Castillo C. Radiology 2006; 238: ;919. by Michael Macari, MD et al. April 2009 Radiology, 251, 77-84. by Camilo Correa-Gallego, Cristina R. Ferrone, Sarah P. Thayer, Jennifer A. Wargo, Andrew L. Warshaw, and Carlos Fern?ndez-del Castillo Pancreatology. 2010 June; 10(2-3): 144-150. by Masao Tanaka et al Pancreatology Volume 12, Issue 3 , Pages 183-197, May 2012 by Dennis ZW Ng et al Ann Acad Med Singapore 2009;38:251-9 by de Jong K et al Clin Gastroenterol Hepatol. 2010 Sep;8(9):806-11 by Michael Macari, MD and Alec J. Megibow, MD, MPH Radiology 2011, 259, 20-23.Marc Engelbrecht, Jennifer Bradshaw and Robin Smithuis Classification Systematic Approach MRI versus CT How to report Management Age and gender Main-duct IPMN Branch-duct IPMN Solid Pseudopapillary Neoplasm Neuroendocrine tumor with cystic degenerationPancreas - Cystic LesionsRadiology department of the Academical Medical Centre, Amsterdam and the Rijnland hospital in Leiderdorp, the Netherlands
abdomen21
1
Peritoneum and Mesentery - Part II - Pathology
by Angela Levy
This review is based on a presentation given by Angela Levy and adapted for the Radiology Assistant by Robin Smithuis. We will discuss the differential diagnosis of cystic and solid peritoneal and mesenteric masses. In Peritoneum and Mesentery - part I: Anatomy the normal anatomy and physiology of the peritoneum and peritoneal cavity are discussed. You can click on images to get an enlarged view. by Angela Levy The first step when diagnosing peritoneal or mesenteric masses is to separate them into cystic and solid. Secondly we have to realize that any loculated fluid collection due to infection, i.e. abscess or as a result of pancreatitis, perforation or bile peritonitis can simulate a cystic mass. Especially fluid collections in the lesser sac can simulate a cystic mass. Lastly we have to know which cystic masses are common and look for specific features of these masses.   Mucinous carcinomatosis is the most common cystic tumor to affect the peritoneal cavity. Usually these metastases arise from mucinous carcinomas of the ovary or of the gastrointestinal tract (stomach, colon, pancreas). The prognosis is poor. However, when low-grade mucinous adenocarcinoma of the appendix spreads to the peritoneal cavity, the consequence is typically pseudomyxoma peritonei, which is a distinct tumor with a better prognosis. In peritoneal carcinomatosis we see tumor nodules along the peritoneal lining (arrow), omental tumor deposits, and bowel obstruction. Pseudomyxoma peritonei is the result of a mucinous adenocarcinoma of the appendix, which presents as a mucocele and spreads to the peritoneal cavity. It is a clinical syndrome, characterized by recurrent and recalcitrant voluminous mucinous ascites due to surface growth on the peritoneum without significant invasion of underlying tissues. A typical feature of pseudomyxoma peritonei is scalloped indentation of the surface of the liver and spleen. Unlike peritoneal metastases, there are no tumor nodules. There may be some calcifications. Pseudomyxoma peritonei (2) On the left another case of pseudomyxoma peritonei. There is hardly any scalloping of the liver. Notice the thickened falciform ligament. There is a mucocele of the appendix (arrow). This finding is only rarely seen. On the left another case of pseudomyxoma peritonei. There is compression of the mesentery resulting in a thickened cake-like hyperdense mesentery (arrow). There are also some calcifications. Pseudomyxoma peritonei is often confused with mucinous carcinomatosis. Unlike carcinomatosis it does not have true omental tumor deposits presenting as omental cake or peritoneal tumor deposits. Mesenteric cyst is a descriptive term for any cystic lesion within the mesentery. Usually it is a lymphangioma. Other mesenteric cysts like enteric duplication cyst, enteric cyst, nonpancreatic pseudocyst and mesothelial cyst are very uncommon and have no specific features. Lymphangioma is a benign lesion of vascular origin. Most lymphangiomas are located in the neck, but 5% of lymphangiomas are abdominal. Lymphangioma has enhancing septa. Unlike in cystic peritoneal metastases, ascites is not a feature of lymphangioma. When you see a septated cystic lesion without ascites the most likely diagnosis is a lymphangioma. Lymphangioma is often closely associated with the small bowel. At surgery it is usually very difficult to separate the tumor from the bowel and in many cases the bowel also has to be resected. The case on the left is also a lymphangioma. Notice that CT does not always appreciate the septations, although the specimen clearly shows multiple septations. Ultrasound or MR depict these septations better than CT. Enteric duplication cyst is a cyst with a wall that has all three layers of the bowel wall, i.e. mucosa, submucosa and muscularis propria. Although we commonly think of duplication cysts when we see a cystic mass adjacent to the bowel, we have to realize, that these are rare lesions. They may occur anywhere in the mesentery, so either adjacent to or away from the bowel. On the left an enteric duplication cyst. It is located in the transverse mesocolon. This patient was suspected of having a cystic pancreatic tumor. The specimen demonstrates all the bowel wall layers. Nonpancreatic pseudocyst is a residual of an old hematoma or infection. Most of these patients have a history of prior abdominal trauma. Often there is a thickened wall and there can be some debris within the lesion. The patient on the left had had a car accident eight months before. This is probably an old mesenteric hematoma as a result of a lap belt injury. Notice the thickened wall on the CT and the debris on the ultrasound. On the left a specimen and CT image of a nonpancreatic pseudocyst. Notice the thick wall. Probably this is an old hematoma or abscess. You can suggest this diagnosis when you have a positive history and you see this thickened wall or debris. These are also mesenteric cysts. They are rare and have nonspecific imaging features. The case on the left was diagnosed as a lymohangioma, simply because a lymphangioma is by far the most common mesenteric cyst. At surgery this was a mesothelial cyst. Also called Multilocular peritoneal inclusion cyst or Benign cystic mesothelioma. This is an uncommon benign primary peritoneal tumor that has no relation with the malignant mesothelioma. It occurs in premenopausal women with prior gynaecological surgery or infection that results in peritoneal scarring. The hormonally active ovaries secrete fluid that becomes loculated in the pelvis. The imaging features of a peritoneal inclusion cyst are non-specific except that it has to be located in the pelvis: On the left a transvaginal ultrasound demonstrating a multicystic pelvic lesion next to the uterus, which proved to be a peritoneal inclusion cyst. Sometimes the ovary is seen 'trapped' with the septate fluid collection (figure). Peritoneal inclusion cyst (2) When these peritoneal inclusion cysts become very large, they may extend into the upper abdomen as is seen in the case on the left. Notice that the left ovary is encircled by the cyst (arrow). There are also some enhancing septa. Peritoneal inclusion cyst (3) On the left another example of a peritoneal inclusion cyst. There is a nice correlation between the MR and the specimen. Peritoneal inclusion cyst (4) On the left images of a male patient, who presented with a lower abdominal mass. There is a multi-cystic mass extending from the pelvis along the right paracolic gutter to the upper abdomen. In a male patient this is a very uncommon diagnosis. These images look quite similar to images of a pseudomyxoma peritonei which was discussed before. In peritoneal inclusion cysts however, you will not see scalloping of the surface of the liver. TB can produce very thick ascites, that can be loculated in distribution. Because of this, it can simulate a cystic lesion.. Usually there is accompanying abnormality of the terminal ileum and lymphadenopathy. The lymph nodes most often are of low attenuation (caseated). So these are the things to look for. On the left another image of the same patient. In TB the peritoneum is usually very thick (arrow). It is unusual for an echinococcal cyst to be located in the peritoneum. It favors the liver, the spleen and even the kidney over the peritoneum. On the left we see cysts in the peritoneum and in the spleen. Notice the daughter cysts as small dark lesions within the large peritoneal cyst (arrows). In the pelvis additional echinococcal cysts are seen. Peritoneal metastases are the most common peritoneal solid masses. Gastrointestinal and ovarian cancers are the most common etiologies. Usually there are omental metastases, i.e. omental cake and ascites. On the left a CT demonstrating omental cake in a patient with ovarian cancer. On the left a patient with a lung carcinoma. This solitary solid mass was found in the pelvis. Based on the history this was suspected to be a metastasis. Other solitary solid tumors like ... can have the same presentation. Biopsy is needed to make the diagnosis. This proved to be a metastasis of the lung carcinoma. NHL is the most common cause of lymphadenopathy. Usually there are other sites with lymphoma. The CT attenuation at diagnosis is very homogeneous in most cases with minimal to no enhancement. Heterogeneous attenuation is seen only in cases with aggressive histology. During treatment the attenuation becomes heterogeneous as a result of necrosis and fibrosis. Calcification may occur. Carcinoid is a slow-growing neuroendocrine tumour most commonly found in the small bowel. Less than 10% of patients with carcinoid will develop the carcinoid syndrome, caused by the overproduction of serotonin, which can lead to symptoms of cutaneous flushing, diarrhea and bronchoconstriction.
 Carcinoid metastasizes to the mesentery, which at times is easier to appreciate than the primary tumor in the small bowel. There is associated bowel wall thickening due to a desmoplastic reaction. On the left a patient with a typical carcinoid with central calcification (blue arrow).
 Notice the bowel retraction and wall thickening. There is a metastasis in the liver (yellow arrow). On the left another patient with a carcinoid. The right image is the octreoscan, which is positive in 85% of carcinoids, so this can be a great help in the differential diagnosis. In this case we can also detect liver metastases on the scan (blue arrows). Notice that there is no activity of a primary tumor in the small bowel. This is often the case because the primary tumor can be quite small. Primary small bowel tumors can extend into the mesentery and the typical example of that is the GIST. You can have a large mesenteric component and such a small attachment to the bowel, that you may not appreciate it. On CT they are of mixed density due to necrosis and hemorrhage and they tend to be well vascularized, so they will enhance like the case on the left. This disease can affect lung, orbit and mesentery. Inflammatory pseudotumor is a diagnosis by exclusion. Usually the diagnosis is made at surgery or biopsy. It is the result of chronic inflammation with an unclear pathogenesis. Probably it is an occult infection due to minor trauma or post surgical. Mesenteric fibromatosis is also known as intra-abdominal fibromatosis, abdominal desmoid or desmoid tumor. On the left a 33-year-old man who complains of an increasing abdominal girth, abdominal fullness, and a palpable abdominal mass. First study the images on the left and continue with the MR. Look for some imaging features that are helpful in the differential diagnosis. First of all this is a well circumscribed lesion with a low density on CT. There is some enhancement around the lesion and there are some small strands of enhancement within the lesion. On MR there is a low signal on T1 as we would expect. On T2 there is relatively high signal. In combination with the low density on CT this tells us that there is mucin within the lesion. This finding is very suggestive of the diagnosis of mesenteric fibromatosis. Mesenteric fibromatosis - Desmoid (2) The enhancement on MR is more intense compared to the enhancement on CT. On CT the low density of the mucin stands out, but on MR we can appreciate the enhancement better. It tells us that the lesion is well vascularized. Mesenteric fibromatosis or desmoid is a benign proliferative process that is locally aggressive and can recur, but it does not metastasize. The small bowel mesentery is the most common site. 13% of patients have familial adenomatous polyposis (FAP). On the left images of another patient with mesenteric fibromatosis. Notice that this lesion is not of low attenuation. This lesion has a more collagenous or fibrous stroma. So there are two distinct patterns. Mesenteric fibromatosis - Desmoid (3) On the left again a more typical case with a low density tumor located in the greater omentum (upper image) and the gastrosplenic ligament (lower image). On the left an unusual location, because normally there is no mesentery deep in the pelvis. This patient had familial adenomatous polyposis. A total colectomy with J-pouch of the ileum was performed. Now accompanying that J-pouch is mesentery in which mesenteric fibromatosis has developed. Notice the low attenuation foci or bands of myxoid stroma within the tumor. In familial adenomatous polyposis the mesenteric fibromatosis is almost always post operative and occurs at the operative sites. It frequently occurs at multiple sites including abdominal wall fibromatosis. These cases can be very aggressive. It usually comes back and when it does, it comes back as a more aggressive tumor. Therfore these patients are treated as conservatively as possible. This disease has multiple synonyms reflecting the wide histologic spectrum: mesenteric panniculitis, fibrosing mesenteritis and mesenteric lipodystrophy. Pathologically it is a chronic inflammation of unknown etiology. This entity is more common than previously thought. The signs and symptoms are variable. Patients present with pain, a palpable mass or bowel complications, but in many cases it is an incidental finding on CT made for other reasons. The image on the left is the form that we most frequently see in patients that are screened for other reasons. This form is mostly named panniculitis mesenterialis. In a more advanced stage you can have significant fibrosis resulting in retraction of the small bowel. Within these masses dystrophic calcifications can be seen as well as lucent areas of fat (arrow) Sclerosing mesenteritis (2) On the left a nice radiological pathological correlation. Notice the retraction of the bowel and also notice the resemblance to carcinoid. In these cases the octreoscan can be a great help in the differential diagnosis. These lesions are situated in the root of the mesentery and this makes a surgical procedure extremely difficult. These lesions are treated conservatively with immunosuppressiva, anti inflammatory drugs and sometimes anti-estrogens as long as possible. Malignant mesothelioma is one of the primary peritoneal malignancies (Table on the left). Suggestive features are a sheet-like peritoneal thickening and absence of lymphadenopathy. Just like pleural mesothelioma, it is associated with asbestos exposure. On the left a patient with malignant mesothelioma. Notice the sheet-like thickening of the peritoneum. The diagnosis was suggested because of the pleural calcifications. Malignant mesothelioma (2) In advanced cases you will see encasement of the intra-peritoneal structures. In the case on the left there is besides encasement of the bowel and the liver, also encasement of the mesentery. On the left a rad-path correlation of the same case. This tumor is also one of the primary peritoneal malignancies. It occurs exclusively in women. This tumor is histologically identical to malignant ovarian surface epithelial tumors. It was once thought to be very rare, but now almost one third of tumors previously diagnosed as ovarian cancer are diagnosed as primary peritoneal serous carcinoma. Consider this diagnosis when: As a radiologist you should consider this diagnosis if you think of metastatic ovarian cancer but the ovaries are normal. On the left a typical case. There is ascites and omental involvement, so your first thought is ovarian cancer, but the ovaries were normal. This tumor is also one of the primary peritoneal malignancies. It is a rare malignancy of uncertain origin. It occurs primarily in young men with a mean age of 19 years. Consider this diagnosis if you see something that looks like peritoneal carcinomatosis in a young man that has no history of a primary malignancy. It is a very aggressive tumor with a poor prognosis. Desmoplastic small round cell tumor begins as a dominant mass and then multiple masses occur within the peritoneum (figure). At this stage it is no different upon imaging to other tumors, however, the age of the patient provides the clue to the diagnosis. NHL would be number one in the differential diagnosis of these images. by Angela D. Levy, COL, MC, USA, Javier Arn?iz, MD, Janet C. Shaw, Lt Col, USAF, MC and Leslie H. Sobin, MD by Angela D. Levy, COL, MC, USA, Janet C. Shaw, Lt Col, USAF, MC and Leslie H. Sobin, MD RadioGraphics March 1, 2009 29:347-373 Expert Differential Diagnoses: Abdomen: Published by Amirsys? by Michael FederleAngela Levy Cystic masses Solid masses Mucinous Carcinomatosis Pseudomyxoma peritonei Mesenteric cyst - Lymphangioma Enteric Duplication Cyst Nonpancreatic Pseudocyst Enteric cyst and mesothelial cyst Peritoneal Inclusion Cyst Tuberculosis Echinococcal Cyst Peritoneal metastases Lymphoma Carcinoid Gastrointestinal Stromal Tumor - GIST Inflammatory Pseudotumor Mesenteric fibromatosis - Desmoid Sclerosing Mesenteritis Malignant mesothelioma Primary Peritoneal Serous Carcinoma Desmoplastic Small Round Cell TumorPeritoneum and Mesentery - Part II - PathologyChief Gastrointestinal Radiology, Department of Radiologic Pathology, Armed Forces Institute of Pathology, Washington DC Associate Professor of Radiology, Uniformed Services University of the Health Sciences, Bethesda, MD
abdomen22
1
Peritoneum and Mesentery - part I: Anatomy
by Angela D. Levy MD
This review is based on a presentation given by Angela Levy and adapted for the Radiology Assistant by Robin Smithuis. We will discuss the normal anatomy and physiology of the peritoneum and peritoneal cavity. In part II we will discuss peritoneal tumors. The illustrations are by Heike Blum Images can be enlarged by clicking on them. by Angela Levy The peritoneum is a serosal membrane, which is composed of a single layer of flat mesothelial cells supported by submesothelial connective tissue. In this subserosal tissue there are fat cells, lymphatics, blood vessels and inflammatory cells like lymphocytes and plasma cells. The visceral peritoneum lines all the organs that are intraperitoneal. The parietal peritoneum lines the anterior, lateral and posterior walls of the peritoneal cavity. The deepest portion of the peritoneal cavity is the pouch of Douglas in women and the retrovesical space in men, both in the upright and supine position. The mesentery is a double fold of the peritoneum. True mesenteries all connect to the posterior peritoneal wall. These are: Specialized mesenteries do not connect to the posterior peritoneal wall. These are: If you remove all of the intraperitoneal bowel, you get a good look at the cut-surface of the mesenteries: Notice that the small bowel mesentery has an oblique orientatien from the ligament of Treitz in the left upper quadrant to the ileocecal junction in the right lower quadrant. These compartments enable the peritoneal cavity to have a normal circulation for peritoneal fluid. In the normal abdomen without intraperitoneal disease, there is a small amount of peritoneal fluid that continuously circulates. The movement of fluid in this circulatory pathway is produced by the movement of the diaphram and peristalsis of bowel. It predominantly flows up the right paracolic gutter which is deeper and wider than the left and is partially cleared by the subphrenic lymphatics. There are watershed regions in the peritoneal cavity that are areas of fluid stasis: When you are staging a patient for gastrointestinal malignancy you have to look for disease in these areas of stasis. Clearly the surgeons do better in finding subtle disease in these areas. 90% of peritoneal fluid is cleared at the subphrenic space by the submesothelial lymphatics. These lymphatics are connected with lymphatics at the other side of the diafragm. The peritoneum is continuous in the male pelvis. In women the peritoneum is discontinuous at the ostia of the oviducts. Through this opening disease can spread from the extraperitoneal pelvis into the peritoneal cavity. For example, pelvic inflammatory disease (PID). The omentum is divided into the greater and lesser omentum. The greater omentum is subdivided into: The lesser omentum is subdivided into:Angela D. Levy MD Peritoneum Mesenteries Peritoneal circulation OmentumPeritoneum and Mesentery - part I: AnatomyChief Gastrointestinal Radiology, Department of Radiologic Pathology, Armed Forces Institute of Pathology, Washington DC Associate Professor of Radiology, Uniformed Services University of the Health Sciences, Bethesda, MD
abdomen23
1
Rectal Cancer - MR imaging
by Max Lahaye, Regina Beets-Tan and Robin Smithuis
The major advancement in the treatment of rectal cancer is total mesorectal excision (TME), which involves complete removal of the tumor along with the mesorectal tissue which contains the lymphatics. The other advancement is the shift from adjuvant to neoadjuvant radiotherapy. Both have dramatically changed the local recurrence rates and survival. The issue is whether a patient with rectal cancer is candidate for TME only or preoperative chemoradiotherapy followed by TME. MRI can answer that question, since it is the most accurate tool for the local staging of rectal cancer. Traditionally rectal cancer surgery consisted of excision of the tumor with a margin of surrounding perirectal fat. This however resulted in high local recurrence rates up to 40%. In 1982 the surgeon Richard John Heald introduced the total mesorectal excision. After many years TME was widely accepted, which caused a drop in local recurrence rates from 40% to 11% (1,2). The role of MRI is to determine whether TME-surgery is possible or whether there is an advanced tumor that should be treated with chemoradiation and followed by TME in a later stage. Total mesorectal excision (TME) is the best surgical treatment for rectal cancer provided that the resection margin is free of tumor. In TME the entire mesorectal compartment including the rectum, surrounding mesorectal fat, perirectal lymph nodes and its envelope, i.e. the mesorectal fascia is completely removed (figure). This minimizes the chance of tumor remnants in the surgical bed. On the left a coronal illustration of the rectum with a tumor extending through the rectal wall into the mesorectal fat and with some lymph nodes. The resection margin along the mesorectal fascia is free of tumor and a TME can be performed. Notice the anal verge (blue arrow). On MRI the mesorectal fat has a high signal intensity on T1- and T2-weighted images. The mesorectal fat is bounded by the mesorectal fascia, which is seen as a fine line of low signal intensity (red arrows). In a TME the mesorectal fascia is the resection plane. The shortest distance from the tumor or lymph nodes to the mesorectal fascia is called the cirumferential resection margin (CRM). It is the most powerful predictor for local recurrence. MR is highly accurate for the prediction of the CRM. A positive resection margin or CRM+ is when there are tumor deposists within 1 mm of the fascia. CRM- is when the distance to the fascia is > 1 mm. On the left an illustration with: Whenever there are lymph nodes within 1 mm of the mesorectal fascia we need to report this , especially when they are large, because the CRM may be involved (blue arrow). MRI has to determine the following: The rectum extends from the anorectal junction to the sigmoid. The rectosigmoid junction is arbitrarily defined as 15 cm above the anal verge. A tumor more than 15 cm above the anal verge is regarded and treated as a sigmoid tumor. Since we cannot detect the anal verge on MR, it is best to measure from the anorectal angle. Rectal cancer can be divided into: Low rectal cancer has a higher local recurrence rate. The distal tapering of the mesorectal fat implies that low rectal cancer more easily invades the surrounding structures and it will be more difficult for the surgeon to get a tumor free resection (see figure). When we know the exact location of the tumor, the next step is to determine the T-stage. MR cannot distinguish tumor growth limited to the submucosa or invasion to the muscularis externa, so it cannot differentiate between T1 and T2 tumors. In most cases these tumors are both treated with TME-surgery, so it is not necessary to make the difference. In a minority of cases a T1 tumor will be treated with local excision. In these cases endorectal US is accurate for staging these superficial tumors. T1 and T2 tumors are limited to the bowel wall and have a good prognosis. They can be accurately identified with MR, because the rectal wall will have an intact black line, i.e. musclaris externa, surrounding the tumor (3). On the left a rectal tumor that is completely surrounded by the black layer of the musculari externa. This is a T2 tumor. A T3 tumor grows through all wall layers and extends into the perirectal fat tissue. In these tumors it is important to determine whether the circumferential resection margin is involved. On the left a tumor that probably infiltrates the mesorectal fat, i.e. T3 (arrow). There is a wide resection margin around the tumor and there are no lymphnodes adjacent to the mesorectal fascia. This tumor is classified as T3 CRM-. In the Netherlands, like in most european countries, this patient will be treated with a short preoperative course of radiotherapy followed by TME. Perirectal stranding MRI has a sensitivity of 82% to detect perirectal tissue invasion. The pitfall is when perirectal stranding is seen. This can be the result of tumor ingrowth or a desmoplastic reaction. To be on the safe side and to avoid understaging, it is advised to stage tumors with perirectal stranding as T3 tumors. On the left two tumors with a similar MR-appearance. In A there was perirectal tumor invasion. In B there was a tumor limited to the bowel wall, i.e. a T2-tumor. The perirectal stranding in this latter case was the result of a desmoplastic reaction. For therapeutic purposes it does not have any consequences to differentiate accurately between a T2 CRM- and a T3 CRM- tumor. Both tumors will be treated with a preoperative low dose radiotherapy of 5x5 Gy followed by TME. On the left a tumor that infiltrates the mesorectal fat with infiltration of the resection margin on the anterior side (arrow). This tumor is classified as T3 CRM+. This patient will be treated with a long course of radiotherapy and chemotherapy preoperatively. If this treatment is succesful, it will be followed by TME. A T4 tumor is an advanced tumor that invades surrounding structures like pelvic wall, vagina, prostate, bladder or seminal vesicles. These patients require a long course of chemoradiation and extensive surgery. For adjacent organ invasion all imaging modalities show similar sensitivity: 70% for transrectal US, 72% for CT and 74% for MR imaging. On the left a T4-tumor with invasion of the prostate. The N-stage is an important risk factor for local recurrence. Unfortunately MR, like any other imaging modality, has a low sensitivity and specificity for the detection of lymph node metastases. When lymph nodes have a short axis of > 5 mm or a spiculated and indistinct border or a mottled heterogeneous appearance, than you can be sure that these nodes contain metastases. However not all positive lymph nodes meet these criteria. Even in T1 and T2 tumors there is a considerable risk for lymph node metastases (Table). The low sensitivity using only size criteria can be explained by the fact that in rectal cancer small lymph nodes still have a high prevalence of malignancy, 9% in 1-2 mm nodes and 17% in 2-5 mm nodes respectively (11). As is demonstrated in the table on the left the majority of positive nodes are 1-5 mm in size. In order not to understage patients, all lymph nodes are regarded as possible malignant. On the left a low rectal cancer with multiple nodes in the perirectal fat on the posterior side. This has a big influence on the prognosis of the patient and based on the advanced stage of the tumor with CRM+ and N+ status, the patient will therefore receive a more aggressive treatment with neoadjuvant chemoradiation. It is important to look beyond the mesorectum for lymph nodes (arrow). These extramesorectal nodes are important, because they can be a cause of local recurrence. When they are detected by MR, the radiation and surgical planning has to be adapted. On the left a patient with extramesorectal nodal recurrence after TME (arrow). During a standard TME procedure these extramesorectal lymph nodes will not be resected. This means that after TME surgery not all tumor deposits will have been removed. The finding of malignant extramesorectal lymph nodes entails that at least a more extensive surgical approach is necessary to remove all the cancer deposits or a boost of radiotherapy to the areas of risk. If not, a nodal recurrence, as shown here, is imminent. On the left axial T2-weighted images of two different rectal cancer patients. These cases illustrate the problems for a radiologist to accurately stage the nodal status. On the far left there is a small extra mesorectal lymph node depicted. On the right there are numerous large mesorectal lymph nodes and also a right extramesorectal lymph node with indistinct borders (red arrow). Although the nodes of these two patients have very different characteristics in size, border and heterogenous appearance, they all proved to be malignant. Rectal cancer is notorious for small nodes ( Only FSE T2 and no Gadolinium The only sequence that is required is a T2-weighted fast spin echo sequence. Gadolinium-enhanced MR does not improve diagnostic accuracy and is therefore not included in the protocol. Images are made in the sagittal, coronal and axial plane. First start with the sagittal series and plan the axial images perpendicular to the rectal wall at the level of the tumor (blue lines). Coronal images are planned perpendicular to the axial images (yellow line). In this way we avoid partial volume artefacts and will be able to accurately evaluate the depth of tumor invasion. It helps if the level of the tumor position is indicated by the referring surgeon for proper planning of the MR-sequences. The cranial border of the field of view (FOV) is L5, the caudal border is below the anal canal. Angulation Axial images have to be abgulated perpendicular to the axis of the tumor to avoid volume averaging. At first the axial images were not properly angulated. This resulted in the false impression, that the circumferential resection margin was involved on the anterior side (red circle). After proper angulation it was noted, that the CRM was not involved (yellow circle). No fat suppression and no bowelpreparation Fat suppression is not helpfull in delineating the tumor. Patients do not need bowel or any other preparation. The use of rectal contrast is not advised, because stretching of the bowel wall may lead to overestimation of an involved CRM. Furthermore the mesorectal nodes in the distal mesorectum are not well appreciated. The radiological report must consist of the following tumor variables: There are differences in rectal cancer treatment between countries and between institutions. Everyone agrees that TME is the best radical treatment for all tumors with free resection margins. In the Netherlands, like in most european countries, a short course of 5x5Gy radiotherapy is given to the majority of patients prior to TME, because additional benefit was seen in the large TME trial. In some institutions this preoperative radiotherapy is not given to tumors that already have a good prognosis, like high-rectal T1N0 and T2N0. Preoperative short course of radiotherapy immediately followed by TME does not result in down-staging and is therefore not suitable for locally advanced tumors. That is why all T4-tumors or tumors with involved resection margins and tumors with suspicious malignant lymph nodes near the resection margin and beyond the first receive high dose chemoradiation. Further action depends on the response to the preoperative treatment. In cases of tumor regression from the mesorectal fascia this will be a less extensive resection. In cases of tumor downstaging to only a small tumor remnant and sterilized nodes (yN0) it can be a local excision. In cases of complete tumor regression and yN0 it can even be a 'wait and see' with omission of surgery. This latter option however is still controversial and not standard practise. case 1 Radiology report: Conclusion: Low-rectal T3N0 tumor with an involved CRM at the pelvic floor muscles on the right side. case 2 Radiology report: Conclusion: Midrectal T3N0 tumor close to the peritoneal fold. case 3 Radiology report: Conclusion: High-rectal T3N2 tumor with an involved CRM on the right ventral side. Recurrence and survival after total mesorectal excision for rectal cancer. Heald RJ, Ryall RD. Lancet 1986; 1:1479- 1482. by R. J. Heald, E. M. Husband, R. D. H. Ryall British Journal of Surgery Volume 69, Issue 10, pages 613-616, October 1982 by Sagar PM, Pemberton JH Br J Surg 83:293-304, 1996 by Regina Beets-Tan and Geerard Beets August 2004 Radiology, 232, 335-346. Accuracy of magnetic resonance imaging in prediction of tumour-free resection margin in rectal cancer surgery. by Beets-Tan RG, Beets GL, Vliegen RF, et al Lancet 357:497-504, 2001 by Bipat S, Glas AS, Slors FJ, Zwinderman AH, Bossuyt PM, Stoker J Radiology 232:773-783, 2004 Brown G, Richards CJ, Bourne MW, et al Radiology 227:371-377, 2003 by Kim JH, Beets GL, Kim MJ, Kessels AG, Beets- Tan RG. Eur J Radiol 2004; 52:78-83 by Fiona Taylor et al AJR 2008; 191:1827-1835 by Kapiteijn E, Marijnen CA, Nagtegaal ID, et al N Engl J Med 2001; 345:638-646 Wang C, Zhou Z, Wang Z, et al Langenbecks Arch Surg. 2005;390:311;318Max Lahaye, Regina Beets-Tan and Robin Smithuis Total mesorectal excision Circumferential resection margin Low rectal cancer T1 and T2 T3 CRM- T3 CRM+ T4Rectal Cancer - MR imagingRadiology Departement of the Maastricht University Hospital and the Rijnland Hospital in Leiderdorp, , the Netherlands
abdomen24
1
Rectum - Dynamic examination
by Tjeerd Wiersma
Dynamic rectal examination (DRE) is also known as defecography or proctography. DRE provides a dynamic assessment of the act of defecation by recording the rectal expulsion of a barium paste that approximates the consistency of feces. DRE provides qualitative and quantitative information on the function of anorectal and pelvic floor function, and the effectiveness of the anal sphincter and rectal evacuation. by Tjeerd Wiersma Indications for dynamic rectal examination are: Two hours prior to the examination the patient ingests 135 ml of liquid barium contrast to opacify the small bowel (figure). Rectal contrast can be administered without rectal preparation. The ideal rectal contrast has to simulate stool in weight and consistency. In our experience, Evacu-Paste? 100 (E-Z-EM Inc., Westbury, NY, USA) is a convenient paste. The barium paste is injected until the patient experiences discomfort due to rectal distension or until about 250 ml has been instilled. In females the vagina is coated with 30 ml amidotrizoic acid 50% solution gel. It is applied by means of a syringe with a soft pediatric enema tip. The use of tampons and gauzes soaked in barium should be avoided, because they can impaire pelvic-function. After sufficient filling of the rectum, the patient is asked to sit on a special commode. The fluoroscopic screening of the rectum and the function of the pelvic musculature and the continence mechanism is assessed. The duration of examination is about 15 minutes. Examinations includes a number of standard images and maneuvers. Initially the patient is screened in lateral projection at rest without consciously contracting any pelvic muscles and a spot film is taken. The patient then maximally contracts the pelvic floor muscles ('squeeze') which results in a more tense muscular diaphragm and in elevation of the entire pelvic floor a spot film or video is taken. Finally the patient is asked to empty the rectum as completely as possible. An estimate of the completeness of defecation and measurement of pelvic floor descent can be made. Morphological abnormalities are usually discernible during this part of procedure. It is important that the patient is sitting during the procedure, since much of physiological nature of defecation is lost when the patient is lying down. Additional oblique or anteroposterior (AP) views should be taken of any unexplained radiographic feature seen on the lateral views (figure). An S-shaped rectum may simulate an intussusception in lateral projection. On the left lateral and an AP-view of a patient with an intussuseption on both views. On the left lateral and an AP-view of a patient with an S-shaped rectum which simulates an intussuseption on the lateral view. The whole procedure of DRE should be recorded on video or DVD. Dynamic recording of the fluoroscopic images enables the examiner to follow the movements of the rectum, facilitating the diagnosis of rectocele, enterocele and intussusception, as well as to evaluate the function of the anal sphincter. Digitization of fluoroscopic image or digital substraction may facilitate direct screen measurements of angles. Some patients give a history of various unusual maneuvers (digital support of vagina or perineum) to aid defecation. Allowing the patients to demonstrate the maneuvers during the examination may facilitate the radiological documentation of the mode of action. Since a number of treatment methods, which successfully can restore the dynamics of rectal evacuation, have been developed speed and completeness of rectal emptying is clinically important and therefore need to be recorded. At rest: distance between vagina and ventral rectal wall A rectocele can be defined as an anterior or posterior bulge of the rectal wall beyond the extrapolated line of the wall (Fig. 1). The formation of an anterior rectocele is often apparent during defecation and may reflect relative weakness of the rectovaginal septum. At the end of the defecation, residual rectal contents may be left in the rectocele ('trapping'). Significantly more anterior rectoceles were found in female patients and in female control subjects than in males. Anterior rectoceles may occur in individuals without complaints of the anorectal region and should therefore particularly in women be considered as a posiible normal phenomenon. The main symptoms associated with a rectocele are usually a feeling of incomplete bowel movement often requiring digital pressure to the vagina or perineum to facilitate emptying, together with aching after a bowel movement. Barium trapping in the rectocele is considered to be important in explaining the repeated sensation of rectal fullness after defecation. In our own series no correlation could be found between the size of the rectocele and the symptoms, so we do not grade rectoceles anymore. It has been suggested that rectoceles may be the result of repeated straining secondary to a preexisting disorder (f.e. spastic pelvic floor syndrome) of defecation rather than to the rectocele being the primary cause of the obstructive symptoms. This may also explain why rectocele repair is often unsuccessful in relieving symptoms. When surgical repair of a rectocele detected at physical examination is considered, preoperative DRE should be performed to exclude other causes of obstructed defecation (intussusception or enterocele). DRE demonstration of a rectal intussusception may change the operative procedure e.g. to correct the intussusception instead of the rectocele. In patients with an anterior rectocele, in whom other causes of obstructed defecation are ruled out, surgical rectocele repair should be considered. In our opinion there are two indications for operating on an anterior rectocele. First: if a patient needs vaginal digital support to facilitate defecation. Second: in cases of disturbed sexual intercourse. Posterior rectoceles are incidental findings and not related to clinical symptoms (figure). Intussusception of the rectum is an invagination of the rectal wall, which begins as a circular fold 6 to 8 cm up in the rectum and develops into a condition in which the entire rectal wall folds in towards the rectal lumen. The intussusception can be intra-rectal, intra-anal or finally extra-anal as a rectal prolapse (figure). In connection with straining this folding inwards progresses and deepens to form a ring pocket, so that it finally fills the entire ampulla. This may reach down to, into or through the anal canal (rectal prolapse). A minimal folding inwards which disappears after the bolus has passed is probably caused by a transient prolapse of the rectal wall and should not be considered pathological. The most common complaint of the patient with intussusception is: - Difficulty in bowel emptying. - Pain, blood loss upon defecation - Incontinence of gas and/or feces - Mucus discharge which often leads to pruritis ani. Upon hard straining the obstructive sensation increases. In order to empty their bowels many patients have to extract the feces manually, while others have to exert pressure with their hands about the anus and perineum. Enemas may be ineffective. Rectal prolapse (extra-anal intussusception) may be transient and difficult to reproduce, while intrarectal intussusception may be overlooked on clinical examination and is seldom revealed by proctoscopy. When the intussusception reaches into the anal canal it leads to maximal dilatation. These patients often complain of fecal incontinence in between defecation while they have the feeling of an obstacle or incomplete emptying during defecation. A longstanding intussusception may lead to the solitary rectal ulcer syndrome. There is seldom doubt regarding the clinical diagnosis of complete rectal prolapse. Lesser grades of prolapse, however, can present a variety of difficulties. Oblique or anteroposterior (AP) views can be necessary too, due to the AP view provides a more reliable image of intra-anal intussusception than the standard lateral view during evacuation. An enterocele is a peritoneal sac that has herniated downwards along the ventral rectal wall. As DRE is routinely performed with small bowel and vaginal contrast, loops of small bowel are then seen to fill the gap between the vagina and the rectum. Grade 1 is maximally reaching down to the distal half of the vagina, and partial or complete reduction of the rectal lumen. Grade 2 is as grade 1, but reaching down to the perineum. Grade 3 is protruding out of the anal canal to form a rectal prolapse. Sometimes the enterocele is identified only at the end of the defecation, after repeated straining. An enterocele may be pressed into the direction of the introitus vaginae. If there is an associated rectocele, this can be pushed downward by the enterocele and finally evacuated (figure). Clinically it can be difficult to diagnose an enterocele. Patients with previous pelvic surgery are predisposed to the formation of an enterocele. In female patients with constipation there is a higher incidence of severe enteroceles in patients with a hysterectomy (22%) compared to the group without hysterectomy (9%). Chronically increased intra-abdominal pressure may cause an enterocele with or without a previous pelvic operation. A sigmoidocele is a prolapse of redundant sigmoid colon into a deep pouch of Douglas. It is less common than an enterocele. On the left a schematic lateral view on the levator ani and external sphincter ani muscles is shown. The puborectal muscle should be contracted at rest (sharp anorectal angle). During defecation the puborectal muscle should relax allowing passage of the stool. Spastic pelvic floor syndrome denotes a persistent contraction of the pelvic floor muscles during defaecation. It represents a functional disorder of the pelvic floor muscles causing an outlet obstruction. The question arises, however, whether persistent contraction is due to the conscious action of an embarrassed patient, thus only occurring during the investigation, or whether it really represents a functional disorder of the pelvic floor muscle resulting in outlet obstruction. The etiology is unknown. Psychological factors may play a role. The anorectal angle (ARA) normally increases on straining as a result of relaxation of the puborectal muscle. The extent of increase may range from 20? to 40?. In a small group of patients with impaired evacuation, DRE demonstrated either an unchanged or decreased ARA on straining or defecation, which findings frequently result from a persistent or paradoxical increase of the puborectal muscle impression. This appearance is often quite persistent and the evacuation may only be achieved after multiple attempts at straining and defecation. Stoker J, Rociu E, Wiersma TjG, Lam?ris JS (2000) Imaging of anorectal disease. Br J of Surg 87; 10-27. Wiersma TjG (1998) Functional anorectal abnormalities: Evaluation with dynamic rectal examination. In: Shahid M Hussain (ed) Imaging of anorectal diseases, chapter 8, pp 104-121, ISBN 1 900 151 367. Wiersma TjG, Mulder CJJ, Reeders JWAJ (1997) Dynamic Rectal Examination: Its significant clinical value. Endoscopy 29: 462-471. Wiersma TjG, Werre AJ, Hartog den G, Thies JE, Tytgat GNJ, Mulder CJJ (1997). Hysterectomy: The anorectal pitfall. A guideline for evaluation. Dynamic Rectal Examination (defecography) (1994).Wiersma TjG, Mulder CJJ, Reeders JWAJ, Tytgat GNJ, Waes PFGM. In: Balli?re Clinical Gastroenterology, Vol. 8, no. 4, chapter 10. ISBN 0-7020-1855-4. Kelvin FM, Maglinte DDT, Hornback JA, Benson JT. Pelvic prolapse: assessment with evacuation proctography (defecography). Radiology 1992; 184: 547-51.Tjeerd Wiersma Indications Technique Recording Rectocele Intussusception Enterocele Spastic pelvic floor syndromeRectum - Dynamic examinationRadiology department of the Rijnstate Hospital, Arnhem, The Netherlands.
abdomen25
1
Rectum - Perianal Fistulas
by Susanne Tonino and Robin Smithuis
Perianal fistula is a common disorder that often recurs because of infection that was missed at surgery. Preoperative MR can help to prevent recurrence. In this review we will address the anatomy, pathogenesis, classification and scanning protocol of perianal fistulas. by Susanne Tonino and Robin Smithuis The anatomical anal canal extends from the perineal skin to the linea dentata. Surgically, the anal canal extends from the perineal skin to the anorectal ring. This is the circular upper border of the puborectal muscle which is digitally palpable upon rectal examination. The anorectal ring lies approximately 1-1,5 cm above the linea dentata. The total length of the surgical anal canal is about 4-5 cm. The anal sphincter is comprised of three layers: The puborectal muscle has its origin on both sides of the pubic symphysis, forming a 'sling' around the anorectum. The puborectal muscle is contracted at rest and accounts for the 80? angulation of the anorectal junction. It relaxes during defecation. On axial and coronal MR-images the different layers of the anal sphincter and the surrounding structures can be displayed perfectly. A perianal fistula is an abnormal connection between the epithilialised surface of the anal canal and the skin. The causes of perianal fistulas: The most widely used classification is the Parks Classification which distinguishes four kinds of fistula: intersphincteric, transsphincteric, suprasphincteric and extrasphincteric. The most common fistulas are the intersphincteric and the transsphincteric. The extrasphincteric fistula is uncommon and only seen in patients who had multiple operations. In these cases the connection with the original fistula tract to the bowel is lost. A superficial fistula is a fistula that has no relation to the sphincter or the perianal glands and is not part of the Parks classification. These are more often due to Crohns disease or anorectal procedures such as haemorrhoidectomy or sphincterotomy. Protocol A localizer in three directions is needed in order to align the T2 sequences axial and coronal to the anal canal. Any localizer that properly displays the anal canal can be used. We use a TRUE FISP, which is the name that Siemens uses for a steady-state precession gradient-echo sequence (GE: FIESTA, Philips: balanced FFE). T2W images without fatsat better display the anatomy, while the fatsat images better depict the fistulas. Reporting When you describe a fistula, it is important to mention the following characteristics: The drawing on the left illustrates the anal clock, which is the surgeon's view of the perianal region when the patient is in the supine lithotomy position (2). This scheme corresponds to the orientation of axial MR images of the perianal region. On the left axial T2W images with and without fat saturation. An intersphincteric fistula is located at 6 o'clock. Continue with coronal images. On the coronal image the fistula runs caudally towards the skin. There is no connection with the external sphincter. On the left coronal images of another patient with an intersphincteric fistula. Use the arrows to scroll through the images. On the left an axial T2WI and T2WI + fatsat of a transsphincteric fistula. The defect through the internal and external sphincter at 6 o'clock is clearly visible and more apparent on the fat sat images. On the left axial T2W-fatsat images of a transsphincteric fistula with the mucosal opening at 11 o'clock. On the left an example of a suprasphincteric fistula. There are two tracts in the ischioanal region. The right sided tract runs over the puborectal muscle (asterisk) and the mucosal opening lies at the level of the linea dentata (black arrow). On the left coronal T2W-images of a small abscess in the left ischioanal fossa, the fistula runs through the levator ani. It is therefore above the sphincter complex and extrasphincteric. On the left an example of a complex fistula. Two tracts in the left buttock form a single tract (no. 1-2). This fistula breaks through the external sphincter (no. 4). In the intersphincteric space it divides again into two tracts (no. 5). One ends blindly in the intersphincteric space (no. 6). The other breaks through the internal sphincter with the mucosal defect at 1 o'clock. On the left a patient with a perianal fistula who has Crohn's disease. Continue with the coronal images. On the coronal images the thickening of the bowel wall is demonstrated. Axial fatsat images depict the transmural inflammation with infiltration of the mesenteric fat. Treatment is focussed on the elimination of the primary and secondary tracts, prevention of recurrence and to retain continence. The treatment given depends on the anatomy of the fistula, if it is a simple fistula with a low mucosal defect is can be probed in the OR to identify the mucosal defect at the linea dentata, then the tract can be opened. This is only possible if the external sphincter is not involved. Seton fistulotomy is a technique where a rubber ligature or vessel loop is pulled through the fistula, it then is tightened every 2 weeks or so in order to obtain pressure necrosis so that the Seton is slowly pulled through the muscle. This has the advantage that the muscle is slowly cut and fibroses at the same time in order to cause as little damage as possible to the sphincter complex. If there is an extrasphincteric fistula, the lower part is opened. The mucosal defect, which is by definition in the rectum, is then surgically closed. This patient was already known to have an intersfincteric fistula, the mucosal defect is at 1 o'clock. In the tract there is a linear structure with a low signal intensity. This is the Seton which was inserted to treat the fistula. Sinus pilonidalis On the left an example of a sinus pilonidalis. There is a small abscess just above the nates. There is no relation with the sphincter complex. Proctitis On the left images of a patient who presented with anal complaints. No fistula was seen. There is, however, a diffuse thickening of the rectal mucosa due to a proctitis. Abscess in the Ischioanal space An abscess in the ischioanal space with no connection to the sphincter complex Halligan, Stoker et al. Radiology 2006;239:18-33 by John Morris Radiographics. 2000;20:623-635 by Karin Horsthuis and Jaap StokerSusanne Tonino and Robin Smithuis Classification Intersphincteric fistula Transsphincteric fistula Extrasphincteric fistula Complex fistula Crohn's diseaseRectum - Perianal FistulasRadiology Department of the Medical Centre Alkmaar and the Rijnland hospital, Leiderdorp, the Netherlands.
breast1
1
Breast - MRI
by Leonard Glassman and Marieke Hazewinkel
This article is based on a presentaton given by Harmien Zonderland and adapted for the Radiology Assistant. Many parts of this article are derived from the Bi-RADS atlas and we encourage anyone who is involved in breast imaging to order this illustrated atlas to get a full knowledge of the Bi-RADS classification (see references). In 2000 the Dutch Institute for Health Care Improvement (CBO) has chosen to use the Breast Imaging Reporting and Data System (BI-RADS) for breast imaging. All other systems that were previously used were abandoned, because unlike the BI-RADS system, they lacked quantification and used very subjective and undefined terms and were not evidence-based. BI-RADS is a quality assurance tool designed to standardize mammography reporting, reduce confusion in breast imaging interpretations, and facilitate outcome monitoring. It contains a lexicon for standardized terminology (descriptors) for mammography, breast US and MRI, aswell as Standard Reporting with Final Assessment Categories and guidelines for Follow-up and Outcome Monitoring. It even enables you to evaluate the quality of your reporting. The reporting system is designed to provide an organized approach to image interpretation and reporting. 1. Describe the indication for the study. 2. Describe the breast composition. 3. Describe any significant finding. 4. Compare to previous studies. 5. Conclude to a final assessment category. 6. Give management recommendations. When you describe a lesion use standard BI-RADS descriptors for Mammography, Ultrasound and MRI (see below). If an additional imaging modality is added, mention type and rationale for each modality. If more than one imaging modality is performed, for instance US with Mammography or with MRI, an integrated report with assessment based on the highest level of suspicion must be used. When you use more modalities, always make sure, that you are dealing with the same lesion. For instance a lesion found with US does not have to be the same as the mammographic or physical findings. Sometimes repeated mammographic imaging with markers on the lesion found with US can be helpful. Mammographic breast composition is described as follows: 1. the breast is almost entirely fat ( 2. scattered fibroglandular densities (25-50%) 3. heterogeneously dense breast tissue (51-75%) 4. extremely dense (> 75% glandular) A 'Mass' is a space occupying lesion seen in two different projections. If a potential mass is seen in only a single projection it should be called a 'Density' until its three-dimensionality is confirmed. Circumscribed (well-defined or sharply-defined) margins: The margins are sharply demarcated with an abrupt transition between the lesion and the surrounding tissue. Without additional modifiers there is nothing to suggest infiltration. Indistinct (ill defined) margins: The poor definition of the margins raises concern that there may be infiltration by the lesion and this is not likely due to superimposed normal breast tissue. Spiculated Margins: The lesion is characterized by lines radiating from the margins of a mass. The normal architecture is distorted with no definite mass visible. This includes spiculations radiating from a point, and focal retraction or distortion of the edge of the parenchyma. Architectural distortion can also be an associated finding. This is a density that cannot be accurately described using the other shapes. It is visible as asymmetry of tissue density with similar shape on two views, but completely lacking borders and the conspicuity of a true mass. It could represent an island of normal breast, but its lack of specific benign characteristics may warrant further evaluation. Additional imaging may reveal a true mass or significant architectural distortion. Due to confusion of the term mass with the term 'density' which describes attenuation characteristics of masses, the term 'density' has been replaced with 'asymmetry'. Amorphous or Indistinct Calcifications: These are often round or 'flake' shaped calcifications that are sufficiently small or hazy in appearance that a more specific morphologic classification cannot be determined. Coarse, Heterogeneous Calcifications: Irregular calcifications with varying sizes and shapes that are usually larger than 0.5 mm in diameter. Fine, Pleomorphic or Branching Calcifications: Fine pleomorphic calcifications are more conspicuous than the amorphic forms. They vary in sizes and shapes and are usually smaller than 0,5 mm. Fine branching calcifications are thin, linear or curvilinear, may be discontinuous and smaller than o,5 mm. Their appearance suggests filling in of the lumen of a duct involved irregularly by breast cancer. Benign Calcifications: Benign calcifications are usually larger than calcifications associated with malignancy. They are usually coarser, often round with smooth margins and are much more easily seen. When you describe an abnormality (mass, architectural distortion, focal asymmetry or calcifications) always use the standard BI-RADS descriptors and mention the lesion size and location. A negative diagnostic examination is one that is negative, with a benign or probably benign finding (BI-RADS 1, 2 or 3). In BI-RADS 3 the radiologist prefers to establish the stability of a lesion by short term follow-up. In the evaluation of your BI-RADS 3 lesions the malignancy rate should be A positive diagnostic examination is one that requires a tissue diagnosis (BI-RADS 4 and 5). In BI-RADS 4 the radiologist has sufficient concern to urge a biopsy (2-95% chance of malignancy). In BI-RADS 5 the chance of malignancy should be > 95%. BI-RADS 0 Need Additional Imaging Evaluation and/or Prior Mammograms For Comparison: BI-RADS 0 is utilized when further imaging evaluation (e.g. additional views or ultrasound) or retrieval of prior films is required. When additional imaging studies are completed, a final assessment is made. Always try to avoid this category by immediately doing additional imaging or retrieving old films before reporting. Even better to have the old films before starting the examination. BI-RADS 1 Negative: There is nothing to comment on. The breasts are symmetric and no masses, architectural distortion or suspicious calcifications are present. BI-RADS 2 Benign Finding: Like BI-RADS 1, this is a normal assessment, but here, the interpreter chooses to describe a benign finding in the mammography report. Involuting, calcified fibroadenomas, multiple secretory calcifications, fat-containing lesions such as oil cysts, lipomas, galactoceles and mixed-density hamartomas all have characteristically benign appearances, and may be labeled with confidence. The interpreter may also choose to describe intramammary lymph nodes, vascular calcifications, implants or architectural distortion clearly related to prior surgery, while still concluding that there is no mammographic evidence of malignancy. BI-RADS 3 Probably Benign Finding - Initial Short-Interval Follow-Up Suggested: A finding placed in this category should have less than a 2% risk of malignancy. It is not expected to change over the follow-up interval, but the radiologist would prefer to establish its stability. Lesions appropriately placed in this category include: The initial short-term follow-up is a unilateral mammogram at 6 months, then a bilateral follow-up examination at 12 months and 24 months after the initial examination. If the findings shows no change in the follow up the final assessment is changed to BI-RADS 2 (benign) and no futher follow up is needed. If a BI-RADS 3 lesion shows any change during follow up, it will change into a BI-RADS 4 or 5 and appropriate action should be taken. The case on the left shows a few amorphous calcifications initially classified as BI-RADS 3. At 12 month follow up more calcifications were noted in a cluster. The findings were now classified as BI-RADS 4. This proved to be DCIS with invasive carcinoma. A solid mass with circumscribed margins, oval shape and horizontal orientation is most likely a fibroadenoma and can be assigned a BI-RADS 3, irrespective if the lesion is palpable or not. As a consequence, solid lesions that do not possess all the typical characteristics of a fibroadenoma should be assigned a BI-RADS 4 and always be biopsied. First control after conservative treatment for breast cancer: new scars and postirradiation thickening of skin and interstitium is assigned BI-RADS 3. 2nd control after Conservative treatment for breast cancer: decrease of sequalae of treatment, BI-RADS category can be changed into BI-RADS 2 (figure) BI-RADS 4 Suspicious Abnormality - Biopsy Should Be Considered: BI-RADS 4 is reserved for findings that do not have the classic appearance of malignancy but have a wide range of probability of malignancy (2 - 95%). By subdividing Category 4 into 4A, 4B and 4C , it is encouraged that relevant probabilities for malignancy be indicated within this category so the patient and her physician can make an informed decision on the ultimate course of action. The case on the left shows another BI-RADS 4 abnormality. The pathologist could report to you that it is sclerosing adenosis or ductal carcinoma in situ. In both cases you as a radiologist would agree. BI-RADS 5 Highly Suggestive of Malignancy. Appropriate Action Should Be Taken: BI-RADS 5 must be reserved for findings that are classic breast cancers, with a >95% likelihood of malignancy. A spiculated, irregular high-density mass, a segmental or linear arrangement of fine linear calcifications or an irregular spiculated mass with associated pleomorphic calcifications are examples of lesions that should be placed in BI-RADS 5. BI-RADS 5 contains lesions for which one-stage surgical treatment could be considered without preliminary biopsy. However, current oncologic management may require percutaneous tissue sampling as, for example, when sentinel node imaging is included in surgical treatment or when neoadjuvant chemotherapy is administered. BI-RADS 6 Known Biopsy Proven Malignancy. Appropriate Action Should Be Taken BI-RADS 6 is reserved for lesions identified on the imaging study with biopsy proof of malignancy prior to definitive therapy. This category was added to the classification because sometimes patients are treated with neo-adjuvant chemotherapy. During the course of the treatment the tumor may be less visible, while still you know you are dealing with cancer (figure). Same case as above. Initial ultrasound shows large tumor. After chemotherapy shrinkage of the tumor. Tutorial by G. Pfarl, MD & T.H. Helbich, MD, Department of Radiology, University of Vienna. by E. Lazarus, M. B. Mainiero, B. Schepps, S. L. Koelliker, and L. S. Livingston Radiology, May 1, 2006; 239(2): 385 - 391. by Wendie A. Berg et al Department of Radiology, University of Maryland School of Medicine, 22 S. Greene St., Baltimore, AJR 2000; 174:1769-1777 The Breast Imaging Reporting and Data System? (BI-RADS?) Atlas is a quality assurance guide to standardize breast-imaging reporting and to facilitate outcome monitoring.Harmien Zonderland Mammographic Breast Composition Mass Architectural distortion Focal asymmetry CalcificationsIntroduction to the Breast Imaging Reporting and Data SystemRadiology department of the Leiden University Medical Centre, Leiden, the Netherlands
breast2
1
Differential diagnosis of Breast Calcifications
by Robin Smithuis and Ruud Pijnappel
This review is based on a presentation given by Leonard Glassman and adapted for the Radiology Assistant by Marieke Hazewinkel. MRI is a powerful tool: it is able to detect cancer not visible on conventional imaging, it can be used as a problem-solving instrument, and it can be applied to screen high-risk patients. Breast MRI is also better at monitoring the response to chemotherapy than other imaging modalities used today. It can change the treatment plan in 15-30% of patients with breast cancer. We will discuss the interpretation of breast MRI by looking at: Enhancing lesions are divided into three main categories: focus/foci, masses, and areas of non-mass enhancement (1). One of the things we run into are 'little bright objects'. These foci are enhancing areas of less than 5mm in diameter and are too small to characterize. They have persistent type 1 curves. These lesions are typically stable on follow-up and are considered to be a part of the normal background enhancement pattern in the breast. Shape A mass can be round, oval, lobulated, or irregular. Lobulated masses have undulating contours. Irregular masses have an uneven shape that cannot be characterized as round, oval, or lobulated. If a mass is irregularly shaped, it has a 32% chance of being malignant. The image on the left shows a large, irregular mass, which proved to be an angiosarcoma. The image on the far left is a juvenile fibroadenoma - it is oval in shape and has smooth margins, i.e. typically benign. The non-enhancing septations are not seen in this case. The image on the right is another example of a fibroadenoma: a lobulated mass with non-enhancing septations. Margin Margins can be described as smooth, irregular, or spiculated . Spiculated margins are frequently a feature of malignant breast lesions and radial scars. If a mass has spiculated margins, it has an 80% chance of being malignant. On the left is an image showing a large, round mass with smooth margins, which turned out to be an epidermal inclusion cyst. The image on the far left shows a spiculated mass, i.e. 80% chance of being malignant. . Next to it the corresponding gross pathologic specimen. You can see the spiculations invading the surrounding tissue in both. Just like on mammography, this lesion is has a high likelihood of malignancy and would be labelled BIRADS 5. The image on the far left shows an irregularly shaped mass with irregular margins, which was an invasive ductal carcinoma. The image on the right shows a similarly irregularly shaped and irregularly marginated lesion, this time an adenoid cystic carcinoma. The image on the left is a classic benign fibroadenoma. It is a lobulated mass with non-enhancing septations. On this image the margins are a bit irregular here and there, which may be a reason to biopsy this lesion anyway. The image on the right is a classic carcinoma. It is an irregularly shaped mass with irregular margins and enhancing internal septations (the enhancement is not well seen on this image). High signal on T1 The pre-contrast T1, non fat-suppressed sequence can show the presence of fat in a lesion. Central high signal on a T1-weighted image can be seen in intramammary lymph nodes or fat necrosis. Fat is also seen in hamartomas. The image on the left shows an example of a fat-containing hamartoma in the breast. Breast lesions containing fat are benign unless they are rapidly growing. Rapidly growing lesions should be biopsied. High signal on T2-fatsat In T2 fat-suppressed images we are looking for water. Lesions that are bright on T2 include cysts, lymph nodes and fat necrosis. These are all benign lesions. Unfortunately there is one malignant lesion that has a high signal intensity on T2 fat-suppressed weighted images. This is the colloid carcinoma. It is the exception to the rule that all things with bright signal on T2 fat-suppressed images are benign. On the image on the left there are multiple rounded areas in both breasts. These are multiple cysts. The image on the far left shows a round lesion with bright signal on T2. This is a a fibroadenoma. On the right is an example of a colloid carcinoma in a breast with dense, glandular tissue. It is the exception to the rule that all things with bright signal on T2 fat-suppressed images are benign. Moderate and low signal on T2-fatsat The T2 fat-suppressed sequences are for detecting lesions with high signal, not moderate or low signal. Moderate and low signal intensities can be caused by cancer. Mass enhancement occurs in six main patterns: Homogeneous enhancement The image on the left shows a homogeneously enhancing lesion. This proved to be an invasive ductal carcinoma. Heterogenous enhancement On the left, the image shows an irregularly shaped mass with spiculations and a heterogeneous internal enhancement pattern, which proved to be an invasive lobular carcinoma. Rim enhancement The image on the left shows rim enhancement of a lesion invading the surrounding tissue in a case of invasive ductal carcinoma. First we look at the initial upslope of the curve during the first one to two minutes. This is either slow, medium or rapid. Then there is the delayed portion - two minutes or more after the injection of contrast. This part of the curve shows either an increase, plateau or washout. The kinetic analysis takes about six minutes of repetitive scanning in total and can lead to three types of curve. Type 1 On the image on the left is a type 1 curve. There is a slow rise and a continued rise with time. A lesion with a type 1 curve has a chance of 6% of being malignant. Type 3 The type 3 curve shows a rapid initial rise, followed by a drop-off with time (washout) in the delayed phase. A lesion with this type of curve is malignant in 29-77%. This is the red on the CAD (Computer Aided Detection). Type 2 Then there is the type 2 curve, which is in the middle: a slow or rapid initial rise followed by a plateau in the delayed phase, which is allowed a variance of 10% up or down. The chance of a lesion with a type 2 curve being malignant lies somewhere between the 6% of the type 1 curve and the 29-77% of the type 3 curve. Many physicians will biopsy lesions with type 2 curves. For non-mass enhancement, kinetics are not very useful. If there is clumped enhancement in a breast it must be biopsied, even though there are no areas with a type 3 curve. Computer Aided Detection is a purely kinetic evaluation. It does not evaluate the anatomy or pathology of the images. CAD looks at the curves and peak enhancements for the contrast (automated kinetics). It also has some very nice features, including motion registration during subtraction, which can correct for a patient's movement during the exam - something not all MRI scanners can do. It can do multiplanar reconstruction and subtraction very well and very quickly – it also has a good measurement package. The CAD shows a large area of red superimposed on the breast lesion in the image on the left. In CAD, red is bad: it means type 3 washout, and probably cancer. The images on the left show a large, abnormally enhancing area in the left breast. The CAD has detected some very small areas with type 3 washout (in red). When you look at CAD images, take note of the worst (red) areas. This was a large invasive ductal carcinoma. Non-mass enhancement is enhancement without three-dimensional characteristics. It is important because it occurs in a significant number of cancers. You need to look at its distribution, its enhancement pattern and its symmetry or asymmetry. The table on the left summarizes the terms used to describe the distribution of non-mass enhancement in the breast. Focal refers to non-mass enhancement in less than 25% of a quadrant of the breast. Ductal involvement is enhancement in a ductal distribution, and is cancer in 60% of cases. Linear enhancement is similar to ductal enhancement, but does not have a ductal orientation. This finding means cancer in 31% of cases. Segmental enhancement refers to multiple ducts and has a 78% chance of being cancer. Regional enhancement is not ductal or segmental but larger than focal and is cancer in 21%. Diffuse non-mass enhancement is typically benign. The image on the left shows focal non-mass enhancement. This proved to be a focal DCIS. The image on the left shows linear non-mass enhancement. This proved to be stromal fibrosis. The image on the left shows a mass as well as areas of linear non-mass enhancement. This proved to be linear DCIS with an invasive ductal carcinoma. On the left examples of segmental and regional non-mass enhancement in DCIS. The image on the far left shows a mass with associated ductal enhancement coming from the mass, which corresponds to anterior and posterior expansion of the tumor in this case of DCIS. The image next to it shows an example of linear non-mass enhancement in a different orientation to that of the ducts in stromal fibrosis. Non-mass enhancement can be termed homogeneous and heterogeneous, just as mass enhancement can. As mentioned earlier, punctate enhancement is usually benign, but it can occur focally. In that case there is a 25% chance of cancer. Clumped enhancement is the most important non-mass enhancing pattern to recognize. It has a 60% chance of cancer (typically DCIS). For non-mass enhancement, kinetics are not very useful. If there is clumped enhancement in a breast it must be biopsied, even if there are no areas with a type 3 curve. On the far left heterogeneous enhancement in an invasive ductal carcinoma. The image next to it shows punctate enhancement in a hamartoma with fibrocystic change (arrows). Clumped enhancement Clumped enhancement is the most important non-mass enhancing pattern to recognize. It has a 60% chance of cancer (typically DCIS). On the left two examples of clumped enhancement in DCIS. Associated findings can be: The image on the left shows a relatively small carcinoma in the right breast, with extensive thickening of the skin. The image on the left shows a large inflammatory carcinoma with diffuse thickening of the skin. The image on the left shows a large enhancing lymph node on the right. Cysts have a high signal on T2 fat-suppressed images. After the injection of gadolinium, they will show up as filling defects, sometimes with rim enhancement. Fibroadenomas are the most common benign breast lesions after cysts. In order to be certain a lesion is a fibroadenoma, certain criteria must be met: A fibroadenoma must have benign spatial characteristics. This means it can not have a spiculated or microlobulated border. On the left an example of a classic fibroadenoma: a round, smoothly marginated lesion with some black or gray areas on the inside, which are the non-enhancing septations. This lesion has a type 1 curve. On the far left is another example of a fibroadenoma with clear non-enhancing septations. These septations are also visible on the gross pathology. The pre-contrast T1, non fat-suppressed sequence can show the presence of fat in a lesion. High signal on a T1-weighted image can be seen in intramammary lymph nodes, fat necrosis and hamartomas. These areas will be dark on fat suppressed images. On the left two classic examples of hamartomas. These lesions have fat-containing areas which are suppressed on these images after the administration of intravenous gadolinium. Kinetics are usually not useful in DCIS, especially not in cases when low-grade. Many cases of DCIS show no washout and usually there is slow initial enhancement. The distribution of the enhancement however is important. DCIS typically shows clumped, ductal, linear or segmental non-mass enhancement. On the left a patient with areas of non-mass enhancement in both breasts (DCIS). There is a small enhancing mass medially in the left breast, which was a small invasive carcinoma. The image on the left shows an enhancing mass in the left breast. This proved to be an invasive carcinoma. Lateral to it is an area of ductal non-mass enhancement, which proved to be DCIS. On the left another case with diffuse, bilateral DCIS. Another case of DCIS, located laterally in both breasts. The cases on the left are more difficult to diagnose . Both of these patients had large homogeneously enhancing areas in the right breast. In both patients this proved to be DCIS. Most invasive carcinoma are ductal, some are lobular, and there is a group of rarer types. Regardless of the type of cancer, they typically appear on breast MRI as an irregularly shaped, spiculated mass with rim- or heterogeneous enhancement after the administration of intravenous gadolinium. On the left two cases. The image on the far left is an invasive ductal carcinoma presenting as a large, heterogeneously enhancing mass. Next to it an example of an invasive ductal carcinoma presenting as a smaller mass with rim-enhancement. The image on the far left shows an irregular mass with some ductal extension, and on the right an irregular mass extending to the chest wall, but not invading it. There is no chest wall enhancement. Invasive lobular carcinoma is one of the types of cancer that does not always show a lot of enhancement on breast MRI, which can make it difficult to diagnose. In these two cases however, this was not a problem. The image on the far left is of a diffuse invasive lobular carcinoma. On the right is a MIP showing a large area of abnormal enhancement, which proved to be a diffuse invasive lobular carcinoma. The image on the left is a T2WI with fat suppression. It is a colloid carcinoma in a breast with dense, glandular tissue. It is the exception to the rule that all things with bright signal on T2 fat-suppressed images are benign. Terminal duct carcinoma On the left a large, irregular, enhancing mass in a male patient. This was a terminal duct carcinoma. Terminal duct carcinoma Sarcoma with osseous differentiation The case on the left is a patient with a sarcoma with osseous differentiation, showing less enhancement. Adenoid cystic carcinoma On the left an image of an irregular enhancing mass which was an adenoid cystic carcinoma. Metaplastic carcinoma On the left an example of a metaplastic carcinoma with rim-enhancement. This is not necessarily a typical presentation. There is a small area of stromal fibrosis laterally in the left breast. by B. Erguvan-Dogan, G. J. Whitman, A. C. Kushwaha, M. J. Phelps, and P. J. Dempsey Am. J. Roentgenol., August 1, 2006; 187(2): W152 - W160. by Mieke Kriege et al NEJM Volume 351:427-437 July 29, 2004 Number 5 Orel et al. Radiology. 1997 Feb;202(2):413-20 Steven G. Lee et al. Radiology 2003;226:773-778 By Gayle F. Tillman, Susan G. Orel, Mitchell D. Schnall, Delray J. Schultz, Jacqueline E. Tan, Lawrence J. Solin Journal of Clinical Oncology, Vol 20, Issue 16 (August), 2002: 3413-3423 by Aliya Qayyum, Robyn L. Birdwell, Bruce L. Daniel, Kent W. Nowels, Stefanie S. Jeffrey, Tony A. Agoston and Robert J. Herfkens. AJR 2002; 178:1227-1232 by Leong CS, Daniel BL, Herfkens RJ, Birdwell RL, Jeffrey SS, Ikeda DM, Sawyer-Glover AM, Glover GH. J Magn Reson Imaging. 2000 Feb;11(2):87-96. by C. Kuhl Radiology, August 1, 2007; 244(2): 356 - 378. by C. K. Kuhl. Radiology, September 1, 2007; 244(3): 672 - 691. Uwe Fischer, MD, Lars Kopka, MD and Eckhardt Grabbe, MD Radiology. 1999;213:881-888 Nicky H. G. M. Peters et al Radiology 2008;246:116-124 K. J. Macura, R. Ouwerkerk, M. A. Jacobs, and D. A. Bluemke RadioGraphics, November 1, 2006; 26(6): 1719 - 1734 by Laura Liberman et al AJR 2002; 179:171-178 REVIEW ARTICLE by Nola Hylton. Journal of Clinical Oncology, Vol 23, No 8 (March 10), 2005: pp. 1678-1684 By Mitchell D. Schnall et al Radiology 2006;238:42-53 American College of Radiology. Breast imaging reporting and data system atlas (BI-RADS atlas). Reston, VA: American College of Radiology, 2003Leonard Glassman and Marieke Hazewinkel Morphology T1-T2 characteristics Enhancement pattern of a mass Temporal Resolution - Kinetic Analysis (Curves) CAD Distribution Internal Enhancement Pattern - Nonmass Cysts Fibroadenoma Fat containing lesions DCIS Invasive ductal carcinoma Invasive lobular carcinoma Colloid carcinoma OthersBreast - MRI
breast3
1
Introduction to the Breast Imaging Reporting and Data System
by Harmien Zonderland
Ductal carcinoma-in-situ (DCIS) represents 25-30% of all reported breast cancers. Approximately 95% of all DCIS is diagnosed because of mammographically detected microcalcifications. In this review we will focus on: by Robin Smithuis and Ruud Pijnappel The basic functional unit in the breast is the lobule, also called the terminal ductal lobular unit (TDLU). The TDLU consists of 10-100 acini, that drain into the terminal duct. The terminal duct drains into larger ducts and finally into the main duct of the lobe (or segment), that drains into the nipple. The breast contains 15-18 lobes, that each contain 20-40 lobules. The terminal ductal lobular unit is an important structure because most invasive cancers arise from the TDLU. It also is the site of origin of ductal carcinoma in situ (DCIS), lobular carcinoma in situ, fibroadenoma and fibrocystic disease, like cysts, apocine metaplasia, adenosis and epitheliosis. Most calcifications in the breast form either within the terminal ducts (intraductal calcifications) or within the acini (lobular calcifications). Lobular calcifications These calcifications fill the acini, which are often dilated. This results in uniform, homogeneous and sharply outlined calcifications, that are often punctate or round. When the acini become very large, as in cystic hyperplasia, 'milk of calcium' may fill these cavities. However when there is more fibrosis, as in sclerosing adenosis, the calcifications are usually smaller and less uniform. In these cases it can be difficult to differentiate them from intraductal calcifications. Lobular calcifications usually have a diffuse or scattered distribution, since most of the breast is involved in the process that forms the calcifications. Lobular calcifications are almost always benign. Intraductal calcifications These calcifications are calcified cellular debris or secretions within the intraductal lumen. The uneven calcification of the cellular debris explains the fragmentation and irregular contours of the calcifications. These calcifications are extremely variable in size, density and form (i.e. pleomorphic from the Greek pleion 'more' and morphe 'form'). Sometimes they form a complete cast of the ductal lumen. This explains why they often have a fine linear or branching form and distribution. Intraductal calcifications are suspicious of malignancy and are classified as BI-RADS 4 or 5. The diagnostic approach to breast calcifications is to analyze the morphology, distribution and sometimes change over time. The form or morphology of calcifications is the most important factor in deciding whether calcifications are typically benign or not. If not, they are either suspicious (intermediate concern) or of a high probability of malignancy. Usually biopsy in these cases is needed to determine the etiology of these calcifications. The form of calcifications is the most important factor in the differentiation between benign and malignant. If calcifications cannot be readily identified as typically benign or as 'high probability of malignancy', they are termed of 'intermediate concern or suspicious'. If a specific etiology cannot be given, a description of the calcifications should include their morphology and distribution using the descriptions given in the BI-RADS atlas (1). In the BI-RADS atlas the following descriptions are given for the distribution of calcifications (1) : Diffuse or scattered distribution is typically seen in benign entities. Even when clusters of calcifications are scattered throughout the breast, this favors a benign entity. Regional distribution according to the BI-RADS atlas would favor a non-ductal distribution (i.e. benignity), while Segmental distribution would favor a ductal distribution (i.e. malignancy). Sometimes this differentiation can be made, but in many cases the differentiation between 'regional' and 'segmental' is problematic, because it is not clear on a mammogram or MRI where the bounderies of a segment (or a lobe) exactly are. Clustered calcifications are both seen in benign and malignant disease and are of intermediate concern. When clusters are scattered througout the breast, this favors a benign entity. A single cluster of calcification favors a malignant entity. Linear distribution is typically seen when DCIS fills the entire duct and its branches with calcifications. There are conflicting data concerning the value of absence of change over time. It is said that the absence of interval change in microcalcifications that are probably benign on the basis of morphologic criteria is a reassuring sign and an indication for continued mammographic follow-up (2). On the other hand in a retrospective study that included indeterminate and suspicious clusters of microcalcifications, stability could not be relied on as a reassuring sign of benignancy (3). In this group of patients with biopsy proven malignancy, 25% of patients had stable microcalcifications for 8-63 months. It seems that the morphology of calcifications is far more important than stability and stability can only be relied on if the calcifications have a probably benign form. In the same study it was shown that the odds for invasive carcinoma versus DCIS are statistically significantly higher among patients with increasing or new microcalcifications. The likelihood that carcinoma will be invasive increases significantly when a suspicious or indeterminate cluster of calcifications is new or increasing. On the left a patient with a few heterogeneous coarse calcifications. They were classified as BIRADS 3 (probably benign with a likelihood of malignancy less than 3%). At six month follow up they had increased in number and DCIS was found at biopsy. Many calcifications can be classified as typically benign and need no follow up (i.e. BI-RADS 1 or 2). Many of these are skin calcifications. These are usually lucent-centered deposits. Atypical forms may be confirmed by tangential views to be in the skin. Usually they are located along the inframammary fold parasternally and in the axilla and areola. When you consider the possibility of dermal calcifications, always study the portion of the skin that is seen en face to look for similar calcifications (arrow). Tatoo sign Skin calcifications may simulate parenchymal breast calcifications and may look like malignant-type calcifications. The cluster calcifications on the left was presented for biopsy. During the vacuum assisted biopsy procedure it was not possible to biopsy these calcifications, because they were out of range. When you look at the oblique and craniocaudal view, notice that the calcifications look exactly the same in configuration. This is called the tattoo sign . Spot views subsequently prooved that these were dermal calcifications. Here another example of the tatoo-sign. First notice that there are some calcifications that are clearly located within the skin (arrows). The cluster calcifications on the MLO-view has the exact configuration as the cluster on the CC-view (next image). On the CC-view the configuration of the microcalcifications is exactly the same. If these calcifications were located in the centre of the breast they should have a different configuration, because the projection is different. Only when calcifications are located within the skin their configuration stays the same. These are linear or form parallel tracks, that are usually clearly associated with blood vessels. Vascular calcifications noted in women On the left typical vascular calcifications. If only one side of a vessel is calcified (arrow), the calcification may simulate intraductal calcification, but usually the diagnosis is straight forward. The classic large 'popcorn-like' calcifications are produced by involuting fibroadenomas. These calcifications usually do not cause a diagnostic problem. When the calcifications in an fibroadenoma are small and numerous, they may resemble malignant-type calcifications and need a biopsy. These are formed within ectatic ducts. These benign calcifications form continuous rods that may occasionally be branching. They are different from malignant-type fine branching calcifications, because they are usually > 1 mm in diameter. They may have lucent centers if the calcium is in the wall of the duct. These calcifications follow a ductal distribution, radiating toward the nipple and are usually bilateral. These secretory calcifications are most often seen in women older than 60 years. Sometimes it is difficult to differentiate these from lineair calcifications as seen in DCIS. Round calcifications are 0.5-1 mm in size and frequently form in the acini of the terminal duct lobular unit. When smaller than 0.5 mm, the term 'punctate' is used. Round and punctate calcifications can be seen in fibrocystic changes or adenosis, skin calcifications, skin talc and rarely in DCIS. Suspect DCIS when the calcifications are small, i.e. punctate , and show some heterogeneity especially when in cluster, linear or segmental distribution. Round and punctate calcifications are classified as: These are round or oval calcifications that range from under 1 mm to over a centimeter. They are the result of fat necrosis, calcified debris in ducts, and occasional fibroadenomas. These are very thin benign calcifications that appear as calcium is deposited on the surface of a sphere. These deposits are usually under 1 mm in thickness when viewed on edge. Although fat necrosis can produce these thin deposits, calcifications in the wall of cysts are the most common 'rim' calcifications. On the left a sharply defined lesion. The low density indicates the presence of fat. This is a typical oil cyst. On a follow up mamogram the wall has calcified resulting in eggshel calcifications. These are benign sedimented calcifications in macro- or microcysts. On craniocaudad views they appear as fuzzy, round or amorphous. Consider magnification spot film with horizontal beam when you think of the possibility of milk of calcium, because on a 90? lateral view they may appear as semilunar, crescent shaped tea cups. Many calcifications representing milk of calcium within microcysts however do not layer on horizontal beam radiographs. The most important feature of these calcifications is the apparent change in shape of the calcific particles on different mammographic projections (craniocaudal versus oblique or 90? lateral). The images show a different shape on the oblique view compared to the mediolateral view. On the mediolateral view there is layering of the calcium. On the craniocaudal image the calcifications are round, fuzzy and ill-defined. On the mediolateral view the calcifications appear as semilunar, crescent shaped tea cups. Click on the image for an enlarged view. They represent calcium deposit on suture material. They are typically linear or tubular in appearance and knots are sometimes visible. These are coarse irregular 'lava-shaped' calcifications. These calcifications are larger than 0.5 mm and often have a lucent center. They are seen in irradiated breast or following trauma. They develop 3-5 years after treatment in about 30% of women. These calcifications are also described as fat necrosis. It is important to differentiate them from a recurrent malignancy. On the left more extensive dystrophic calcifications. If calcifications are not typically benign, they are either called 'Suspicious or of Intermediate Concern' or they are called 'High Probability of Malignancy'. We will first discuss suspicious calcifications. These calcifications have either an amorphous or coarse heterogeneous form. Usually these calcifications are biopsied to determine their exact nature. Amorphous or indistinct calcifications are defined as 'without a clearly defined shape or form'. These calcifications are usually so small or hazy in appearance, that a more specific morphologic classification cannot be determined. On the left amorphous and pleomorphic calcifications. Based on the morphology these calcifications were classified as BI-RADS 4. Biopsy revealed fibrocystic changes (FCC) Amorphous calcifications (2) Many benign and malignant breast diseases may present with amorphous calcifications (Table). About 20% of amorphous calcifications turn out to be malignant. Usually it is low grade DCIS. Amorphous calcifications (3)  On the left amorphous calcifications within a denser area of the breast. This was classified as Bi-RADS 4 (3-95% chance of malignancy). Biopsy revealed DCIS with invasive ductal carcinoma. Coarse heterogeneous microcalcifications, formerly called coarse granular, are irregular, conspicuous calcifications that are generally larger than 0.5 mm. They are considered to be of intermediate concern, along with amorphous microcalcifications. They have to be differentiated from fine pleomorphic microcalcifications, formerly called fine granular, that vary in size and shape, are usually less than 0.5 mm in diameter and are considered to be of higher probability of malignancy, along with the fine linear microcalcifications (1). Coarse heterogeneous microcalcifications tend to coalesce but are not the size of the larger irregular dystrophic calcifications. On the left coarse heterogeneous calcifications They were classified as Bi-RADS 4. Biopsy revealed DCIS. The differential diagnosis of coarse heterogeneous calcifications includes: Multiplicity and bilaterality of such calcifications favors a benign etiology. DCIS is considered when these calcifications have a clustered, linear or segmental distribution. On the left a patient in whom new calcifications were detected during follow up for breastcancer in the contralateral breast. There are coarse heterogeneous calcifications in a segmented distribution. These calcifications were classified as Bi-RADS 4. Biopsy showed calcifications within fibrous stroma. There was no sign of malignancy. Calcifications with a higher probability of malignancy are fine pleomorphic and fine linear or fine linear branching. These calcifications vary in size and shapes and are usually They are more conspicuous than the amorphic calcifications. There is a 25-40% risk of malignancy. On the left fine pleomorphic calcifications in a segmental and linear distribution. These were classified as BI-RADS 5. Biopsy revealed high grade DCIS. On the left a mammogram demonstrating two forms of calcifications. There are some round typically benign calcifications. The most conspicious calcifications however are the fine pleomorphic calcifications. They have a segmental distribution. Even without the presence of the mass these calcifications would be classified as Bi-RADS 5. Biopsy demonstrated an extensive high grade DCIS with an invasive carcinoma. The calcifications on the left were detected on the first mammogram in a screening program. There is a cluster of amorphous and fine pleomorphic calcifications. These calcifications were classified as BI-RADS 4. A biopsy was performed and only fibrocystic changes were found. On the left a case that looks quite similar to the one above. New calcifications were detected during follow up in a screening program. These are fine pleomorphic calcifications in a cluster. These calcifications were classified as Bi-RADS 4. This proved to be DCIS. The message is that with these calcifications you cannot tell whether they are malignant or not and they have to be biopsied. These are thin, linear or curvilinear irregular calcifications. They may be discontinuous. Usually they are Their appearance suggests filling of the lumen of a duct, i.e. 'casting' calcifications. These calcifications are classified as Bi-RADS 5. On the left calcifications in a segmental distribution. Some have a linear distribution and some have a branching morphology. This is highly suggestive of malignancy (Bi-RADS 5). On the left fine linear and branching calcifications in a segmental distribution highly suggestive of malignancy (Bi-RADS 5). Extensive high grade DCIS was found at biopsy. On the left a patient with new calcifications detected in a screening program. These are fine pleomorphic and fine linear calcifications. The distribution is linear. On the basis of the morphology and distribution these calcifications were classified as BI-RADS 5. At biopsy this was high grade DCIS. On the left artifacts within a cassette that simulate fine pleomorphic calcifications. A repeat exam with a different cassette did not show any calcifications. Image of the cassette. The image on the left shows the same artifacts. On the image on the right DCIS. The Breast Imaging Reporting and Data System? (BI-RADS?) Atlas is a quality assurance guide to standardize breast-imaging reporting and to facilitate outcome monitoring. by Sickles EA. Radiology 1991; 179: 463-468. by AS Lev-Toaff, et al Radiology, Vol 192, 153-156 Diagnostic Imaging BREAST First Edition by Wendie E. Berg Amirsys by Wendie A. Berg et al Department of Radiology, University of Maryland School of Medicine, 22 S. Greene St., Baltimore, AJR 2000; 174:1769-1777 by Elizabeth S. Burnside et al Radiology 2007;242:388-395. by R M Pijnappel et al British Journal of Radiology (2004) 77, 312-314 by G M Tse et al Journal of Clinical Pathology 2008;61:145-151 by Rina L. Loffman Felman Radiology 2002;223:481-482. by E. Lazarus, M. B. Mainiero, B. Schepps, S. L. Koelliker, and L. S. Livingston Radiology, May 1, 2006; 239(2): 385 - 391. by Dianne Georgian-Smith et al AJR 2001; 176:1255-1259 by Linda Moy, Priscilla J. Slanetz, Eren D. Yeh, Richard Moore, Elizabeth Rafferty, Kathleen A. McCarthy, Deborah Hall and Daniel B. Kopans AJR 2001; 177:173-175 by SS Linden and EA Sickles American Journal of Roentgenology, Vol 152, Issue 5, 967-971 Abstract and PDF Constance D. Lehman, M.D., Ph.D., Constantine Gatsonis, Ph.D., Christiane K. Kuhl, M.D. et al NEJM Volume 356:1295-1303 March 29, 2007 Number 13Robin Smithuis and Ruud Pijnappel Terminal ductal lobular unit Morphology Distribution Change over time Skin Calcifications - Tatoo sign Vascular Calcifications Coarse or 'Popcorn-like' Large Rod-like, Plasma cell mastitis Round and punctate calcifications Lucent-Centered Eggshell or Rim Calcifications Milk of Calcium Suture calcifications Dystrophic calcifications Amorphous calcifications Coarse Heterogeneous Fine Pleomorphic Fine Linear or Fine Linear BranchingDifferential diagnosis of Breast CalcificationsRadiology department, Rijnland Hospital, Leiderdorp and Martini Ziekenhuis, Groningen, the Netherlands.
breast4
1
Pathology of the Male Breast
by Leonard M. Glassman
This review is based on a presentation given by Leonard Glassman and adapted for the Radiology Assistant by Robin Smithuis. We will discuss: by Leonard M. Glassman When you do breast imaging in a male, always stick to the following rule: 'If it is not normal, gynecomastia or classically benign, it needs a biopsy'. So, first we're going to give you a couple of examples of the normal male breast, just to get used to what normal looks like. Then we will discuss gynecomastia and finally discuss some specific tumors. Obviously, the lesions in this last category will need a biopsy, unless you are sure that it is classically benign, like for instance a lymph node or lipoma. On the left two examples of a normal male mammogram. The left image shows normal skin, a nipple and a small amount of connective tissue behind the nipple. The image on the right shows a bit more connective tissue, but this is still normal. On the left another normal male breast. There is more fatty tissue and there are a number of blood vessels. There is a small amount of fibrous connective tissue, but basically most of this breast is just fat. On the left a MRI of the normal male breast. There is a small amount of connective tissue behind the nipple. Gynecomastia is the most common abnormality in the male breast. Clinically, it presents as a soft mobile tender subareolar mass. Every word in this sentence is critical: soft - mobile - tender - subareolar . So it has to be soft and mobile. It is tender in the acute phase, but not in the chronic phase. Gynecomastia must be subareolar! Any mass that is not subareolar is not gynecomastia. Imaging pattern There are three imaging patterns: On the left a male breast with a nodular glandular pattern of gynecomastia. There is a fan shaped density radiating from the nipple. It may be more prominent in the upper outer quadrant and, more importantly, it blends into the surrounding fat. If you think about the mammogram on the left as the breast of a woman instead of a man, than you might say that there is an ill-defined mass and you might conclude that this is a malignancy . However, in a man this indistinct border is a sign of a gynecomastia. Gynecomastia is simultaneous proliferation of ducts and stroma without encapsulation, so it must blend into the surrounding fat tissue. There is no proliferation of lobuli, like there is in women. So you will not see tumors that start in the lobuli, for example lactating adenomas, fibroadenomas, phyllodes tumors, and also invasive lobular carcinomas are extremely rare. On the far left a mammogram of a male with gynecomastia and next to it a mammogram of an eight year old girl with juvenile hypertrophy. Notice that they look very much the same. On the left unilateral gynecomastia. This was an incidental finding on a CT-scan done for some other reason. By definition gynecomastia is 2 cm or more of subareolar tissue in a non obese male. It is a common 'normal' finding, that is seen in 55% of men at autopsy. The peak incidence is 60 - 69 years. It is significant if it is new or symptomatic. In elderly males gynecomastia makes up 65% of all breast lesions.
 25% is carcinoma and 10% are other lesions. The nodular pattern of gynecomastia is seen in the florid early phase. It begins as an increased number of ducts and epithelial proliferation with edema and cellular fibroblastic stroma. This phase is reversible. On the left a mammogram and an ultrasound image of a patient with a nodular glandular pattern of gynecomastia. Notice that it is situated underneath the nipple. The ultrasound image shows the typical appearance of gynecomastia: a hypoechoic mass with lobulation or even spiculation. If this was seen in a woman, you would say that this is a mass with microlobulation and spiculation, i.e. Birads IV or V. In a man this is typical for gynecomastia. On the left the same ultrasound image, but now in the normal position. Notice how 'malignant' it looks. On the left a T2W-image with fatsat and a T1W-image after Gadolinium with fatsat. A radiologist who was not used to looking at 'male' mammograms ordered the MR for problem solving. Obviously this MR was performed for the wrong reason. MR should not be used to solve a problem that can be solved with mammography. Anyhow the MR shows gynecomastia of the nodular pattern. The dendritic pattern is seen in the fibrotic or late phase. There are dilated ducts, moderate epithelial proliferation and fibrosis. On the left a mammogram showing a subareolar density with prominent extensions into the fat. Usually the density is smaller than in the nodular pattern. On the mammogram on the left we can imagine, that there is fibrosis with extension into the fat. This is different from the glandular edema-like appearance in the acute phase of gynecomastia. The ultrasound shows a spiculated appearance. These cases clearly demonstrate that gynecomastia can have an appearance which we would call malignant in a woman. Unfortunately some of the malignant lesions in a man can look benign and we will show some examples in the next chapter. This pattern is seen in males with very high estrogen levels. The images on the left simply look like small female breasts. This was a patient who was on estrogen therapy for prostate carcinoma. This is usually bilateral and there is no palpable mass. Remember that gynecomastia presents clinically as a soft, mobile, tender, subareolar mass. Pseudogynecomastia results from excessive fat deposition in the breast area. It is seen as a normal variant, in obesity and in neurofibromatosis. Let's first start with a list of lesions that should not be diagnosed in male patients, because they simply do not get these lesions. A man does not get a lactating adenoma, because it is only seen in pregnancy. Because there are very few lobules in a man, lobular tumors are extremely rare. There are only a few invasive lobular carcinomas reported in men. Fibroepithelial lesions are also extremely rare because they too start in the lobules. So do not diagnose a fibroadenoma in a man, even if it looks like a fibroadenoma. When you get a biopsy result that says fibroadenoma, get another pathologist. On the left lesions that do occur in males. Exept for gynecomastia and pseudogynecomastia, in most of these lesions we will not get a diagnosis from imaging. We just report that there is a Birads IV lesion and do a biopsy. Myofibroblastoma is an interesting lesion because it is the only one lesion that is more common in men than in women. It presents as a freely moveable, solitary, palpable, firm mass.
 There are no calcifications. The mean age is in the late 50's. On the left a large lesion, that looks like a fibroadenoma. Now we know that men do not get fibroadenomas. The pathology diagnosis was myofibroblastoma and the lesion was treated with local excision. On the left another myofibroblastoma. It presents as a circumscribed lobulated mass without calcification. Notice that the lesion is eccentric to the nipple. It is the lobulated mass that needs to be biopsied, not the retromamillary gynecomastia. On the left another myofibroblastoma. The nipple is marked and the lesion is not retromamillary. On the ultrasound image the lesion is difficult to differentiate from the surrounding fat. On the left another myofibroblastoma. Even if this lesion was located behind the nipple, you would not diagnose this as gynecomastia, because it is a lobulated mass. This is a benign tumor of neural origin. They occur anywhere in the body. 6% occur in the breasts. They are typically seen in men in their 30's. Usually they present as circumscribed in males, but sometimes they have a spiculated appearance. Notice that the lesion on the left has a indistinct border as is usually seen in gynecomastia, but on the mammogram it is not located directly under the skin. So this is not gynecomastia and a biopsy is necessary. On the left another Granular cell tumor. Epidermal inclusion cyst is a skin lesion. It presents as a round well circumscribed dense mass. On the left a small epidermal inclusion cyst. Notice how it raises the skin. On the left a large epidermal inclusion cyst. On the left a T2W image demonstrating the cystic nature and the pathology specimen. Most are idiopathic. Specific causes must be excluded like TB, Sarcoid and fat necrosis. On the left is a male breast that looks like the breast of a female. These lesions can sometimes be spiculated. On the left a lesion that looks like a cyst, but remember that cysts originate in the lobules and men do not have lobules, unless they take estrogen. This is a varix and if you puncture it, you get a big red surprise. On the left a lesion, that looks like a fibroadenoma, but men do not get fibroadenomas. It is a solid encapsulated mass and at biopsy it happened to be a leiomyoma. If there are more than 2 mitoses per high power field the pathologist calls it a leiomyosarcoma. Malignant disease in men just looks like malignant disease in women. In the USA there are about 1700 new cases each year, which is 1% of all breast cancer. There is a higher incidence in people from China and Africa due to hyperestrogenism secondary to parasitic liver disease. These cancers present as a unilateral painless subareolar mass. This subareolar location is just like in gynecomastia, but usually it is eccentric to the nipple. It sometimes presents with bloody nipple discharge, which is unlike gynecomastia. Usually it is invasive ductal cancer. As stated above invasive lobular cancer is extremely rare. Also DCIS is rare because there is no screening program for men, so they will present when there is a palpable mass. On the left an eccentric irregular mass with spiculae. If this was a women you would have no trouble in diagnosing this as a cancer. In a man it is the same. Male breast cancer presents as a round, oval or irregular mass. Calcifications are rare, but when they occur they are coarser than in women. On the left a small invasive ductal carcinoma. It is subareolar and central, but it is also encapsulated. This is not gynecomastia. Paget's disease of the nipple and skin ulceration are more common than in women. On the left an invasive ductal carcinoma with skin retraction. There is a long list of carcinoma risk factors and they are the same as in women: On the left a small eccentric encasulated invasive ductal carcinoma. On the left a huge invasive ductal carcinoma with some coarse benign looking calcifications. Malignancies other than ductal carcinoma are uncommon. On the left a list of all malignancies in men. Metastases from prostate cancer are the most common metastases in males. It results from hematogenous spread and is usually seen in patients with widespread disease. It presents as round or lobulated non-calcified masses. On the left a patient with two metastases of a small cell lung carcinoma. A liposarcoma is a rare sarcoma. It presents as a slowly enlarging painful mass. It is usually of water-density and is not typically fatty. On the CT on the left you can see the density of the lesion that proved to be a liposarcoma. In conclusion we can say, that male breast disease either presents as mass, pain or nipple discharge. Gynecomastia and invasive ductal cancer are the most common lesions in the male breast, but there are other rarer benign and malignant lesions. Gynecomastia and carcinoma can usually be differentiated, but biopsy is sometimes necessary to separate them. All lesions eccentric to the nipple need biopsy unless they are characteristically benign, i.e.contain fat or typical lymph node. On the left a list of characteristics of gynecomastia versus carcinoma. Notice that there are many similarities. Both gynecomastia and carcinoma occur mostly at the age of 60 and can be soft, mobile, subareolar and unilateral. So that does not help. Carcinoma is usually eccentric, while gynecomastia is never eccentric. Gynecomastia has to have extensions into the surroundig fat. Carcinoma sometimes may have spiculations, that can look the same. Actually we call it extension into the fat, if we think it is gynecomastia and spiculation, if we think it is a carcinoma. On the left two cases, that demonstrate, that it can be difficult to differentiate gynecomastia from carcinoma on a mammogram. The carcinoma on the right is a little bit more encapsulated than the gynecomastia on the right. In less than 10% of the cases a biopsy can be needed to make the differentiation. On the left two more cases. On the far left there is diffuse gynecomastia. On the right a huge cancer which is encapsulated. The last cases on the left look very similar to each other. Based on the mammogram these two can not be differentiated. In those rare instances a biopsy is needed.Leonard M. Glassman Nodular pattern Dendritic Pattern Diffuse glandular pattern Pseudogynecomastia Myofibroblastoma Granular Cell Tumor Epidermal inclusion cyst Granulomatous Mastitis Varix Leiomyoma Metastases Liposarcoma Gynecomastia versus CarcinomaPathology of the Male Breast
cardio4
1
Contrast-enhanced Periferal MRA
by Tim Leiner
Contrast-enhanced MR angiography (CE-MRA) is more sensitive and specific for diagnosis and preinterventional work-up of Periferal Arterial Disease (PAD) compared to Duplex (1). CE-MRA detects more patent arteries than IA-DSA in patients with chronic critical ischemia and can modify the choice of therapeutic strategy in these patients (2). It is important to distinguish between patients with intermittend claudication and patients with chronic critical ischemia because they need a different imaging approach. In this overview guidelines are given how to tailor the MRA-examination by optimizing the use of surface coils, k-space filling, spatial resolution and contrast delivery. by Tim Leiner Department of Radiology, Maastricht University Hospital, the Netherlands Intermittend claudication is a benign form of periferal arterial occlusive disease. Typically these patients have 'single level' disease usually an isolated stenosis in the iliac or femoral artery. Mostly these patients are treated with risk factor management and exercise training. There is only a relative indication for invasive therapy in order to relief the symptoms ( usually PTA, sometimes surgery). Only a limited number of these patients will progress to more severe disease i.e. critical ischemia. When MRA is performed in these patients perfect imaging of lower leg and feet is not the issue since no surgeon will perform a infragenual procedure in patients with these complaints. A one-step examination from aorta to the lower leg will usually be sufficient and venous filling is usually no issue. In patients with chronic critical ischemia however, there is rest pain and/or tissue loss. They typically have 'multi level' disease with bilateral, severe stenoses or occlusions in multiple arteries and segments. In critical ischemia there is an absolute indication for invasive therapy. The goal is wound healing and limb saving. In these patients it is the job of MRA to find patent arteries in the lower leg or feet for bypass surgery with or without PTA. The utmost attempt must be made to find vessels in the lower legs and feet because if no arteries are found amputation will be unavoidable in most cases. A two-step examination is necessary. First you have to focus on the lower leg and feet with the best spatial resolution possible. Secondly the iliac and femoral level can be imaged. Venous enhancement from the first serie will usually not be a problem at these levels. Centric k-spacing provides the capability to acquire the optimal part of the central k-space during the arterial-phase of contrast enhancement in a short time-period. The remaining time is used for increasing spatial resolution. Venous enhancement will not be of much of a problem because if the outer k-space is filled with data this does not add much to the contrast in the image. The arterio-venous (AV) window will be enlarged allowing longer scan time. If centric k-space filling is not available use linear k-space filling (see tips and tricks).  Surface coils Although some advocate the use of the bodycoil for imaging all three station, surface coils will dramatically improve image quality. Especially t in the lower leg and feet a surface coil is mandatory. a 3-station coil is optimal for MRA from the aorta to the feet. If not available use as many surface coils as possible. Synergy body and synergy spine coils are very helpfull, As a rule of thumb you need at least 3 pixels per vessel-diameter in order to reliably differentiate between 50% stenosis. So at different stations a different spatial resolution is needed (figure). The CE-MRA series can be planned on a rough TOF-serie that gives you a good idea where the vessels of interest are located. Do not use the same box for every station, but adapt the size and the angulation of the boxes at iliac, femoral and crural level. Especially at the femoral level a small box usually is sufficient. This will save time and allows you to speed up going to the lower leg. At the crural level especially if the pedal arch has to be included a large box will be needed. The spatial resolution at this level has to be high. This results in more thin slices and a longer scan time. How to beat venous enhancement in the lower leg when there is a longer scantime is explained later. In patients with intermittent claudication a one-step examination with imaging of 3 sequential stations is optimal. In patients with critical ischemia first the lower legs with the pedal arch included are examined. Secondly a separate contrast injection is necessary for the examination of the Aorto-Iliac and Femoral station. Depending on the speed of the MR-system contrast can be delivered at a higher rate. 1. Prolong the arterio-venous (AV) window by venous compression. Use a midfemoral compression with a pressure cuff at 50-60 mmHg. A normal blood pressure cuff without the metal parts usually works fine. 2. Use centric k-space filling if available. If ontrast appears in the veins, this will not add much to the contrast in the image as at that moment the periferal lines of k-space are scanned, which mostly add to he resolution in the image. 3. In patients with critical ischemia do a biphasic examination to concentrate first on lower leg and feet. At the iliac level centric k-spacing is not necessary. Linear filling of k-space works good at this level and provides the advance that the sequence can be started in front of the arrival of the contrast-bolus. If centric k-spacing is not available at your MR-machine, use linear filling of k-space at all levels. At the level of the lower leg and feet the arterial contrast in the images will be less optimal and the risk of venous enhancement will be greater. A 3-station coil is optimal for MRA from the aorta to the feet. If not available use as many surface coils as possible. Synergy body and synergy spine coils are very helpfull, Use the best surface coil that you have for imaging the lower leg and feet and consider to do a biphasic examination. The most important issue in MRA of the aorta and iliac arteries is that the patient manages to hold his breath. A lower resolution breath-hold scan is superior to a higher resolution scan with breathing artifacts. Before the actual series start you need to practise the breath-hold with the patient. If adequate breath-hold is not possible than you have shorten scantime by lowering the matrix-size and increasing the slice thickness at the expense of in-plane resolution. Radiology 2005;235:699-708. by Tim Leiner, MD, PhD et al. Investigative Radiology. 39(7):435-444, July 2004. by Leiner, Tim MD, PhD et al.Tim Leiner Intermittend claudication Chronic critical ischemia Equipment requirements Spatial resolution Planning the series Contrast bolus-timing Injection protocol How to beat venous enhancement in lower legs Centric k-spacing not available 3-station coil not available Problems with breath-holdContrast-enhanced Periferal MRADepartment of Radiology, Maastricht University Hospital
cardio3
1
Coronary anatomy and anomalies
by Robin Smithuis and Tineke Willems
In this article we describe the anatomy of the coronary arteries of the heart and some of the anomalies with illustrations and CT-images. This article is an update of an article that appeared earlier in the Radiology Assistant. by Robin Smithuis and Tineke Wilems Radiology department of the Rijnland Hospital Leiderdorp and the University Medical Centre Groningen, the Netherlands. On the left an overview of the coronary arteries in the anterior projection.  On the left an overview of the coronary arteries in the right anterior oblique projection.  On the left an overview of the coronary arteries in the lateral projection.  The left coronary artery (LCA) is also known as the left main. The LCA arises from the left coronary cusp. The aortic valve has three leaflets, each having a cusp or cup-like configuration. These are known as the left coronary cusp (L), the right coronary cusp (R) and the posterior non-coronary cusp (N). Just above the aortic valves there are anatomic dilations of the ascending aorta, also known as the sinus of Valsalva. The left aortic sinus gives rise to the left coronary artery. The right aortic sinus which lies anteriorly, gives rise to the right coronary artery. The non-coronary sinus is postioned on the right side. The LCA divides almost immediately into the circumflex artery (Cx) and left anterior descending artery (LAD). On the left an axial CT-image. The LCA travels between the right ventricle outflow tract anteriorly and the left atrium posteriorly and divides into LAD and Cx. On the image on the left we see the left main artery dividing into On volume rendered images the left atrial appendage needs to be removed to get a good look on the LCA. In 15% of cases a third branch arises in between the LAD and the Cx, known as the ramus intermedius or intermediate branch. This intermediate branche behaves as a diagonal branch of the Cx. The LAD travels in the anterior interventricular groove and continues up to the apex of the heart. The LAD supplies the anterior part of the septum with septal branches and the anterior wall of the left ventricle with diagonal branches. The LAD supplies most of the left ventricle and also the AV-bundle. Mnemonic: Diagonal branches arise from the LAD. The diagonal branches come off the LAD and run laterally to supply the antero-lateral wall of the left ventricle. The first diagonal branch serves as the boundary between the proximal and mid portion of the LAD (2). There can be one or more diagonal branches: D1, D2 , etc. The Cx lies in the left AV groove between the left atrium and left ventricle and supplies the vessels of the lateral wall of the left ventricle. These vessels are known as obtuse marginals (M1, M2...), because they supply the lateral margin of the left ventricle and branch off with an obtuse angle. In most cases the Cx ends as an obtuse marginal branch, but 10% of patients have a left dominant circulation in which the Cx also supplies the posterior descending artery (PDA). Mnemonic: Marginal branches arise from the Cx and supply the lateral Margin of the left ventricle. The right coronary artery arises from the anterior sinus of Valsalva and courses through the right atrioventricular (AV) groove between the right artium and right ventricle to the inferior part of the septum. In 50-60% the first branch of the RCA is the small conus branch, that supplies the right ventricle outflow tract. In 20-30% the conus branch arises directly from the aorta. In 60% a sinus node artery arises as second branch of the RCA, that runs posteriorly to the SA-node (in 40% it originates from the Cx). The next branches are some diagonals that run anteriorly to supply the anterior wall of the right ventricle. The large acute marginal branch (AM) comes off with an acute angle and runs along the margin of the right ventricle above the diaphragm. The RCA continues in the AV groove posteriorly and gives off a branch to the AV node. In 65% of cases the posterior descending artery (PDA) is a branch of the RCA (right dominant circulation). The PDA supplies the inferior wall of the left ventricle and inferior part of the septum. On the image on the far left we see the most common situation, in which the RCA comes off the right cusp and will provide the conus branch at a lower level (not shown). On the image next to it, we see a conus branch, that comes off directly from the aorta. The large acute marginal branch (AM) supplies the lateral wall of the right ventricle. In this case there is a right dominant circulation, because the posterior descending artery (PDA) comes off the RCA. Coronary anomalies are uncommon with a prevalence of 1%. Early detection and evaluation of coronary artery anomalies is essential because of their potential association with myocardial ischemia and sudden death (3). With the increased use of cardiac-CT, we will see these anomalies more frequently. Coronary anomalies can be differentiated into anomalies of the origin, the course and termination (Table). The illustration in the left upper corner is the most common and clinically significant anomaly. There is an anomalous origin of the LCA from the right sinus of Valsalva and the LCA courses between the aorta and pulmonary artery. This interarterial course can lead to compression of the LCA (yellow arrows) resulting in myocardial ischemia. The other anomalies in the figure on the left are not hemodynamically significant. On the left images of a patient with an anomalous origin of the LCA from the right sinus of Valsalva and coursing between the aorta and pulmonary artery. Sudden death is frequently observed in these patients. On the left images of a patient with an anomalous origin of the LCA from the pulmonary artery, also known as ALCAPA. ALCAPA results in the left ventricular myocardium being perfused by relatively desaturated blood under low pressure, leading to myocardial ischemia. ALCAPA is a rare, congenital cardiac anomaly accounting for approximately 0.25-0.5% of all congenital heart diseases. Approximately 85% of patients present with clinical symptoms of CHF within the first 1-2 months of life. Myocardial bridging is most commonly observed of the LAD (figure). The depth of the vessel under the myocardium is more important that the lenght of the myocardial bridging. There is debate, whether some of these myocardial bridges are hemodynamically significant. On the image on the left we see a large LAD giving rise to a large septal branch that terminates in the right ventricle (blue arrow). by Carl Jaffe and Patrick J. Lynch by M. Abdulla This site includes instructional movies, 3-D animation, panoramic views, online quiz, interactive video-clips, interactive heart sounds & murmurs and interactive echocardiograms. by G.J. de Jonge et al European Radiology, Volume 18, Number 11 / November, 2008, 2425-2432Robin Smithuis and Tineke Willems Interarterial LCA ALCAPA Myocardial bridging FistulaCoronary anatomy and anomaliesRadiology department of the Rijnland Hospital Leiderdorp and the University Medical Centre Groningen, the Netherlands.
cardio2
1
Ischemic and non-ischemic cardiomyopathy
by Wouter van Es, Hans van Heesewijk, Benno Rensing, Jan van der Heijden and Robin Smithuis
In this presentation we will discuss the MRI features of ischemic cardiomyopathy and non-ischemic cardiomyopathies and the role of late enhancement imaging in differentiating between the various types of cardiomyopathy. Images can be enlarged by clicking on them. If a video doesn't work, just click the stop button and then the play button once more. For proper printing you may have to adjust the print settings of your internet browser. Myocardial segments with abnormal enhancement or wall motion disturbances are named and localized according to the 17 segments model of the American Heart Association (37). Individual myocardial segments can be assigned to the 3 major coronary arteries with the recognition that there is anatomic variability. 17 segments model (2)    Administration of Gadolinium results in uptake of the contrast agent into both normal and injured myocardium. In normal myocardium there will be early wash out of contrast. In injured myocardium the wash out is very slow resulting in delayed enhancement after 10 - 15 minutes compared to the normal myocardium. Delayed enhancement of myocardial tissue is seen in many pathophysiologic scenarios: The causes of cardiomyopathy (CM) can be divided into ischemic and non-ischemic (1-5). Ischemic CM is defined as dysfunction of the left ventricle as a result of a chronic lack of oxygen due to coronary artery disease. Delayed enhancement MR images will show fibrosis, which appears as high signal intensity in an area of coronary artery distribution. Since all infarctions start subendocardially and may progress to transmural, the subendocardial region is always involved. Non-ischemic CM has a variable etiology, i.e. genetic, toxic, metabolic, infectious and idiopathic. In nonischemic myocardial disease the delayed enhancement usually does not occur in a coronary artery distribution and is often midwall or epicardial rather than subendocardial or transmural. Infarcted myocardium is bright on late-enhancement images. All patients with ischemic cardiomyopathy demonstrate delayed enhancement in a typical 'CAD' pattern, one in which the subendocardium is always involved. When a coronary artery is occluded the infarction always starts subendocardially and progresses towards the epicardium depending on the duration of the occlusion [6]. Both acute and chronic infarctions enhance. In acute infarctions the contrast enters the damaged myocardial cells due to myocyte membrane disruption. In chronic infarctions the late enhancement is a result of retention of contrast material in the large interstitial space between the collagen fibers in the fibrotic tissue [7]. No reflow phenomenon is the failure of blood to reperfuse an ischemic area after the physical obstruction has been removed or bypassed. No reflow zones are identified on late-enhancement images as a dark core surrounded by an enhancing rim. This finding indicates the presence of damaged microvasculature in the core of an area of infarction The presence of a 'no reflow' zone is associated with worse functional outcome, larger infarcts and adverse clinical outcome [8,9]. Both acute and chronic infarctions demonstrate delayed-enhancement, but an acute infarction can often be distinguished by the presence of a 'no reflow' zone and high signal on T2 weighted images. Visit the website to view the videos. Cine imaging in combination with delayed-enhancement MR allows identification of: Stunning is defined as postischemic myocardial dysfunction that persists despite restoration of normal blood flow. Over time there can be a gradual return of contractile function depending on the transmurality of the ischemia [10]. If the degree of transmurality as seen on the delayed enhancement images is less than 50%, the myocardial function is likely to recover [11]. On the left a long axis cine 6 days after revascularization of an acute inferior wall infarction. First study the video and then continue reading. Continue with the delayed enhancement image. On the left the long axis delayed enhancement image of the same patient. There is less than 50% enhancement of the myocardium. This is a good prognostic sign and we can expect a restoration of some of the contractile function. Continue with the cine-view four months later. Visit the website to view the videos. On the left the same patient four months after the inferior infarction and revascularization. First study the video and then continue reading. The long axis cine shows improved function of the inferior wall. Now it can be concluded that the hypokinesia was due to stunning. Myocardial regions that demonstrate little or no evidence of hyperenhancement (i.e. infarction) have a high likelihood of recovery, whereas regions with transmural hyperenhancement have virtually no chance of recovery. Visit the website to view the videos. Hibernation is a state in which some segments of the myocardium exhibit abnormalities of contractile function at rest [10]. This phenomenon is highly significant clinically because it usually manifests itself in the setting of chronic ischemia, that is potentially reversible by revascularization. The reduced coronary blood flow causes the myocytes to enter a low-energy 'sleep mode' to conserve energy. There is an inverse relationship between the transmural extent of hyperenhancement, and the likelihood of wall motion recovery following revascularization. If the transmural extent of late enhancement is less than 50% the function is likely to improve after revascularization [12]. On the left long axis cine-images of a patient with a severe stenosis of the LAD. First study the video and then continue reading. The cine images show: On the left the long axis late enhancement image in the same patient. Noice the following: Visit the website to view the videos. After PTCA there is improvement of the function of the anterior wall. The ejection fraction improved from 17 to 49%. Non Ischemic cardiomyopathy is defined as a myocardial disorder in which the heart muscle is structurally and functionally abnormal, in the absence of other causes of heart dysfunction, like coronary artery disease, hypertension, valvular disease and congenital heart disease. We will discuss the cardiomyopathies listed in the table on the left. Hypertrophic cardiomyopathy (HCM) is characterized by a hypertrophied left ventricle, defined as diastolic wall thickness 15mm or more, without any identifiable cause such as hypertension or valvular disease. Normal ventricular septal measurement is 8-12 mm. Usually there is asymmetric thickening of the wall most prominently involving the ventricular septum without abnormal enlargement of the ventricular cavities. It is a genetic myocardial disorder with a prevalence of 1:500. In about 25% of patients there is obstruction of the left ventricular outflow tract (LVOT) due to hypertrophy of the basal septum and a systolic anterior motion of the mitral valve (SAM). In these cases the term HOCM or hypertrophic obstructive cardiomyopathy is used. The systolic anterior motion of the mitral valve is probably the result of the increased flow velocity and decreased pressure above the valve caused by the hypertrophied interventricular septum (the Venturi effect). In the vast majority of patients the systolic anterior motion of the mitral valva is the mechanism of obstruction in HCM and also the cause of the mitral regurgitation. On an end-systolic image the following findings can be depicted (figure): HOCM (2) On the left an end-diastolic image. The arrow points to the hypertrophic basal septum. Continue with the 3-chamber view movie. Visit the website to view the videos. On the left the 3-chamber view movie of the same patient. First study the video and then continue reading. The video nicely demonstrates: HOCM (3) On the far left a 3-chamber late enhancement image which nicely demonstrates the enhancement of the hypertrophic basal septum (arrow). Next to it a short axis late enhancement image which demonstrates the typical enhancement at the anterior and posterior right ventricular insertion points (arrows). The therapy of HOCM is pharmacological, surgical myotomy or alcohol ablation [15]. The results of the alcohol ablation are very well depicted with MRI [19]. On the left a 3-chamber late enhancement image before and after alcohol ablation. Note the transmural infarction of the basal septum (arrow). Continue with the 3-chamber movie pre-alcohol ablation. Visit the website to view the videos. HOCM (4) On the left a 3-chamber movie of the same patient before the alcohol ablation. Notice the systolic anterior movement of the anterior leaflet of the mitral valve and the mitral regurgitation. Visit the website to view the videos. On the left the 3-chamber movie post-alcohol ablation with thinning of the basal septum and normalization of the function of the mitral valve. Visit the website to view the videos. The most common cause of restrictive cardiomyopathy is amyloidosis [20]. Amyloid deposits in the myocardium cause abnormal diastolic function with biatrial enlargement, concentric thickening of the left ventricle and reduced systolic function of usually both ventricles. Cardiac involvement in systemic amyloidosis occurs in up to 50% and has a poor prognosis with a median survival of 6 months [3]. On the left a 4-chamber movie of a patient with amyloidosis. There is diffuse hypokinesia of the left and right ventricle. Visit the website to view the videos. Same patient, short axis movie. Late enhancement image shows enhancement over the entire subendocardial circumference, variably extending into the neighboring myocardium [21]. Sometimes it is difficult to find the optimal inversion time for nulling the normal myocardium [1]. On the left the 4-chamber and short axis late enhancement images. There is circumferential subendocardial enhancement extending into the neighboring myocardium. The most important differential diagnosis of restrictive cardiomyopathy is constrictive cardiomyopathy. MRI can differentiate between those two diagnoses: Visit the website to view the videos. On the left the 4-chamber movie in a patient with constrictive CM. Notice the diastolic septal bounce which is typical for constrictive cardiomyopathy. Visit the website to view the videos. Same patient, short axis movie. Visit the website to view the videos. Dilated cardiomyopathy is defined as dilatation with an end diastolic diameter greater than 55mm measured on the left ventricular outflow image and an ejection fraction Patients with idiopathic dilated cardiomyopathy show either no enhancement or linear midmyocardial enhancement [24]. This enhancement is explained by the presence of fibrosis. This indicates a poorer prognosis. Patients with midmyocardial enhancement are at higher risk of sudden cardiac death and arrhythmias [25]. On the left a 4-chamber view of a patient with idiopathic cardiomyopathy. Notice the mitral regurgitation. Continue with the late enhancement image. The late enhancement image does not show any enhancement. This is compatible with idiopathic dilated cardiomyopathy. Visit the website to view the videos. Dilated cardiomyopathy (2) The differentiation between idiopathic dilated cardiomyopathy and ischemic dilated cardiomyopathy is important, as ischemic cardiomyopathy might be treated with revascularization and idiopathic disease not. Late enhancement MRI will show subendocardial enhancement in patients with ischemic cardiomyopathy. On the left a 4-chamber movie of a patient with dilated cardiomyopathy. Continue with the late enhancement image. The late enhancement MRI shows subendocardial enhancement in this patient. So we can conclude that this is dilated cardiomyopathy as a result of ischemia. Visit the website to view the videos. Dilated cardiomyopathy (3) In patients with dilated cardiomyopathy it is important to determine the ejection fraction. According to the guidelines of ACC/AHA/HRS 2008 [26] there is an indication for an automated implantable cardioverter-defibrillator (AICD) if: On the left the 4-chamber view of a patient with the idiopathic dilated cardiomyopathy. The ejection fraction was measured to be 28%. Visit the website to view the videos. Same patient with the idiopathic dilated cardiomyopathy: short axis view. Notice the poor contraction. On the left the late enhancement images of the same patient. There is midmyocardial septal enhancement consistent with fibrosis. Arrhythmogenic right ventricular cardiomyopathy (ARVC) is an inherited cardiomyopathy whose hallmark is fibrofatty replacement of the RV myocardium. The left ventricle is also involved in at least 15% of patients. The patients develop progressive RV failure and present with ventricular arrhythmias which can cause sudden cardiac death especially in young people. Morphologically the right ventricle can have regional wall thinning, hypertrophy, dilatation and microaneurysms. Functionally cine images are evaluated for RV dysfunction, microaneurysm formation, and focal areas of RV dyskinesia. MR scans may be overinterpreted since the RV has substantial normal variations including variable trabeculation and small outward bulges near the insertion of the moderator band. On the left axial black-blood images of a patient with fatty ARVC. Visit the website to view the videos. There are two variants of ARVC: fatty and fibro-fatty. The fatty form is characterized by fatty replacement of the myocardium without thinning of the ventricular wall. The fibro-fatty form is associated with significant thinning of the right ventricular wall. The sites of involvement are mostly found in the subtricuspid area, the right ventricular apex, and the infundibulum, the 'triangle of dysplasia' [4]. On the left a 4-chamber movie in a patient with ARVC. Notice the dilated right ventricle with severe segmental hypo- and dyskinesis resulting in small aneurysms. Visit the website to view the videos. On the left a short axis movie in a patient with ARVC. Notice the dilated right ventricle with severe segmental dyskinesis resulting in small aneurysms. ARVC (2) MRI can show segmental hypokinesis, dilatation, fatty infiltration in the right ventricular myocardium, small aneurysms and late enhancement of the myocardium [5,27]. Fat infiltration is seldom the only abnormality seen on MRI in ARVC, it should coincide with right ventricular regional dysfunction [28]. The diagnosis ARVC cannot be made on MRI findings alone. Visit the website to view the videos. On the left a 4-chamber movie of a patient with ARVC. There is a dilated right ventricle with severe segmental hypokinesis and dyskinesis. ARVC (3) The diagnosis is based on major and minor Task Force criteria, many of which involve clinical and laboratory information [29]. Major criteria demonstrated by MRI are: Minor criteria shown by MRI include [27] : Myocarditis is often caused by a viral infection. Acute myocarditis can be a cause of sudden cardiac death. Most patients spontaneously recover, however 5-10% of the patients will develop a dilated cardiomyopathy [30]. Acute myocarditis may clinically mimic an acute myocardial infarction with chest pain. Abnormal laboratory findings and ECG changes may also suggest an acute coronary syndrome. The MRI findings however are discriminatory between those two diagnoses. The late enhancement images are key, as the late enhancement in myocarditis is subepicardially or midmyocardially located, and does not originate from the subendocardium [30]. On the left a patient with myocarditis. Notice the midmyocardial enhancement of the lateral wall. Same patient with myocarditis. Notice that the midmyocardial enhancement of the lateral wall has diminished. Visit the website to view the videos. Myocarditis (2) Most lesions with myocarditis occur in the lateral free wall. Wall motion abnormalities may or may not be present. There is a potential relationship between the location of late enhancement, the etiologic virus and the prognosis [31]. On the left a patient with myocarditis. The 4-chamber movie demonstrates hypokinesia of the lateral wall of the left ventricle. Continue with the movie 10 months later. Visit the website to view the videos. 4-chamber movie 10 months later. The lateral wall is now normokinetic. Tako-Tsubo cardiomyopathy or apical ballooning syndrome is a transient cardiomyopathy affecting postmenopausal women after physical or emotional stress. Patients present with symptoms mimicking an acute myocardial infarction. The ECG changes and abnormal laboratory findings may also mimic an infarction. However, coronary angiography is usually normal, but if a left ventricle angiogram is performed, marked hypokinesia of the apical cardiac segments is noted (figure). The Japanese word takotsubo means octopus pot. This pot was used to capture octopus and resembles the shape of the left ventricle during systole in these patients Visit the website to view the videos. These apical wall motion abnormalities are well seen with MRI. The motion abnormalities are transient and return to normal within weeks. On the left a patient with Tako-Tsubo cardiomyopathy. Notice the hypokinesia of the apex. The apical wall motion abnormalities were transient and returned to normal within weeks. Continue with the late enhancement image. Tako-Tsubo cardiomyopathy (2) Typically there is no late enhancement, which distinguishes it from an infarction [4]. The pathogenesis is unknown, but it is probably caused by the release of catecholamines. The modified Mayo Clinic criteria for diagnosis of takotsubo cardiomyopathy: Visit the website to view the videos. On the left an angiogram of a patient with Tako-Tsubo cardiomyopathy. Mahrholdt H, Wagner A, Judd RM, Sechtem U, Kim RJ. Delayed enhancement cardiovascular magnetic resonance assessment of non-ischaemic cardiomyopathies. Eur Heart J 2005; 26:1461-1474 Vogel-Claussen J, Rochitte CE, Wu KC, Kamel IR, Foo TK, Lima JA, Bluemke DA. et al Radiographics 2006; 26:795-810 White JA, Patel MR. The role of cardiovascular MRI in heart failure and the cardiomyopathies. Magn Reson Imaging Clin N Am 2007; 15:541-564 Harris SR, Glockner J, Misselt AJ, Syed IS, Araoz PA. Cardiac MR imaging of nonischemic cardiomyopathies. Magn Reson Imaging Clin N Am 2008; 16:165-183 Shehata ML, Turkbey EB, Vogel-Claussen J, Bluemke DA. Role of cardiac magnetic resonance imaging in assessment of nonischemic cardiomyopathies. Top Magn Reson Imaging 2008; 19:43-57 Jackson E, Bellenger N, Seddon M, Harden S, Peebles C. Ischaemic and non-ischaemic cardiomyopathies - cardiac MRI appearances with delayed enhancement. Clinical Radiology 2007; 62:395-403 Reddy GP, Pujadas S, Ordovas KG, Higgins CB. MR imaging of ischemic heart disease. Magn Reson Imaging Clin N Am 2008; 16:201-212 Nijveldt R, Beek AM, Hirsch A, Hofman MB, Umans VA, Algra PR, van Rossum AC. 'No-reflow' after acute myocardial infarction: direct visualisation of microvascular obstruction by gadolinium-enhanced CMR. Neth Heart J 2008; 16:179-181 Pineda V, Merino X, Gispert S, Mah?a P, Garcia B, Dom?nguez-Oronoz R. No-reflow phenomenon in cardiac MRI: diagnosis and clinical implications. AJR Am J Roentgenol 2008; 191:73-79 Gowda RM, Khan IA, Vasavada BC, Sacchi TJ. Reversible myocardial dysfunction: basics and evaluation. Int J Cardiol 2004; 97:349-353 Beek AM, K?hl HP, Bondarenko O, Twisk JW, Hofman MB, van Dockum WG, Visser CA, van Rossum AC. Delayed contrast-enhanced magnetic resonance imaging for the prediction of regional functional improvement after acute myocardial infarction. J Am Coll Cardiol 2003; 42:895-901 Kim RJ, Wu E, Rafael A, Chen EL, Parker MA, Simonetti O, Klocke FJ, Bonow RO, Judd RM. The use of contrast-enhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 2000; 343:1445-1453 Maron BJ, Towbin JA, Thiene G, et al. Contemporary definitions and classification of the cardiomyopathies: an American Heart Association Scientific Statement from the Council on Clinical Cardiology, Heart Failure and Transplantation Committee; Quality of Care and Outcomes Research and Functional Genomics and Translational Biology Interdisciplinary Working Groups; and Council on Epidemiology and Prevention. Circulation 2006; 113:1807-1816 Elliott P, Andersson B, Arbustini E, et al. Classification of the cardiomyopathies: a position statement from the European Society Of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J 2008; 29:270-276 Nishimura RA, Holmes DR, Jr. Clinical practice. Hypertrophic obstructive cardiomyopathy. N Engl J Med 2004; 350:1320-1327 Hansen MW, Merchant N. MRI of hypertrophic cardiomyopathy: part 1, MRI appearances. AJR Am J Roentgenol 2007; 189:1335-1343 Moon JC, McKenna WJ, McCrohon JA, Elliott PM, Smith GC, Pennell DJ. Toward clinical risk assessment in hypertrophic cardiomyopathy with gadolinium cardiovascular magnetic resonance. J Am Coll Cardiol 2003; 41:1561-1567 Adabag AS, Maron BJ, Appelbaum E, Harrigan CJ, Buros JL, Gibson CM, Lesser JR, Hanna CA, Udelson JE, Manning WJ, Maron MS. Occurrence and frequency of arrhythmias in hypertrophic cardiomyopathy in relation to delayed enhancement on cardiovascular magnetic resonance. J Am Coll Cardiol 2008; 51:1369-1374 Hansen MW, Merchant N. MRI of hypertrophic cardiomyopathy: part 2, Differential diagnosis, risk stratification, and posttreatment MRI appearances. AJR Am J Roentgenol 2007; 189:1344-1352 Vanden Driesen RI, Slaughter RE, Strugnell WE. MR findings in cardiac amyloidosis. AJR Am J Roentgenol 2006; 186:1682-1685 Vogelsberg H, Mahrholdt H, Deluigi CC, Yilmaz A, Kispert EM, Greulich S, Klingel K, Kandolf R, Sechtem U. Cardiovascular magnetic resonance in clinically suspected cardiac amyloidosis: noninvasive imaging compared to endomyocardial biopsy. J Am Coll Cardiol 2008; 51:1022-1030 Giorgi B, Mollet NR, Dymarkowski S, Rademakers FE, Bogaert J. Clinically suspected constrictive pericarditis: MR imaging assessment of ventricular septal motion and configuration in patients and healthy subjects. Radiology 2003; 228:417-424 Francone M, Dymarkowski S, Kalantzi M, Rademakers FE, Bogaert J. Assessment of ventricular coupling with real-time cine MRI and its value to differentiate constrictive pericarditis from restrictive cardiomyopathy. Eur Radiol 2006; 16:944-951 McCrohon JA, Moon JC, Prasad SK, McKenna WJ, Lorenz CH, Coats AJ, Pennell DJ. Differentiation of heart failure related to dilated cardiomyopathy and coronary artery disease using gadolinium-enhanced cardiovascular magnetic resonance. Circulation 2003; 108:54-59 Assomull RG, Prasad SK, Lyne J, Smith G, Burman ED, Khan M, Sheppard MN, Poole-Wilson PA, Pennell DJ. Cardiovascular magnetic resonance, fibrosis, and prognosis in dilated cardiomyopathy. J Am Coll Cardiol 2006; 48:1977-1985 Epstein AE, DiMarco JP, Ellenbogen KA, et al. ACC/AHA/HRS 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices) developed in collaboration with the American Association for Thoracic Surgery and Society of Thoracic Surgeons. J Am Coll Cardiol 2008; 51:e1-62 Jain A, Tandri H, Calkins H, Bluemke DA. Role of cardiovascular magnetic resonance imaging in arrhythmogenic right ventricular dysplasia. Cardiovasc Magn Reson 2008 20; 10:32 Tandri H, Macedo R, Calkins H, Marcus F, Cannom D, Scheinman M, Daubert J, Estes M 3rd, Wilber D, Talajic M, Duff H, Krahn A, Sweeney M, Garan H, Bluemke DA. Role of magnetic resonance imaging in arrhythmogenic right ventricular dysplasia: insights from the North American arrhythmogenic right ventricular dysplasia (ARVD/C) study. Am Heart J 2008; 155:147-153 McKenna WJ, Thiene G, Nava A, Fontaliran F, Blomstrom-Lundqvist C, Fontaine G, Camerini F. Diagnosis of arrhythmogenic right ventricular dysplasia/cardiomyopathy. Task Force of the Working Group Myocardial and Pericardial Disease of the European Society of Cardiology and of the Scientific Council on Cardiomyopathies of the International Society and Federation of Cardiology. Br Heart J 1994; 71:215-218 Mahrholdt H, Goedecke C, Wagner A, Meinhardt G, Athanasiadis A, Vogelsberg H, Fritz P, Klingel K, Kandolf R, Sechtem U. Cardiovascular Magnetic Resonance Assessment of Human Myocarditis. A Comparison to Histology and Molecular Pathology. Circulation 2004; 109:1250-1258 Mahrholdt H, Wagner A, Deluigi CC, Kispert E, Hager S, Meinhardt G, Vogelsberg H, Fritz P, Dippon J, Bock CT, Klingel K, Kandolf R, Sechtem U. Presentation, patterns of myocardial damage, and clinical course of viral myocarditis. Circulation 2006; 114:1581-1590 by Srijita Sen-Chowdhry et al J Am Coll Cardiol, 2006; 48:2132-2140 by Heiko Mahrholdt et al European Heart Journal 2005 26(15):1461-1474 A Statement for Healthcare Professionals From the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association Circulation. 2002;105:539Wouter van Es, Hans van Heesewijk, Benno Rensing, Jan van der Heijden and Robin Smithuis 17 segments model Enhancement patterns Ischemic versus non-ischemic Infarction and delayed enhancement No reflow phenomenon Stunning Hibernation Hypertrophic cardiomyopathy Restrictive cardiomyopathy - Amyloidosis Constrictive cardiomyopathy Dilated cardiomyopathy ARVC Myocarditis Tako-Tsubo cardiomyopathyIschemic and non-ischemic cardiomyopathyRadiology and Cardiology department of the St. Antonius Hospital in Nieuwegein and the Rijnland hospital in Leiderdorp, the Netherlands
cardio5
1
Thoracic Aorta - the Acute Aortic Syndrome
by Ferco Berger, Robin Smithuis, Otto van Delden
The term Acute Aortic Syndrome (AAS) is used to describe three closely related emergency entities of the thoracic aorta: classic Aortic Dissection (AD), Intramural Hematoma (IMH) and Penetrating Atherosclerotic Ulcer (PAU). Clinically these conditions are indistinguishable. CT is the most accurate imaging modality for the initial diagnosis, differentiation and staging. This review will discuss the imaging features and important pitfalls. by Ferco Berger, Otto van Delden and Robin Smithuis Radiology department of the Academical Medical Centre, Amsterdam and the Rijnland Hospital, Leiderdorp, the Netherlands. Image protocol will be based on the type of scanner that is available. Our imaging protocol is based on a 4 slice helical CT-scanner. For the evaluation of patients with suspected AAS, we use 4x2,5 mm collimation technique with 5 mm axial reconstructions and coronal, sagittal and oblique MPRs. A non-enhanced scan of the thoracic aorta is included for the detection of an intramural hematoma (IMH). This is followed by a contrast-enhanced scan of the aorta in the arterial phase with bolus triggering and in the venous phase (100 ml Visipaque, 3ml/sec). Contrast differences between arterial and venous phase can be helpful in differentiating true and false lumen. The iliac tract is included for evaluation of endovascular treatment possibilities. The branches of the arch are visualized to evaluate the extend of dissection and awareness of possible neurological complications. Important reduction of artifacts and thus decreasing pitfalls, can be achieved by: Classic Aortic Dissection (AD), Intramural Hematoma (IMH) and Penetrating Atherosclerotic Ulcer (PAU) are distinct entities, but closely related. This is reflected upon in their identical therapeutical strategies. The main goal for the radiologist is not only to detect which entity is causing the clinical problem, but more importantly to differentiate between type A and B! The Acute Aortic Syndrome (AAS) is classified according to Stanford. Stanford Type A lesions involve the ascending aorta and aortic arch and may or may not involve the descending aorta. Stanford Type B lesions involve the thoracic aorta distal to the left subclavian artery. The Stanford classification has replaced the DeBakey classification (type I= ascending, arch and descending aorta: type II= only ascending aorta: type III= only descending aorta). Treatment options for the 2 subgroups of the acute aortic syndrome (AAS) are very different: - Stanford type A will be treated with surgery or endovascular therapy. - Stanford Type B will be treated medically. Classic Aortic Dissection is the most common entity causing an acute aortic syndrome (70%). Management decisions are based on the following information: In Aortic dissection an intima flap is seen in only 70% of cases. When there are 2 lumina, these will spiral around each other (figure). On the left consecutive images are seen of a Type B dissection. The true lumen is surrounded by calcifications. The true lumen is smaller, as the false lumen wedges around the true lumen due to permanent systolic pressure (so called Beak-sign). Thrombus material invariably is located in the false lumen, which enhances later than the true lumen. True lumen: False lumen: On the left an aortic dissection is seen with a large false lumen. The compressed true lumen is seen on the inner side and is brighter than the false lumen. Thrombus formation within the false lumen. The true lumen usually is smaller as the false lumen wedges around the true lumen due to permanent systolic pressure. The false lumen usually adheres to the outer curvature of the aortic arch, as is seen in this case. Collageneous media-remnants (cobwebs) are only seen in the false lumen. The same holds true for thrombusmaterial. If one of the lumina is surrounded by the other, it invariably is the true lumen. This almost only occurs in type A dissections. The figures on the left both show a type A dissection with clear entry points in the ascending aorta. The true lumen is surrounded by the false lumen, which is bigger and wedges around the true lumen due to permanent systolic pressure. Carefully sort out which branches of the aortic arch are involved. Make sure from which lumina they arise. The celiac trunc, SMA and right renal artery flow usually originates from the true lumen. Left renal artery flow mostly originates from the false lumen. Impaired perfusion of end-organs can be due to 2 mechanisms: 1) static = continuing dissection in the feeding artery (usually treated by stenting) 2) dynamic = dissection flap hanging in front of ostium like a curtain (usually treated with fenestration). This may be hard to discern, MPR's can be helpfull. Look for the re-entry point, usually to be found in the iliac tract. Provide information about tortuisity and calcifications of the iliac tract if endovascular procedures are being considered. When no end-organs are compromised and there is sufficient perfusion, dissection can be left alone. This may persist for a long time without clinical consequence, as is seen in the patient on the left with follow-up of 2 years. Some dissections remained unchanged during a follow up of more than 5 years. Even the slightest amount of fluid in pericardium, mediastinum or pleural cavity is suggestive of rupture of the dissection. The cases on the left show evident rupture, with presence of extensive hematoma in above mentioned locations. Note extreme hematothorax and hematomediastinum, causing shift of the mediastinum and compression on the pulmonary veins and even aorta. No pericardial effusion visible. The case on the left is a patient who presented with a fully thrombosed false lumen. 5 days after initial presentation this patient complained of acute chest pain mimicking the earlier episode. Re-examination showed recurrence of flow in the false lumen, locally contained, but with alarming adhering pleural effusion. The patient could not undergo surgical or endovascular repair for various reasons and was treated consevatively. It can be difficult to differentiate an aneurysm with thrombus from a dissection with a thrombosed false lumen. If there are intima calcifications this will be very helpfull. A false lumen displaces the intimal calcifications. Brief facts: What the clinician needs to know On the left a Intramural hematoma, hyperdense on a NECT. Same case contrast enhanced CT. Note that the IMH does not spiral around the true lumen, like in classic AD, helping to differentiate both. Essentially, this is not important, therapeutical decision will be made by whether this IMH is classified as Type A or Type B IMH! Note that there is no pericardial effusion. IMH thickness stays below 2 cm, making regression of this Type B IMH likely (up to 80%). PAU is defined as an ulceration of an atheromatous plaque that has eroded the inner elastic layer of the aortic wall. It has reached the media and produced a hematoma within the media. Brief facts: What the clinician needs to know  The complications of a Penetrating Atherosclerotic Ulcer include:  However most patients have a poor prognosis because of generalized atherosclerosis leading to diffuse organ failure. E. Casta?er et al, Radiographics 2003; 23:S93- S110 T.T. Tsai et al. Acute Aortic Syndromes. Circulation 2005;112;3802-3813 D. Mukherjee et al. Aortic Dissection. An Update. Current Problems in Cardiology 2005;30:287-325 H. Hayashi et al. Penetrating Atherosclerotic Ulcer of the Aorta: Imaging Features and Disease Concept. Radiographics 2000;20:995-1005 K.R. Cho et al. Penetrating atherosclerotic ulcer of the descending thoracic aorta and arch. J Thorac Cardiovasc Surg 2004;127:1393-401 F.Ganaha et al. Prognosis of Aortic Intramural Hematoma With and Without Penetrating Atherosclerotic Ulcer. A Clinical and Radiological Analysis. Circulation. 2002;106:342-348. Y. von Kodolitsch et al. Intramural Hematoma of the Aorta Predictors of Progression to Dissection and Rupture. Circulation. 2003;107:1158-1163 S. Willoteaux et al. Imaging of aortic dissection by helical CT. Eur Radiol, 2004;14:1999?2008 J-K Song. Diagnosis of aortic intramural haematoma Heart 2004;90:368-371. A. Evangelista et al. Acute Intramural Hematoma of the Aorta. A Mystery in evolution. Circulation 2005;111:1063-1070 N. Mangat et al. Multi-detector row computed tomography: Imaging in Acute Aortic Syndrome Clin Rad 2005;60:1256-1267Ferco Berger, Robin Smithuis, Otto van Delden Stanford classification Imaging features Dissection into brachiocephalic arteries Dissection into abdominal arteries Rupture into pericardium and thoracic cavity Aneurysm with thrombus versus thrombosed dissection Imaging features ComplicationsThoracic Aorta - the Acute Aortic SyndromeFrom the Radiology Department of the Academical Medical Centre, Amsterdam and the Rijnland Hospital, Leiderdorp, the Netherlands
cardio6
1
Vascular Anomalies of Aorta, Pulmonary and Systemic vessels
by Marilyn J. Siegel and Robin Smithuis
This review is based on a presentation by Marilyn Siegel and was adapted and illustrated for the Radiology Assistant by Robin Smithuis. Marilyn Siegel is specialized in pediatric and chest radiology. The second edition of her book entitled Pediatric Body CT will be out next week. In this review we will discuss the most common vascular anomalies of the aorta, pulmonary vessels and systemic veins in the chest. Most of these anomalies are found in children, but sometimes they are discovered later in adulthood. Many of these anomalies are asymptomatic or 'leave alone' lesions, but some of these anomalies are symptomatic and need to be treated. As a radiologist we have to be familiar with these anomalies. A simple mouse click on an item on the left will bring you directly to this subject. by Marilyn J. Siegel and Robin Smithuis Mallinckrodt Institute of Radiology, Washington University School of Medicine in St. Louis, USA Radiology department of the Rijnland Hospital in Leiderdorp, the Netherlands.             When you look at these illustrations, you have to realize, that these are views from above, while CT-images have a 'view from feet'. On a CT-image the ascending aorta will be on the upper part of the image and the descending aorta will be on the lower part. This is the mirror-image variety of the left arch. On the left a 2 year old girl with wheezing and coughing. Study the images and then continue. You have to realize, that axial CT-images are viewed from the feet, while the illustrations above are viewed from above On the axial image there is a right arch On the volume rendered image there is mirror image branching of the brachiocephalic arteries, no aberrant subclavian artery, so this is a right arch mirror image. This anomaly is asymptomatic, because there is no obstructing ring. Almost all of these patients however come to our attention because they have associated congenital heart disease in 98% of cases. This patient had a mirror image aortic arch and a VSD. On the left an adult who was operated in his childhood for a Tetralogy of Fallot (pulmonary stenosis, right ventricular hypertrophy, VSD, overriding aorta). At surgery the VSD was patched and the pulmonary outflow tract was enlarged. Notice that there is also a right arch. In the United States there are now more than one million adults who have survived their congenital heart disease.
 In the ER you will see these patients because they age and get chest pain like many adults do and so you will see these anomalies more frequently. The Right Aortic Arch with an aberrant left subclavian is an obstructing arch anomaly. The first branch of the aorta is the left common carotid, followed by the right subclavian artery and the left common carotid. This also is a true ring. The ligamentum ductus arteriosus between the arch at the level of the left subclavian artery and the left pumonary artery completes the ring. If this ligament is very short, there will be a lot of compression. On the left a patient with a right arch with an aberrant left subclavian (indicated by the yellow arrow). Scroll through the images on the left. Again you have to realize that the axial CT-images have a 'view from feet'. Which vessels are indicated by the yellow and green arrow? There is a right arch and the left subclavian artery is the last branch of the aortic arch, indicating that this is an aberrant left subclavian. Medially to the left subclavian artery we see the left common carotid, that originates from the right side and has an oblique course to the left. The yellow arrow indicates the azygos vein. The green arrow indicates the left superior intercostal vein, a normal variant, that we will discuss later. Same patient. Posterior oblique view of volume rendered image to show the aberrant left subclavian artery. In a mirror type right arch, the left subclavian is the first brach and forms the left innominate together with the left common carotid. On the left images of a symptomatic child. On the axial image there is a right arch with the left subclavian artery that comes off on the posterior side and runs behind the trachea and the esophagus. The compression of the trachea is demonstrated on the volume rendered view. On the left a chest film of a 6-month old boy with stridor and cough. The trachea is deviated to the left, otherwise the chest film is normal. So there is some mass effect on the right side. On the left the reconstructions demonstrating a double aortic arch. There are branches coming off the right arch and branches coming off the left arch. The right arch is typically larger and higher than the left. There is a complete ring that encircles the esophagus and the trachea and usually there is stridor or dysphagia. Two brachiocephalic arteries arise on each side separately (four vessel sign). On the left a chest film of a young adult with a cough. There is a right paratracheal mass. The differential diagnosis is tumor, adenopathy or vessel (right arch, dilated azygos vein, dilated aberrant right subclavian artery). On the left axial images and posterior view of volume rendered reconstruction. Describe the findings and then continue. The findings are: The narrowing of the trachea is seen on the axial images, but better appreciated on the MPR and Volume Rendered image. On the left preoperative and postoperative MDCT studies of a 2-month-old female infant with double aortic arch presenting with stridor and repeated apnea. The smaller left arch is partially resected. Double Arch with Atretic Segment Occasionally the double arch can have an atretic segment. You should not confuse it for a right arch. The left arch is just very small and there is still a four vessel sign. On the left a dominant right arch and a small left arch. The atretic segment is marked by the arrow. Notice the four vessel sign. On a posterior view the interruption is nicely demonstrated. Remember that there is still a ring, so there is still obstruction. Another case on the left. Do not call this a right arch. It still is a double arch and there is a atretic fibrotic segment on the posterior side of the left arch, that completes the ring. Notice the four vessel sign. Same patient. Always look at the airways. On the recoonstruction the impression on the trachea is better appreciated.  On the left a young patient, who has a CT for another reason. Study the images and then continue. Notice that there is a left arch, but the right subclavian artery is the last brachiocephalic artery to branch off the arch. Only rarely these patients become dysphagic (dysphagia lusoria) , when the origin of the right subclavian artery becomes dilated. On a barium study of the esophagus you will see a posterior impression with an oblique course directed towards the right shoulder. On the left a 78 year old woman with dysphagia. There is consolidation in the right upper lobe, maybe due to aspiration. There is a dilated vessel that compresses the esophagus and it originates from the left-sided aorta, i.e. an aberrant right subclavian artery. On the left the same patient with dilated aberrant right subclavian artery. Coronal reconstruction. On the left another patient with an aberrant right subclavian. Scroll through the images. When you follow the artery from inferior to superior, it starts on the left side of the arch and travels obliquely behind the esophagus to go to the right. On the left a sagittal scanogram, axial image and sagittal reconstruction of a 5 year old girl with noisy breathing and occasional episodes of cyanosis. First look at the images then continue. The findings are: The diagnosis is the innominate artery compression syndrome. In infants the innominate artery arises more to the left than in adults, so it's got to go in front of the trachea. It may compress the trachea, leading to stridor, cough and dyspnea. This compression decreases with age and these patients will outgrow it. The compression in the innominate artery compression syndrome is located on the right anterior side and at the level of the thoracic inlet. This is much higher than in the double arch or Right Aortic Arch with Aberrant left subclavian On the left another case with mild compression on the trachea.  On the left a 2 month old boy with heart failure. First study the image, then continue The findings are: The diagnosis is coarctation, which is nicely demonstrated on the posterior view of the reconstruction. There are two types of coarctation. The type we usually see is the post-ductal type, which is distal to the left subclavian artery. The uncommon pre-ductal type is seen in neonates. They present with severe heart failure, mostly within the first week of life, usually on the first day. The occlusion is in front of the left subclavian. First study the axial image followed by the sagittal reconstruction, then continue. The findings are: The intercostal collaterals typically occur between the 3rd and the 8th rib. On the left two neonates with the pre-ductal type of coarctation. The stenosis is in front of the left subclavia and there is arch hypoplasia. Collaterals do not occur, probably because they don't have time to develop. Coarctation is treated with angioplasty, stent placement or patch aortoplasty. The image on the far left is the result after angioplasty. Next to it a patient who was treated with a stent. Notice that the stent is obstructing the orfice of the left subclavian artery. On the far left a patient who was treated with a stent. The stent ruptured causing restenosis. Next to it two patients with pseudo-aneurysm. One after angioplasty and another who developed a pseudo-aneurysm after stent placement. They have to be repaired because they will rupture. Pseudo-aneurysms are seen in They most common anomalies of the pulmonary arteries are listed in the table on the left.  On the left a young adult, who had cyanotic spells as a child. She is now in good health and comes in for another reason. On the chest film the differential is atelectasis, pneumonia or maybe a tumor. The CT shows, that he right lung is not developed and the space around the atresic pulmonary artery is filled with fibrofatty tissue with collaterals. So this is pulmonary agenesis. If many collaterals develop there will also be some development of the lung. On the left another case of absent pulmonary artery with absence of lung development. On the CT the left lung is absent. These patients may be totally asymptomatic. On the left a 4 month old girl with abnormal echo, benign heart murmur and no respiratory or feeding difficulties. The sagittal reconstruction shows an anomalous vessel on the posterior side of the trachea. There is a little mass effect on the trachea. In pulmonary sling the left PA originates from the right PA and courses between the esophagus and the trachea, where it compresses the right main bronchus. Pulmonary sling is seen more frequent in children as it is more symptomatic than in adults, because the chest is smaller, but you can also encounter it in adults. On the left images of a child with wheezing and dyspnea. The left PA comes off the right PA and runs between the esophagus (with nasogastric tube) and the trachea. Some of these patients also have long segment stenosis in the trachea because of cartilagenous rings. On the left an adolescent with a murmur. On axial image and reconstruction the patent ductus arteriosus is seen. The ductus arteriosus is the communication between the pulmonary artery and the proximal descending aorta. It shunts blood in utero from the right ventricle to the aorta to bypass the non-functioning lungs. On the first day of life there is a functional closure and an anatomic closure with fibrosis in the first two weeks. If it does not close these patients come to attention either with a murmur or later with pulmonary hypertension. On the left a young adult with a murmur. The cardiologists are not interested in the flow direction, but just want to confirm the diagnosis. Notice the connection between the pulmonary artery and the descending aorta. When the duct closes it may also calcify. This a normal variant. The most common features of Partial Anomalous Venous Return are listed in the table on the left. The anomalous veins drain into the following structures: Right upper lobe anomalous venous return On the left a 2 month old, who is asymptomatic but has a murmur on physical examination. There is a connection between the SVC and a pulmonary vein, so this is an anomalous venous return. All these partially anomalous pulmonary venous returns are left to right shunts, but when small, they are clinically insignificant. When there is a significant shunt, they may cause (late) pulmonary hypertension as seen in the case on the left. The chest film in this adult shows large pulmonary arteries and a large right atrium and ventricle as a result of pulmonary hypertension. Right upper lobe anomalous return (2) On the left a patient with a murmur. Study the images and then continue. On the left a similar case. Notice the anomalous return of the right upper lobe vein into the VCS and the additional ASD at a lower level. Right lower lobe anomalous return On the left a right lower lobe anomalous return. The vein drains into the IVC. The anomalous vein gently curves to the right cardiophrenic angle and is shaped like a Turkish sword ('Scimitar') On the left another right lower lobe anomalous return. The vein drains into the azygos vein. Upper lobe veins may also drain into the azygos vein. On the left a 10 year old girl suspected of having pneumonia. Study the images carefully, because there are three findings and then continue reading. The findings are: This patient has a scimitar syndrome and also a right arch. So the lesson is, that when you see one anomaly, look for another one. The features in scimitar syndrome are listed in the table on the left. On the left another patient with a scimitar syndrome. There is a hypoplastic right lung with mediastinal shift and there is anomalous venous return. Notice that on the coronal MIP you can nicely see the difference in vascularization of the lungs with hypovascularity on the right. Scroll through the images on the left. Notice how the left upper lobe vein runs from the hilum cranially into the brachiocephalic vein. The differential diagnosis of a left upper lobe anomalous venous return into brachiocephalic veins is a left Superior Vena Cava (SVC). A left SVC however drains into the coronary sinus.  Describe the images on the left and then continue reading. On the left side there is a vascular structure, that runs inferiorly below the level of the left hilum and enters into a dilated coronary sinus. The diagnosis is left or double superior vena cava. This is an anastomosis between the accessory hemiazygos vein and the left brachiocephalic vein. It courses along the lateral margin of the aortic arch ('aortic nipple'). It is a normal variant and if you look for this structure you will frequently notice it. Catheters or pacemaker leads may course along left side of mediastinum. On the left a patient with a left superior intercostal vein. Notice the 'aortic nipple sign'. On the left another example of a left superior intercostal vein. It courses along the lateral margin of the aortic arch from the the accessory hemiazygos vein to the left brachiocephalic vein. Summary of left paramediastinal structures  Ideally a 64 slice scanner is used, but even a 4-slice scanner will suffice for studying vascular anomalies. The technique for these anomalies in the chest is the same as we use for pulmonary embolus detection. Thin collimation is used in combination with a fast table speed in order to get the highest resolution with the lowest radiation exposure. Usually a pitch of 1.5 is used. In children we preferably do not use thin collimation, because of the higher radiation exposure, but these anomalies can be very small (voorbeeld dia 18), so thin collimation is necessary. mAs and kVp In a child with a weight of less than 10Kg 40mAs will work in the chest. In children with a weight more than 45 Kg adult protocols are used with 100 mAs or more. In small children under 50 kg you can decrease the kVp to 80 and that works very well in the chest. Remember in the chest there is inherent contrast from the lungs and by dropping the kVp you enhance this contrast. On the left a 3-year old. Non-breath hold images with 50mAs and 80 kVp on a 16 row detector. Although the axial images are a little bit grainy, the reconstructions are just fine. Do these patients need sedation? Well most of the time they don't. If you can get the patient on the table and they are relatively still, even if they are breathing, you will get good studies. If you can't get the patient on the table, because they prefer the floor, you've got to sedate. In about 20-25% of pediatric studies we use sedation. If the catheter is not in the antecubital vein, hand injection is preferred. Scan Initiation Time Bolus tracking is used and the trigger is set at 120 HU. This may not always work, because in small children the amount of contast may be too small to trigger or due to breathing the cursor may fall to the lungs. If bolus tracking does not trigger, start the scan at 15 seconds. Positioning the ROI Post-processing Multiplanar reconstructions (MPR), volume rendered techniques (VRT) and maximum intensity (MIP) are very helpful. There is no role for shaded surface or mini-IP's. On the left an external and internal rendering which provides in contrast to MPR real 3D information. In volume rendering the posterior view is preferred to get a good look at the arch and descending aorta. If you want to study peripheral vessels you will need thick slab maximal intensity projections. For instance if you study arteriovenous malformations or discrepancy in lung flow. Notice that on the coronal MIP you can nicely see the difference in vascularization of the lungs with hypovascularity on the right in a patient with scimitar syndrome. 3D reconstructions are helpful when there are short focal lesions like coarctation or when vessels course obliquely (figure). Adding 3D reconstructions to axial images will increase the sensitivity from 90% to 100% (Lee, Siegel AJR 182:777-784 ) On the left a 17-day old boy with minimal cyanosis, mild heart failure and upper extremity hypertension. On axial images you will have trouble diagnosis coarctation, because it is focal and in the same plane as the axial image. If you want to read more about pediatric body CT, buy: Pediatric Body CT, 2nd edition. Lippincott Williams & Wilkins, Philadelphia. 2008 (3) by Marilyn Siegel. by Edward Y. Lee, Marilyn J. Siegel, Charles F. Hildebolt, Fernando R. Gutierrez, Sanjeev Bhalla and Juliet H. Fallah of the Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S Kingshighway Blvd., St. Louis, MO 63110. AJR 2004; 182:777-784 Diagnostic Imaging: Chest By Jud Gurney, MD et al AMIRSYS Title, ISBN: 1416023348, ISBN-13: 9781416023340 This book represents today's best single source of guidance on chest diagnostic imaging! Great vessels. In: Pediatric Body CT, 2nd edition by Marilyn Siegel Lippincott Williams & Wilkins, Philadelphia. 2008 Chan MSM, Chu WCW, Cheung KL, Arifi AA, Lam WWM AJR 2005; 185:11248-1251Marilyn J. Siegel and Robin Smithuis Embryology Right Arch Mirror Image Right Arch with Aberrant left subclavian Double Aortic Arch Left Arch Aberrant Right SCA Innominate artery compression syndrome Aortic Coarctation Pulmonary agenesis Pulmonary Sling Patent Ductus Arteriosus Partial Anomalous Venous Return Scimitar syndrome Left Superior Vena Cava Left Superior Intercostal Vein. Azygos Continuation of IVCVascular Anomalies of Aorta, Pulmonary and Systemic vesselsMallinckrodt Institute of Radiology, Washington University School of Medicine in St. Louis, USA and the Rijnland Hospital in Leiderdorp, the Netherlands
headneck1
1
Infrahyoid neck
by Frank Pameijer, Erik Beek, Frank Joosten and Robin Smithuis
In this article we will focus on: The infrahyoid neck is the region of the neck extending from the hyoid bone to the thoracic inlet. Traditionally the anatomy of the infrahyoid neck has been subdivided into a group of surgical triangles whose borders are readily palpable bones and muscles (figure). These triangles have a cranial-caudal orientation and therefore are difficult to correlate with cross-sectional imaging. Another approach to the anatomy of the neck is the so-called 'spatial approach', which we shall use in this review. Notice the musculus omohyoideus in the illustration. This is one of the 'strap' muscles, an important landmark in the neck. It is a group of four pairs of muscles in the anterior part of the neck: the sternothyroid, sternohyoid, thyrohyoid and omohyoid muscles. They are all attached to the hyoid bone and look like a strap. The other strap-muscles are not drawn in this illustration. In the spatial approach to the anatomy of the infrahyoid neck, the cross-sectional anatomy is described as a series of spaces defined by the various layers of the deep cervical fascia. This facilitates the understanding and interpretation of cross-sectional imaging modalities like CT and MRI (1). Some of these infrahyoid spaces are continuous with the suprahyoid neck and some are continuous with the superior mediastinum. The infrahyoid neck is divided into 5 major anatomical compartments or spaces by the various layers of the cervical fascia (2). These spaces are well recognized in the axial plane and therefore suited for analysis on axial CT or MR.  On the left a CT image of a patient with massive subcutaneous emphysema after a motor vehicle accident. Air has dissected along the layers of the cervical fascia. Notice that you are able to find all five spaces - they are now outlined by air. The systematic approach to pathology in the infrahyoid neck is a three-step procedure: The visceral space extends from the hyoid to the anterior mediastinum and does not extend into the suprahyoid space. On the left the normal contents and derived pathology of the visceral space. The CT section is at the level of the supraglottic larynx and the thyroid cartilage. Anterior to the thyroid cartilage are the 'strap' muscles: sternothyroid, sternohyoid, thyrohyoid and omohyoid . They are all connected to the hyoid and depress the hyoid bone and larynx during swallowing and speaking. These muscles are long and flat much like a strap. We will now continue with a few cases. Although we have provided the diagnosis in these cases, we still want you to follow the 3-step approach. Step 1: which space On the left a patient with a swelling on the right side of the neck. Study the image and decide in which space the lesion is located. Then continue reading. The swelling is centered within the borders of the thyroid cartilage. Therefore this must be pathology arising in the visceral space. Step 2: normal contents On the left an additional image is shown at a slightly lower level. Study the images and decide which structures in the visceral space are present at this level and which are not. Then continue reading. The CT section shows the lesion present at the level of the supraglottic larynx and the thyroid cartilage. So we are much too cranial for trachea, thyroid gland, parathyroid glands and recurrent laryngeal nerve, which lies in the tracheo-esophageal groove. Paratracheal Level VI lymph nodes are located around the larynx and not within the larynx, so they can be ruled out. The hypopharynx is posterior to the lesion and has a normal appearance. Embryological remnants like thyroglossal duct cyst can be considered, but these are typically embedded in the laryngeal strap musculature and therefore should be located anterior to the thyroid cartilage. So the only normal anatomy from which this lesion could have arisen is the larynx. Step 3: Pattern recognition This lesion presents as a cystic lesion with sharply defined, enhancing boundaries. The lesion is located in the supraglottic larynx in the right paraglottic space and also has an extralaryngeal component, which explains the lump on the right side of the neck. At endoscopy a large submucosal swelling on the right was seen in the larynx. Squamous cell cancer, which is a mucosal disease, can therefore be dismissed. When we think about the radiological appearance of the four submucosal entities mentioned in the table on the left, we can make the following remarks: Laryngocele (2) When a larynocele is suspected you always have to search carefully for an underlying cause. Primary laryngocele has no underlying cause. Secondary laryngocele arises due to pathology in the laryngeal ventricle, which is a slit-like opening between the true and the false vocal cords. A secondary laryngocele is frequently caused by a squamous cell carcinoma, as in this case. At endoscopy the tumor may be obscured by the laryngocele itself. Mechanism of a laryngocele The laryngeal ventricle (v) is a slit-like opening between the false and true vocal cords (image far left). It is the anatomic landmark between supraglottis and glottis. The ventricle extends laterally and then cranially into the paraglottic space. When the opening of the laryngeal ventricle is completely obstructed by tumor, the mucosa in the paraglottic space continues to produce fluid. This results in a fluid-filled internal laryngocele. Eventually the paraglottic space becomes filled up and the internal laryngocele will become external by extending outside of the larynx through the thyro-hyoid membrane. When the opening of the laryngeal ventricle is partially obstructed, a pressure-valve mechanism may result in an air-containing internal laryngocele which may, eventually, become external (right image, red arrow). On the left, a CT-image at the level of the thyroid cartilage. There is an irregular mass centered in the right piriform sinus. This mass is in the visceral space. In this region the most common tumor is a squamous cell carcinoma. This was proven at biopsy. Notice the retropharyngeal space (yellow arrow). This is a virtual space containing only some fat. Squamous cell carcinoma (2) On the left, contiguous slices in a craniocaudal direction at the level of the larynx. Study this case, which is quite similar to a previously discussed case and then continue reading.  Step 1: Which space On the left a patient with a swelling on the left side of the neck, which has existed for years. The swelling is adjacent to the left lamina of the thyroid cartilage. The strap musculature seems to be draped over the lesion (blue arrow). Therefore this lesion lies within the visceral space. Step 2: Normal contents Analysis of the normal anatomical contents of the visceral space rules out many possible tissues and organs from which this pathology may arise: By exclusion we can say that this mass arises either from the thyroid gland or the parathyroid glands. Step 3: Pattern recognition and clinical information On the chest film we notice a displacement of the trachea to the right by an upper mediastinal mass. So the mass is located within the visceral space and extends into the anterior mediastinum, since the trachea is located within the anterior mediastinum. It is well-defined towards the surrounding fat and there are a few scattered coarse calcifications. When we combine these findings, we recognize the radiological pattern of a benign multinodular goiter. This diagnosis is compatible with the clinical information that the swelling in the neck has been present for years. On the left axial T1- and T2-weighted images at the level of the hyoid bone. There was no enhancement on the post Gadolinium study (not shown) It is a midline cystic lesion, party external and partly internal to the hyoid bone and located in the visceral space. The lesion is embedded in the strap musculature. It is unlikely that this lesion arises from the trachea, thyroid gland, parathyroid glands or recurrent laryngeal nerve, since these structures are located more caudally. Lymph nodes are usually seen in the subcutaneous fat around the larynx. By exclusion a thyroglossal duct cyst is the most likely diagnosis. Thyroglossal duct cyst (2) Key facts On the left an example of a paramedian thyroglossal duct cyst. This lesion not in the midline, but the key finding is that this lesion is cystic and embedded in the strap musculature. Thyroglossal duct cyst (3) When the diagnosis thyroglossal duct cyst is made, always check if there is a thyroid in the normal position. The thyroid anlage may never travel along the thyroglossal duct. In that case it stays at the tongue base. In these rare cases, the patient has a so-called lingual thyroid. On the left another paramedian thyroglossal duct cyst On the left, a child with a lingual thyroid. This is the only functioning thyroid tissue that this child has. It would be a disaster if such a 'lesion' were to be excised. On the left images of a three-year old girl with a slowly enlarging tumor in the midline. On ultrasound a hypoechoic ovoid smooth bordered lesion is seen at the level of the hyoid bone and slightly right off midline (left image). During US examination, the lesion moves simultaneously with extrusion of the tongue. Cystic nature and close relation to the hyoid bone makes thyroglossal duct cyst the most likely diagnosis. Notice that a normal thyroid gland is present (right image). The carotid space extends from the skull base to the aortic arch. It transverses the suprahyoid and infrahyoid neck into the anterior mediastinum. On the left the normal contents of the carotid space and the derived pathology. Step 1: Which space On the left a patient with a swelling on the left side of the neck. Study the MR-image at the level of the supraglottic larynx and decide in which space the lesion is located. Then continue reading. The swelling is centered between the external and internal carotid artery. Notice that these vessels are compressed. Evidently this lesion must be located in the carotid space. Please note that there is a smaller, but identical, lesion present, located in the right carotid space. Step 2: Normal contents Now we must try to figure out the normal anatomical source that has caused this pathology. Once again, we use exclusion: Therefore it is very likely that the bilateral swellings of this patient are coming from these neural structures. Now we are down to a fairly limited and space-specific differential diagnosis (see next image).  Paraganglioma (2) On the left images of a 21-year old female with a mass on the right. This lesion is located between the internal and external carotid artery and therefore is a neural tumor. The differential diagnosis is limited to tumors arising from the vagus nerve and sympathetic plexus. On CT and color doppler the mass is clearly hypervascular and the only possible diagnosis is a paraganglioma. Paraganglioma (3) Key facts On the left images of a 28-year old female with a nontender mass at the left mandibular angle. Step 1 The mass is located in carotid space. Step 2 Anatomical contents: carotid artery, internal jugular vein, vagus nerve, sympathetic plexus, lymph nodes (Level 2-4) and congenital remnants of the 2nd branchial cleft . Step 3 Therefore it is very likely that this mass has a neural origin: Thrombosis of the internal jugular vein is an under-diagnosed condition that may occur as a complication of head and neck infections, surgery, central venous access, and intravenous drug abuse. An infected jugular vein thrombus caused by extension of an oropharyngeal infection is referred to as Lemierre's syndrome. This is a bacterial infection that may have severe morbidity or even fatal outcome, as eventually septic emboli may spread to the lungs. On the left a patient who had undergone a total laryngectomy several years previously. The present complaint is a painful swelling on the left side of the neck since one day. Step 1 Contrast-enhanced CT at mid-neck level shows the neopharynx with some surgical clips on the left and the enhancing right thyroid lobe which is still in situ. In addition there is a round, hypodense lesion in the left carotid space. Step 2 In this case, analysis of the normal anatomical contents of the carotid space can be short. When we compare left to right it is clear that we are dealing with an internal jugular vein which is enlarged and does not enhance. An image at a higher level shows the same appearance of the internal jugular vein (arrow). Step 3 In combination with the clinical information of a painful swelling on the left side of the neck, there is only one possible diagnosis: Acute thrombosis of the left internal jugular vein. Lemierre' s syndrome When you diagnose an acute thrombosis of the internal jugular vein, always look for pulmonary symptoms, i.e. Lemierre' s syndrome. The oral infection spreads into the neck and causes internal jugular vein thrombophlebitis with subsequent septic emboli. On the left images of a 36-year old female with a progressive swelling on the left side of the neck. She had recently recovered from a peritonsillar abscess. ENT-examination showed a fistula orifice in the left tonsil. The position of the mass on the CT indicates that it is located in the carotid space. Analysis based on normal anatomical contents: Second branchial cleft cyst (2) Key facts On the left a longitudinal and transverse image fo a large second branchial cleft cyst in a 12-year old girl situated between the parotid gland (left image) and the submandibular gland (right image). It is superficial to the carotid artery and jugular vein (arrow). The cyst contents is hypoechoic with freely moving debris. The MR of the same patient confirms the presence of a cystic mass in the right carotid space. The lesion is situated between the submandibular gland and the anterior margin of the sternocleidomastoid muscle, which is the classic position of a second branchial cleft cyst. The lesion shows edge enhancement post-Gadolinium. Notice that these lesions may contain small amounts of lymphoid tissue which is a possible explanation for the small area of enhancement inside the cyst wall (arrow). Coronal STIR image of the same cyst. The retropharyngeal space extends superiorly to the base of the skull and inferiorly to the posterior mediastinum at the level of the tracheal bifurcation. In normal circumstances, the retropharyngeal space is a virtual space and contains the retropharyngeal lymph nodes superiorly as well as some fatty tissue. Infections of the mouth can spread through this space into the posterior mediastinum. There are two other spaces in close proximity to the retropharyngeal space: the danger space and the prevertebral space. They are often confused with the retropharyngeal space. The danger space actually lies between the alar fascia, which forms the posterior border of the retropharyngeal space, and the prevertebral fascia. It extends from the cranial base above to the level of the diaphragm. The prevertebral space is bounded anteriorly by the prevertebral fascia and posteriorly by the longus colli muscles of the spine. It extends down the mediastinum and continues to the insertion of the psoas muscles. All we need to realize is that pathology in this region can extend into the posterior mediastinum and we should not stop imaging until the tracheal bifurcation is reached. On the left an axial contrast enhanced CT-image of an infant with fever. The child cannot swallow. This pathology is located in the retropharyngeal space. The prevertebral muscles are pushed towards the vertebral body. If this were a lesion located in the perivertebral space, these muscles would be pushed anteriorly. The retropharyngeal space is expanded and contains multiple pockets of material with fluid density. Obviously this is a retropharyngeal infection with multiple abscesses. This is an emergency situation because the abscesses will expand and may eventually obstruct the airways. Usually these deep abscesses require surgical drainage. On the left an adult patient with a retropharyngeal abscess after surgical drainage. The drainage catheters run from left to right through the retropharyngeal space. The retropharyngeal space can become infected in two ways. The 'classic' retropharyngeal abscess observed in pediatric patients occurs when an upper respiratory infection like pharyngitis or adenoiditis spreads to retropharyngeal lymph nodes. Penetrating trauma, i.e. foreign bodies, fish bones or iatrogenic causes such as endoscopy or intubation, can also be involved in retropharyngeal space infection. On the left two images of a patient with a piriform sinus carcinoma (shown earlier). On the far left the pirifrom sinus tumor before radiotherapy. The image next to it shows complete response after the radiotherapy. The retropharyngeal space is now distended and shows increased density of the fatty tissue due to post-radiotherapy edema. On the left a table with the normal contents of the posterior cervical space and subsequent pathology. MPNST is short for malignant peripheral nerve sheath tumor, which is the proper name for a malignant schwannoma. Step 1: Which space On the left a patient with bilateral swelling in the neck. CT image at the level of the hyoid bone shows multiple rounded lesions medial to the sternocleidomastoid muscles and dorsal to the internal jugular veins. These bilateral multiple lesions are located in the posterior cervical space. Step 2 Analysis of the normal anatomical components of the posterior cervical space: Step 3 By exclusion we can conclude that these are bilaterally enlarged lymph nodes with homogeneous enhancement. Homogeneous enhancement is typical for lymphoma. Central necrosis is more typical for squamous cell carcinoma metastases. Lymph node biopsy in this patient revealed B-cell Non-Hodgkin lymphoma. Lymphoma (2) On the left images of a 67-year old woman who had a history of Non-Hodgkin lymphoma. She had recently noticed a swelling on the left side of the neck. Step 1 CT image at the level of the true vocal cords shows a mass, which is clearly located in the posterior cervical space. Step 2 The mass is well-defined and isodense to muscle. Coronal reformation shows the mass to be elongated and extending towards the axilla following rhe course of the cervico-brachial plexus.. Continue with the MR images. The lesion originates at the left neural foramina and grows along the course of the brachial plexus (red arrow). In fact, we are looking at a grossly thickened plexus. Step 3 The radiological pattern confirms the neurogenic origin of the mass. Combined with the history the final diagnosis is diffuse infiltration of the left brachial plexus by recurrent NHL. On the left images of a patient with a swelling posteriorly on the left side of the neck. MR image at the level of the hyoid bone. The lesion is located in the posterior cervical space. Analysis of the normal anatomical components of the posterior cervical space can be short in this case. The mass has the signal intensity of fat on a T1-weighted image and the signal is completely suppressed with fat suppression. There was no enhancement (not shown), so we can conclude that this is a lipoma. On the left T1- and T2-weighted images of another patient with a lipoma. On the left an axial T2-weighted image with fatsat and a coronal T1-weighted image of a 12-year old girl who presented with a soft swelling in the neck. Step 1 A multiloculated lesion is present in the posterior cervical space. Step 2 Analysis of the anatomical components: Step 3 On the T2-weighted image with fatsat the lesion is multiloculated and has a fluid intensity. There is no enhancement on the T1-weighted image. These findings, in combination with the fact that the swelling is soft and is present in a child, is specific for the diagnosis of a lymphangioma, also known as cystic hygroma. Lymphangioma (2) Key facts On the left a table with the normal contants of the perivertebral space and subsequent pathology. On the left a contrast enhanced CT image through the upper neck of a patient who complained of a slowly growing swelling posteriorly on the left side of the neck. Step 1: Which space There is a large soft tissue mass adjacent to the vertebral body centered in the perivertebral space. Step 2 Analysis of the normal anatomical components of the perivertebral space: Step 3 The normal fat planes between the individual muscles have disappeared. The imaging characteristics are otherwise non-specific. This leaves us with a fairly large differential diagnosis of muscle pathology: sarcoma, fibromatosis, lymphoma and infection. The clinical information of a slow-growing mass favors a malignant process. Biopsy of the mass revealed sarcoma. On the left a 75-year old male with a slowly enlarging mid-line mass. Step 1 The lesion is located in the perivertebral space. Step 2 The vertebral body and vertebral vessels are not involved. Lesions coming from the cervico-brachial plexus are expected to be found in more paraspinal locations. There is vivid enhancement of the mass. Centrally flow voids are present, indicating a hypervascular nature. The imaging characteristics are otherwise non-specific. Step 3 The clinical information of a slow-growing mass favors a malignant process. A biopsy taken before excision revealed a benign fibrous tumor. Handbook of Head and Neck Imaging by H. Ric Harnsberger, 2d ed. Mosby 1995 Infrahyoid neck: CT and MR imaging versus histopathology by Becker M, Kurt AM. Eur. Radiol. 2001; 11 (Suppl. 2):S23-S38 by Amparo Castellote et al May 1999 RadioGraphics, 19, 583-600. Online powerpoint presentation Infrahyoid Neck by Sigal R. Radiological Clinics of North America 1998;36:781-799. Head and neck radiology; a teaching file. Lippincott, Williams&Wilkins 2002. by Anthony A. Mancuso, Hiroya Ojiri, Ronald G. Quisling. Atlas of Head and Neck Imaging, Thieme 2004. by Suresh Mukherji, Vincent ChongFrank Pameijer, Erik Beek, Frank Joosten and Robin Smithuis Surgical triangles Spaces of the infrahyoid neck Laryngocele Squamous cell carcinoma Multinodular goiter Thyroglossal duct cyst Paraganglioma Schwannoma Jugular vein thrombosis Second branchial cleft cyst Retropharyngeal abscess Retropharyngeal edema Lymphoma Lipoma Lymphangioma or Cystic hygroma Sarcoma Benign fibrous tumorInfrahyoid neckRadiology department of the University Medical Centre of Utrecht, the Rijnstate Hospital in Arnhem and the Rijnland hospital in Leiderdorp, the Netherlands
headneck2
1
Orbita - pathology
by by David Youssem
This review is based on a presentation given by David Yousem and adapted for the Radiology Assistant by Robin Smithuis. David Yousem is currently the Director of Neuroradiology and Professor of Radiology at the Johns Hopkins Hospital. He is also the editor of the book 'Neuroradiology: the Requisites'. In this article a systematic approach to orbital pathology is presented based on division of the orbit into the following compartments: The first thing you do when you see a lesion in the orbit, is to decide whether it is an ocular lesion or a non-ocular lesion, i.e. is it involving the globe or involving the structures outside the globe. If it is a non-ocular lesion, the next question is whether the lesion is located within the intraconal space, i.e. within the space bounded by the cone formed by the extraocular muscles, or whether it is located within the conal or extraconal space? Once you have decided where the lesion is located, consider the differential diagnostic possibilities using the mnemonic VITAMIN C and D. We will first describe the anatomic spaces of the orbit and summarize the pathology within these spaces, even if some of these pathologies are not visible radiologically. Then we will discuss the radiological findings in certain orbital diseases. Ocular space The eye has the following well defined anatomic spaces: Anterior chamber When we move from anterior to posterior the first area is the anterior chamber. It is bounded by the cornea anteriorly and the lens and iris posteriorly. Specific pathologies within the anterior chamber are: Posterior chamber This is a very small area posterior to the iris, which we cannot discern on imaging. Specific pathologies in this area are: glaucoma, uveitis and ciliary melanoma. Vitreous body The larger area posterior to the lens is the vitreous body. Specific pathologies within the vitreous body are: The vitreous body is surrounded by the membranes of the retina, the choroid and the sclera. Retina pathology: Choroid pathology: Sclera pathology: Intraconal space The ocular muscles within the orbit form a muscle-cone. These ocular muscles are connected via the annulus of Zin, which is a fibrous connective tissue sheet and together they form the conal space. It separates the intraconal from the extraconal space. Intra-orbital pathology which is non-ocular is either in the intraconal, conal or extraconal space. Intraconal space pathology: Conal space The conal space is formed by the ocular muscles and an envelope of fascia. Conal space pathology: Extraconal space The extraconal space is the area outside the muscle cone. Extraconal space pathology: Orbital appendages The lacrimal gland is located superolaterally in the orbit. Diseases of the lacrimal gland can be divided into granulomatous, glandular and developmental (see Table). Secretions go medially across the globe and are collected in the punctum and then go into the lacrimal sac. From the lacrimal sac secretions travel inferiorly to the nasal lacrimal duct, which drains under the inferior terminate into the nose. In children congenital obstructions of the valves in the lacrimal duct can lead to cystic areas medially in the orbit also known as dacryocystoceles. In adults obstruction is more often due to strictures from ethmoid sinusitis or stones blocking the nasolacrimal duct. This will result in epiphera or increased tearing. Drainage can be improved with balloon dilatation. In adults the most common intraorbital calcifications occur at the tendinous insertion of the ocular muscles. Other common calcifications are at the optic nerve head within the eye, also called 'optic disc drusen'. These are usually asymptomatic, but when the ophtomologist inspects the eye, there is the impression of papilledema, i.e. pseudo-papilledema. In children calcifications in the globe means retinoblastoma until proven otherwise even if it is bilateral. On the left an image of an adolescent with bilateral retinoblastoma. As you can see in the table on the left, retinoblastoma is a one of the more common tumor in the first year of life. The other tumors in this age group are neuroblastoma, Wilm's tumor, leukemia and teratoma. All bilateral cases are hereditary and result from a deficient tumor suppression gene on chromosome 13. The diseases that are listed in the differential diagnosis are all uncommon. On the left images are of a 13 month old female with bilateral lesions as a result of bilateral retinoblastoma. Small retinoblastomas are treated with different kinds of therapy (cryoablation, laser photocoagulation, chemothermotherapy, brachytherapy, plaque radiotherapy) in order to save the eye and avoid enucleation. If the patent is treated with radiation, there is a 30% chance of a second malignancy within the radiation field, due to the radiation but also due to the deficient tumor suppression gene. Outside the radiation field there is an 8% chance of malignancy. In order of frequency: Osteosarcoma > other sarcoma > melanoma > carcinoma. These patients are also at risk for pineal tumors and parasellar PNETs. The pineal gland is considered as the third eye and the third testicle. Meaning, you can develop retinoblastoma in the pineal gland, i.e. trilateral retinoblastoma, but also germinoma. Always examine the brain in these patients and remember that at the age of 0-4 years, which is the peak age for retinoblastoma, the pineal gland does not calcify, so any calcification in this region is suspicious of retinoblastoma. On the left images are of another patient with retinoblastoma. This tumor presents as a large calcification. When a retinoblastoma occupies more than half of the globe, as in this case, the eye has to be enucleated. Leukocoria Usually, when a light shines through the iris, the retina appears red to the observer. In leukocoria (white pupil) the retina abnormally appears white. Retinablastoma is usually detected through leukocoria as it occurs in two third of patients with retinoblastoma. These children are usually too young to present with visual complaints. There are many causes of leukocoria as listed in the table on the left. On the left images of an adult with an ocular mass. The most common intraocular lesion in an adult is melanoma (as in this case). Number two is metastases and others like hemangioma, leiomyoma and osteoma are uncommon. On the left another cause of leukocoria. This is persistent hyperplastic primary vitrous (PHPV). There is a persistent hyaloid canal when the hyaloid artery does not integrate. On the images we see a persistent canal that goes from the optic nerve to the lens. There is also retinal detachment (occurs in 30-55%) and notice the microphtalmia. PHPV is the second most common cause of leukocoria. These patients also develop glaucoma and cataract. Coats' disease is a rare eye disorder of unknown cause, leading to full or partial blindness, characterized by abnormal development of blood vessels behind the retina. On the left images of a patient who presented in the ER with post-traumatic orbital swelling. This patient has globe rupture and specifically rupture of the anterior chamber. As radiologists we are used to looking at the vitreous body if we think of globe rupture, but that is not enough. Notice that the depth of the anterior chamber is decreased. There is increased density anteriorly as a result of hyphema (blood in the anterior chamber). Also notice that the lens on the right side is blurred and slightly less dense. This is called a traumatic cataract. Maybe you would have expected the lens to be more dense, but that is usually not the case. On the left CT images of a patient who had a left eye trauma. Study the images for 5 findings and then continue reading. The findings are : Globe rupture is seen most commonly at the anterior chamber. Blood can be located in the following locations: Retinal detachment can be distinguished from choroidal detachment, because the retinal epithelium ends at the ora serrata (figure). Evidently a retinal detachment will not go beneath this point. Retinal detachment with haemorrhage is seen mostly in adults with diabetes mellitus and hypertension. In young infants it can be seen as part of a shaken baby syndrome. In choroidal detachment recent intraocular surgery is the most common association followed by trauma. On the far left a CT of a choroidal detachment going beyond ten and two o'clock (with the lens at twlve o'clock) and evidently more anteriorly to the ora serrata. It looks as if the detachment ends at the optic nerve but, if you look carefully, the choroidal detachment actually crosses the optic nerve. That would be very unusual for a retinal detachment, but is sometimes seen in choroidal detachment. On the right a T1WI of a retinal detachment. It ends at the optic nerve and at the ora serrata. On the left an image of another case of choroidal detachment. Coloboma is a congenital malformation in which part of the eye does not form due to failure of fusion of an embryonic structure called the intraocular fissure. Often there is microphtalmia and the eye protrudes inferiorly. In 10% there are other CNS anomalies. On the left images of a patient with bilateral colobomas. Coloboma can be part of the CHARGE syndrome: Coloboma can also be part of the COACH syndrome: On the left images of a patient with a small coloboma at the entrance of the optic nerve. The patient on the left had a coloboma and also agenesis of the corpus callosum with an associated midline lipoma. Devic's syndrome is also known neuromyelitis optica. Let's first look at the images and then discuss it in more detail. On the left image there is a normal optic nerve on the right side. Notice that the optic nerve is white matter tract. It has the same signal intensity as the white matter in the brain. On the contralateral side there is high signal intensity in the optic nerve. This is therefore extra-ocular intraconal disease and we will be thinking of neoplastic versus demyelinating diseases. Continue with the next image. On the left a FLAIR image with fat-sat. Notice the abnormal signal intensity and the fact that the optic nerve is not enlarged, which argues against the possibility of a tumor. Images of the cervical spinal cord show a long segment of non-space occupying disease. Based on these images the differential diagnosis is MS and Devic's syndrome (also caled neuromyelitis optica). Since MS is far more common, this would be the most likely diagnosis, but this happened to be Devic's syndrome. Some consider Devic's syndrome as a form of MS, but Devic's syndrome differs from MS: On the left images of a different patient, who also has optic neuritis. There is high signal in the optic nerve and in the brain there are multiple lesions as a result of MS. These lesions did not occur at the same time, so there is dissemination in time and in place, which is specific for MS.  On the left images of another patient with extra-ocular intraconal disease. First look at the images, describe them and come up with a differential diagnosis (for a moment disregard the fact that the title of this paragraph is meningioma). The optic nerves are normal, but there is abnormal mass-like enhancement of the optic nerve sheath on the left. So this is probably a neoplasm and of the neoplasms meningioma is by far the most common optic nerve sheath tumor. Meningiomas present with visual disturbances early in the course of the disease as a result of ischemic neuropathy due to venous obstruction. Clinically this presents as a pale disk. Abnormal enhancement of the optic nerve sheath On the left a table with the differential diagnosis of abnormal enhancement of the optic nerve sheath, also called optic nerve tram track sign. Meningioma of nerve sheath is a result of subdural growth leading to progressive visual loss, papilledema, optic atrophy. There is a strong association with NF-2. The pale disk is due to venous outflow impairment. Calcifications are seen in 20-50%. Seeding into the subarachnoid space is another cause of abnormal ennhancement of the optic sheath due to the fact, that the sheet of the optic nerve communicates with the intracranial compartment of the subarachnoid space). First look at the images on the left. Which side is abnormal and what is the most likely diagnosis? There is sphenoid wing hypoplasia on the right and on the left the optic nerve near the chiasma is enlarged (visible on the MR). So the diagnosis is neurofibromatosis type I with sphenoid wing hypoplasia and an optic pathway glioma. The term optic nerve glioma is a misnomer. Actually the tumor can present anywhere along the optic tract from the occipital region to the chiasm and the optic nerve. The term glioma is also rather non-specific. These tumors are juvenile pilocytic astrocytomas WHO type 1, which is the most benign form of astrocytoma. They make up 4% of all orbital tumors. More than 50% of patients who have an optic nerve glioma have NF1, but in NF1 only about 10% have optic nerve glioma. They are less commonly cystic in NF than in non-NF. The mean age at diagnosis is 4-5 years and only 20% of these patients have visual symptoms, because the glioma does not affect the optic nerve early and because these small children do not complain of vision problems. On the left another case with a more typical example of optic nerve glioma also in a patient with NF1. The criteria for the diagnosis of NF1 are met in an individual if 2 or more of the following signs are found: Take a look at the images on the left, describe them and come up with a differential diagnosis and again disregard the title of this paragraph. The diagnosis is thyroid eye disease and the differential diagnosis is pseudotumor of the orbit. In a moment we will discuss how to differentiate these two diseases. In the past the term Grave's ophtalmopathy was used. This however suggested that the patient is hyperthyroid. Nowadays we know that patients that are treated for Graves disease can be euthyroid or even hypothyroid and still develop thyroid eye disease and therefore nowadays we use the term thyroid eye disease. The great danger of thyroid eye disease is compressive optic neuropathy either due to direct compression by the muscles or ischemic by compression of the vessels. The key feature to look for is the orbital apex. If you do not see any fat around the orbital apex, there is a great chance of compression. These patients are treated with decompression through an endoscopic procedure in which the medial wall of the orbit i.e. the lamina papyracea is crushed. Take a look at the images on the left. This is a case of pseudotumor. Pseudotumor is idiopathic inflammation of the orbit. It can affect every part of the orbit: muscles, tendons, fat, optic nerve, nerve sheet, lacrimal gland etc. The key distinction between pseudotumor and thyroid eye disease is the fact that in pseudotumor not only the muscles, but also the tendons are involved. These patients feel pain when they are moving their eyes as the tendons get irritated. On the image on the far left notice the tapering of the swollen muscle at the point of the tendinous insertion in a patient with thyroid eye disease. Next to it the images of the patient with pseudotumor. Notice that the swelling includes to the tendinus insertion. On the left nonenhanced CT-images of a patient with a evident periosteal or periorbital abscess. The teaching point to make is the following: Do not wait for peripheral enhancement to call it an abscess! In every other location you wait for nice rim enhancement to call it an abscess and if not you say it is a phlegmone. The treatment is the treatment of the sinusitis. The other teaching point to make is the following: In children be very careful about extension outside the sinuses! Any change outside the sinus should be called an abscess. In children the periorbita is far more fenestrated and disease will easily spread. So be aggressive in calling small abnormalities an abscess. Periorbital abscess can lead to venous thrombosis of the superior and inferior ophtalmic vein. In certain fungal sinusitis (e.g.aspergillosis) you can even get cavernous sinus thrombosis and cavernous-carotid fistula On the left images of a patient who presented in the ER with a 'red hot eye' and proptosis. Now the difference between orbital and periorbital cellulitis is an important one and is based on an anatomic structure, which is called the orbital septum. If a patient comes in the ER with a red hot eye and the inflammation includes the orbital septum and everything superficial to it, the diagnosis is periorbital cellulitis and the patient is treated with oral antibiotics on an outpatient basis. In the case on the left however the structures posterior to the septum are also involved. This patient has an orbital cellulitis and will have to stay in the hospital to receive antibiotics intravenously. On the left a CT image of a patient with proptosis due to a sphenoid wing lesion. There are four sphenoid wing lesions that can cause proptosis: Lacrimal gland lesions are listed in the table on the left. Inflammatory conditions are by far the most common lesions of the lacrimal gland (i.e. Sj?gren's, TB, fungus, pseudotumor). These conditions do not cause masses. The most common mass of the lacrimal gland is lymphoma followed by pleomorphic adenoma. Epithelial tumors including adenoid cystic tumors are uncommon. Vascular malformations can be intraconal, extraconal or multicompartment and that is the reason why they are not discussed in the categories above. Before we show you some cases, we have to discuss the Mulliken and Glowacki system for the categorization of vascular anomalies in the head neck region, which is widely accepted now. The first lesion in the Mulliken & Glowacki system is the capillary hemangioma. Capillary hemangiomas have the following characteristics: The second lesion in the Mulliken & Glowacki system is the venous vascular malformation. On the left an image of a venous vascular malformation. There is a lesion in the intraconal compartment with a phlebolith. Most are unilocular, but this one is multilocular. Venous vascular malformations have the following characteristics: The next entity is the lymphatic or veno-lymphatic malformation. These are little cystic areas, that often bleed after minor trauma. They may have high signal intensity on T1WI due to high protein or hemorrhage. they usually do not enhance unless there is a venous component, that may show enhancement. This lesion is in the 'family' of cystic hygroma. More characteristics are: On the left images of a patient with an orbital varix, who had noticed that during straining there was a propulsion of the left eye . The upper image is during rest and the lower image is during valsalva at the moment of sneezing. During valsalva the varix shows extreme dilation (red arrow). Notice that during valsalva also on the normal side the superior ophthalmic vein dilates (blue arrow). Important issues to remember are: And think of VITAMIN C and D by David Youssem Calcifications Retinoblastoma Melanoma Persistent hyperplastic primary vitrous (PHPV) Coats' disease Globe rupture Retinal and choroidal detachment Coloboma Devic's syndrome MS Meningioma Optic nerve glioma Thyroid eye disease Pseudotumor Periorbital abscess Orbital and periorbital cellulitis Sphenoid wing lesions Lacrimal gland lesions Capillary hemangioma Venous Vascular Malformation Lymphatic malformations Orbital varixOrbita - pathologyNeuroradiology department of the Johns Hopkins Hospital in Baltimore
headneck3
1
Paranasal Sinuses - MRI
by Laurie Loevner and Jennifer Bradshaw
This article is based on a presentation given by Laurie Loevner and adapted for the Radiology Assistant by Jennifer Bradshaw. This review focusses on the complimentary roles that CT and MR play in the assessment of: CT is of value for determining anatomic landmarks and variants. This information is of vital importance to the ENT-surgeon. In addition, we need it to identify erosive processes and acquired developmental deficiencies of the bone. CT is also excellent for determining whether there is intraorbital extension of sino-nasal disease in the ventral 2/3 of the orbit. When pathology approaches the orbital apex, an MRI study is necessary to assess spread to the cavernous sinus and intracranial compartment. CT is performed without contrast medium. If additional imaging is necessary, orbital MRI is the next step. The real value of unenhanced CT is the following: if you see an opacified sinus with hyperdense contents, it is usually a sign of benign disease. Tumor is not hyper-dense. The hyperdensity is due to one or a combination of the following: On the left you see a case that was initially interpreted as a tumor. There is hyperdense material in the posterior right ethmoid, the bilateral spheno-ethmoidal recesses, the sphenoid sinus and there is involvement of the clivus. The hyperdensity is a good prognostic sign, indicating a benign process. This is an example of allergic fungal sinusitis. Usually it is more anteriorly located. On the left another, more characteristic, example of allergic fungal sinusitis. There is bilateral opacification of the nasal cavities, usually a sign of an inflammatory process or polyps. Note the concentric lamellated appearance of alternating hyper- and hypodensity in the maxillary sinusses. The hyperdensity is due to inspissated secretions and fungal elements. The hypodensity reflects cysts, mucosal disease, and granulation tissue. In the ethmoidal region some of the hyper-density reflects periostitis and neo-osteogenesis along the septae. MRI is extremely helpful in complicated sinonasal disease. MRI can discern secretions and mucosa from masses. When you understand the signal characteristics, you are readily able to distinguish soft tissues masses from inspissated secretions. The signal intensity of secretions can vary and mainly depends on the ratio of water to protein and the viscosity. Different protein contents result in different signal intensities on T1 and T2W-images (figure). Fungus usually has a high protein content of more than 28% and can mimic an aerated sinus because it is low on T1- and T2WI. You need CT to make the distinction! MRI is also useful for determining invasion of the skull base. Involvement of the skull base is seen as replacement of the high signal of the fatty marrow on T1WI by hypointense signal of the tumor. Also look for foraminal extension, whether by perineural spread or direct invasion of the tumor. MRI is also the study of choice for detecting intracranial extension of sinonasal disease. Role of CT and MR (2) On the left a T2W-image in an immuno-compromised patient with fever. Initially a MRI was performed to rule out sinusitis. Notice the low signal intensity of the left sphenoid sinus, which also had a low signal intensity on the T1W-image (not shown). Continue with the CT. The CT clearly shows the opacified sinus, which is slightly hyperdense. The signal characteristics on MRI and the attentuation on CT are a result of the high protein content of fungus. This is a good example of the pitfall of the 'pseudo-pneumatized sinus' . This is an example of an Actinomyes infection. So, when invasive fungal infection is suspected, start with a CT, then move on to MRI to rule out spread to the eye, cavernous sinus and intracranial compartment! In general bright signal on T2 is a sign of benign disease, since fluid and mucosal disease usually have a high water content. Secretions do not have solid enhancement. If you have an enhancing mass, you must rule out tumor. On the left an example of infectious sinonasal disease. On the pre-contrast scan you see relatively high signal content of the maxillary sinusses due to proteineous material. After the administration of i.v. contrast there is only enhancement of the circumferential mucosa and no solid enhancement. Role of CT and MR (3) In complicated cases both CT and MR are needed to demonstrate the extention of the disease. On the left CT-images of a patient post-lung transplant with fever and multiple rapidly progressing cranial nerve palsies. We will show you CT- and MR-images of this patient. The diagnosis lymphoma was made through biopsy. First study the images to study the extention od the disease. Then continue reading. On the CT-images the findings are: The image on the right is more cranial. There is opacification of the sphenoid sinus with destruction of and osteopenia of the sphenoid bone. CT nicely demonstrates the bone destruction and some of the soft tissue involvement. Continue with the MR-images. On the left the corresponding MRI. First study the images, then continue reading. The findings are: Continue with the coronal images. Coronal images of the same patient: T1 pre-and post-contrast. Normal aspect of the right Meckel's cave, tissue in the left Meckel's cave extending into the cavernous sinus (blue arrow). The red arrow points to the dural margin of the cavernous sinus: there is enhancement on both sides of the dura. The disease wraps around the temporal lobe (green arrow) and extents downward in the foramen ovale (yellow arrow) and into the masticator space. The asterix indicates normal non-enhancing tissue in the masticator space. This patient had a lymphoma. Nine out of ten times an immunocompromised patient will have a fungal infection, in one out of ten it will be a lymphoma. CT and MR have a complimentary role in this case, but finally a biopsy is called for to differentiate between these two diagnoses, because of different treatment. Role of CT and MR (4) On the left images of a 64-yrs-old, immuno-competent patient, who had a follow-up scan for left-sided vestibular neuroma. On the image on the left hypointense tissue is seen in the pterygo-palatine fossa and videan canal (yellow arrow). On the image on the right, which is more cranial, there is hypointense tissue in the pterygo-maxillary fissure and pterygo-palatine fossa. Continue with the contrast-enhanced T1W-image. There is solid enhancement of the abnormality. The differential diagnosis again consists of 2 catagories: neoplasm and chronic invasive fungal infection. In an immuno-competent patient, a neoplasm is much more likely. Continue with the CT-images. This is the corresponding CT, performed not to make the diagnosis, but to assess the condition of the adjacent bony structures, especially the sphenoid sinus. Also, it serves to guide the endoscopist for intraoperative biospy. There is extensive destruction of the skull base. The coronal image illustrates a normal foramen rotundum on the left (yellow arrow), which on the right has been obliterated by soft tissue. There is extensive bone destruction, and a possible area for biopsy is indicated by the blue arrow. At biopsy the diagnosis of a spindle cell carcinoma was made. When assessing the complications of sinusitis, CT is excellent for imaging of subperiostial abscesses or orbital extension into the ventral 2/3 of the orbit. MRI is necessary for assessing intracranial complications, such as brain or epidural abscesses, subdural empyema or sinus thrombosis. On the left images of a patient was initially diagnosed with a glioblastoma multiforme. There are abnormalities in both frontal lobes. Notice however the abnormal tissue in the frontal sinus (yellow arrow), subperiosteal abscess (red arrow) and the fluid-fluid level (green arrow) in the large intracranial lesion which has ring enhancement. All abnormalities are continuous meaning there is frontal bony destruction. The restricted diffusion also supports the diagnosis of brain abscess. This is a subperiosteal abscess and osteomyelitis of the frontal bone, usually with a soft tender swelling of the forehead. This is also called Pott's puffy tumor after Sir Pott, an English surgeon who first described this entity. Brain abscess (2) On the left images of another patient, who had recently been treated for sinusitis and now presented with a seizure. The CT shows an abnormality in the left temporal lobe with shaggy thick rim enhancement, and a large amount of vasogenic edema. This is also a brain abscess, most probably due to reflux of bacteria into cranial veins and the venous plexus around the cavernous sinus. On the left images of a patient with acute sinusitis and ethmoid air cell disease. He presented with blurred vision. First study the images, then continue reading. Notice the fluid in the left anterior clinoid process. The optic nerve runs medial to it. Continue with the coronal images. The coronal T2WI shows expansion of the clinoid process. The T1WI shows loss of normal fat compared with the right side, and extension into the orbital apex (red arrow). This is a mucocele of the anterior clinoid with secondary involvement of the optic nerve. Left is an axial T1WI, right is a coronal T2WI. There is an abnormality on the left side, but to a lesser degree also on the right. Try to determine which structures are involved. The yellow arrows point to the naso-lacrimal ducts. The naso-lacrimal sac connects with the duct, which then drains into the inferior meatus. On the left there is peri-orbital pre-septal soft tissue swelling. On the coronal image there is bilateral high signal at the junction of the nasolacrimal duct and sac, indicative of a fluid collection. On the left side there is also edema of the surrounding tissue. Post-contrast T1WI, axial and coronal. Lateral to the naso-lacrimal ducts on both sides, there are the fluid collections which now show peripheral enhancement. The additional images (T2WI) show mucosal disease of the right maxillary sinus and a fluid level in the left maxillary sinus, in addition to extensive ethmoidal and sphenoidal sinus disease. This patient had acute sinusitis which was complicated by orbital cellulitis and dacrocystitis with abscesses. Developmental or inflammatory narrowing of the naso-lacrimal duct is a risk factor for developing dacrocystitis. A rare complication of FESS is seen on the images on the left. The Hounsfields Units of the tiny abnormalities that the asterisk points to were around -120. First study the axial images. Then continue with the coronal images. There is a bone defect at the fovea ethmoidalis (red arrow). Also there are post-operative changes indicating that the patient had undergone FESS. The intracranial air is a complication of FESS. With this complication, usually the patient goes home feeling fine, and then shows up approximately two weeks later with CSF leak and meningitis, due to the defect in the bone and dura. Tension pneumocephalus occurs when air in the head acts like a mass: there is a bony defect which lets air in but not out (valve-like function). Every time the patient sneezes, air is forced through the defect into the intra-cranial space, and remains trapped there. At a certain moment the amount of air is sufficient to cause mass effect on the surrounding intra-cranial structures. Role of CT and MR When it comes to imaging of neoplasms of the paranasal sinuses, CT and MRI play complementary roles. It is not about the histology but about answering the question 'is it tumor or not?' and then determining the extent of the disease, for example intracranial or orbital extension. Use MRI to differentiate inspissated secretions from neoplasms. Scanning down to the hyoid bone allows for examination of the levels I and II lymph nodes: about 10% of paranasal neoplasms have nodal metastases at presentation. Coronal T2WI of the patient on the left show an abnormal structure in the right nasal cavity. When you've decided what it is, then stop to think about whether the abnormality is developmental or acquired. This patient has an encephalocele. There are two findings on the images that let you know that this is developmental. First of all, notice the smaller encephalocele on the left side (green arrow). Acquired encephaloceles are more often than not unilateral. The second clue is the cortical dysplasia (yellow arrow) as part of a migrational abnormality. Acquired encephaloceles (ie after surgery) tend to lead to dead gliotic brain, which would have a high signal intensity on T2WI. The strange looking structure in the left image (red arrow) is surgical packing, placed there after the involuntary encounter with brain tissue. The blue arrow points to the sphenoid sinus which is hyperintense, due to mucoid impaction as a result of obstruction by the encephalocele. Mucoceles are benign, locally expansile paranasal sinus masses most commonly found in the frontal sinus. Secondary to obstruction of the sinus ostia, there is accumulation of fluid within a mucoperiosteal lined cavity, resulting in erosion and remodelling of the surrounding bone. The most common causes of mucoceles are chronic infection, allergic sinonasal disease, trauma and previous surgery. The most common location of a mucocele is the fronto-ethmoidal sinus, followed by the sphenoid sinus. The least common location is the maxillary sinus. On the left a patient with an uncommon cause of a mucocele. Notice the obstructing solid mass at the frontal ethmoidal junction (yellow arrows). Pre- and post-contrast MRI of the same patient. The mucocele shows high signal intensity on T1W (benign finding). The mass in the ethmoidal region is hypointense and solidly enhancing. Mucocele (2) The case on the left shows two classic complications of frontal facial trauma. The bilateral nasal fractures are the clue to a traumatic etiology. Looking at the CT scan on the far left you will notice a convex soft tissue mass in the frontal sinus. The corresponding MRI shows a round hyperintense structure in the same location. Study the images on the left and then continue reading. Now look again carefully at the far left CT. Did you notice the bony defect on the left side, at the lateral border of the ethmoid air space (yellow arrow)? The MRI shows that there is brain tissue at the site of the defect. This patient had both a mucocele and an acquired encephalocele. The two most common causes of mucoceles are trauma and chronic inflammation due to blockage of the ostia. Mucocele (3) This companion case nicely demonstrates bilateral mucoceles. This patient has chronic sinusitis with sino-nasal opacification, mucoid impaction in the maxillary sinusses and huge bilateral mucoceles. The CT shows hyperdensity and the MRI shows hyperintensity on T2WI, both of which you will remember are benign signs in sino-nasal disease, indicating a proteinaceous substance. There is smooth bone remodelling and elevation of the frontal sinusses, and although it looks as if there is bony destruction at the orbital boundary of the frontal sinus, usually the surgeon will still see a fine line of bone in place. Inverted papilloma is characterized by inversion of the neoplastic epithelium into the underlying stroma. It presents as a unilateral nasal polyp arising from the lateral nasal wall, usually in the region of the middle meatus and middle turbinate. Extension into the maxillary and ethmoid sinuses is common. It causes non-specific symptoms like nasal congestion or epistaxis. Biopsy is necessary to make the diagnosis and because more than 10% of inverted papillomas harbor a squamous cell carcinoma. When you want to differentiate inspissated secretions from neoplasms it is important to have pre- and post-contrast images. If you were to just look at the post-contrast study on the right, you might be tempted to think that there was solid enhancement of the mass in the nasal cavity (asterisk) as well as in the ethmoidal and maxillary sinus on the right. Looking at the pre-contrast study, however, you will notice that the contents of the ethmoidal and maxillary sinuses are hyperintense as opposed to the mass in the nasal cavity (the middle meatal region), because the sinuses are filled with inspissated secretions. This solidly enhancing mass is a tumor until proven otherwise. The imaging findings are non-specific and the differential diagnosis includes a polyp or a carcinoma. Biopsy revealed an inverted papilloma. Inverted papilloma (2) On the left another patient who presented with nasal stuffiness. Study the images on the left. Decide for yourself whether you are looking at a solid mass, inspissated secretions, a combination of both or something entirely different. The pre-contrast T1WI shows a hyperintense area within the maxillary sinus, corresponding to a proteinaceous substance. Medial to it is an area with hypointense signal similar to the signal in the orbital globes (so probably cystic). The majority of the soft tissue in the right maxillary sinus is relatively hypointense on the pre-contrast T1WI, but solidly enhances, meaning tumor. The T2W-image on the left confirms the cystic element (yellow arrow). The coronal CT nicely demonstrates remodelling of the bone and expansion (arrowheads). This proved to be an inverting papilloma. The localisation is rather typical. Malignant tumors of the sinonasal tract are extremely rare. The clinical presentation is non-specific and often mimics benign disease. As a result a delay in diagnosis is common. 75% of all paranasal sinus tumors are Stage T3 or T4 at the time of diagnosis. Perineural spread is a manifestation of advanced disease and indicates a poor prognosis. On the left an axial MR-image showing a mass in the ethmoids. The MRI shows no intracranial extension. What is the next step? A CT is necessary to determine the integrity of the adjacent bone. Notice the bony destruction of the fovea ethmoidalis and planum sphenoidale. This indicates that this is a malignant lesion and biopsy demonstrated an adenocarcinoma. If the patient is a surgical candidate, frontal endoscopic sino-nasal surgery won't be enough and a cranio-facial take-down will also be required. A meningioma can spread transcranially. On the left is a patient with a meningioma, which spreads along the anterior clinoid (arrow). Continue with a coronal image more anteriorly. MRI nicely demonstrates how the meningioma spreads down into the sino-nasal cavity. First look at the images on the left. From where is this lesion arising? The lesion is expansile with bone remodelling and there is an obvious relation to a tooth. It is very important to determine whether or not sinus pathology has an odontogentic origin, simply because the surgical approach is different. If odontogentic, the surgery will be done preferably by a maxillofacial surgeon. If approached by a sino-nasal route it won't be removed entirely. This is a keratocyst On the left another case. This patient presented with facial pain. On this contrast-enhanced MRI we see a non-enhancing expansile lesion in the right maxillary sinus. As it doesn't enhance, we know we aren't dealing with tumor. It is tempting to call this a retention cyst and move on to the next case, but a CT is called for to make the correct diagnosis. The corresponding CT shows elevation of the maxillary bone (blue arrow). The red arrow in the right image indicates the cyst dissecting around the root of a tooth. This is also a keratocyst. On the left a patient who presented with asymmetric eyes. First study the images and try to describe what is going on. Then continue reading. This film was originally read as ethmoid sinusitis and post-surgery. However, this patient had never undergone sino-nasal surgery. What you in fact see, is adhesion of the middle right turbinate (red arrow) and the uncinate process to the floor of the orbit. There is also volume loss of the right maxillary sinus. This is called the silent sinus syndrome, which consists of painless facial asymmetry and enophthalmos caused by chronic maxillary sinus atelectasis. The most characteristic imaging feature of the silent sinus syndrome is the inward retraction of the sinus walls into the sinus lumen with associated decrease in sinus volume and enlargement of the middle meatus (2). In many cases the infundibulum is occluded due to lateral retraction of the uncinate process. Fibro-osseous lesions are very common incidental findings and often misinterpreted as tumor. CT is usually diagnostic, so when you see a bizarre lesion at the skull-base, think of a fibro-osseous lesion and get an unenhanced CT. The most common skull base lesion is fibrous dysplasia, followed by osteomas. Osteomas are usually located in the sino-nasal cavities. Fibromas are less common and can be ossifying or non-ossifying. Malignancies are rare. On the left images of a patient who was thought to have a chondrosarcoma. On the T2W-images there is a hypointense lesion (yellow arrow) with a cystic component (red arrow). On the pre- and post-contrast T1W-images there is solid enhancement of a mass with peripheral enhancement of the cystic portion. Question: what should be the next step? Next step: get a CT! On CT this is classic fibrous dysplasia (FD) with cortical sparing and ground-glass appearance. Many of the fibrous dysplasia lesions in the clivus, skull base or sino-nasal cavity and in children may have large cystic components, so don't let that dissuade you from the diagnosis! Fibrous dysplasia (2) On the left another example of fibrous dysplasia. This lesion originates from the middle turbinate. Fibrous dysplasia (3) On the left another patient read as having a soft tissue tumor (yellow arrow) anterior to the temporal bone. These are the corresponding axial CT images showing classic fibrous dysplasia of the sphenoid wing. The differential diagnosis is a meningioma. Fibrous dysplasia (4) On the left images of a different patient with soft tissue in the sphenoid sinus on the left, and an abnormality of the left sphenoid wing, which is about 3 times its normal size whilst maintaining its normal shape. On the post-contrast images there is solid enhancement. Again the diagnosis is fibrous dysplasia. Fibrous dysplasia is a very vascular lesion and can enhance avidly. In contrast, this is a patient with osteitis of the middle turbinate and ethmoid septae. Note the laminated high density, due to chronic inflammation and recurrent periostal reaction with neo-osteogenesis around the septae. This is a patient who had been having brain MRI for the past 1,5 yrs for frontal headaches. On the MRI (not shown) it looked as if the patient had a little mucosal disease of the frontal sinus. The sinus CT clearly shows an osteoma with a bony defect (arrow) indicating progressive growth. This lesion requires surgical excision. Note: multiple para-nasal osteomas are found in Gardner's syndrome, which also includes cutaneous and soft tissues tumors in addition to colonic polyps with a predilection to malignant degeneration. Imaging can also be used to monitor treatment response. This patient had a skull-base B-cell lymphoma. Pre-treatment we see the extent of the lesion, which is FDG-avid. Three months into therapy the skull-base mass has resolved. There is no more pathological FDG uptake, and there is remineralisation of the bone. The patient on the left had a resection of a carcinoma. Now there is a recurrence (blue arrow on MR and fusion image). The green arrow, however points to post-radiation tissue changes. On the left a patient who had a prior cranio-facial resection (yellow arrows) for an undifferentiated carcinoma. This patient was treated with chemoradiation. The T2WI on the left shows tumor recurrence intracranially. Continue with the contrast enhanced images. In cases like this a recurrence tend to show bizarre patterns such as these extensive dural implants. in eMedicine by Gerard Domanowski. by Anna Illner et al. AJR 2002; 178:503-506Laurie Loevner and Jennifer Bradshaw Signal characteristics of secretions Pseudo-pneumatized sinus Enhancement Brain abscess Mucocele Orbital Cellulitis and Abscess Complication of FESS Encephalocele Mucocele Inverted papilloma Malignant tumors of the sinonasal tract Meningioma Keratocyst Silent sinus Fibrous dysplasia Osteitis OsteomaParanasal Sinuses - MRIRadiology department of the University of Pennsylvania, USA and the radiology department the Medical Centre Alkmaar, the Netherlands
headneck4
1
Swallowing disorders - interpretation of radiographic studies
by Robin Smithuis
In this overview a simple method of interpreting swallowing disorders is presented. We will concentrate on four basic findings: Visit the website to view the videos. Swallowing is a complex movement. It requires the coordination of nerves and muscles in the buccolabial area, the tongue, the palate, the pharynx, the larynx and finally the esophagus. Look at the video on the left. It is clear that analyzing swallowing videos can be quite difficult even when they are in slow motion . We will discuss a systematic approach on how to analyse these videos, but first a short introduction on normal swallowing is given. In the oral phase food is prepared for swallowing and then transported to the pharynx. This is a preparatory phase in which the food is held within the mouth while the base of the tongue and the soft palate close the oral cavity posteriorly to prevent food spilling into the open larynx and trachea. A bolus is formed in the central portion of the tongue and then pushed posteriorly toward the pharynx with an anterior-to-posterior tongue elevation. As the bolus enters the pharynx the actual swallow or pharyngeal reflex is triggered. This phase is a reflex action. The bolus passes through the pharynx quickly and then enters the esophagus. This takes place in less than a second. The initiation of this process starts when the bolus passes the anterior faucial arch and reaches the posterior pharyngeal wall. Elevation of the soft palate prevents material from entering the nasal cavity. This stage is followed by the pharyngeal constrictor muscles pushing the bolus further into the pharynx, toward the cricopharyngeal sphincter. The larynx prevents material from entering the trachea by respectively closing the true vocal cords, false vocal folds, and aryepiglottic folds. Contraction of the lower pharyngeal constrictor is followed by relaxation of the cricopharyngeal muscle, allowing the bolus to pass into the esophagus. The most important images of the swallowing study are those taken of the lateral view. Click through the images 1-8 on the left.  The AP-view is less important than the lateral view in the analysis of patients with dysphagia. It is especially important to look for asymmetry. Once a good series of the pharyngeal phase has been acquired, follow the contrast bolus all the way down to the gastroesophageal junction.  Indications for a swallowing study are dysphagia, globus sensation and aspiration. Dysphagia is a general term used to describe the inability to move food properly from the mouth to the stomach. Globus sensation is a term to describe the feeling, that there is something in the throat, that is in the way or needs to be swallowed. Aspiration is the most severe form of a swallowing disorder and can result in aspiration pneumonia, chronic lung disease or choking. In some patients with chronic lung disease a swallowing study can be performed to look for 'silent aspiration'. When starting a study, try to find out exactly what the patients problem is, so you can customize the series. Is there a risk of aspiration (i.e. wet voice, recurrent pneumonia, aspiration)? If so, do not start the examination with barium contrast, but instead use non-ionic water-soluble contrast. If, during the first few swallows no aspiration takes place, you can switch back to barium, as this gives better quality images. When solid food is the problem, you may want to add a solid substance to the barium (for instance biscuits or bread). The examination of patients with a possible swallowing disorder consists of: We start with one or two lateral swallows followed by a lateral double-contrast view of the pharynx (see later). Then an AP-swallow is recorded followed by an AP double-contrast view of the pharynx. Next the passage through the esophagus is recorded, followed by double-contrast views of the gastroesophageal junction. Before we start the examination, the procedure is explained to the patient and we practise certain manoeuvres (i.e. modified Valsalva). The act of swallowing is recorded on video or some sort of digital recording. We use grabs from the fluographic images and store them into the PACS-system, where we can play them back and forth for analysis. It is very important to always start with the lateral view first, because when the patient aspirates, the lateral view will tell you how and why it happened. If you have to stop the examination, the most important images will already have been recorded. If you start with an AP-view and the patient aspirates, you may have to stop the examination and you will never find out why aspiration occurred and what the real problem is. Use only a small amount of barium for the first swallow and if the patient is doing fine, coninue with larger portions. Aspiration of a small amount of barium is usually not a big problem, but you don't want a lot of barium filling the bronchi.  For the lateral view, ask the patient to sing an ???, as this will move the tongue in an anterior position and give a better view on the oro- and hypopharynx. In Dutch this will be the letter ???, as it is pronounced the same as the english ???. For the AP-view the modified Valsalva manoeuvre is performed. The patients has to blow air through the tightened lips as in trumpet-playing, while relaxing the neck region. Always practise this manoeuvre prior to the examination, so the patient knows what to do. On the left DC views of the pharynx. Outpounching of the lateral wall of the pharynx is normal and can be quite severe (Dizzy Gillespie). These are called 'lateral pharyngeal ears'. Always follow the passage of barium through the esophagus until it enters the stomach. Disorders of the gastroesophageal junction are often experienced as a problem within the throat. The rationale for this is that in patients with a distal obstruction, gastroesophageal reflux or a motility disorder, the cricopharyngeal muscle has to work very hard to prevent foodspillage back into the pharynx - along with its risk of aspiration. This increased muscle tone gives the patient the sensation that there is a problem in the throat. The patient on the left complained of globus sensation. This was due to severe reflux and subsequent increased tone of the cricopharyngeal muscle. A complex paraesophageal hernia is seen. Excellent views of the gastroesophageal junction can be achieved by doing the following: A simple way to analyze a swallowing study is to concentrate on four easily detectable findings. These are: These findings are mostly already apparent during the examination, but analysis of all the images will clarify the mechanisms that cause these abnormal findings. These imaging findings may be isolated findings or may be related to each other. For instance, premature closure of the cricopharyngeal muscle may lead to stasis of contrast in the pharynx, which may result in aspiration as the larynx opens at the end of swallowing. Asymmetric swallowing on an AP-view is usually the result of an asymmetric tilting of the epiglottis. Sometimes it is caused by rotation of the head, but in many cases no real explanation is found. Even when the head is not rotated, the epiglottis can tilt asymmetrically when it hits the posterior pharyngeal wall. This is more likely to occur when only a small bolus is given,as the pharynx will not fully distend. An asymmetric swallow may be followed by a symmetric swallow in the same patient when a larger bolus is given. In the case on the left rotation of the head closes the side to which the head is turned (Figure). If a patient has a unilateral pharyngeal paresis, turning of the head towards the affected side will help the patient in preventing aspiration. By turning the head towards the affected side, this side will be closed preventing stasis on this side and possible secondary aspiration. However before you decide that it is a normal finding you have to exclude a pharyngeal tumour or unilateral paresis. The double-contrast views can be helpful in these cases. In unilateral paresis the paralysed side will protrude during the modified Valsalva. When a tumour is present in the pharynx, it is usually visible on the DC views. Sometimes endoscopy is necessary to solve the problem of an asymmetric swallow. Asymmetry (2) The case on the left is an odd case, but it nicely demonstrates the difficulty that sometimes exists in determining the cause of asymmetry. On the far left asymmetry is seen on the fluorographic study (green arrow). A tumour in the right pyriform sinus has to be excluded. On the DC view on the right the piryform sinus is normal (green arrow), but at the level of the vallecula on the right a lobulated proces is seen (yellow arrow) and at a higher level a smooth indentation of the oropharynx is seen (blue arrow). The lobulated tumor at the level of the valleculae proved to be remnants of the tongue tonsil, which is a common finding and sometimes difficult to differentiate from cancer of the tongue base. In some cases endoscopy is needed to differentiate the two. Next image is the CT of this patient. Asymmetry (3) On the left we see a CT image of the same patient. The smooth indentation of the oropharynx on the right was caused by an elongated internal carotid artery. So indeed this is a very uncommon case in which on the fluorographic study a tumor was suspected in the piriform sinus. Finally a proces was found within the oropharynx and a proces compressing the wall from outside at a higher level. Due to these processes there was an asymmetric passage of contrast in the hypopharynx simulating a proces in the piriform sinus. Stasis is the result of insufficient cleansing of the pharynx, either due to an obstruction (i.e. dysfunction of the cricopharyngeus) or due to insufficient contraction of the pharyngeal constrictors. Insufficient contraction is the result of pharyngeal paresis resulting from a neuromuscular disorder. Excessive movements of the tongue base and larynx are sometimes seen on lateral fluorographic studies to compensate for the loss of function of the pharyngeal constrictors. When the patient resumes breathing, aspiration can occur (Figure). Stasis (2) On the left a set of images demonstrating a patient with a paresis of the pharynxconstrictors. This is usually associated with a disturbed relaxation of the cricopharyngeal muscle. In this example we can see how the patient tries to compensate for the loss of pharyngeal constriction by excessive movement of the tongue and the head. This patient is in extreme stress, because he knows that when he starts breathing and the throat is not empty, he will aspirate.  In some of these cases cricopharyngeotomy is the only solution to facilitate the passage of food to the esophagus. Insufficient opening and premature closure are the most common problems of the cricopharyngeal muscle. Normally you should not see an impression of the cricopharyngeus during passage of the bolus, but a small non-obstructive indentation is sometimes seen and is not clinically significant (Figure). It can however sometimes explain the symptoms of the patient. It is assumed that the passage of food irritates the mucosa that covers the cricopharyngeal muscle resulting in a globus sensation. Cricopharyngeal dysfunction (2) Premature closure of the cricopharyngeus results in an increased pressure in the hypopharynx, just above the cricopharyngeus, as the pressurewave of the pharyngeal constrictors pushes the bolus downwards. This increased pressure can result in an outpouching at a weak spot in the posterior pharyngeal wall (Killian's dehiscence). First this will result in a small pouch, that in time can increase and form a true Zenker's diverticulum (Figure). A Zenker's diverticulum is always the result of cricopharyngeal dysfunction. Cricopharyngeal dysfunction (3) On the left a patient who complained of globus sensation in the throat and difficulty of swallowing with regurgitation of undigested food. The digital recordings nicely demonstrate the filling of a large Zenker's diverticle, just above the contracted cricopharyngeal muscle (yellow arrow). The contraction of the lower pharyngeal constrictors is indicated by the blue arrow. The contraction of the lower pharyngeal muscle against the closed cricopharyngeal muscle causes the posterior outpouching. Visit the website to view the videos. Cricopharyngeal dysfunction (4) On the left a video of a patient with a Zenker diverticulum. Notice that the insufficient opening of the cricopharyngeal muscle is the cause of the diverticulum. There are three instances when aspiration can occur: before, during or after the actual swallow.  Webs usually occur at the level of the hypopharynx or the upper esophagus, producing dysphagia for solids. Liquids usually pass well, but in many cases a 'jet' is seen. The passage of solid food may produce irritation or damage to the mucosa, resulting in a globus feeling. They are best diagnosed on the lateral projection of the barium swallow. Webs are frequently overlooked at esophagoscopy unless special attention is given to this area. During esophagoscopy they may rupture, but in many cases recurrence is seen. Aspiration before swallowing When tongue or soft palate are unable to prevent spillage of food into the pharynx, aspiration may occur since the larynx is still open. Weakness of these muscles in the mouth and the throat is due to paralysis or myopathy. Visit the website to view the videos. On the left a video of a patient who aspirates prior to the actual swallowing reflex. On the left the same video images step by step.  We have to assume that a failure of the sensory nerves in the pharynx is the problem. Notice also that there is no coughing reflex. This patient probably has silent aspiration. Aspiration during swallowing This is due to an insufficient closure of the larynx when it should be closed. Closure of the larynx is a result of anterosuperior lifting of the larynx which allows the true cords, false cords and finally, the aryepiglottic folds to contract, followed by a backwards folding of the epiglottis over the closed larynx. The aryepiglottic folds are the main gatekeepers, while the epiglottis plays only a minor role in preventing aspiration. Both failure of these intrinsic muscles of the larynx as well as failure of the extrinsic muscles (i.e. muscles that lift the larynx) may lead to aspiration during swallowing. Weakness of the extrinsic muscles is seen after radiotherapy, in neurologic disorders and in recurrens nerve paralysis (i.e. neuromuscular dysfunction). Aspiration during swallowing (2) On the left a digital recording of a patient who complained of a sore throat and a wet voice. Notice that when the contrast enters the throat, the swallowing reflex is not triggered immediately. This allows for barium contrast to enter the larynx, where it sits on the vocals cords. Once the swallowing reflex is initiated, the larynx closes properly, but contrast is already in. Notice also that while the contrast enters the larynx, it does not initiate a cough reflex. Althoug the digital recording perfectly explains the complaints of the patient, it is difficult to say what causes this problem. Usually when there is aspiration during swallowing, the problem is at the level of the larynx. Mostly it is the larynx that is unable to close either due to weakness of the intrinsic laryngeal muscles or weakness of the extrinsic muscles that lift the larynx and allow it to contract (for instance after neck radiation or surgery). In this case the problem is probably at the level of the sensory nerves in the pharynx, who are not triggered properly. You could also argue, that this maybe is aspiration before swallowing, because it happens before the actual swallowing reflex. Maybe a more forcefull push of the bolus posteriorly from the mouth into the oropharynx could help in triggering these nerves instead of allowing the bolus to slide over the back of the tongue. Aspiration after swallowing This is the result of stasis of contrast in the pharynx due to insufficient contraction of the pharyngeal constrictors or insufficient opening of the cricopharyngeal muscle. When the larynx opens the contrast may leak into the trachea. On the left a patient who aspirates after swallowing. There is severe stasis. Aspiration occurs as the patient inhales to start breathing again.  The total absence of pharyngeal contraction due to paresis is the primary problem in this patient. As a result there is stasis of contrast and secondary aspiration as the patient starts to breath again. The results of the swallowing examination help in establishing a final diagnosis. Based on this examination alone however, a specific diagnosis usually cannot be made, since most severe swallowing disorders are the result of a complex neuromuscular dysfunction. Hence the swallowing study should be regarded as part of the total evaluation of the patient by gastroenterologist, neurologist and speech therapist. The strength of the fluoroscopic examination is, that it is the only examination that can show us, what is really going on during swallowing and can therefore lead to a rehabilitation plan. Swallowing rehabilitation is a specialty on its own. Here we will make some brief comments on rehabilitation as it may help you to better understand the dynamics of swallowing. In unilateral pharyngeal paralysis stasis can be prevented by closing down one of the lateral food channels by turning the head towards the affected side or by manually compressing it. Patients with aspiration before swallowing due to insufficient closure of the mouth, can be helped by flexing their head during chewing and thus holding the food in the anterior part of the oral cavity. In patients with aspiration during or after swallowing the 'supraglottic swallow' may help. Before swallowing a deep breath is taken. Air is prevented to leak out of the airways by compressing the vocal cords. Immediately after swallowing the patient has to cough forcefully in order to clear the airways and the throat from any residual food. Some patients only aspirate when they ingest fluids. These patients can be helped by changing their fluid intake into jelly-like liquids. Logemann JA: In: Manual for the Videofluorographic Study of Swallowing. 2nd ed. Austin, TX: Pro-Ed Inc; 1993. by Nam-Jong Paik by WJ Dodds, JA Logemann, ET Stewart American Journal of Roentgenology, 1990 - Am Roentgen Ray Soc Page 1. 965 Review ArticleRobin Smithuis Oral phase Pharyngeal phase Fluoroscopic imaging Fluorographic study of the actual swallowing Double contrast images of the pharynx Examination of the esophagus Asymmetry Stasis Cricopharyngeal Dysfunction Aspiration Diagnosis Swallowing RehabilitationSwallowing disorders - interpretation of radiographic studiesRadiology department of the Rijnland Hospital in Leiderdorp, the Netherlands
headneck5
1
Temporal bone - Anatomy
by Erik Beek and Robin Smithuis
Updated version: 21-2-2007 In this review we present the normal coronal and axial anatomy of the temporal bone. Learn the anatomy by scrolling through the images. The middle ear consists of the tympanic cavity and the antrum. The antrum is a large aircell superior and posterior to the tympanic cavity and connected to the tympanic cavity via the aditus ad antrum. The epitympanum or attic is the upper portion of the tympanic cavity above the tympanic membrane, and contains the head of the malleus and the body of the incus. The tympanic membrane, the malleus, incus and stapes transfer soundwaves to the stapes footplate, which is attached to the base of the cochlea in the oval window.  At the most inferior level we see the facial nerve passing inferiorly to finally reach the stylomastoid foramen (not shown in this image). The carotid artery is shown within the carotid canal. Also at this level is the top of the jugular bulb. The manubrium of the malleus (yellow arrow) is connected to the tympanic membrane. At this level we can see the manubrium of the malleus (yellow arow) anterior to the long process of the incus. The round window is indicated by the blue arrow. The round window dissipates the pressure generated by the fluid vibrations within the cochlea and thus serves as a release valve. The base of the stapes rocks in and out against the oval window. The vibrations are transmitted via the endolymph to the hair cells of the organ of Corti of the cochlea. Within the cochlea the movement of the hair cells convert the sound-vibrations into nerve impulses, that travel over the cochlear nerve to the auditory cortex of the brain, which interprets the impulses as sound. . The head of the malleus is seen anterior to the head of the incus (yellow arrow). In this image at the level of the internal auditory canal, the tympanic segment of the facial nerve is seen just medial and parallel to the wall of the epitympanum. The head of the malleus (yellow arrow) is seen anterior to the head and the short process of the incus. At this level the aditus ad antrum is seen. This is the connection between the tympanic cavity and the antrum. The labyrinthine segment of the facial nerve coming from the internal auditory canal angles sharply forward, nearly at right angles to the long axis of the petrous bone, to reach the geniculate ganglion. At the ganglion the facial nerve makes a U-turn (first genu of the facial nerve) to run posteriorly as the tympanic segment along the medial wall of the epitympanum. At this level the antrum is seen surrounded by smaller mastoid aircells just lateral to the superior semicircular canals . The three semicircular canals lie perpendicular to each other to sense acceleration and deceleration movements in each of the 3 spatial planes. Static head position is sensed by the vestibule, which contains the position hair cells. Different head positions produce different gravity effects by small calcium carbonate particles (otoliths) on these hair cells. The petrous bone is positioned in an oblique orientation from posterolateral to anteromedial. As a result most structures will be sectioned obliquely on coronal images. The following coronal images go from anterior to posterior. First we will see the tympanic membrane with the ossicles, followed by the cochlea, antrum and semicircular canalls. Finally the most posterior image will show the point where the facial nerve exits the temporal bone at the stylomastoid foramen. The scutum (yellow arrow) is a sharp bony spur formed by the lateral wall of the tympanic cavity and the superior wall of the external auditory canal. It is usually the first bony structure to erode as a result of a cholesteatoma, that is formed by medial retraction of the pars flaccida of the tympanic membrane into the epitympanum. If the retraction continues it will result in ossicular destruction. If the cholesteatoma passes posteriorly through the aditus ad antrum into the mastoid itself, erosion of the tegmen mastoideum, with exposure of the dura and erosion of the lateral semicircular canal with deafness and vertigo, may result. On the left the most anterior point of the facial nerve is seen (white arrow). At this point the nerve makes a U-turn. It is named the genu or geniculum and represents the geniculate ganglion. The malleus is seen connected to the tympanic membrane and to the incus. In many illustrations you will see the incus connecting medially to the malleus, but this is not correct. On the coronal reconstruction on the left it is clearly demonstated that the incus is positioned posterolaterally to the malleolar head. The long crus of the incus subsequently runs inferomedially to the stapes. A coronal image slightly more posteriorly will show the facial nerve twice. The medial portion is the part that exits the internal auditory canal and runs towards the geniculate ganglion (medial white arrow). The lateral portion is the part that courses in a posterior direction, coming from the U-turn of the first genu. The facial nerve is seen in the internal auditory canal and entering the temporal bone (medial white arrow). The lateral white arrow represents the tympanic segment of the facial nerve running in the facial canal and curving around the oval window niche. At this point, the nerve runs in a horizontal plane in a posterior direction superiorly to the oval window . The incus (orange arrow) is seen connecting to the stapes (blue arrow). Coronal scan showing the facial nerve (white arrow) above the oval window and below the lateral semicircular canal. The antrum is a large aircell superior and posterior to the tympanic cavity and connected to the tympanic cavity via the aditus ad antrum. It is surrounded by smaller mastoid aircells. On this last posterior coronal image the facial nerve assumes a vertical position to exit the petrous bone through the stylomastoid foramen. Interactive Digital Education (part I), F.J.A. Beek, radiologist, Radiology department of the Wilhelmina Children's Hospital and the University Medical Centre of Utrecht, the Netherlands Lemmerling M, Kollias SS, eds. Radiology of the Petrous Bone. Springer 2003. Ch. 1, p. 1-14Erik Beek and Robin Smithuis Scroll through the axial anatomy from inferior to superior Axial anatomy from inferior to superior Tympanic membrane Stapes Cochlea Tympanic segment of the facial nerve Geniculate ganglion of the facial nerve Antrum Scutum Facial nerve canal AntrumTemporal bone - AnatomyRadiology department of the University Medical Centre of Utrecht and the Rijnland Hospital, Leiderdorp, the Netherlands
headneck6
1
Temporal bone - Pathology
by Eric Beek and Frank Pameijer
The aim of this presentation is to demonstrate imaging findings of common diseases of the temporal bone. CT is the imaging modality of choice for most of the pathologic conditions of the temporal bone, especially for those of the middle ear. MRI is more useful for diseases of the inner ear. Disease processes in the pontine angle and in the internal acoustic meatus are not discussed. There are several normal variants which may simulate disease or should be reported because they can endanger the surgical approach. 
 Variants which may simulate disease: Variants which may pose a danger during surgery: On the left an illustration of a cholesteatoma. This will be discussed later. A cochlear cleft is a narrow curved lucency extending from the cochlea towards the promontory. It is often visible in infants and children but can also be seen in adults. It can be mistaken for a fracture line or an otosclerotic focus. On the left an example of bilateral cochlear cleft in a one-year old boy with congenital hearing loss. The petromastoid canal or subarcuate canal connects the mastoid antrum with the cranial cavity and houses the subarcuate artery and vein. Its diameter is around 0.5 mm. It can be confused with a fracture line. On the left a 40-year old female with a sclerotic mastoid. The petromastoid canal is easily seen. (arrow) On the left a well-pneumatized mastoid. The petromastoid canal is difficult to discern (arrow). On the left another patient with a sclerotic mastoid. The petromastoid canal is well seen. If this patient would be a trauma victim, the canal could easily be confused with a fracture line (arrow). The cochlear aqueduct connects the perilymph with the subarachoid space. The cochlear aqueduct is a narrow canal which runs towards the cochlea in almost the same direction as the inner auditory canal, but situated more caudally. It is a point where infected cerebrospinal fluid can enter the inner ear. This can happen in patients with meningitis and cause labyrinthitis ossificans. On the left a 58-year old male. The blue arrow indicates the cochlear aqueduct coursing towards the cochlea. This could be mistaken for a fracture line (arrow). Note there is also opacification of the tympanic cavity and mastoid air cells. On the left axial and coronal images of a 64-year old male. The jugular bulb rises above the lower limb of the posterior semicircular canal (arrows). The jugular bulb is often asymmetric, with the right jugular bulb usually being larger than the left. If it reaches above the posterior semicircular canal it is called a high jugular bulb. If the bony separation between the jugular bulb and the tympanic cavity is absent, it is termed a dehiscent jugular bulb. Rarely an outpouching is seen – this is known as a jugular bulb diverticulum. On the left axial and coronal images of a 50-year old male. Incidental finding of a jugular bulb diverticulum (arrows). The sigmoid sinus can protrude into the posterior mastoid. It can be accidentally lacerated during a mastoidectomy and therefore should be mentioned in the radiological report when present. On the left an axial image of a 43-year old male, post-mastoidectomy. The sigmoid sinus bulges anteriorly The vestibular aqueduct is a narrow bony canal (aqueduct) that connects the endolymphatic sac with the inner ear (vestibule). Running through this bony canal is a tube called the endolymphatic duct. A large vestibular aqueduct is associated with progressive sensorineural hearing loss. This progression is reportedly associated with minor head trauma, which exposes the inner ear to pressure waves via the large vestibular aqueduct. The large vestibular aqueduct is associated with an absence of the bony modiolus in more than 90% of patients. On the left a patient with a bilateral large vestibular aqueduct. Notice that the bony modiolus is not visible. On the left a 5-year old boy with bilateral progressive hearing loss. A large vestibular aqueduct is seen (black arrow). The cochlea has no bony modiolus. (white arrow). In external ear atresia the external auditory canal is not developed and sound cannot reach the tympanic membrane. A conductive hearing loss is the result. It is important to note whether the atretic plate is composed of soft tissue or bone. The extent of ossicular chain malformation can vary from a fusion of the mallear head and incudal body to a small clump of malformed ossicles, which is often fused to the wall of the tympanic cavity. The mastoid portion of the facial nerve canal can be located more anteriorly than normal and this is important to report to the ENT surgeon in order to avoid iatrogenic injury to the nerve during surgery. On the left a 2-year old boy with bilateral bony external auditory canal atresia. The malleus and incus are fused (arrow). The cochlea is normal. The cochlea develops between 3 and 10 weeks of gestation. Early developmental arrest leads to an inner ear that consists of a small cyst, the so-called Michel deformity. Developmental arrest at a later stage leads to more or less severe deformities of the cochlea and of the vestibular apparatus. An incomplete partition of the cochlea is called a Mondini malformation Instead of the normal two-and-one-half turns, there is only a normal basal turn and a cystic apex. On the left a 2-year old girl. The images are of a CT-examination is done prior to cochlear implantation. A minor deformity of the cochlear apex is visible – there is no separation of the second and third turn and the bony modiolus is absent. The vestibular aqueduct is normal. Malformations of the vestibule and semicircular canals vary from a common cavity to all these structures to a hypoplastic lateral semicircular canal. During embryogenesis the lateral semicircular canal is the last structure to form, thus in malformations of the semicircular canals the lateral canal is most commonly affected. On the left a 10-year old boy, scheduled for cochlear implantation. There is a widening and shortening of the lateral semicircular canal. The vestibule is relatively large (arrow). On the left a 16-year old boy, examined preoperatively for a cholesteatoma of the right ear. As a coincidental finding, there is a plump lateral semicircular canal (yellow arrow) and an absence of the superior canal (blue arrow). In the expected position of the superior canal only a bump is seen. The posterior canal is normal. For the ENT-surgeon the differentiation between chronic otitis media and cholesteatoma is important. Both diseases often occur in poorly pneumatized mastoids. An important finding which can help differentiate the two conditions is bony erosion. Erosion of the lateral wall of the epitympanum and of the ossicular chain is common in cholesteatoma (around 75%). Erosion can occur in chronic otitis, but reportedly in less than 10% of patients. Displacement of the ossicular chain can be seen in cholesteatoma, not in chronic otitis. Cholesteatoma can present with a non-dependent mass while chronic otitis shows thickened mucosal lining. However, in both diseases the middle ear cavity can be completely opacified, obscuring a cholesteatoma. On the far left a 54-year old male with a normally pneumatized mastoid with aerated cells.
 Next to it a 69-year old female. The mastoid is completely sclerotic - no air cells are present. On the left a 14-year old boy. The eardrum is thickened. A small amount of soft tissue (arrow) is visible between the scutum and the ossicular chain but no erosion is present. This favors the diagnosis of chronic otitis media. On the left an 11-year old girl with bilateral ear infections. There is calcification of the eardrum (white arrow) and calcific deposits on the stapes and the tendon of the stapedius muscle (black arrow). On the left a 37-year old female who was admitted with a peritonsillar abces. She also suffered from chronic otitis media. CT shows a tympanostomy tube (yellow arrow) and almost opacification of the tympanic cavity and mastoid air cells with soft tissue. Calcification is visible around the head of the stapes (blue arrow). No erosions are present. On the left a coronal reconstruction of the same patient. CT shows the tympanostomy tube (yellow arrow) and complete opacification of the tympanic cavity and mastoid air cells with soft tissue. Cholesteatoma is believed to arise in retraction pockets of the eardrum. It gradually enlarges over time due to exfoliation and encapsulation of the tissue. Most cholesteatomas are acquired, but some are congenital. The ENT surgeon often states that cholesteatoma is a clinical diagnosis. Scraps of cholesteatoma are visible in the external auditory canal. On CT a small cholesteatoma presents as a soft tissue mass. In more extensive disease erosions may be present. Large cholesteatomas can erode the auditory ossicles and the walls of the antrum and extend into the middle cranial fossa. The most affected structures are: On the left a 20-year old woman with recurrent otitis. There were granulations on the left ear drum. CT demonstrates a soft tissue mass between the ossicular chain and the lateral tympanic wall, which is eroded. this favors the diagnosis of cholesteatoma. On the left the coronal images of the same patient as above. Notice how the cholesteatoma has eroded the scutum (arrow). There are two patterns of spread: On the left a large cholesteatoma in the right middle ear with destruction of the lateral wall of the tympanic cavity. The body of the incus, which is lateral to the mallear head is also eroded (arrow). CT signs of cholesteatoma are: Erosion of the facial nerve canal is difficult to distinguish because the wall is often so thin that it is not visible at CT. On the left a 50-year old male with hearing loss on the left side. There is a soft tissue mass with erosion of the long process of the incus. This location is typical of a pars tensa cholesteatoma. On the left images of a cholesteatoma, which has eroded the ossicular chain and the wall of the lateral semicircular canal (arrows). The thickened ear drum is perforated. On the left images of a 6-year old boy. A large cholesteatoma has resulted in a so called 'automastoidectomy', with severe erosion of the lateral tympanic cavity wall and destruction of the ossicular chain. On the left images of a 50-year old man who presented with a left- sided retraction pocket and otorrhoea. CT shows erosion of the long proces of the incus and of the stapedial superstructure. All these findings favor the diagnosis of a cholesteatoma, but at surgery, chronic mastoiditis was found and no cholesteatoma was identified. A minority of patients with chronic mastoiditis show bony erosions. On the left coronal images of the same patient. The scutum is blunted (arrow). On the left an image of a 53-year old man complaining of vertigo. He had undergone several ear operations in the past. The CT shows erosion of the wall of the lateral semicircular canal (arrow) due to cholesteatoma. On the left a 22-year old man suffering from persistent otitis. The right ear shows a soft tissue mass medial to the ossicular chain with lateral displacement of the incus with erosion of its lenticular process and of the stapes, compatible with a pars tensa cholesteatoma (arrow). On the left coronal images of the same patient. Cholesteatomas are of mixed intensity on T1-weighted pulse sequences and of high intensity on T2-weighted pulse sequences. MRI is particularly useful for evaluating the extension of a cholesteatoma into the middle and/or posterior fossa, and for demonstrating possible herniation of intracranial contents into the temporal bone - especially after surgery. After intravenous contrast MRI can distinguish granulation tissue from effusions. Diffusion weighted MR can differentiate between a cholesteatoma, which has a restricted diffusion, and other abnormalities - especially granulation tissue - which have normal diffusion characteristics (figure). Otosclerosis is a genetically mediated metabolic bone disease of unknown etiology. It is sometimes called otospongiosis because the disease begins with an otospongiotic phase, which is followed by an otosclerotic phase when osteoclasts are replaced by osteoblasts and dense sclerotic bone is deposited in areas of previous bone resorption. When this process involves the oval window in the region of the footplate, the footplate becomes fixed, resulting in conductive hearing loss. Conductive hearing loss develops early in the third decade and is considered to be the hallmark of the disease. However, involvement of other portions of the otic capsule can result in mixed sensorineural hearing loss. The process starts in the region of the oval window, classically at the fissula ante fenestram, i.e. in front of the oval window (fenestral otosclerosis). It can also occur around the cochlea (retrofenestral otosclerosis). On the left a transverse CT-image of a 23-year old female with conductive hearing loss. There is a subtle otosclerotic focus in the characteristic site: the fissula ante fenestram (arrows). On CT the detection of otosclerosis can be difficult to the inexperienced eye because the spread of the disease is often symmetrical. A small lucency at the fissula ante fenestram is typical for otosclerosis. In more severe cases lucencies are also present around the cochlea. Sometimes the whole otic capsule is surrounded by these 'otospongiotic' foci, forming the so-called fourth ring of Valvassori. In a minority of patients the disease is unilateral. This is virtually always limited to a lucency at the fissula ante fenestram. On the left a 49-year old male with left sided conductive hearing loss. There is a lucency anterior to the oval window (arrow) and between the cochlea and the internal auditory canal. This is combined fenestral and retrofenestral otosclerosis. Same patient. Notice that the otosclerosis is seen on both sides. On the left a patient with a well-positioned metallic stapedial prosthesis: medially it touches the oval window and laterally it connects with the long process of the incus. Notice the lucency between vestibule and cochlea as a manifestation of otosclerosis (arrow). On the left images of a 68-year old woman who experienced a traumatic head injury 50 years ago. Three years ago she was diagnosed with total hearing loss of the right ear. The image shows a subluxation of the incudomallear joint (arrow). Left ear for comparison. Fractures of the temporal bone are associated with head injuries. The consequences of the intracranial injuries dominate in the early period after the trauma. A temporal bone fracture can manifest itself with acute signs like bleeding from the ear or acute facial paralysis. Hearing loss is of course not a life-threatening event. Temporal bone fractures can be classified as longitudinal or transverse. Longitudinal fractures generally spare the inner ear, which is more often breached by transverse fractures. However, many temporal bone fractures are neither longitudinal nor transverse and a comprehensive description of the structures which are crossed by the fracture is needed. On the left images of a woman who had fallen down from the stairs three days earlier. She suffered from severe sensorineural hearing loss on the left side. A longitudinal fracture is visible, which courses anteriorly to the cochlea through the region of the geniculate ganglion (arrows). There were no signs of facial nerve paralysis. No fracture line could be seen across the inner ear. Opacification of the middle ear, likely as a result of a hematotympanum. Right ear for comparison. Posttraumatic conductive hearing loss can be caused by a hematotympanum or a tear of the tympanic membrane. In these cases the hearing loss usually resolves spontaneously. In persistent conductive hearing loss there is usually a disruption of the ossicular chain. The most common disruption is a dislocation of the incudostapedial joint which is often a subtle finding. Disruptions can occur at the incudomallear joint. Fractures of the long process of the incus or the crura of the stapes are difficult to diagnose. Fractures of the inner ear are seen in posttraumatic sensorineural hearing loss. Careful inspection is required in order to pick out these thin fracture lines. Cochlear concussion with blood in the cochlea can be visualized with MRI. On the left images of a man who had suffered a traumatic head injury two months previously. He complained of intermittent tinnitus. There is a longitudinal fracture (yellow arrow) coursing through the mastoid towards the region of the geniculate ganglion. There is a dislocation of the incus with luxation of the incudo-mallear and incudo-stapedial joint (blue arrow). No involvement of the inner ear. Left ear for comparison. On the left images of a 54-year old male several years after head trauma, followed by left-sided hearing loss. There is a transverse fracture through the vestibule and facial nerve canal (arrows). The lateral semicircular canal is partially filled with dense material, compatible with labyrinthitis ossificans. Facial nerve paralysis can be acute or delayed. In acute posttraumatic paralysis a fracture line through the facial nerve canal - usually in the tympanic part - can be observed, sometimes with a bony fragment impinging on the canal. In delayed facial paralysis the nerve is probably edematous and fracture lines can be absent. On the other hand, a fracture line may be seen to cross the facial nerve canal without any associated nerve dysfunction. Several normal structures may be mistaken for fractures: A vascular anomaly can be suspected if the patient complains of pulsatile tinnitus or when there is a reddish or bluish mass behind the eardrum. Vascular anomalies are: Non-vascular anomalies which can also manifest as a retrotympanic mass: In patients with an aberrant internal carotid artery the cervical part of the internal carotid artery is absent. It is replaced by the ascending pharyngeal artery which connects with the horizontal part of the internal carotid artery. It courses through the middle ear. On the left coronal images of the same patient. On the right side the internal carotid artery is separated from the middle ear (blue arrow). On the left side the internal carotid artery courses through the middle ear (red arrow) On the left a dehiscent jugular bulb (blue arrow). This can be dangerous during myringotomy. Note also the bulging sigmoid sinus (yellow arrow). Tumors of the temporal bone are rare. The following tumors can be seen: Schwannomas will not be discussed On the left bilateral bony lesions of the external auditory canal, typical of exostoses. Exostoses of the external auditory canal are usually multiple, sessile, and bilateral and can cause severe narrowing of the external auditory canal. Exostoses are caused by contact with cold water and mostly seen in swimmers and surfers. Osteomas are less common and mostly unilateral and pedunculated. On the left images of a 57-year old male with a slowly progressive glomus jugulotympanicum tumor, visible as a mass on the floor of the tympanic cavity (arrow). Mucus is seen in the meso- and epitympanum. At CT, the glomus jugulotympanic tumor manifests as a destructive lesion at the jugular foramen, often spreading into the hypotympanum. The bone can be permeated by tumor. The glomus tympanicum tumor is typically a small soft tissue mass on the promontory. Large tumors have a 'salt and pepper' appearance at MRI due to their rich vascularity with flow voids. They enhance strongly after i.v. contrast. Embolisation can diminish intra-operative blood loss. On the left angiographic images of the left external carotid artery before embolisation and the common carotid artery after embolisation (blue arrow). Only a faint blush remains. Glomus tumors arise from paraganglion cells which are present in the jugular foramen and on the promontory of the cochlea around the tympanic branch of the glossopharyngeal nerve. Elderly persons are most commonly affected with a female predominance. The presenting symptoms are conductive hearing loss, tinnitus, and pain. At otoscopy a blue ear drum is seen. Glomus tumors of the jugular foramen (also called glomus jugulotympanicum tumors) are more common than tumors which are confined to the middle ear (glomus tympanicum tumor) ELST is a rare entity. These tumors originate from the endolymphatic sac. We will discuss them because their CT appearance is very typical. At CT a destructive process is seen on the dorsal surface of the petrosal part of the temporal bone with punctate calcifications. On MRI there is usually strong enhancement. The amount of destruction in this case would be atypical for a meningioma. On the left a large destructive process of the dorsal temporal bone. Parts of the tumor show strong enhancement. There is a cystic component on the dorsal aspect which does not enhance. Notice the cystic component of the tumor on a T2W-image The postoperative ear is often difficult to describe. A previous CT-examination, if present, can be a lot of help. The best one can do is to describe the extent of the previous operation, the state of the ossicular chain (if present), and the aeration of the postoperative cavity. This cavity can be filled with swollen mucosa, recurrent disease or with some tissue implanted during the operation. On the left images of a 15-year old girl with chronic otitis media, who was treated with an attico-antrotomy. An entry into the antrum is created, but most of the mastoid air cells are still present. The ossicular chain is preserved. On the left coronal images of the same patient. Notice the thickened and calcified eardrum. On the left images of a 42-year old male who was treated with a mastoidectomy. The posterior wall of the external auditory canal and the ossicular chain are intact. Almost all of the mastoid air cells are removed. On the left coronal images of the same patient. The posterior wall of the external auditory canal and the ossicular chain are intact. Almost all the mastoid air cells are removed. If the Eustachian tube is assumed to be dysfunctioning, tympanostomy tubes can be inserted into the eardrum to facilitate the drainage of middle ear fluid. In young children the course of the Eustachian tube between the middle ear and the nasopharynx runs more horizontally than in adults, predisposing to stasis of fluid in the middle ear and secondary infection. After a while tympanostomy tubes are extruded by the eardrum and can be seen to lay in the external auditory canal. The image on the left shows a dislocated tube lying in the external auditory canal. Stapes prostheses are inserted in patients with otosclerosis to replace the native stapes, which is fixed in the oval window. Steel stapes prostheses are easily visible. Prostheses made of Teflon can be almost invisible. One should describe the position of the prosthesis in the oval window and the integrity of its connection with the long process of the incus. On the left images of a metallic stapes prosthesis. The prosthesis is in a good position. Medially it lies in the oval window, laterally it connects to the long process of the incus. On the left a patient with a stapes prosthesis. The metallic prosthesis is dislocated and lies in the vestibule. A re-operation was performed and a new prosthesis was inserted. Notice the small lucency at the fissula ante fenestram, a sign of otosclerosis (arrow). On the left images of a patient with a synthetic stapes prosthesis. It is connected to the long process of the incus (yellow arrow). The tip lies in the oval window (blue arrow). A remodelled incus can be used to repair the ossicular chain. Most often it is inserted between the eardrum and the stapes superstructure. The interposed incus can either be the patient’s own or one from a cadaver. Alternatively, a Partial Ossicular Replacement Prosthesis (PORP) or Total Ossicular Replacement Prosthesis (TORP) can be used. On the left images of a 13 -year old boy. Five years earlier a cholesteatoma was removed. The following year the ossicular chain was reconstructed with a donor incus (arrow). The malleus handle is present. On the left coronal images of the same patient. On the left axial images of a patient with a reconstruction of the ossicular chain with an autologous incus (arrow) between the ear drum and the stapes. In postoperative imaging look for dehiscence of the bony covering of the sigmoid sinus and for interruption of the tegmen tympani. A herniation of cranial contents can be present. If the tegmen is disrupted and continuous soft tissue is present between the middle ear and the cranial contents, MRI can be used to demonstrate if there is a postoperative meningo (encephalo)cele. On the left images of a 24 year old female. She was operated at the age of 8 for chronic otitis media. Since one year progressive hearing loss of the right ear. CT shows a rounded mass (arrow) in the attico-antrotomy with erosion of the tegmen tympani. An MRI depicts a mass in the mastoid abutting the dura. At operation a large cholesteatoma was removed. The dura was intact. The defect was closed with a flap of temoral muscle and a chain recontruction was done. On the left an MRI image of the same patient. The MRI depicts a mass in the mastoid abutting the dura. The dura is intact. Cochlear implantation is performed in patients with sensorineural deafness due to degeneration of the organ of Corti. After implantation of a multichannel electrode a wide array of electrical pulses can be produced to stimulate the acoustic nerve. The electrode is inserted into the scala tympani of the cochlea via the round window or via a drill hole directly into the basal turn. Post-operatively its position can be evaluated with plain films or with CT. A well-inserted electrode is positioned with all its channels in the cochlea and extends up to the top of the cochlea. On the left an image of an one-year old girl with unsatisfactory hearing tests after cochlear implantation. CT shows that the electrode is inserted next to the internal carotid artery in the carotid canal. Criteria for cochlear implantation are: MRI can demonstrate fibrous obliteration of the cochlea, something which is not appreciated on CT. MRI, on the other hand, can show a fluid-filled cochlea while CT depicts small calcifications. Therefore, a combination of both modalities can be used. MRI can also demonstrate absence of the 8th nerve, which precludes cochlear implantation. Labyrinthitis ossificans is seen after meningitis. It is a condition in which the inner ear is filled with fibrotic tissue, which calcifies. It mostly affects the cochlea, but the vestibule and semicircular canals can also be involved. On the left images of a 56-year old male, who is a candidate for cochlear implantation. Small calcification in basal turn of cochlea as a result of labyrinthitis ossificans (arrows). On the left images of a 14-year old boy with bilateral sensorineural hearing loss. Calcification of superior semicircular canal on the left (yellow arrow). Right ear for comparison (blue arrow). On the left coronal images of the same patient. by Vercruysse JP, De Foer B, Pouillon M, Somers T, Casselman J, Offeciers E. Eur Radiol 2006; 16:1461-1467Eric Beek and Frank Pameijer Cochlear cleft Petromastoid canal Cochlear aqueduct High jugular bulb Jugular bulb diverticulum Bulging sigmoid sinus Large vestibular aqueduct External auditory canal atresia Cochlear deformities Lateral semicircular canal malformation Aberrant internal carotid artery Dehiscent jugular bulb Exostoses Glomus tumor EndoLymphatic Sac Tumor (ELST) Attico-antrotomy Mastoidectomy Tympanostomy tubes Stapes prosthesis Incus interpostion Postoperative meningoceleTemporal bone - PathologyRadiology department of the University Medical Centre of Utrecht, the Netherlands
msk1
1
Ankle fracture - Mechanism
by Robin Smithuis
The ankle is the most frequently injured joint. Management decisions are based on the interpretation of the AP and lateral X-rays. In this article we will focus on: Fracture mechanism and Radiography by Robin Smithuis The ankle joint has to be flexible in order to deal with the enormous forces applied exerted on the talus within the ankle fork. . The medial side of the joint is quite rigid because the medial malleolus - unlike the lateral malleolus - is attached to the tibia and the medial collateral ligaments are very strong. On the lateral side there is a flexible support by the fibula, syndesmosis and lateral collateral ligaments. This lateral complex allows the talus to move laterally and dorsally in exorotation during forward motion and subsequently pushes it back into its normal position. The fibula has no weight-bearing function, but merely serves as a flexible lateral support. The syndesmosis is the fibrous connection between the fibula and tibia formed by the anterior and posterior tibiofibular ligaments - located at the level of the tibial plafond (French for ceiling) - and the interosseus ligament, which is the thickened lower portion of the interosseus membrane. The anterior and posterior tibiofibular ligaments are often referred to as anterior and posterior syndesmosis. There are two positions of the foot in which the flexible ankle joint becomes a rigid and vulnerable system: extreme supination and pronation. In these positions forces applied to the talus within the ankle mortise can result in fractures of the malleoli and rupture of the ligaments. In 80% of ankle fractures the foot is in supination. The injury starts on the lateral side, since that is where the maximum tension is. In 20% of fractures the foot is in pronation with maximum tension on the medial side. The injury starts on the medial side with either a rupture of the medial collateral ligaments or an avulsion of the medial malleolus. The shape of a fracture indicates which forces were involved. An oblique or vertically oriented fracture indicates 'push-off'. A transverse or horizontal fracture is the result of a 'pull-off'. On the left image the lateral malleolus is pushed off by exorotation of the talus. On the right image the medial malleolus is pulled off by the medial collateral ligament due to pronation of the foot. The ankle can be thought of as a ring in which bones as well as ligaments play an equally important role in the maintenance of joint stability. If the ring is broken in one place the ring remains stable. When it is broken in two places, the ring is unstable and may dislocate. Now anyone can figger out, that an ankle is unstable when both the medial and the lateral malleoli are fractured. It becomes more problematic when there is a combination of a fracture and a ligamentous rupture, because the ligamentous rupture is not detectable on the X-ray. In some fractures there may even be a proximal fibular fracture - which is not visible on the ankle radiographs - in combination with ligamentous ruptures at the level of the ankle. It is important to realize that in these cases the radiographs of the ankle may be normal, while there still is an unstable ankle injury. There is also an ring of stability in the axial plane. When the anterior and posterior syndesmosis rupture or avulse, then the ankle mortise is unstable. There are many combinations of avulsion fractures and ligamentous ruptures that can produce an unstable ring in the axial plane. A Anteriorly the anterior syndesmosis (or antior tibiofibular ligament) is one of the first structures to rupture. When the posterior syndesmosis also ruptures, then the ankle is unstable. B Less commonly the anterior syndesmosis avulses from the tibial attachment - Tilleaux fracture. C On the posterior side frequently the malleolus tertius avulses. Sometimes these fractures are difficult to detect, as we will discuss in a moment. D After the injury the bones frequently align again. Stability (2) It is important to realize, that for the stability of the ankle it doesn't matter whether there is a rupture of a ligament or an avulsion at the insertion. Almost every ligamentous rupture has a fracture equivalent. Stability (3) On the left image a Weber A or SA-fracture. This ankle is stable because there is only an avulsion fracture of the lateral malleolus below the level of the syndesmosis. The ring is broken in only one place. On the right image there is an unstable fracture. The ring of the ankle is broken in two places. There is a lateral fracture and on the medial side there is a rupture of the collateral ligament allowing the talus to dislocate laterally. Stability (4) The medial clear space should not exceed 4 mm and is usually equal to the distance between the tibial plafond and the talus. Widening of the medial joint space up to 6 mm or more requires disruption of the medial collateral ligament. Stability (5) The lateral clear space is measured from the medial border of the fibula to the lateral border of the posterior tibia 1cm above the tibial plafond. It is less well defined because its width varies with positioning. Evident widening of the lateral clear space indicates syndesmotic rupture. Some state that a width of 5.5 mm is abnormal. It is very important to realize that a normal lateral or medial clear space does not exclude ligamentous rupture. It simply means that there is no dislocation, but there can still be instability. The case on the left shows a Weber B fracture. On these images the ankle fork is normal. Both the medial and lateral clear spaces are prominent, but within normal limits. We can conclude that there is no dislocation, but we do not know if there is rupture of the medial collateral ligaments or of the syndesmosis. Continue with the images post surgery. Following osteosynthesis there is obvious widening of the medial and lateral clear spaces (image on the far left). This indicates that there is a syndesmotic rupture and medial collateral ligament rupture. The ring is still broken in two places. The ankle joint is unstable and dislocated. Resurgery was necessary with placement of a syndesmotic screw to stabilize the ankle joint. Stability (6) On the left another case. There is a Weber B fracture. Both the medial and lateral clear spaces are widened, indicating instability. The talus is displaced laterally. Patient was scheduled for osteosynthesis of the fibular fracture and placement of a syndesmotic screw if necessary. After osteosynthesis of the fibula, the ankle was tested in the operating room and found to be stable. There was no indication for a syndesmotic screw. It was concluded that the syndesmosis was only partially ruptured, as is usually the case in Weber B fractures. The ring was broken in two places and after repairing one of them, the ring was stable. These rules are used to determine the need for radiographs in patients with an ankle injury. Ankle X-ray series are only required in case of: Pain in the malleolar zone and any one of the following: A basic radiographic examination of the injured ankle consists of an AP-view, a Mortise-view and a lateral view. The Mortise-view is an AP-view taken with a 15-25? endorotation of the foot. The technologist turns the foot inwards until the lateral malleolus is at the same height as the medial malleolus. This view visualizes both the lateral and medial joint spaces. On a true AP-view the talus overlaps a portion of the lateral malleolus, obscuring the lateral aspect of the ankle joint. Many think that for a good lateral view the distal fibula should be in the center of the distal tibia. However, since the fibula is positioned more dorsally, the fibula should project over the posterior part of the distal tibia (arrow). Malpositioning of the lateral view is the most common mistake in radiography of the injured ankle. Because the patient is in pain, the technologist is afraid to let the patient turn the ankle fully lateral. This is one of the reasons why we miss so many fractures of the malleolus tertius. The CT demonstrates a large tertius fracture. On the lateral view and also on the AP- and Mortise views, which will be shown in the paragraph on tertius fractures, this fracture was not visible. The explanation is that on the lateral radiograph the fibula projects in the middle of the tibia. The x-ray beam is not parallel to the fracture line. Since the fracture line of a tertius fracture always has this orientation, we must insist on a true lateral view. On a well positioned lateral view the tertius fracture is obvious (red arrow). This was the only fracture that was seen on the x-rays of the ankle and this patient turned out to have an unstable Weber-C fracture and went for surgery. The x-ray beam has to be centered on the malleoli. Notice the exorotation of the foot for a proper lateral view. Fractures of the ankle, combined experimental-surgical and experimental-roentgenologic investigations by N. Lauge-Hansen (1948) Die verletzungen des oberen sprunggelenkes by B.G. Weber (1966) East Lancashire Foot and Ankle Hyperbook The AO Surgery Reference is a huge online repository of surgical knowledge, consisting of more than 7000 pages.Robin Smithuis Normal flexibility of the ankle Vulnerable positions of the foot Pull-off or Push-off fractures Stability Mortise view Lateral view Malpositioning of the Lateral viewAnkle fracture - MechanismRadiology Department of the Rijnland Hospital, Leiderdorp, the Netherlands
msk2
1
Ankle fracture - Special cases
by Robin Smithuis
The ankle is the most frequently injured joint. Management decisions are based on the interpretation of the AP and lateral X-rays. In this article we will focus on detection of fractures, that may not be so obvious at first sight. Before you read this article, you need to understand the classification of ankle fractures and exorotation injuries that were highlighted in Ankle - Fractures 1 and 2. Almost all fractures of the malleolus tertius are part of a rotational injury resulting in a Weber B or Weber C fracture. The tertius fracture is stage 3 in Weber B and stage 4 in Weber C (figure). In some cases the tertius fractures are easily seen on the x-rays, but frequently they can be difficult to detect. It is important to find these fractures, since a tertius fracture can be the only clue to an unstable ankle injury. When we study the radiographs of a patient with an ankle injury, we have to study the region of the malleolus tertius very carefully. In many cases there is only a small gap between the fracture parts and detection depends on optimal radiography and a high level of suspicion. The images show an obvious Weber B fracture. On the AP-view the linear lucency is the clue to a tertius fracture (red arrow). It results from subtle malalignment of the fracture fragment. Likewise in some cases malalignment can result in a linear density. In this case there is a Weber B fracture with avulsion of the medial malleolus. The bright line on the AP-view indicates a large tertius fracture fragment. This tertius fracture can also be seen on the lateral view, but in many cases we need all the information of both the lateral and AP-view to diagnose a tertius fracture. Here more examples of the bright line that indicates a tertius fracture. In some cases a fracture of the malleolus tertius is barely or not visible on the radiographs and can only be seen on CT. First study the radiographs and then continue with the CT. By the way....there are two fractures. You can enlarge the images by clicking on them. The CT shows an avulsion of the tertius at the insertion of the posterior syndesmosis (red arrows). The alignment is so perfect, that you do not see the fracture on the radiographs. Maybe the fracture is seen on the AP-view as indicated by the red arrows, but this is questionable. Notice that there is also an avulsion at the tibial insertion of the anterior syndesmosis, i.e. Tilleaux fracture. This combination of findings implicate that the ankle is unstable. A syndesmotic screw has to be inserted. Here we have images of an extremely difficult case. This woman had a distortion of the ankle and had pain on both medial and lateral side. She was referred to the radiology department by her general practitioner. The technician made the standard AP-, Mortise- and lateral view and showed them to the radiologist, who was a little bit puzzled. First study the images and then continue reading. The findings are: The radiologist decided first to order a CT to find out if there really was a tertius fracture. Continue with the CT and be amazed. Scroll through the images. It is amazing, that such a large tertius fragment is so difficult to see on the radiographs. Also notice the soft tissue swelling on the medial side indicating rupture of the medial collateral ligaments (arrow). Do you have an idea what kind of injury this is? Medial soft tissue swelling and a tertius fracture are both indications of a Weber C or Pronation Exorotation injury. Since there is no fibula fracture seen on the x-rays of the ankle, there must be a high fibular fracture. At physical exam there was some swelling on the medial side and although the patient did not complain of any pain higher in the lower leg, there was some tenderness when the fibula was palpated. This spot was marked and a fracture was found. This case illustrates the importance of medial soft tissue swelling aswell as the finding of a tertius fracture. According to Lauge Hansen we can conclude that this patient first had a rupture of the medial collateral ligaments (stage 1), followed by a rupture of the anterior syndesmosis (stage 2) and a high fibula fracture (stage 3) and finally an avulsion of the malleolus tertius, i.e. PE stage 4. At surgery the ankle was found to be unstable and syndesmosis screws were inserted. There was an indication for fixing the posterior malleolar fracture, since the fragment involved more than 25% of the articular surface of the distal tibia. This patient had a twisted ankle and the only abnormality is seen on the lateral view. This was thought to be an avulsion of the malleolus tertius. Knowing that this can be the only clue to a high Weber C, additional radiographs were taken. Continue with the images of the lower leg. A subtle high fibula fracture is seen (arrow). Final diagnosis is a Weber C fracture or according to Lauge Hansen: Pronation Exorotation injury stage 4. A fracture of the malleolus tertius as an isolated finding is very uncommon. It is seen when someone's foot hits the ground and a fragment of the malleolus tertius is pushed off by the talus. The size of this fragment depends on the direction of the force (figure). The Salter-Harris classification describes fractures that involve the epiphyseal plate or growth plate. The most common is type II, which accounts for 75%.  These Salter-Harris fractures can be easily missed. In many cases there is only minimal or no displacement. The fracture through the growth plate is usually obscure and difficult to differentiate from normal variations of the growth plate. And finally we tend not to look carefully at the epiphysis. Type I Salter-Harris fractures tend to occur in younger children (5). It is a transverse fracture through the cartilage of the growth plate or physis. Often, x-rays of a child with a type I growth plate fracture will appear normal. Most type I growth plate injuries are treated with a cast. Healing of type I fractures tends to be rapid and complications are rare. A type II growth plate fracture starts across the growth plate, but the fracture then continues up through the metaphysis. This is the most common type of growth plate fracture, and tends to occur in older children. Often type II growth plate fractures must be repositioned under anesthesia, but healing is usually quick and complications are uncommon. Type III is a fracture through the growth plate and epiphysis sparing the metaphysis. A type III fracture also starts through the growth plate, but turns and exits through the end of the bone, and into the adjacent joint. These injuries can be concerning because the joint cartilage is disrupted by the fracture. Proper positioning is essential after a type II growth plate fracture. These injuries also tend to affect older children in whom the growth plate is partially closed. Study the images and then scroll to the next images. The fracture through the epiphysis can be easily missed (blue arrow). The fracture through the growth plate is only seen on CT. Continue with the CT images. The CT-images nicely display the fracture through the growth plate and the epiphysis. Study the images and then scroll to the next images. This is also a Salter-Harris type III fracture. Notice that there is also a Tilleaux fracture. We will discuss these fractures in a moment. Type IV is a fracture through all three elements of the bone, the growth plate, metaphysis and epiphysis. Notice that the epiphyseal fracture is in the sagittal plane, the fracture through the growth plate is in the axial plane and the metaphyseal fracture is in the coronal plane. These fractures are also named triplane fractures. These are discussed in the next chapter. Proper positioning is also essential with type IV growth plate fractures, and surgery may be needed to hold the bone fragments in proper position. Type V growth plate injuries occur with the growth plate is crushed. Type V growth plate fractures carry the most concerning prognosis as bone alignment and length can be affected. These types of fractures may permanently injure the growth plate, requiring later treatment to restore alignment of the limb. This fracture is named triplane because it occurs in the coronal, sagittal and axial plane. It is actually a Salter-Harris type IV. It is seen exclusively in young adolescents in the period, when the medial tibial epiphysis is closed, while the lateral portion is still open leaving it vulnerable to injury. As the force cannot continue into the medial part of the growth plate since this is already closed, the epiphysis will fracture. As in most ankle fractures the mechanism is external rotation. The injury results in: Study the images and then continue reading. At first this looks like a Weber B fracture with an oblique fracture in the fibula as seen on the lateral view (black arrows). Notice however that this fracture line stops at the level of the epiphyseal plate. So this is the fracture of the metaphysis in the coronal plane. On the AP-view there is a lucency within the epiphysis, which is the epiphyseal fracture in the sagittal plane. Notice also that the medial epiphysis is already closed, while the lateral portion is still open(blue arrows). We have to assume that there is an epiphysiolysis of this lateral portion. Here another example. There is only a small metaphyseal fragment, which is usually the case (red arrow). The fracture through the epiphysis is indicated by the blue arrow. In 1840 Maisonneuve described a frature of the proximal shaft of the fibula, which was caused by exorotation force applied to the ankle. It is a high Weber C fracture. These fractures are easily overlooked because the patients rarely complain of pain in the region of the proximal fibula, since the ankle is most painful. There are three situations in which we should suspect a high Weber C or Maisonneuve fracture: Isolated fracture of the medial malleolus According to Lauge-Hansen this is the first stage of a pronation exorotation injury, which results in a Weber C fracture. So we have to look for higher stages. The injury can continue to the following: In all these subsequent stages, purely ligamentous injury will not be visible on the radiographs of the ankle. So even in a Weber C stage 4 sometimes only a fracture of the medial malleolus will be visible. In the illustration we see the fractures and ligamentous injury on the left and the resulting x-rays on the right. Isolated fracture of the malleolus tertius Truely isolated fractures of the malleolus tertius are very uncommon. Most fractures of the malleolus tertius are part of a complex ankle injury, either Weber B or Weber C. A Weber B fracture is easily detected because of the characteristic oblique fracture. So if there is a tertiu sfracture and no sign of a Weber B fracture, then we have to start looking for a high Weber C fracture. In that case we have the following combination: An isolated tertius fracture on the ankle radiographs indicates the presence of an unstable ankle injury. Any medial painful swelling or hematoma Normal radiographs do not rule out a Weber C fracture. We may have the following combination: Example 1 On the left images of a patient with a hematoma on the medial side.   This case demonstrates that there can be an unstable ankle injury that needs surgery even when the radiographs of the ankle do not show a fracture. In any patient with an ankle injury you should always ask yourself the question......can I exclude a high Weber C fracture or do I need additional imaging. Example 2  Example 3 In this case no fracture is seen, but only soft tissue swelling on the medial side. In such a case, you have to rule out a Maisonneuve fracture, which is a high Weber C fracture. Additional x-rays of the lower leg were taken. There is a high fibula fracture. External rotation injury of the ankle is the most common ankle injury and can lead to a Weber B or Weber C fracture. One of the first stages in this injury is rupture of the anterior tibiofibular ligament (or anterior syndesmosis). Less frequently it leads to an avulsion of the anterolateral tibial epiphysis. Whenever you see such a fracture, you have to look for higher stages of this exorotation injury. The x-ray shows a subtle Tilleaux fracture, which is better appreciated on the CT-images. Study these images carefully and remember the stages of an exorotation injury. What is going on here? There is a Tilleaux fracture due to avulsion of the anterolateral part of the distal tibia by the anterior syndesmosis. This can be a stage 2 of a Weber C fracture. Stage 1 is rupture of the medial collateral ligaments and stage 3 is a fibula fracture above the level of the syndesmosis. So now we start looking for stage 4, which is rupture or avulsion of the posterior syndesmosis. Do you now see the tertius fracture on the axial CT-image? This patient has an unstable ankle injury and a syndesmotic screw needs to be inserted. Another Tilleaux in a patient with a strange combination of findings. There is an avulsion of the lateral malleolus, a Tilleaux and a medial malleolar fracture. A Tilleaux fracture is more commonly seen in adolescents at the age of 12 -15 years. At that age it is a fracture through the growth plate and is then called a juvenile Tilleaux. It occurs before the distal tibial epiphysis has completely fused. The fracture occurs when the medial epiphysis has fused and the lateral part becomes avulsed at the attachment of the anterior tibiofibular ligament (or syndesmosis). Study the images and then continue reading. You can enlarge the images by clicking on them. There is a subtle widening of the lateral part of the growth plate of the right ankle. There is also a very subtle fracture through the epiphysis. Continue with the CT. The CT-images show a epiphysiolysis fracture Salter Harris type 3. This juvenile Tilleaux is especially seen in young athletes. Always look for higher stages of an exorotation injury. Fractures of the ankle, combined experimental-surgical and experimental-roentgenologic investigations by N. Lauge-Hansen (1948) East Lancashire Foot and Ankle Hyperbook The AO Surgery Reference is a huge online repository of surgical knowledge, consisting of more than 7000 pages. By Jonathan Cluett, M.D., About.com GuideRobin Smithuis Isolated Tertius fracture Type I Type II Type III Type IV Type V juvenile TilleauxAnkle fracture - Special casesRadiology Department of the Rijnland Hospital, Leiderdorp, the Netherlands
msk3
1
Ankle fracture - Weber and Lauge-Hansen Classification
by Robin Smithuis
Classification of ankle fractures is important in order to estimate the extent of the ligamentous injury and the stability of the joint. The Weber classification focuses on the integrity of the syndesmosis, which holds the ankle mortise together. The Lauge-Hansen system focuses on the trauma mechanism. Adding the stages of Lauge-Hansen to the Weber system will help you to predict ligamentous injury and instability. This article will help you to correctly stage ankle injuries and to detect fractures, that are not obvious at first sight. Basically there are three main types of ankle fractures. Weber classified them as: These fractures are identical to the fractures described by Lauge-Hansen as supination-adduction, supination-exorotation and pronation-exorotation. We will first give a short overview of these fractures and then discuss them in more detail. Once you understand the trauma mechanism as described by Lauge-Hansen and the sequence of events that take place in stages, then you know where to look for fractures and ligamentous injuries. Occurs below the syndesmosis, which is intact. According to Lauge-Hansen, it is the result of an adduction force on the supinated foot. Scroll through the images. Notice that the fibular fracture is transverse, because it is an avulsion or pull-off fracture. The tibial fracture is vertical or oblique, because it is a push-off fracture. This is a transsyndesmotic fracture with usually partial - and less commonly, total - rupture of the syndesmosis. According to Lauge-Hansen, it is the result of an exorotation force on the supinated foot. Scroll through the images. Notice the oblique or vertical orientation of the push-off fibular fracture. This is a fracture above the level of the syndesmosis. Usually there is a total rupture of the syndesmosis with instability of the ankle. According to Lauge-Hansen, it is the result of an exorotation force on the pronated foot. Scroll through the images Weber A fractures are usually not a problem. Weber B and C are more difficult and it is essential to understand the sequence of events in these injuries, which are both exorotation injuries. This implies that 75-80% of ankle injuries are exorotation injuries. Weber B starts anterolaterally and the sequence is: Weber C starts medially and the sequence is: Another important thing to remember is, that a ligament can rupture or cause an avulsion fracture at the insertion. Every ligamentous rupture has it's avulsion fracture counterpart.  Instability is seen in: In daily practice most use the Weber system, which is easy to memorize, while the Lauge-Hansen seems rather difficult at first glance. Combining the simplicity of Weber with the explanation of the trauma mechanism given by Lauge-Hansen has the advantage that you still use a simple system, but now you really know what is going on. For instance if you see a fracture that is a stage 2 in the Lauge-Hansen system, then you know that there also is a stage 1 injury and you will study the radiographs with a high suspicion for signs of stage 3 and 4. This can best be demonstrated by giving an example.   Additional radiographs of the lower extremity were ordered and they demonstrate a high fibular fracture, i.e. Weber C stage 3 also known as a Maisonneuve fracture. This is un unstable ankle injury that needs surgical repair. This example is an every day case. The point that I want to make is, that when you understand the sequence of injuries to the ankle, then you know where to look for fractures and soft tissue swelling indicating ligamentous injury. We will now discuss the Weber classification and add the stages of the Lauge-Hansen system. Weber A is seen in 20-25% of all ankle fractures. The diagnosis as well as the treatment is usually no problem. According to Lauge-Hansen the fracture results from an adduction force on the supinated foot. The lateral side is under extreme tension with stretch on the ligaments which results in an avulsion fracture. Almost always the avulsion is seen as a horizontal fracture. This is called a pull off type of fracture in contrast to a push off type, which is seen as an oblique or vertical fracture. The images show the usual Weber type A fractures. These are all stage-1 fractures. Stage-2 is extremely uncommon. Notice the horizontal orientation of the fracture lines. These are pull off type fractures as a result of avulsion. Stage 2 is uncommon and easy to detect. More adduction force results in the medial malleolus being pushed off in a vertical or oblique way. Stage 2 is unstable because the ring of the ankle is broken in two places. Notice the horizontal orientation of the lateral malleolus fracture and the vertical orientation of the fracture of the medial malleolus. Enormous forces must have pushed off the medial malleolus. More on the ring of the ankle and instability in the article 'Ankle fracture - Mechanism'. Weber B is the most common type of ankle fracture and occurs in about 60 %. According to Lauge-Hansen the fracture results from an exorotation force on the supinated foot. Stage 1 is usually not visible on x-rays. What we normally see is a stage 2 oblique fracture through the syndesmosis and we have to assume that there is also a rupture of the anterior tibiofibular ligament, which is stage 1. According to Lauge Hansen the first injury is on the lateral side, which is under maximum tension. In stage 2 the talus exorotates further and since the foot is in supination, the lateral malleolus is held tightly in place by the lateral collateral ligaments. The lateral malleolus cannot move away without breaking. As a result more rotation of the talus will fracture the fibula in an oblique or spiral fashion because the lateral malleolus is pushed off from anteromedially to posterolaterally. The images show a Weber B fracture. The oblique course of the fracture is typical for Weber B and results from the exorotation of the talus that pushes against the fixed lateral malleolus. The malleolar fracture usually starts medially at the level of the talar dome, but can also start a few centimeters above this level. Stage 3 More posterior displacement of the lateral malleolus fragment by the talus results in tension on the posterior syndesmosis with rupture or avulsion of the malleolus tertius. Stage 4 Further posterior movement of the talus will result in extreme tension on the medial side and the deltoid ligament will either rupture or pull off the medial malleolus in the transverse plane. The sequences in a Weber B fracture or Lauge-Hansen supination exorotation injury take place in a clockwise manner: Immediately after the injury the injured parts may again align, which can make it difficult to detect the injuries. The radiographs show a typical Weber B fracture. First study the images and then continue reading. Do you see what stage this is? This is a Weber B stage 4 injury. Notice that all 4 stages are visible:  These images show another typical Weber B fracture stage 4. There is an oblique fracture of the fibula. There is an avulsion of the malleolus tertius and an avulsion of the medial malleolus. Here another typical Weber B fracture stage 4. First notice the oblique fibular fracture, which is best seen on the lateral view. This is stage 2 and we have to assume, that the anterior syndesmosis is ruptured. On the lateral view a small tertius fragment is seen indicating stage 3. Now you start looking for stage 4 and you will notice the subtle lucency in the medial malleolus on the AP view (green arrow). Knowing the stages of Lauge Hansen this must be a fracture. Here a more subtle case. At first impression there is a Weber B fracture stage 2. Now we start looking for stage 3, which is a tertius fracture. The small linear density on the AP-view is enough to diagnose a tertius fracture. The soft tissue swelling on the medial side is probably a rupture of the medial collateral band , i.e. stage 4. Weber C is seen in approximately 20% of ankle fractures. It is the most difficult fracture to diagnose and the Lauge-Hansen system will help you to understand the fracture-mechanism, as this will be an enormous help. According to Lauge-Hansen the fracture results from an exorotation force on the pronated foot. Stage 1 The first injury will occur on the medial side, which is under maximum tension due to the pronation. It will lead to rupture of the medial collateral ligament or avulsion of the medial malleolus. Now the injury can stop and there will only be a rupture of the medial collateral ligaments or avulsion of the medial malleolus. Lauge Hansen calls this PE stage 1. We can not cathegorize this in the Weber classification, since there is no fibular fracture. In many cases the injury progresses to a higher stage. The talus rotates externally and moves laterally because it is free from its medial attachment. Due to the pronation, the lateral ligaments are not under tension and the fibula can move away from the tibia. This causes rupture of the anterior syndesmosis. This is stage 2. Continuous force will twist the fibula and displace it distally, while proximally it is fixed to the tibia. Finally the interosseus membrane will rupture up to the point where the fibular shaft fractures. This is stage 3. This is always above the level of the syndesmosis. In many cases it is visible on the radiographs of the ankle, but in some cases the fracture is located high and will only be visible on a radiograph of the lower leg. This last type of fracture is also called Maisonneuve fracture. Here we see the different stages in the axial plane. Scroll through the images. The radiographs shows a Weber C fracture. There is an avulsion fracture of the medial malleolus and a fibula fracture above the level of the syndesmosis. According to Lauge-Hansen this is stage 3 pronation exorotation injury and so the anterior syndesmosis (stage 2) must also be ruptured. We do not see a tertius fracture, which would indicate stage 4, but there may be a rupture of the posterior syndesmosis. Here an example of a Weber C fracture with a proximal fibula fracture. Notice that on the radiograph of the ankle no fracture is seen. You might misdiagnose this as only some soft tissue swelling. In fact this is an unstable ankle fracture, since there also must be a rupture of the medial collateral ligament (stage 1) , so the ring is broken in two places leading to instability. According to Lauge Hansen we are probably dealing with: Finally the posterior syndesmotic ligament ruptures, or there is an avulsion of the posterior malleolus, also known as malleolus tertius fracture (red arrow). The medial clear space is only slightly widened, but based on the stages of Lauge Hansen there must be a collateral band rupture. Start with a first impression and look for fractures and signs of ligamentous rupture. This impression will direct you to both a Weber as well as a Lauge-Hansen classification. The Lauge-Hansen classification will give you the fracture mechanism and the preliminary stage of the ankle injury. Now re-examine the films to make sure that you do not overlook a higher grade ankle injury. After this re-examination you can make a final report. In the final report the fracture is described according to Weber and/or Lauge-Hansen. Describe the number of malleoli involved and whether there are signs of instability or dislocation. Example 1  Example 2 Example 3 Example 4 Anatomy of the distal tibiofibular syndesmosis in adults: a pictorial essay with a multimodality approach by John J. Hermans, Annechien Beumer, Ton A. W. de Jong and Gert-Jan Kleinrensink. J. Anat. (2010) 217, pp633-645 Animation on YouTube by Dr Glass.Robin Smithuis Weber A Weber B Weber C Exorotation injury Ligamentous rupture or Avulsion Weber and Lauge-Hansen summary Stage 1 Stage 2 Stage 1-2 Stage 3-4 Stage 1 Stage 2-3 Stage 4Ankle fracture - Weber and Lauge-Hansen ClassificationRadiology Department of the Rijnland Hospital, Leiderdorp, the Netherlands
msk4
1
Bone tumor - Systematic approach and Differential diagnosis
by Henk Jan van der Woude and Robin Smithuis
In this article we will discuss a systematic approach to the differential diagnosis of bone tumors and tumor-like lesions. The differential diagnosis mostly depends on the review of the conventional radiographs and the age of the patient. Abbreviations used:  The most important determinators in the analysis of a potential bone tumor are: It is important to realize that the plain radiograph is the most useful examination for differentiating these lesions. CT and MRI are only helpful in selected cases. Read more about bone tumors in the articles 'Bone tumor - Well-defined osteolytic tumors and tumor-like lesions','Bone tumor - Ill-defined osteolytic tumors and tumor-like lesions' and 'Bone tumor A-G'. Most bone tumors are osteolytic. The most reliable indicator in determining whether these lesions are benign or malignant is the zone of transition between the lesion and the adjacent normal bone (1). Once we have decided whether a bone lesion is sclerotic or osteolytic and whether it has a well-defined or ill-defined margins, the next question should be: how old is the patient? Age is the most important clinical clue. Finally other clues need to be considered, such as a lesion’s localization within the skeleton and within the bone, any periosteal reaction, cortical destruction, matrix calcifications, etc. In the table on the left the morphology of a bone lesion is combined with the age of the patient. Notice the following: In order to classify osteolytic lesions as well-defined or ill-defined, we need to look at the zone of transition between the lesion and the adjacent normal bone. The zone of transition is the most reliable indicator in determining whether an osteolytic lesion is benign or malignant (1). The zone of transition only applies to osteolytic lesions since sclerotic lesions usually have a narrow transition zone. Small zone of transition A small zone of transition results in a sharp, well-defined border and is a sign of slow growth. A sclerotic border especially indicates poor biological activity. In patients In patients > 30years, and particularly over 40 years, despite benign radiographic features, metastasis or plasmacytoma also have to be considered On the left three bone lesions with a narrow zone of transition. Based on the morphology and the age of the patients, these lesions are benign. Notice that in all three patients, the growth plates have not yet closed. In patients > 40 years metastases and multiple myeloma are the most common bone tumors. Metastases under the age of 40 are extremely rare, unless a patient is known to have a primary malignancy. Metastases could be included in the differential diagnosis if a younger patient is known to have a malignancy, such as neuroblastoma, rhabdomyosarcoma or retinoblastoma. Wide zone of transition An ill-defined border with a broad zone of transition is a sign of aggressive growth (1). It is a feature of malignant bone tumors. There are two tumor-like lesions which may mimic a malignancy and have to be included in the differential diagnosis. These are infections and eosinophilic granuloma. Both of these entities may have an aggressive growth pattern. Infections and eosinophilic granuloma are exceptional because they are benign lesions which may seem malignant due to their aggressive biologic behavior. These lesions may have ill-defined margins, but cortical destruction and an aggressive type of periosteal reaction may also be seen. EG almost always occurs in patients Infections have to be included in the differential diagnosis of any bone lesion at any age. Age is the most important clinical clue in differentiating possible bone tumors. There are many ways of splitting age groups, as can be seen in the first table. Some prefer to divide patients into two age groups: 30 years. Most primary bone tumors are seen in patients In patients > 30 years we must always include metastases and myeloma in the differential diagnosis. A periosteal reaction is a non-specific reaction and will occur whenever the periosteum is irritated by a malignant tumor, benign tumor, infection or trauma. There are two patterns of periosteal reaction: a benign and an aggressive type. The benign type is seen in benign lesions such as benign tumors and following trauma. An aggressive type is seen in malignant tumors, but also in benign lesions with aggressive behavior, such as infections and eosinophilic granuloma. Benign periosteal reaction Detecting a benign periosteal reaction may be very helpful, since malignant lesions never cause a benign periosteal reaction. A benign type of periosteal reaction is a thick, wavy and uniform callus formation resulting from chronic irritation. In the case of benign, slowly growing lesions, the periosteum has time to lay down thick new bone and remodel it into a more normal-appearing cortex. Aggressive periosteal reaction This type of periostitis is multilayered, lamellated or demonstrates bone formation perpendicular to the cortical bone. It may be spiculated and interrupted - sometimes there is a Codman's triangle. A Codman's triangle refers to an elevation of the periosteum away from the cortex, forming an angle where the elevated periosteum and bone come together. In aggressive periostitis the periosteum does not have time to consolidate. Aggressive periosteal reaction (2)  Fibrous dysplasia, Enchondroma, NOF and SBC are common bone lesions. They will not present with a periosteal reaction unless there is a fracture. If no fracture is present, these bone tumors can be excluded. Cortical destruction is a common finding, and not very useful in distinguishing between malignant and benign lesions. Complete destruction may be seen in high-grade malignant lesions, but also in locally aggressive benign lesions like EG and osteomyelitis. More uniform cortical bone destruction can be found in benign and low-grade malignant lesions. Endosteal scalloping of the cortical bone can be seen in benign lesions like FD and low-grade chondrosarcoma. The images on the left show irregular cortical destruction in an osteosarcoma (left) and cortical destruction with aggressive periosteal reaction in Ewing's sarcoma. Ballooning is a special type of cortical destruction. In ballooning the destruction of endosteal cortical bone and the addition of new bone on the outside occur at the same rate, resulting in expansion. This 'neocortex' can be smooth and uninterrupted, but may also be focally interrupted in more aggressive lesions like GCT. Cortical destruction (3) In the group of malignant small round cell tumors which include Ewing's sarcoma, bone lymphoma and small cell osteosarcoma, the cortex may appear almost normal radiographically, while there is permeative growth throughout the Haversian channels. These tumors may be accompanied by a large soft tissue mass while there is almost no visible bone destruction. The image on the left shows an Ewing's sarcoma with permeative growth through the Haversian channels accompanied by a large soft tissue mass. The radiograph does not shown any signs of cortical destruction. Location within the skeleton The location of a bone lesion within the skeleton can be a clue in the differential diagnosis. The illustration on the left shows the preferred locations of the most common bone tumors. In some locations, such as in the humerus or around the knee, almost all bone tumors may be found. Top five location of bone tumors in alphabethic order  Differentiating between a diaphyseal and a metaphyseal location is not always possible. Many lesions can be located in both or move from the metaphysis to the diaphysis during growth. Large lesions tend to expand into both areas.   Calcifications or mineralization within a bone lesion may be an important clue in the differential diagnosis. There are two kinds of mineralization: a chondroid matrix in cartilaginous tumors like enchondromas and chondrosarcomsa and an osteoid matrix in osseus tumors like osteoid osteomas and osteosarcomas. Chondroid matrix Calcifications in chondroid tumors have many descriptions: rings-and-arcs, popcorn, focal stippled or flocculent.  Osteoid matrix Mineralization in osteoid tumors can be described as a trabecular ossification pattern in benign bone-forming lesions and as a cloud-like or ill-defined amorphous pattern in osteosarcomas. Sclerosis can also be reactive, e.g. in Ewing’s sarcoma or lymphoma.  Most bone tumors are solitary lesions. If there are multiple or polyostotic lesions, the differential diagnosis must be adjusted. Polyostotic lesions NOF, fibrous dysplasia, multifocal osteomyelitis, enchondromas, osteochondoma, leukemia and metastatic Ewing' s sarcoma. Multiple enchondromas are seen in Morbus Ollier. Multiple enchondromas and hemangiomas are seen in Maffucci's syndrome. Polyostotic lesions > 30 years Common: Metastases, multiple myeloma, multiple enchondromas. Less common: Fibrous dysplasia, Brown tumors of hyperparathyroidism, bone infarcts. Mnemonic for multiple oseolytic lesions: FEEMHI: Fibrous dysplasia, enchondromas, EG, Mets and myeloma, Hyperparathyroidism, Infection.  If you encounter printing problems with the margins of the document, simply adjust the margins or the scale of the document in the print settings. On the left an example with the print settings on a 70% scale. Fundamentals of Skeletal Radiology, second edition by Clyde A. Helms W. B. Saunders company 1995 by Mark J. Kransdorf and Donald E. Sweet AJR 1995;164:573-580 Online teaching by the Musculoskeletal Radiology academic section of the University of Washington Online teaching by the Musculoskeletal Radiology academic section of the University of Washington Online teaching by the Musculoskeletal Radiology academic section of the University of Washington by Theodore Miller March 2008 Radiology, 246, 662-674 by Henri de Groot by Jack Edeiken by Nancy M. Major, Clyde A. Helms and William J. Richardson. AJR 2000; 175:261-263 Radiological atlas of bone tumours of the Netherlands Committee on Bone Tumors by Mulder JD, et al. Amsterdam: Elsevier, 1993.Henk Jan van der Woude and Robin Smithuis Zone of transition Age Periosteal reaction Cortical destruction Location: epiphysis - metaphysis - diaphysis Location: centric - eccentric - juxtacortical Matrix Polyostotic or multiple lesionsBone tumor - Systematic approach and Differential diagnosisRadiology department of the Onze Lieve Vrouwe Gasthuis, Amsterdam and the Rijnland hospital, Leiderdorp, the Netherlands
msk5
1
Bone tumor - Well-defined osteolytic tumors and tumor-like lesions
by Henk Jan van der Woude and Robin Smithuis
In the article Bone Tumors - Differential diagnosis we discuss a systematic approach to the differential diagnosis of bone tumors and tumor-like lesions. In this article we will discuss the differential diagnosis of well-defined osteolytic bone tumors and tumor-like lesions. Abbreviations used: by Henk Jan van der Woude and Robin Smithuis On the left the most common well-defined bone tumors and tumor-like lesions. These lesions are sometimes referred to as benign cystic lesions, which is a misnomer since most of them are not cystic, except for SBC and ABC. It is true that in patients under 30 years a well-defined border means that we are dealing with a benign lesion, but in patients over 40 years metastases and multiple myeloma have to be included in the differential diagnosis. On the left a table with well-defined osteolytic bone tumors and tumor-like lesions in different age-groups. Notice the following: Most bone tumors present as well-defined osteolytic lesions, sometimes referred to as 'bubbly lesions'. It is important to have a good differential diagnostic approach to these lesions. You can use the table above, but another way to look at the differential diagnosis of well defined osteolytic bone lesions is to use the mnemonic Fegnomashic, which is popularized by Clyde Helms (1). Some prefer to use the mnemonic Fogmachines, which is formed by the same letters, but is a real word.  Fibrous dysplasia is a benign disorder characterized by tumor-like proliferation of fibro-osseus tissue and can look like anything. FD most commonly presents as a long lesion in a long bone. FD is often purely lytic and takes on ground-glass look as the matrix calcifies. In many cases there is bone expansion and bone deformity. The ipsilateral proximal femur is invariably affected when the pelvis is involved. When FD in the tibia is considered, adamantinoma should be in the differential diagnosis. Discriminator: More on Fibrous dysplasia in the article 'Bone Tumor A-G'. Enchondroma is a benign cartilage tumor. Frequently it is a coincidental finding. In the phalanges of the hand it frequently presents with a fracture. It is the most common lesion in the phalanges, i.e. a well-defined lytic lesion in the hand is almost always an enchondroma. In some locations it can be difficult to differentiate between enchondroma and bone infarct. It is almost impossible to differentiate between enchondroma and low grade chondrosarcoma based on radiographic features alone. Ollier's disease is multiple enchondromas. Maffucci's syndrome is multiple enchondromas with soft tissue hemangiomas. Features that favor the diagnosis of a low-grade chondrosarcoma: Discriminators :  EG is a non-neoplastic proliferation of histiocytes and is also known as Langerhans cell histiocytosis. It should be included in the differential diagnosis of any sclerotic or osteolytic lesion, either well-defined or ill-defined, in patients under the age of 30. The diagnosis EG can be excluded in age > 30. EG is usually monostotic, but can be polyostotic.  Discriminator: Giant cell tumor is a lesion with multinucleated giant cells. In most cases it is a benign lesion. Malignant GCT is rare and differentiation between benign or malignant GCT is not possible based on the radiographs. GCT is also included in the differential diagnosis of an ill-defined osteolytic lesion, provided the age and the site of the lesion are compatible. Discriminators: NOF is a benign well-defined, solitary lesion due to proliferation of fibrous tissue. It is the most common bone lesion. NOF is frequently a coincidental finding with or without a fracture. NOF usually has a sclerotic border and can be expansile. They regress spontaneously with gradual fill in. NOF may occur as a multifocal lesion. The radiographic appearance is almost always typical, and as such additional imaging and biopsy is not warranted. Discriminators: Osteoblastoma is a rare solitary, benign tumor that produces osteoid and bone. Consider osteoblastoma when ABC is in the differential diagnosis of a spine lesion (figure). A typical osteoblastoma is larger than 2 cm, otherwise it completely resembles osteoid osteoma. Discriminator: Metastases are the most common malignant bone tumors. Metastases must be included in the differential diagnosis of any bone lesion, whether well-defined or ill-defined osteolytic or sclerotic in age > 40. Bone metastases have a predilection for hematopoietic marrow sites: spine, pelvis, ribs, cranium and proximal long bones: femur, humerus. Metastases can be included in the differential diagnosis if a younger patient is known to have a malignancy, like neuroblastoma, rhabdomyosarcoma, retinoblastoma. Most common osteolytic metastases: kidney, lung, colon and melanoma. Most common osteosclerotic metastases: prostate and breast. Discriminator: Multiple myeloma must be included in the differential diagnosis of any lytic bone lesion, whether well-defined or ill-defined in age > 40. The most common location is in the axial skeleton (spine, skull, pelvis and ribs) and in the diaphysis of long bones (femur and humerus). Most common presentation: multiple lytic 'punched out' lesions. Multiple myeloma doe not show any uptake on bone scan. Discriminator: Multiple Myeloma (2) Differential diagnosis: On the left a CT-image of a patient with multiple myeloma. Notice the numerous osteolytic lesions and permeative cortical destruction pattern. In the left sacral wing there is a larger lesion with a high density due to replacement of fatty bone marrow by myeloma (red arrow). ABC is a solitary expansile well-defined osteolytic bone lesion, that is filled with blood. It is named aneurysmal because it is expansile. ABC is thought to be the result of a reactive process secondary to trauma or increased venous pressure. Sometimes an underlying lesion like GCT, osteoblastoma or chondroblastoma can be found. ABC can occur almost anywhere in the skeleton. Discriminators: More on ABC in the article 'Bone Tumor A-G'. Solitary bone cyst, also known as unicameral bone cyst, is a true cyst. Many well-defined osteolytic lesions are often called cystic, but this is a misnomer. SBC frequently presents with a fracture. Sometimes a fallen fragment is appreciated. Predilection sites: proximal humerus and femur. Usually less expansion compared with ABC. Differential diagnosis: ABC, FD when cystic. SBC may migrate from metaphysis to diaphysis during growth of the bone. Discriminators: Brown tumors can occur in any bone and present as osteolytic lesions with sharp margins. Septa and ridges may be seen. Differential diagnosis: ABC, metastases and GCT depending on location and age. On the left a patient who had a nefrectomy for renal cell carcinoma and who was on dialysis. Multiple well-defined osteolytic lesions were found on a follow up CT scan. The differential diagnosis included metastases and Brown tumors in hyperparathyroidism. Biopsy revealed Brown tumor. Discriminators: Infection or osteomyelitis is the great mimicker of bone tumors. It has a broad spectrum of radiographic features and occurs at any age and has no typical location. In the chronic stage it can mimic a benign bone tumor (Brodies abscess). In the acute stage it can mimic a malignant bone tumor with ill-defined margins, cortical destruction and an aggressive type of periostitis. Only when there is a thick solid periosteal reaction we can recognize the non-malignant underlying process. Discriminators:  The patella, carpal and tarsal bones can be regarded as epiphysis conceirning the differential diagnosis. On the left a chondroblastoma located in the patella. Discriminators : Chondromyxoid Fibroma is a rare lesion. CMF resembles NOF. Preferential sites: proximal tibia and foot. Although the name suggests that CMF is a chondroid lesion, calcifications are usually not seen. On the left images of a CMF. There is an eccentric osteolytic lesion in the metaphysis of the proximal tibia. On the inner side there is a sclerotic margin. On the outer side there is a regular cortical destruction with peripheral bone layer. The MR also shows a sclerotic margin with low signal intensity. Discriminators : On the left a summary of things to look for in well-defined osteolytic lesions. Fundamentals of Skeletal Radiology, second edition by Clyde A. Helms W. B. Saunders company 1995Henk Jan van der Woude and Robin Smithuis Fibrous dysplasia Enchondroma Eosinophilic granuloma Giant cell tumor NOF Osteoblastoma Metastases Multiple Myeloma Aneurysmal Bone Cyst Solitary Bone Cyst Hyperparathyroidism Infection Chondroblastoma Chondromyxoid FibromaBone tumor - Well-defined osteolytic tumors and tumor-like lesionsRadiology department of the Onze Lieve Vrouwe Gasthuis, Amsterdam and the Rijnland hospital, Leiderdorp, the Netherlands
msk6
1
Bone tumor - ill defined osteolytic tumors and tumor-like lesions
by Henk Jan van der Woude and Robin Smithuis
In the article Bone Tumors - Differential diagnosis we discussed a systematic approach to the differential diagnosis of bone tumors and tumor-like lesions. In this article we will discuss the differential diagnosis of ill-defined osteolytic bone tumors in alphabetic order. You can click on an item on the left. On the left the most common ill-defined bone tumors and tumor-like lesions. An ill-defined zone of transition is seen in: On the left a table with the most common bone tumors and tumor-like lesions in different age-groups. In the middle column common ill-defined osteolytic lesions. Notice the following: Key facts On the left a partially ill-defined osteolytic lesion with endosteal scalloping. There are cloud-like calcifications indicating a chondroid matrix. These imaging findings and the size of the lesion favor the diagnosis of a chondrosarcoma. On the left two other lesions that proved to be a chondrosarcoma. Notice that calcifications are not essential in chondrosarcoma. More on chondrosarcoma in the article 'Bone Tumor A-G'. On the left a lobulated partially ill-defined lytic lesion of the proximal humerus. The presence of calcifications suggest that this is a chondroid tumor. The lytic parts with cortical involvement and expansion should raise the suspicion of a high grade chondrosarcoma. Key facts On the left some examples of EG with ill-defined borders. There are single or multiple layered periosteal reactions. On the left a typical presentation of EG in the skull as an ill-defined osteolytic lesion. More on Eosinophilic granuloma in the article 'Bone Tumor A-G'. Key facts On the left a patient with a Ewing's sarcoma in the femur. Notice the ill-defined osteolysis. There is an aggressive periosteal reaction. On the left an ill-defined lytic lesion in the femur of a young patient. There is a permeative destruction pattern with irregular cortical destruction. There is an aggressive periosteal reaction (arrow). This is also called sunburst appearance. On the left an ill-defined lytic lesion of the right iliac bone in a young patient which can easily be overlooked. Final diagnosis: Ewing's sarcoma. More on Ewing's sarcoma in the article 'Bone Tumor A-G'. Key facts On the left a giant cell tumor of the distal radius with ill-defined margins, destruction of the subchondral bone plate and extension towards the soft tissues. On the right a giant cell tumor in the proximal tibia with somewhat better defined margin and non-interrupted cortical bone. More on Giant cell tumor in the article 'Bone Tumor A-G'. Key facts The plain radiograph on the left shows an ill-defined lytic lesion of the humerus diaphysis. Notice tunneling of the cortical bone (red arrows). On the MR notice the linear abnormalities within the cortical bone and the circumferential soft tissue mass. Differential diagnosis (depending on age): Ewing's sarcoma, osteomyelitis and bone lymphoma. Biopsy revealed Non-Hodgkin lymphoma Key facts On the left a 60 year old patient with a known malignancy. There is a lesion in the distal femur, that you could easily overlook and think that it was focal osteopenia. The lesion presents as a large ill-defined osteolytic mass extending into the epiphysis and almost abuttting the articular surface. In a patient 20-40 years of age GCT would be a possible diagnosis. It proved to be a metastasis. Key facts Key facts On the left a subacute form of osteomyelitis. There is an eccentric ill-defined lesion seen on both sides of the physeal plate in the proximal tibia. This is highly suggestive for osteomyelitis. Other lesions do not cross the growth plate in general. On the right coronal T1-weighted MR image reveals a well-defined epi-metaphyseal lesion. There is a dark peripheral zone of reactive sclerosis, and extensive edema with low signal Intensity in the metaphysis. On the left an ill-defined osteolytic lesion in the proximal metaphysis of the tibia with extensive reactive sclerosis and solid periosteal reaction. Key facts of Osteosarcoma On the left a mixed osteolytic and sclerotic lesion in the proximal humerus with irregular cortical destruction. There is an aggressive periosteal reaction and a soft tissue mass.Henk Jan van der Woude and Robin SmithuisBone tumor - ill defined osteolytic tumors and tumor-like lesionsRadiology department of the Onze Lieve Vrouwe Gasthuis, Amsterdam and the Rijnland hospital, Leiderdorp, the Netherlands
msk7
1
Bone tumor A-G
by Henk Jan van de Woude and Robin Smithuis 
In the article Bone Tumors - Differential diagnosis we discussed a systematic approach to the differential diagnosis of bone tumors and tumor-like lesions. . In this article, which is the first in a series of three, we will discuss the most common bone tumors and tumor-like lesions in alphabethic order. key facts: In the spine osteoblastoma may mimic ABC. In the proximal humerus or femur of young children there is frequently a differential diagnosis of ABC, SBC and fibrous dysplasia. Cavities filled with blood can also be found in giant cell tumor, osteoblastoma and chondroblastoma (i.e with secondary ABC). ABC (2) On the left a typical location for an ABC in the posterior elements of the spine. Notice the well-defined osteolytic presentation with multiple fluid-fluid levels on MR with the patient in supine position. The differential diagnosis based on the CT is: ABC, Osteoblastoma and Tuberculosis (1). ABC (3) On the left images of an aneurysmal or expansile well-defined osteolytic bone lesion in the fibula. The T2-weighted MR-image shows the fluid content and on the T1-weighted image there is a subtle fluid-fluid level. Differential diagnosis: SBC or fibrous dysplasia with cystic changes ABC (4) On the left images of an aneurysmal or expansile well-defined osteolytic bone lesion in the proximal phalanx. Notice the expansion and the enhancing rim. ABC (5) On the left an expansile well-defined osteolytic lesion with a sclerotic margin in the talus. Axial PD-weighted image shows lobulated contours and cystic appearance with fluid-fluid level (arrow). Most likely diagnosis: ABC. Differential diagnosis: chondroblastoma with secondary ABC. ABC (6) - atypical case On the left two different patients with an intracortical or subperosteal osteolytic well-defined lesion in the tibia. The lesion on the far left was thought to be an adamantinoma because of the localisation in the anterior tibial cortex. At biopsy it proved to be an ABC. The image on the right is an adamantinoma. Continue with the additional examinations in this ABC. On the left the MR, bone scan and the sonogram of the same patient. On the axial T2 WI with fat saturation subtle sedimentation is seen. The bone marrow is completely normal. ABC (7) On the left another ABC in the talus. It is barely visible on the X-rays and was initially missed. MR shows typical flui-fluid levels. ABC (8) On the left another ABC, located in the distal femur. The plain radiograph shows a layered periosteal reaction and Codman triangle in direct relationship to an expansile lytic lesion with a thin peripheral bone shell. CT also reveals the subperiosteal origin of the lesion with secondary involvement of the cortical bone. Axial T2-weighted image with fatsat and contrast enhanced T1-weighted image with fat sat show multiple fluid-fluid levels with rim enhancement of the cavities filled with blood. This is typical for an aneurysmal bone cyst. key facts: Adamantinoma (2) Young patient with a lobulated lytic lesion within the anterior cortical bone of the proximal tibia. There is a second lucency separately more proximal within the cortical bone. Axail CT image prior to biopsy demonstrates the lytic appearance of the lesion within the thickened cortical bone. In the differential diagnosis could have been chondromyxoid fibroma or fibro-osseous lesion, however, the separate cortical lesion strongly suggests adamantinoma, which is almost exclusively found in the tibia and often multicentric. Adamantinoma (3) Adamantinoma of the tibia in another patient. There is an ovoid osteolytic lesion within the anterior cortical bone. Lobulated high signal intensity on axial T2-weighted image. There is no extension into the bone marrow. Homogeneous enhancement on T1-weighted image after Gd-DTPA. Adamantinoma (4) On the left a radiograph and CT-image of another typical adamantinoma. CT was performed prior to biopsy. A bone island consists of well-differentiated mature bone tissue within the marrow, also referred to as enostosis. Usually it is seen as a coincidental finding. In patients with breast- or prostate cancer a bone island can be mistaken for an osteoblastic metastasis. A bone island normally does not show increased uptake on a bone scan. Bone island (2) On the left a well-defined compact sclerotic lesion in the proximal humerus. Most likely diagnosis: bone island or enostosis. In patient with a known malignancy consider: osteoblastic metastasis. key facts: On the left a patient who had a nefrectomy for renal cell carcinoma and who was on dialysis. Multiple well-defined osteolytic lesions were found on a follow up CT scan. The differential diagnosis included metastases and Brown tumors in hyperparathyroidism. Biopsy revealed Brown tumor. Brown tumor (2) On the left images of a 30-year-old male with well-defined lytic lesion of the olecranon. On the radiograph several ridges can be seen and a pathologic fracture (arrow). The T2-weighted image with fat saturation demonstrates fluid-levels due to sedimentation. Most likely diagnosis: giant cell tumor and ABC. Biopsy revealed brown tumor. key facts: The lesion can be lobulated, usually with a sclerotic margin. Frequently a regular benign periosteal reaction is present. MR imaging usually shows prominent bone marrow and soft tissue edema. The differential diagnosis of an epiphyseal lesion in young patients, besides chondroblastoma, includes osteomyelitis and ganglion cyst. In the foot the differential diagnosis is longer. More on bone tumors in the foot in the article 'Bone tumor - Systematic approach and Differential diagnosis'. Chondroblastoma (2) Well-defined lytic lesion located posteriorly in the proximal epiphysis of the tibia. There is some reactive sclerosis surrounding the lesion. There is no matrix formation. On the coronal T2-weighted image with fat suppression the lesion has a high SI and subtle internal ridges. There is edema of the entire epiphysis. On a sagittal T1-weighted image there is a discrete sclerotic margin. If there were signs of osteoarthrosis, the differential diagnosis would be a geode. Chondroblastoma (3) The talar or calcaneal bone is one of the predilectional sites of chondroblastoma. Note the thin sclerotic margin and surrounding edema. Other diagnostic possibilities in the young population: SBC (no edema), ABC (fluid levels), or osteoblastoma. Chondroblastoma (4) On the left a lesion that is located in the epi- and metaphysis of the proximal humerus. The lesion is predominantly calcified. Coronal T1-weighted image shows lobulated margins and peripheral low SI due to the calcifications. Notice the surrounding decreased signal intensity of the bone marrow, consistent with edema. Edema almost always accompanies chondroblastoma, but is unusual in other chondroid tumors, like enchondroma or low-grade chondrosarcoma. Chondroblastoma (5) On the left an eccentric well-defined lytic lesion in the proximal femur. CT image shows a lobulated and sclerotic border. The location in the epiphysis is typical for a chondroblastoma. Chondroblastoma (6) Edema is almost always present in chondroblastoma. On the left some examples: Chondroblastoma (7) The images show a chondroblastoma in the patella. Notice the extensive edema (blue arrow) key facts: Chondromyxoid fibroma (2) Eccentric well-defined lytic lesion in the metaphysis of the proximal tibia in a young child. Contours are somewhat lobulated, narrow transitional zone. Differential diagnosis, because of the eccentric manifestation: non-ossifying fibroma or chondromyxoid fibroma. Axial T2-weighted image nicely shows the low SI sclerotic margin, high SI of the intrinsic part of the lesion, bone marrow edema and some soft tissue reaction. There is thick peripheral enhancement on sagittal T1-weighted after Gd-DTPA. The MR features support the diagnosis of CMF Chondromyxoid fibroma (3) On the left a well-defined lytic lesion at the base of the 2nd metatarsal bone. There is no mineralisation. T1-weighted images before and after Gd-DTPA demonstrate some expansion and lobulation. On the T1-weighted image before contrast there is a nonspecific intermediate signal intensity. After the administration of Gadolinium there is thick peripheral enhancement. Differential diagnosis based on plain radiograph: giant cell tumor or chondroid lesion, i.e. enchondroma, low grade chondrosarcoma or CMF. Biopsy revealed CMF. Although this is a rare lesion, the foot is one of the preferential sites of origin. Chondromyxoid fibroma (4) On the left a diaphyseal cortically based lytic lesion with expansion and a thin peripheral bone shell. There is peripheral enhancement on the axial T1-weighted image with fat saturation. The lobulated morphology with high SI on the T2-weighted image with fat saturation suggests that this is a cartilaginous lesion. Fianl diagnosis: chondromyxoid fibroma key facts: On plain radiographs the differential diagnosis with enchondroma can be difficult. Think of chondrosarcoma instead of enchondroma if there is one or more of the following features: Chondrosarcoma (2) The differential diagnosis on the plain radiographs in all these three cases is enchondroma. Additional MR imaging and bone scintigraphy may be helpful to make the diagnosis of chondrosarcoma more or less likely. Chondrosarcoma (3) On the left a massive chondrosarcoma of the skull base with extension to the nasal and paranasal cavities and orbita. CT demonstrates irregular calcifications produced by the tumor. A T1-weighted image after Gd shows typical septal and nodular enhancement. Chondrosarcoma (4) On the left a chondrosarcoma in the proximal tibia diaphysis. The tumor is recognized by subtle calcifications in the proximal part. The distal border is not well defined. Notice endosteal scalloping at the medial side which is a hallmark of chondrosarcoma. MR better defines the extension of the lesion. MRI also demonstrates the endosteal scalloping. Chondrosarcoma (5) Partially calcified chondroid tumor in the proximal tibia. Based on the imaging findings it is not possible to differentiate between an enchondroma or a low grade chonrosarcoma. On the coronal T1-weighted MR the typical chondroid morphology is seen. There is no evidence of endosteal scalloping. Continue with bone scan and dynamic MR. Bone scintigraphy of the same patient shows increased uptake in the lesion. This increased uptake in a chondroid tumor is in favor of the diagnosis of a low grade (grade I) chondrosarcoma. The fast dynamic contrast enhanced MR image with subtraction revealed early and progressive enhancement, which is also in favor of the diagnosis of a grade I chondrosarcoma. Chondrosarcoma (6) On the left a patient with a lesion in the proximal humerus with typical popcorn calcification. The most likely diagnosis is enchondroma, however progression to a low-grade chondrosarcoma cannot be excluded based on the plain radiograph alone. There is increased activity on the nuclear bone scan, which is more in favor of the diagnosis of a chondrosarcoma. This proved to be a low-grade chondrosarcoma. Chondrosarcoma (7) On the left a patient with a calcified lesion in the proximal diaphysis of the tibia. On the sagittal T1-weighted CE image. The calcifications with low SI are present in the proximal part. On the MR, the lesion is far more extensive than suspected on the plain radiograph at first glance. On second inspection you will notice the subtle endoteal scalloping of the tumor on the radiogrph. The large diameter and the scalloping favor the diagnosis of a chondrosarcoma. Chondrosarcoma (8) On the left a patient with a chondrosarcoma of the right acetabulum. On the CT expansion and subtle calcifications are present. On the coronal T2-weighted image the tumor is seen as a large lobulated mass with very high SI, which is typical for chondroid tumors. The T1-weighted image after Gd shows typical peripheral nodular enhancement. In the center there is no enhancement. This is probably due to a large myxoid component. . Peripheral chondrosarcomas can arise from osteochondromas. Measuring the cartilage cap will help in distinguishing benign osteochondromas from chondrosarcomas. With 2 cm used as a cutoff for distinguishing benign osteochondromas from chondrosarcomas, the sensitivities and specificities were 100% and 98% for MR imaging and 100% and 95% for CT, respectively (3). Chondrosarcoma (9) On the left a patient with a calcified mass arising from the proximal fibula. The size of the lesion and the high uptake on the bone scan suggest that this is a chondrosarcoma. Chondrosarcoma (9) continued On the left the axial T2 WI. Notice that the tumor arises from an osteochondroma, which is shown in the center part of the image (arrow). Chondrosarcoma (10) On the left a patient with a broad-based osteochondroma with extension of the cortical bone into the stalk of the lesion. Notice the lytic peripheral part with subtle calcifications. This part corresponds to a zone of high SI on T2-WI with FS on the right. This represents a thick cartilage cap. This is an example of progression of an osteochondroma to a peripheral chondrosarcoma. On the left a typical broad-based osteochondroma arising from the proximal humerus. The major part of the bony protrusion consists of fatty bone marrow. There is a thin peripheral cartilage cap, i.e. no suspicion for malignant degeneration. On the left a lesion consisting of rings-and-arcs calcifications in the proximal humerus. The differential diagnosis is enchondroma or low grade chondrosarcoma. The CT shows the calcifications with subtle endosteal thinning of the cortical bone (arrows). Final diagnosis: low grade chondrosarcoma. On the left a lytic ill-defined lesion in the distal diaphysis of the femur in an old patient. Notice the cortical thickening, but also endosteal scalloping (blue arrow). The differential diagnosis should include metastasis and myeloma. At closer look, there are also faint calcifications present (arrowhead). The T1WI+Gd with fatsat demonstrates the presence of a solid enhancing part and perilesional edema. In combination with the ill-defined appearance on the plain radiograph this suggest that this is a high grade chondrosarcoma. key facts: On the left lytic lesions of the C2 and C3 vertebrae with cortical destruction posteriorly. The differential diagnosis based on the CT-findings includes primarily metastases and myeloma. The sagittal T2-weighted image with fat saturation demonstrates continuity between the abnormalities with soft tissue extension and compression of the myelum. Now our differential should also include chordoma which has its origin in the neural axis. Although it most commonly is seen in the lumbar-sacral area, it may occur in the other parts of the spine. Chordoma (2) Chordoma does not metastasize, but local recurrence is commmon. Chordomas in the sacrococcygeal region may be cured by radical en bloc excision. Chordomas in the base of the skull are usually inaccessible to surgery but may respond to radiation therapy. On the left a large tumor in the sacral region which proved to be a chordoma. key facts: Degenerative cyst (2) On the left a patient with arthrosis of the knee and a large well-defined osteolytic lesion in the epiphysis of the tibia. In young patients the differential diagnosis would include chondroblastoma, intraosseous ganglion and giant cell tumor. In this elderly patient with arthrosis this lesion is most probably a degenerative cyst. Degenerative cyst (3) On the left a well-defned lytic lesion with sclerotic margins in a patient with features of arthrosis. Most likely diagnosis: degenerative cyst. No change during follow up. On the left a well-defined lucent lesion in the epiphysis of the proximal tibia in young patient. On the sagittal T2-WI with FS, the lesion has high SI, but there is no extensive edema, which makes the diagnosis chondroblastoma less likely. In an older patient with arthrosis the likely diagnosis would be a degenerative cyst. Degenerative cyst versus intraosseus ganglion (2) On the left another intraosseus ganglion. The radiograph shows a well-defined osteolytic lesion in the proximal tibia epiphysis. It is well-defined with high SI on T2 WI with FS (right), without reactive edema. key facts: Left: Well-defined lytic lesion with sclerotic margin and some expansion. There are fine calcifications. This patient presented with a fracture, which is a common first presentation of an enchondroma. Middle: Well-defined lytic lesion without a sclerotic border also with a fracture. Right: Lesion with irregular cortical bone destruction. Low-grade chondrosarcoma has to be included in the differential diagnosis, but is uncommon at this specific location. Enchondroma (2) Lytic lesion within the phalanx with irregular cortical bone destruction and soft tissue extension should raise the suspicion of a chondrosarcoma. A lytic lesion within a metacarpal bone with expansion with or without calcifications, with or without cortical bone destruction is not infrequently due to chondrosarcoma. Enchondroma (3) On the left another enchondroma. here is a well-defined eccentric osteolytic lesion. The location is typical for enchondroma. Notice the lack of calcifications. Enchondroma (4) On the left a well-defined lytic lesion with some expansion of the rib. The differential diagnosis based on the radiograph is: fibrous dysplasia, enchondroma, and less likely eosinophilic granuloma or hemangioma. The coronal T1-WI after Gd with fatsat shows a lobulated lesion with peripheral enhancement consistent with the diagnosis of an enchondroma. Multiple enchondromatosis is known as Ollier's disease. Multiple enchondromas and hemangiomas of soft tissue are known as Maffucci's syndrome. In both conditions there is a 30% risk of malignant transformation. Enchondroma (5) On the left a patient with multiple eccentric lytic lesions in the metacarpal bones and phalanges of the left hand. On T2-WI with FS there is homogeneous high SI of the lesions. On the left another patient with multiple well-defined lytic lesions in a central and eccentric localisation in the phalanges. On the right the MR of a different patient with multiple lobulated chondroid lesions with high SI on T2-WI with fatsat. key facts: Eosinophilic granuloma (2) On the left some examples of EG demonstrating the various more or less aggressive presentations as ill-defined osteolytic lesions (blue arrow) and even as a less common sclerotic lesion (red arrow). Eosinophilic granuloma (3) On the far left a well-defined osteolytic lesion. A zone of sclerosis can be seen surrounding the lytic lesion. Sometimes a so called button-sequestrum is found in the central part. The case on the right shows multiple ill-defined lesions in the parietal and frontal bone. Histology revealed eosiniphilic granuloma. In a young child with multiple lytic lesions of the neurocranium EG is the most likely diagnosis. Eosinophilic granuloma (4) On the left an ill-defined lytic lesion in the shaft of the femur with a solid layered periosteal reaction. Differential diagnosis: EG, osteomyelitis and Ewing sarcoma. Ewing sarcoma usually shows an irregular and interrupted periosteal reaction, however cannot be excluded. Axial T2-WI with FS shows high SI of the bone marrow and a soft tissue mass envelopping the bone. Eosinophilic granuloma (5) On the left another case of EG presenting as a well-defined osteolytic lesion in a young child. Both inner and outer table of the skull are affected. Eosinophilic granuloma (6) On the left the sclerotic lesion in the clavicle that was shown above. Notice on the T2-weighted MR-image the edema surrounding the clavicle. Based on these imaging findings differentiation from a malignant bone tumor or infection is not possible. key facts: Ewing sarcoma (2) On the left some examples of a Ewing sarcoma.  Ewing sarcoma (3) On the left images of a large lytic tumor arising from the right iliac bone. On the plain film it is very hard to appreciate the lesion because of the permeative destruction pattern. Scintigraphy shows extensive uptake within the iliac bone. Sometimes, a cold spot is found in Ewing sarcoma. MR reveals the intra- and extraosseous tumor extension. Ewing sarcoma (4) This case was also shown above. There is a Ewing sarcoma with permeative growth through the haversian channels accompanied by a large soft tissue mass as seen on the MR. The radiograph does not show any sign of cortical destruction. In fact it looks normal, although someone might argue that the cortex is somewhat ill-defined and that there maybe is some periosteal reaction. Ewing sarcoma (5) On the left a mixed lytic-sclerotic lesion within the diaphysis of the femur. There is a permeative pattern of destruction with a spiculated periosteal reaction and soft-tissue extension. The final diagnosis was a Ewing sarcoma. Based on these images osteosarcoma should be in the differential. Ewing sarcoma (5) continued The fat suppressed T1-weighted enhanced MR image demonstrates the permeative cortical destruction and enhancing soft tissue mass. Ewing sarcoma (6) On the left a similar case. On the radiograph there is a subtle interruped periosteal reaction of the humeral diaphysis with otherwise normal appearing cortical bone. Axial T2-weighted MR image with FS shows large accompanying soft tissue mass with almost normal appearing cortical bone. This suggests a malignant small-cell tumor like Ewing sarcoma or lymphoma. Final diagnosis: Ewing sarcoma Ewing sarcoma (7) Treatment for Ewing's sarcoma includes surgery, radiation and chemotherapy. On the left a Ewing sarcoma, presenting as a large pleural soft tissue mass (same case as above). On the right, the soft tissue mass has completely resolved after neoadjuvant chemotherapy. key facts: Fibrous dysplasia (2) On the left a polyostotic manifestation of FD. There are multiple lesions, which are partially lytic and partially mixed lytic/sclerotic. The radiographic appearance is determined by the extent of dysplastic bone and the amount of bone produced and the degree of calcifications and ossifications. Fibrous dysplasia (3) On the left images of a patient with fibrous dysplasia of the rib with remarkable expansion. Fibrous dysplasia can cause huge deformaties of bones. CT shows expansion and a peripheral zone of sclerosis. No internal matrix formation. Axial T2-weighted image shows mixed pattern of low and high SI within the expansile lesion. Fibrous dysplasia (3) Fibrous dysplasia can be monostotic or polyostotic. In this patient there are identical lesions within the proximal femur and left acetabulum. There is a groundglass appearance with focal areas of calcifications. Fibrous dysplasia (4) The appearance of FD may vary from entirely lytic (probably due to cystic degeneration) to entirely sclerotic. On the left images of a patient with polyostotic fibrous dysplasia, with lucent lesions in the proximal and mid-diaphyseal femur, and lesion with groundglass density and calcifications in the fibula. Fibrous dysplasia (5) Bone scintigram in 40-year old patient in the tibia shaft. Plain radiograph shows well-defined lesion with ground glass density and sclerotic margin. Features are not characteristic for adamantinoma, histology revealed fibrous dysplasia. Fibrous dysplasia (6) On the left images of a patient with polyostotic FD. On the far left a well-defined lytic lesion with groundglass appearance in the proximal femur diaphysis, consistent with fibrous dysplasia. T2-weighted MR image with FS reveals cytic degeneration of the fibrous dysplasia, which is a common finding. This patient has a second lesion in the shaft of the humerus witha pathologic fracture. Fibrous dysplasia (7) The radiograph on the left shows a mixed lytic-sclerotic lesion of the left iliac bone. Axial CT image on the right shows some broadening of the iliac bone with a ground glass appearance and no cortical destruction. key facts: Giant cell tumor (2) On the left a giant cell tumor presenting as an eccentric lytic lesion in the medial epi- and metaphysis of the distal femur. There is a small transitional zone resulting in well-defined borders. Continue with the MR-images. Giant cell tumor (2) continued MR-images of the same patient. Sagittal T1-weighted TSE images before and after Gd. The tumor extends to the subchondral bone plate with endosteal cortical involvement. There is inhomogeneous enhancement. On T2-weighted image, the tumor has a remarkable low SI, which is commonly seen in GCT. There is surrounding edema with high SI. Giant cell tumor (3) On the left a typical giant cell tumor of the distal radius. Notice the aggressive appearance with ill-defined borders, extension to the soft tissues and destruction of subchondral bone plate. The localisation in the epiphysis and metaphysis is in favor of diagnosis of GCT. On the right a coronal T1-weighted CE image. There is diffuse heterogeneous enhancement, and extension to the radiocarpal joint and surrounding edema in bone and soft tissues. Giant cell tumor (3) continued On the left images of the same patient during follow-up after curettage and cement-placement. There is soft tissue swelling with osteolysis of the bone adjacent to the cement. This indicates recurrent or residual disease. On the left another GCT presenting as an ill-defined expansile lesion with a very subtle peripheral bone shell arising from the spinous processes C5 and C6. The sagittal T2-weighted MR-image with FS demonstrates multiple small cavities within the tumor, due to secondary ABC. Giant cell tumor (4) On the left a large lytic lesion originating from the sacral bone. CT prior to biopsy shows a lesion, that is entirely lytic without minerelization. Differential diagnosis in a patient of 50 years of age: metastasis, plasmacytoma, chordoma, chondrosarcoma (no calcifications present), giant cell tumor. Giant cell tumor (4) continued On the coronal T1WI after Gd , the tumor shows diffuse enhancement of the intra- and extraosseous component, with the exception of multiple cystic cavities, which have a low SI on T1 and high SI on T2. These cysts are not uncommonly encountered in giant cell tumor. On the left more examples of GCT around the knee. Notice that most of these lesions are well-defined and located in the epiphysis and extend into the metaphysis. Some extend onto the articular surface (yellow arrow and small red arrows). The lesion on the upper right has an ill-defined border with a broad zone of transition (blue arrow). by Clyde A. Helms W. B. Saunders company 1995 by Henry DeGroot by Stephanie A. Bernard, MD, Mark D. Murphey, MD et al Radiology 2010, 256Henk Jan van de Woude and Robin Smithuis Chondrosarcoma arising from osteochondroma low grade vs high grade Degenerative cyst versus intraosseus ganglion Multiple enchondromas Polyostotic FDBone tumor A-GRadiology department of the Onze Lieve Vrouwe Gasthuis, Amsterdam and the Rijnland hospital, Leiderdorp, the Netherlands
msk8
1
Diabetic foot - MRI examination
by Ivo Schoots, Mario Maas and Robin Smithuis
Diabetes-related foot problems like osteomyelitis and Charcot neuro-osteoarthropathy are associated with a high morbidity and high healthcare costs. A red hot foot in a patient with diabetic neuropathy is a diagnostic problem. In this overview we will focus on two questions: Osteomyelitis Osteomyelitis in a diabetic with neuropathy is infection of the bone that usually results from contiguous spread of a skin ulcer. Consequently, the most common location for osteomyelitis is not in the midfoot, but at the pressure points of the forefoot (metatarsal heads, IP joints) and in the hindfoot at the plantar aspect of the posterior calcaneus. To determine whether osteomyelitis is present, is to place a marker on the ulcer or sinus tract and track it down to the bone and evaluate the MR- signal intensity of the marrow (1). Active Charcot Unlike osteomyelitis, Charcot neuro-osteoarthropathy is primarily an articular disease, which is most commonly located in the midfoot. In the early stage radiography will not demonstrate bone abnormalities, but MRI will show subchondral bone marrow edema. The subcutaneous soft tissues are not much involved. Signal intensities on MRI will not discriminate between active Charcot Joint or osteomyelitis. Location, i.e. bone or joint and ulcer or not, are the clues to the right diagnosis. Chronic stage of CharcotThe chronic stage of Charcot no longer shows a warm and red foot, but the edema usually persists. Joint deformity, subluxation and dislocation of the metatarsals lead to a rocker-bottom type deformity in which the cuboid becomes a weight-bearing structure. The deformity of the foot with abnormal pressure distribution on the plantar surface with reduced or loss of sensation, makes the foot vulnerable and leads to callus and blister formation and to foot ulceration. Charcot with superimposed osteomyelitis Foot ulceration can subsequently lead to infections, such as cellulitis and osteomyelitis, and this may eventually lead to amputation. The simplest method to determine whether osteomyelitis is present is to follow the path of an ulcer or sinus tract to the bone and evaluate the signal intensity of the bone marrow (1). Osteomyelitis in chronic Charcot is usually located in the midfoot, while osteomyelitis in diabetic neuropathy without Charcot is usually in the forefoot and hindfoot. While diagnosing osteomyelitis is important, it is unfortunately also difficult. Clinical and laboratory signs and symptoms are generally unhelpful. The clinical diagnosis relies on the identification and characterization of an associated foot ulcer, a method that is often unreliable. It is important to mark the skin or subcutaneous abnormality, i.e. ulcer or sinus tract and to find its relation to the area of bone abnormality. The probe-to-bone test, i.e. palpation of bone with a sterile blunt metal probe in the depths of infected pedal ulcers was thought to be highly correlated to ostemyelitis. In later studies however it had a relatively low positive predictive value (7). On plain radiographs bone infection may not show up on the first 2 weeks and in a later stage the radiographic characteristics of neuro-osteoarthropathy and osteomyelitis overlap. In both cases there will be demineralization, destruction and periosteal reaction of the bones, particularly when neuro-osteoarthropathy presents in a later stage. On the left images of a patient with a small cutaneous defect with subcutaneous edema at the metatarsals. As secondary sign an abscess is shown in the forefoot, with high signal intensity on STIR, low or intermediate signal on intensity T1W, and ring-enhancement of the borders showing high signal intensity on T1+Gd. Charcot neuro-osteoarthropathy is a degenerative disease with progressive destruction of the bones and joints. It is seen in patients with neurological disorders with sensory loss of the feet, including tabes dorsalis, leprosy, diabetic neuropathy, and other conditions involving injury to the spinal cord. In 1868 Jean-Martin Charcot gave the first detailed description of the neuropathic aspect of this condition in a patient with syphilis. Today diabetes mellitus is the most common etiology associated with Charcot osteoarthropathy, with the joints of the foot and ankle being most commonly affected. On the left an illustration with the key MR-features of acute Charcot neuro-osteoarthropathy: The exact nature of Charcot arthropathy is unknown. The neurotraumatic theory states that Charcot arthropathy is caused by an unperceived trauma to an insensate foot. The sensory neuropathy renders the patient unaware of the osseous destruction that occurs with continuous ambulation. The neurovascular theory suggests that the underlying condition leads to the development of autonomic neuropathy, causing the extremity to receive an increased blood flow, which in turn results in a mismatch in bone destruction by increased osteoclastic activity and bone synthesis (1). On the left progressive neuro-osteoarthropathy of the tarsometatarsal joints (Lisfranc dislocation) with subchondral cysts, erosions, joint distention and dislocation. Acute active Charcot neuro-osteoarthropathy is defined by clinical signs. There should be neuropathy and a warm and swollen foot. The skin temperature should be 2?C or more at the site of maximum deformity of the affected foot compared with a similar site on the contralateral foot. Osteomyelitis should be excluded and fever is not present. Serum C-reactive protein level is normal or only a slightly elevated. The differential diagnosis is infection (osteomyelitis, cellulitis, septic arthritis), inflammation (gout, rheumatoid arthritis) and deep vein thrombosis. In this early stage radiographic abnormalities are not present. The acute stage of Charcot neuro-osteoarthropathy shows rapid and progressive bone and joint destruction within days or weeks. Immobility by total contact casting can prevent further bone and joint destruction. On the left a radiograph of a patient with diabethic neuropathy and a red hot foot. In the acute stage the radiographs are normal and may not exclude the diagnosis of acute Charcot neuro-osteoarthropathy. Within 4 months there is progressive decrease of calcaneal inclination with equinus deformity at the ankle. There is destruction of the tarsometatarsal joint with the typical rocker-bottom deformity. There is some bony debris on the dorsal side. In the acute stage MRI shows only subchondral bone marrow edema. On the left MRI images of a patient with acute Charcot neuro-osteoarthropathy. The bone marrow edema typically is not restricted to one or two bones, but is seen in the entire midfoot. Bone marrow edema and its enhancement are typically centered in the subchondral bone suggesting articular disease. The subcutaneous tissues are relatively normal and there is no ulcer or other signs of infection. The chronic inactive stage no longer shows a warm and red foot. The edema usually persists. Crepitus, palpable loose bodies and large osteophytes are the result of extensive bone and cartilage destruction. Joint deformity, subluxation and dislocation of the metatarsals lead to a Rocker-bottom type deformity in which the cuboid becomes a weight-bearing structure. This results in excessive skin callus formation, blisters and foot ulceration. At the stage of chronic inactive Charcot bone healing and change of active periosteal reaction will proceed into inactive periosteal reaction and sclerotic borders. The classic radiographic description of neuro-osteoarthropathy is that of the five D' s. Debris may be present and effusions may decompress along fascial planes, carrying bony debris far from the joint. Dislocation is the result of ligamentous laxity. On the far left a normal radiograph in the acute stage of Charcot. Subsequently progressive Charcot neuro-osteoarthropathy is seen with dislocation in the Lisfranc joint. To determine whether osteomyelitis in a Charcot foot at MR imaging is present is to follow the path of an ulcer or sinus tract to the bone and evaluate the signal intensity of the bone marrow. If there is bone marrow edema, osteomyelitis is very likely. If there is bone marrow edema in the absence of a cutaneous defect, active Charcot may be present. If it is normal, both active Charcot as well as osteomyelitis is not likely. On the left a typical rocker-bottom deformity of the foot due to collapse of the longitudinal arch. Abnormal pressure on the cuboid has led to ulceration. In a patient with Charcot neuro-osteoarthropathy and a rocker-bottom foot, the cuboid bone is an important location of osteomyelitis. If the T1-weighted image at that location shows low signal intensity in combination with a cutaneous defect, osteomyelitis is extremely likely. On the left STIR and T1-weighted images of a patient with active Charcot neuro-osteoarthropathy with a plantar ulcer at the bony protuberance of the cuboid. There is abnormal signal intensity in the cuboid bone next to the ulcer indicative of osteomyelitis. On the left the contrast enhanced images with and without fat saturation. Enhancement of the cuboid bone and adjacent soft tissues on postcontrast images, together with the plantar ulcer, makes osteomyelitis very likely. On the left a patient with Charcot neuro-osteoarthropathy with a subcutaneous fistula tract (arrow). This patient has subcutaneous edema and swelling. When we follow the fistula tract to the bony protuberances of the cuboid, there is no marrow edema at the midfoot. This makes yet osteomyelitis unlikely. Ghost sign The ghost sign is indicative of neuro-osteoarthropathy with superimposed osteomyelitis. The 'gost sign' refers to poor definition of the margins of a bone on T1-weighted images, which become clear after contrast administration. On the left a patient with neuro-osteoarthropathy and superimposed osteomyelitis. The areas of osteomyelitis are more pronounced on the contrast-enhanced T1-weighted image as compared to the native T1-weighted image. The bone marrow edema which is of low signal intensity on the T1-weighted image without contrast enhances and becomes as bright as normal bone marrow. The MRI examination includes special attention for positioning of the foot. It must be placed in the centre of the magnet, to obtain homogeneous fat suppression. Markers have to be placed over ulcers or sinus tracts. T1 and STIR or T2 fatsat sequences are needed. Because of the curvature of the foot, fat suppression is more uniform with the use of STIR than with T2- weighted imaging with chemical fat saturation. However, SIR cannot be combined with contrats administration. As an alternative to spectral fat saturation technique, Dixon chemical shift imaging is described (8). Sagittal views are for evaluation of midfoot involvement, the plantar surface and the posterior calcaneus. A view parallel to the toes is adequate for imaging the metatarsophalangeal and interphalangeal joints. Contrast is used to better depict devitalized regions, abscesses, sinus tracts and joint or tendon involvement. by Andrea Donovan, MD and Mark E. Schweitzer, MD May 2010 RadioGraphics, 30, 723-736. by Byron M Perrin et al Australian Family Physician Vol.39 no.3 march2010 by Ivo G. Schoots et al Semin Musculoskelet Radiol 2010;14:365-376. By Robert Bem et al Diabetes Care, Volume 29, number 6, june 2006 Benjamin Lipsky et al Clin Infect Dis. 2004, 39 (7): 885-910 by Lawrence A. Lavery et al Diabetes Care February 2007 vol. 30 no. 2 270-274 by Maas M, Dijkstra PF, Akkerman EM. Radiology. 1999 Jan; 210(1):189-93.Ivo Schoots, Mario Maas and Robin Smithuis Osteomyelitis versus Charcot Acute Charcot Chronic CharcotDiabetic foot - MRI examinationRadiology department of the AMC in Amsterdam and the Rijnland hospital in Leiderdorp, the Netherlands
msk9
1
Elbow - Fractures in Children
by Updated version by Robin Smithuis
Elbow fractures are the most common fractures in children. The assessment of the elbow can be difficult because of the changing anatomy of the growing skeleton and the subtility of some of these fractures. In this review important signs of fractures and dislocations of the elbow will be discussed. Before reading this article you can try one of the cases in the menubar. You can test your knowledge on pediatric elbow fractures with these interactive cases. On some of the images you can click to get a larger view. This does not work for the iPhone application If you want to use images in a presentation, please mention the Radiology Assistant. Injury to the elbow joint is usely the result of hyperextension or extreme valgus due to a fall on the outstretched arm. Scroll through the images on the left to see how hyperextension leads to a supracondylar fracture. The hemarthros will result in a displacement of the anterior fat pad upwards and the posterior fat backwards. The other important fracture mechanism is extreme valgus of the elbow. The normal elbow already has a valgus positioning. When a child falls on the outstrechted arm, this can lead to extreme valgus. On the lateral side this can result in a dislocation or a fracture of the radius with or without involvement of the olecranon. When the forces have more effect on the humerus, the extreme valgus will result in a fracture of the lateral condyle. On the medial side the valgus force can lead to avulsion of the medial epicondyle. Sometimes the medial epicondyl becomes trapped within the joint. Because of the valgus position of the normal elbow an avulsion of the lateral epicondyle will be uncommon. When looking at radiographs of the elbow after trauma a methodical review of the radiographs is needed . You should ask yourself the following important questions. Is there a sign of joint effusion? After trauma this almost always indicates the presence of hemarthros due to a fracture (either visible or occult). Is there a normal alignment between the bones? In children dislocations are frequent and can be very subtle. Are the ossification centres normal? Is the piece of bone that you're looking at a normal ossification centre and is this ossification centre in the normal position. . Look especially for the position of the radial epiphysis and the medial epicondyle (figure). Is there a subtle fracture? Some of the fractures in children are very subtle. So you need to be familiar with the typical picture of these fractures. . Normally on a lateral view of the elbow flexed in 90? a fat pad is seen on the anterior aspect of the joint . This is normal fat located in the joint capsule. On the posterior side no fat pad is seen since the posterior fat is located within the deep intercondylar fossa. Positive fat pad sign Distention of the joint will cause the anterior fat pad to become elevated and the posterior fat pad to become visible. An elevated anterior lucency or a visible posterior lucency on a true lateral radiograph of an elbow flexed at 90? is described as a positive fat pad sign (figure). Hemarthros results in an upward displacement of the anterior fat pad and a backward displacement the posterior fat. Positive fat pad sign (2) Any elbow joint distention either hemorrhagic, inflammatory or traumatic gives rise to a positive fat pad sign. If a positive fat pad sign is not present in a child, significant intra-articular injury is unlikely. A visible fat pad sign without the demonstration of a fracture should be regarded as an occult fracture. These patients are treated as having a nondisplaced fracture with 2 weeks splinting. Skaggs et al repeated x-rays after three weeks in patients with a positive posterior fat pad sign but no visible fracture. They found evidence of fracture in 75%. They concluded that in trauma displacement of the posterior fat pad is virtually pathognomonic of the presence of a fracture. Displacement of the anterior fat pad alone however can occur due to minimal joint effusion and is less specific for fracture. Notice that the elbow is not positioned well. Try to find out what went wrong in the chapter on positioning. There are two important lines which help in the diagnosis of dislocation and fracture . These are the Radiocapitellar line and the Anterior humeral line. Radiocapitellar line A line drawn through the centre of the radial neck should pass throught the centre of the capitellum, whatever the positioning of the patient, since the radius articulates with the capitellum (figure). In dislocation of the radius this line will not pass through the centre of the capitellum. On the left we see, that the radiocapitellar line goes through centre of the capitellum on every radiogragh even though C and D are not well positioned. Notice supracondylar fracture in B. On the left more examples of the radiocapitellar line. The right lower image shows an obvious dislocation of the radius. Anterior humeral line. A line drawn on a lateral view along the anterior surface of the humerus should pass through the middle third of the capitellum.. This line is called the Anterior Humeral line . In cases of a supracondylar fracture the anterior humeral line usually passes through the anterior third of the capitellum or in front of the capitellum due to posterior bending of the distal humeral fragment. On the left the anterior humeral line passes through the anterior third of the capitellum. This indicates that the condyles are displaced dorsally (i.e. supracondylar fracture). First study the images on the left. Then continue reading. The radiocapitellar line ends above the capitellum. This means that the radius is dislocated. Did you also notice the olecranon fracture? Whenever the radius is fractured or dislocated, always study the ulna carefully. There are 6 ossification centres around the elbow joint. They appear and fuse to the adjacent bones at different ages. It is important to know the sequence of appearance since the ossification centers always appear in a strict order. This order of appearance is specified in the mnemonic C-R-I-T-O-E (Capitellum - Radius - Internal or medial epicondyle - Trochlea - Olecranon - External or lateral epicondyle). The ages at which these ossification centres appear are highly variable and differ between individuals. It is not important to know these ages, but as a general guide you could remember 1-3-5-7-9-11 years. The Trochlea has two or more ossification centres which can give the trochlea a fragmented appearance. On a lateral view the trochlea ossifications may project into the joint. They should not be mistaken for loose intra-articular bodies (arrow). Error 1: Shoulder higher than elbow For a true lateral view the shoulder should be at the level of the elbow. If the shoulder is higher than the elbow, the radius and capitellum will project on the ulna. The solution is either to lift the examination table which will lift the elbow or to lower the shoulder by placing the patient on a smaller chair. Error 2: Wrist lower than elbow On the left two examples of a 'low wrist positioning' leading to rotation of the humerus. The low position of the wrist leads to endorotation of the humerus. The lateral structures like the capitellum and the radius will move anteriorly, while a medial structure like the medial epicondyle will move posteriorly. The wrist should be higher than the elbow to compensate for the normal valgus position of the elbow. The hand should be with the 'thumb up'. These fractures account for more than 60% of all elbow fractures in children (see Table). More than 95% of supracondylar fractures are hyperextension type due to a fall on the outstretched hand. The elbow becomes locked in hyperextension. The olecranon is pushed into the olecranon fossa causing the anterior humeral cortex to bend and eventually break. If the force continues both the anterior and posterior cortex will fracture. Supracondylar fractures (2) If there is only minimal or no displacement these fractures can be occult on radiographs. The only sign will be a positive fat pad sign. Usually there is some displacement and the anterior humeral line will not pass through the centre of the capitellum but through the anterior third or even anterior to the capitellum (figure). Supracondylar fractures (3) Supracondylar fractures are classified according to Gartland. Gartland Type I fractures are often difficult to see on X-rays since there is only minimal displacement. Most of these fractures consist of greenstick or torus fractures. The only clue to the diagnosis may be a positive fat pad sign. These patients are treated with casting. In Gartland type II fractures there is displacement but the posterior cortex is intact. There may be some rotation. These fractures require closed reduction and some need percutaneous fixation if a long-arm cast does not adequately hold the reduction. Gartland type III fractures are completely dislocated and are at risk for malunion and neurovascular complications (figure). They require reduction by closed or if necessary open means. Stabilisation is maintained with either two lateral pins or medial lateral cross pin technique. Supracondylar fractures (4) Malunion will result in the classic 'gunstock' deformity due to rotation or inadequate correction of medial collaps. Posterolateral displacement of the distal fragment can be associated with injurie to the neurovascular bundle which is displaced over the medial metaphyseal spike. Nerve injurie almost always results in neuropraxis that resolves in 3-4 months. Vascular injurie usually results in a pulseless but pink hand. Conservative management and vascular intervention have the same outcome. A pulseless and white hand after reduction needs exploration. Supracondylar fractures (5) Flexion-type fractures are uncommon (5% of all supracondylar fractures). They are caused by direct impact on the flexed elbow. Ulnar nerve injury is more common. Compared to extension types, they are more likely to be unstable, so more likely to require fixation. This fracture is the second most common distal humerus fracture in children. They occur between the ages of 4 and 10 years. These fractures occur when a varus force is applied to the extended elbow. They tend to be unstable and become displaced because of the pull of the forearm extensors. Since these fractures are intra-articular they are prone to nonunion because the fracture is bathed in synovial fluid. Lateral condyle fractures are classified according to Milch. They are Salter-Harris IV epiphysiolysis fractures. Most are Milch II fractures that travel from the lateral humeral metaphysis above the epiphysis and exit through the lateral crista of the trochlea leading to an unstable humeral ulnar articulation. Lateral Condyle fractures (2) The problem with the Milch-classification is the fact that the fracture fragments are primarily cartilaginous. The fracture line through the cartilage is not visible on radiographs, so the radiographic interpretation concerning classification is difficult. Treatment strategies are therefore based on the amount of displacement (see Table). Undisplaced fractures are treated with a long arm cast. These fractures must be carefully monitored as they have a tendency to displace. At follow up both AP and Oblique views are taken after removal of the cast. Once displaced fractures consolidate in a malunited position, treatment is difficult and fraught with complications. For this reason surgical reductions is recommended within the first 48 hours. Open reduction is indicated for all displaced fractures and those demonstrating joint instability. Lateral Condyle fractures (3) . The diagnosis of a lateral condyle fracture can be challenging. Fracture lines are sometimes barely visible (figure). Remembering the fact that the lateral condyle fracture is the second most common elbow-fracture in children and because you know where to look for will help you Lateral Condyle fractures (4) . Since most of the structures involved are cartilageneous, it is very difficult to know the exact extent of the fracture. Sometimes the fracture runs through the ossified part of the capitellum. In those cases it is easy. The case on the left shows a lateral condyle fracture extending through the ossified part of the capitellum. This is a Milch I fracture. The elbow is stable. There is too much displacement so osteosynthesis has to be performed. MRI can be helpfull in depicting the full extent of the cartilaginous component of the fracture. The case on the left shows a fracture extending into the unossified trochlear ridge. The fracture through the trochlear cartilage is so far medial that the ulna is only supported on the medial side. This means that the elbowjoint is unstable. Lateral Condyle fractures (5) In lateral condyle fractures the actual fracture line can be very subtle since the metaphyseal flake of bone may be minor. The fracture fragment is often rotated. An oblique view can be helpfull, but usually these are not routinely performed (figure). Lateral Condyle fractures (6) . Overprojection of the capitellum on the humeral metaphysis may simulate a lateral condyle fracture (figure). Lateral Condyle fractures (7) . On the left a couple of examples of lateral condyle fractures. The medial epicondyle is an apophysis since it does not contribute to the longitudinal growth of the humerus. It is located on the dorsal side of the elbow. On a lateral view especially if the arm is endorotated it can project so far posteriorly that one could suggest an avulsion (figure). However avulsions are located more distally and anteriorly. Since the medial epicondyle is an extra-articular structure a fracture or avulsion will not automatically produce a positive fat pad sign. Medial Epicondyle avulsion (2). 80% of avulsion fractures occur in boys with a peak age in early adolescence. The mechanism is an acute valgus stress due to a fall on the outstretched hand or sometimes due to armwrestling. Chronic injuries do occur in young athletes (little league elbow). The mechanism that causes these stressfractures on the medial side is the same mechanism that causes a osteochondritis of the capitellum due to impaction on the lateral side. Medial Epicondyle avulsion (3). There is a 50% incidence of associated elbow dislocations. When the elbow is dislocated and the medial epicondyle is avulsed, it may become interposed between the articular surface of the humerus and the olecranon (figure). In every dislocation the first question should be 'where is the medial epicondyle'. On reducing the elbow the fragment may return to it's original position or remain trapped in the joint. This may severely damage the articular surface. So post-reduction films should be studied carefully. Medial Epicondyle avulsion (4). Due to the extreme valgus force the joint may temporarily open. The avulsed fragment may become entrapped in the joint even when there is no dislocation of the elbow. Medial Epicondyle avulsion (5). An avulsed fragment that is located within the joint can give diagnostic problems. On an AP-view this fragment may be overlooked (figure). When the trochlea is not yet ossified the avulsed fragment may simulate a trochlear ossification centre. Medial Epicondyle avulsion (6). Treatment Non-displaced fractures are treated with 1-2 weeks cast or splint. There is disagreement about the amount of displacement of the medial epicondyle that requires operative fixation. There is support for both operative aswell as non-operative management of medial epicondyle fractures with 5-15mm displacement. Medial Epicondyle avulsion (7). If the history or the radiographs suggest that the elbow was or is dislocated, greater soft tissue injurie is likely to be present requiring need for early motion. Medial Epicondyle avulsion (8). Study the images. You can click on the image to enlarge. There are three findings, that you should comment on.  Continue with the MRI. The MR shows the small medial epicondyle with tendon attachement trapped within the joint. The avulsed medial epicondyl was found within the joint and repositioned and fixated with K-wires. In adults fractures usually involve the articular surface of the radial head. In children however it's the radial neck that fractures because the metaphyseal bone is weak due to constant remodelling. Usually it is a Salter Harris II fracture. If there is no displacement it can be difficult to make the diagnosis (figure). If there is less than 30? tilt of the radial head patients are treated with a collar. It is important to realize that there is normally some angulation of the radial head ( up to 15?). If there is more than 30? tilt closed reduction is performed. Whenever closed reduction is unsuccesfull in restoring tilt or when it is not possible to pronate and supinate up to 60?, a K-wire is inserted to maintain reduction. Radial neck fractures aswell as radial head dislocations are in 50% of the cases associated with other elbow injuries. The most common is a fracture of the olecranon. When the radial epiphysis is yet very small a slipped radial epiphysis may be overlooked (figure). If these fractures are not recognized or reduction is unsuccesfull radial head overgrowth can be the result. A short radius may also be the result since the epiphysis of the radius contributes to the length growth of the radius. Dislocations of the radial head can be very obvious. It is however not uncommon that these dislocations are subtle and easily overlooked. In all cases one should look for associated injury. In the original discription of Monteggia there is a radial dislocation in combination with a proximal ulnar shaft fracture. However fractures anywhere along the ulna have been reported. Especially associated fractures of the olecranon are very common (figure). Radius Pulled Elbow (Nursemaid's elbow) In children When the forearm is pulled the radial head moves distally and the ligament slips over the radial head and becomes trapped within the joint. The X-ray is normal. The condition is cured by supination of the forearm. Sometimes this happens during positioning for a true lateral view (which is with the forearm in supination). Olecranon fractures in children are less common than in adults. As discussed above they are associated with radial neck fractures and radial dislocations. Olecranon fractures (2) Do not mistake the apophysis or its separate ossification centres for a fracture. The apophysis has undulating faintly sclerotic margins. The growth plate usually has a different oblique course compared to a fracture-line. Olecranon fractures (3) On the left some examples of fractures of the olecranon. Notice how subtle some of these fractures are. Whenever you study a radiograph of the elbow of a child, always look for: Elbow and forearm injuries in children by T. David Cox, MD, and Andrew Sonin, MD, in Radiology of Skeletal trauma Third edition Editor Lee F. Rogers MD A site developed for Postgraduate Orthopaedic Trainees preparing for the FRCS Examination in the United Kingdom. A site with detailed information on fractures and therapy.Updated version by Robin Smithuis Hyperextension Extreme valgus Methodical review Fat Pad Sign and Joint effusion Alignment Ossification centres Common errors in positioning Supracondylar fractures Lateral Condyle fractures Medial Epicondyle avulsion Proximal fractures of the Radius Dislocations of the Radial head Olecranon fracturesElbow - Fractures in ChildrenRadiology department, Rijnland Hospital Leiderdorp, the Netherlands.
msk10
1
Hip Arthroplasty
by Iain Watt, Susanne Boldrik, Evert van Langelaan and Robin Smithuis
Radiography is the primary imaging method for the evaluation of Total Hip Arthroplasty. This overview focusses on the normal findings and complications of cemented versus non-cemented hip arthroplasties. Interactive cases are presented in the menubar to test your knowledge on hipprostheses.  Modern Total Hip Arthroplasty (THA) systems are modular. This means that the femoral stem, head, acetabular shell and liner are separate pieces. This modularity allows for greater flexibility in customizing prosthesis sizing and fit. The acetabular part is usually a polyethylene liner with or without metal backing. Fixation is with cement, spikes, screws or cementless with porous coating for bone ingrowth. The femoral part is composed of a metal stem (chromium cobalt or titanium) and a femoral head of metal or ceramic. Stem-fixation is also either with cement or cementless with porous coating for bone ingrowth. Most modern non-cemented THA have a femoral stem with only proximal coating, as this results in a better longterm outcome than fully coated (less loosening). Some of the non-cemented THA have femoral stems with additional hydroxyapatite coating which further improve bone ingrowth. This coating is not visible on radiographs. Hybrid total hip replacements are a combination of cement- and cementless fixation. As cemented acetabular components have a tendency to loosen over time, the combination of a cementless acetabular component with a cemented femoral component is sometimes used. Overall there is a tendency to use preferably non-cemented THA, which have better logterm results. On the left we see a hybrid THA with bone-ingrowth acetabular cup and cemented femoral component and next to it a non-cemented bone ingrowth THA. The initial films serve as a baseline study and are used as reference films for comparison with all future studies, since sequential radiography is the most valuable method for detecting complications. The initial postoperative films are obtained to look for possible dislocation or fracture and to see if the prosthesis is good positioned. Dislocation Dislocation can occur as a late complication in prostheses that are not well positioned, but it is most common in the immediate postoperative period (incidence 3%). Periprosthetic fractures Fractures may be seen postoperatively in patients with poor bone stock and long stem revision prostheses or when the anatomy is abnormal as in hip dysplasia. or prior surgery. They are also more common in non-cemented femoral stems, as these have to fit exactly and can cause a fracture during insertion. The incidence of fractures ranges from 0.1 to 1.0 percent for cemented components and 3 to 18 percent for uncemented components. Most intraoperative fractures occur on the femoral side. Cement extrusion When the acetabulum is prepared for placement of the cup a perforation may occur. This defect is filled with bone chips, cement or bone transplant. Cement extrusion is usually asymptomatic. Rare complications include bowel fistulas, encasement of neurovascular structures and bladder wall burn. Acetabular and femoral component positioning should mimic normal anatomy. The distance from center of the femoral head to teardrop (or other identifiable landmark) should be equal bilaterally. This is called the horizontal center of rotation. Excessive lateral positioning of the acetabular component increases the risk for dislocation and may cause limping. The transischial line is used as a reference to measure the lateral inclination of the acetabular cup (30-50?). This line is also used to measure any leg length discrepancy. Leg length discrepancy up to 1 cm is well tolerated. A high positionened cup is better tolerated than a lateral positioned cup. The anteversion of the acetabular cup should be 5-25?. Exact measurement of this angle on a cross-table or true lateral radiograph is not possible , since the apparent degree of angulation on a radiograph is affected by pelvic or thigh rotation (figure). Measurement with CT is more accurate, but you still have to compensate for pelvic angulation. The following conditions predispose to dislocation: - Increased lateral inclination of the acetabular cup. - Decreased or increased anteversion of the cup. - Excessive lateral positioning of the acetabular cup - Increased or decreased anteversion of the femoral stem. Due to increased forces on the superolateral margin of the cup, increased lateral inclination of the acetabular component also may increase the risk of polyethylene wear of the acetabular liner (see figure). The preferred position of the femoral component is with the stem centered in the femoral canal. The center of rotation of the femoral head should be at the level of the tip of the greater trochanter. Varus position of the femoral stem predisposes to loosening and fracture. Normal findings in cemented-THA are different from non-cemented prostheses as the native bone shows more reactive changes to non-cemented prostheses. In cemented THA ideally you would not expect any lucencies at the bone-cement or cement-prosthesis interface, but even in stable cemented prostheses they do occur. A lucency at the metal-cement interface along the proximal lateral aspect of the femoral stem may be seen on the initial postoperative radiograph as a reflection of suboptimal metal-cement contact at the time of surgery. A stable lucent zone is good, but if the lucency enlarges or develops at the metal-cement interface during follow up, then it is a sign of loosening (figure). Ideally there is only a 3-4mm layer of cement around the prosthesis. Abundant cement packing leads to loosening. At the bone-cement interface a thin fibrous layer may form as responce to local necrosis of osseous tissue due to the heat of the cement-polymerization. It becomes stable by 2 years. On radiographs this layer is seen as a lucent zone that should be Especially in acetabular zone I a 1-2 mm lucency is frequently seen at the bone-cement interface, this is a normal finding provided it is stable. If also other zones are involved and the lucency widens, it is however a sign of loosening. In your report always indicate which zones are involved (figure). In the acetabulum you have three zones marked I-III. It is quite common to see a radio lucent line in zone I, but you shouldn't see it in zone II and III. Similarly in the femur there are zones 1 - 7. It is very common to see radiolucency in zone 1, occasionally in zone 7, but it should not occur in the subtrochanteric region zones 2-6. The implantation of a bone ingrowth prosthesis results in altered stress distribution to the native bone, especially in the older models with non tapered and fully coated femoral stems. Stress shielding proximally may result in proximal osteoporosis and calcar resorption. Stress loading distally may result in cortical thickening and bridging sclerosis at the tip of the prosthesis ( called pedestal). In an effort to avoid these changes, most modern cementless prosthesis only have fixation proximally, so you usually will not find proximal stress shielding. The distal part of the femoral prosthesis is not 'loaded', so there will be no distal stress loading. In stable non-cemented hip arthroplasties lucent zones at the metal-bone interface do occur, as it usually is a combination of bone ingrowth and fibrous tissue ingrowth, that provides the fixation in most cases. This fibrous tissue presents as a lucent zone at the interface. Again it should be stable and well within a range of 1 -2 mm. The figure on the left sums all the findings in some of the non-cemented prostheses, that can be normal. You have to be familiar with the normal and abnormal changes in the types of prostheses, that are used by your orthopaedic surgeons. Thin lucent zones along the bone-metal interface due to fibrous tissue are therefore common (80%). They should be less than 2mm and accompanied by a sclerotic line parallel to it. If they stay stable for 2 years than fixation by a strong fibrous tissue has taken place. Stress shielding or bone resorption is seen in areas that are relatively unstressed. The forces are transmitted through the relative stiff femoral stem and is seen as osteoporosis in the proximal femur with thinning of the cortex and bone resorption of the femoral neck. This is seen medially as calcar resorption, as the calcar has lost it's function (figure). It is also called calcar round off. There are many complications in THA. Radiographic follow up and comparison with the oldest films available is the most valuable method of detecting these complications. The most important complications are mechanical loosening, particle disease and infection. These complications however may have similar imaging findings and overlap exists. Mechanical loosening presents as diffuse lucency. Particle disease presents as focal lucency. Evidence of polyethylene wear, which appears as asymmetric positioning of the femoral head within the acetabular cup, often coexists with particle disease. . Infection presents as irregular lucency with periosteal reaction, but may be difficult to differentiate from loosening and particle disease.In typical cases the imaging findings of loosening, particle disease and infection are straight forward (figure). Infection is often low grade and is difficult to detect with any imaging method. In more agressive cases there will be irregular osteolysis, no sclerotic border, cortical bone resorption and a periosteal reaction. Mechanical loosening remains the most common indication for revision. Patients are usually symptomatic, although asymptomatic radiographic changes may be seen. The most common radiographic manifestation of loosening are: - Lucent zone > 2 mm at interface (indicative) - Component migration (diagnostic). A lucent zone of more than 2 mm at the bone-prosthesis interface or at the bone-cement interface is very indicative of loosening. Especially if more zones are involved and if there is progression. A lucent zone Component migration is diagnostic for loosening. It is seen as tilting or cranial migration of the acteabular cup or as subsidence (>10mm) and varus tilting of the femoral stem. The case on the left shows progressive subsidence, which is diagnostic for loosening, with subsequent break of the screws. Loosening (2) As migration can be very subtle, it requires carefull comparison with the initial postoperative films. Do not just compare to the prior examination. The case on the left shows migration of the acetabular cup, which is better appreciated if a reference point is used (see next figure) If we look at the same radiographs and we use the tear drop figure as a landmark, the migration becomes more evident. Migration of the cup in cranial direction has resulted in a fracture in the acetabular wall (blue arrow). Migration of acetabular components is never acceptable. It is seen as upward movement or tilting of the cup (figure) The case on the left is for several reasons not ideal : - High and very lateral positioning of the cup. - Too much lateral inclination. - Abundant cement packing. - Screws are positioned too horizontally (too much stress). - Lucency in zone II and III > 2 mm. Especially lucency in these zones is very indicative of loosening. During follow up upward migration with increased tilting is seen causing the fixation screw to break. Originally this was called cement disease or aggressive granulomatosus. It is a histiocytic response that occurs as a result of macrophage reaction to any of the components, that are shed of the surface of the components of the arthroplasty. Nowadays it is mostly seen in non-cemented hips as a reaction to small polyethylene wear particles. Radiographically these aggressive granulomatous lesions present as focal radiolucencies around the prosthesis. The condition tends to occur between 1 and 5 years after surgery and is associated with smooth endosteal scalloping. The key feature is that it produces no secondary bone response. These characteristics help to distinguish small particle disease from infection, which often has more aggressive features, although the distinction is not always possible. Although particle disease is a result of polyethylene wear, you will not always see evident findings of polyethyleen wear in the acetabular cup, but whenever you see an eccentric position of the femoral head within the cup, look for focal lucencies. Large focal defects may be seen while the prosthesis is still stable. Particle Disease is relentlessly progressive with loosening, fracture and destruction of bone. Sometimes revision of a stable THA is needed because more bone loss would make revision surgery impossible. Particle Disease (2) The small wear-particles of the polyethylene liner are shed into the joint fluid and can be transported around the prosthesis through small channels even in stable hips. They have a tendency to be transported through screw holes (figure). This is why surgeons are more and more reluctant to use screws for the fixation of acetabular cups. Another case on the left. Again there is focal osteolysis around screws after migration of wear particles through the screwholes. Eccentric position of femoral head within acetabular cup as a result of polyethylene wear. Normal loading of the polyethylene cup comes up the femoral shaft, along the femoral neck towards the lumbar spine. So it is normal to see slight thinning in the area of the weight bearing as the plastic moulds itself. This remoulding of the cup is called creep. Abnormal loading leads to pressure more lateral, resulting in polyethylene wear on the supero-lateral side. Radiologic findings in patients with low grade infection may be unremarkable or may mimic loosening or small particle disease. With more aggressive organisms, progression can be rapid, with bone destruction and sinus tract formation, resulting in radiological findings as listed in the table on the left. Uniform criteria for the diagnosis of infection associated with prostheses have not been established. In several studies infection was diagnosed if at least one of the following criteria was present: - Same microorganism in two cultures of synovial fluid. - Purulence of synovial fluid at the implant site - Inflammation on pathological examination of periprosthetic tissue. - Presence of a sinus tract communicating with the prosthesis. On the left the typical radiographic findings of infection with irregular bone destruction and periosteal reaction. In many cases however the infection is really low grade and difficult to establish. Radionuclide bone scans are very sensitive for infection, but not specific as they may show findings similar to those occurring in loosening. Negative findings on a bone scan suggest that no infection exists. The role of dedicated radionuclide techniques for infection such as gallium scanning or indium-labeled WBC or immunoglobulin G is not clear, but they tend to be a bit more specific compared to normal Technetium bone scan. Most researchers advocate fluoroscopic or sonographic guided joint aspiration to assess infection. Several samples should be taken to minimize confusion caused by skin contaminants. Infections up to one year after the insertion of the prosthesis are acquired during implantation. The risk of intraoperative infection is less than 1% due to the use of antimicrobial prophylaxis and laminar airflow surgical environment. Late infections are acquired by hematogenous seeding from respiratory tract, dental and urinary tract infections. Incidence post-operative: - cemented THA: 0.4% - press fit prosthesis: 2.5% - revision hip arthroplasty: 7.2% Usually it does not affect outcome, but may require cerclage cables. Sometimes a control perforation is placed by the surgeon during revision to aid in removal of the previously placed femoral component. Fractures during follow up are a result of loosening, particle disease, infection or severe cases of stress shielding. As discussed above, dislocation or subluxation of the components may occur because of patient factors including poor muscle tone or trauma or because of surgical factors such as a posterior (rather than lateral) surgical approach. Another factor is difficulty in achieving ideal angulation of the acetabular component. This is usually the result of severe degenerative changes or dysplasia. Dislocation can be in posterior, anterior or lateral direction. On the left another case with dislocation as a result of tilting of the cup due to loosening. Component fracture is uncommon. The case on the left is probably secondary to severe polyethylene wear resulting in cup and cement fracture. Component dissociation, as opposed to component fracture, most commonly develops when the plastic liner of the acetabulum slips from its backing. The case on the left shows severe wear and fracture of the polyethylene liner. The metal backing is still intact. The case on the left shows a fracture of the metal head of the femoral component. The classification of heterotopic ossification includes four grades based on an AP radiograph of the pelvis and hip. Grade I = islands of bone within soft tissues. Grade II = bone spurs leaving > 1 cm between opposing bone surfaces. Grade III = bone spurs leaving Grade IV = radiographic ankylosis of the hip. Heterotopic Ossification occurs when primitive mesenchymal cells in the surrounding soft tissues are transformed into osteoblastic cells, that form mature lamellar bone. It typically occurs around the femoral neck and adjacent to the greater trochanter and occurs in 15-50% of patients. Many patients with radiographically low-grade heterotopic ossification are asymptomatic. If it becomes symptomatic, hip stiffness is the most common complaint and pain is rarely a problem. Arthrography and infection Arthrography plays a role in the evaluation of possible infection. Its value in the evaluation of possible loosening and painful hips is limited. The arthrogram is used to confirm intra-articular position of needle and fluid is aspirated for aerobic and anaerobic culture. The sensitivity for infection is 66-90%. Arthrography and loosening Lack of abnormal contrast extension does not exclude loosening as fibrosis and cells may fill the interfaces preventing contrast passage. In non-cemented THA arthrography is not accurate for the detection of loosening, as small channels between bone ingrowth may persist allowing contrast-passage in stable hips. In cemented THA contrast extension at the bone-cement interface can indicate prosthesis loosening. Communication with the trochanteric bursa, which is common, further reduces sensitivity as a good intra-articular pressure cannot be established (figure). Arthrography and painfull hips Sometimes arthrography is used to find out, if the patients symptoms subside by putting in a long lasting local anaesthetic. This is to see if the painful hip is due to the prosthesis and not something else. Imaging of total hip replacement ,BN Weissman, Radiology 1997; 202: 611. From the RSNA refresher courses. Total hip arthroplasty: radiographic evaluation, BJ Manaster, RadioGraphics 1996; 16: 645. Glatt, A. E., Melamed, E., Cohen, I., Robinson, D., Zimmerli, W., Trampuz, A. (2005).. N Engl J Med 352: 95-97 S Ostlere, FRCR and S Soin, MB BChir. Nuffield Orthopaedic Centre and Oxford Radcliffe Hospital, Oxford, UK Complications of total hip arthroplasty. Saleh, KJ, Kassim, R, Yoon, P, Vorlicky, LN. Am J Orthop 2002; 31:485Iain Watt, Susanne Boldrik, Evert van Langelaan and Robin Smithuis Alignment and Positioning Cemented THA Non Cemented THA Loosening Particle Disease Polyethylene wear Infection Fractures Dislocation Component fracture Heterotopic OssificationHip Arthroplastyfrom the Radiology Departments of the Leids University Hospital, Leiden; the Medical Centre Alkmaar, Alkmaar and the Orthopedic and Radiology Department of the Rijnland Hospital, Leiderdorp, the Netherlands
msk11
1
Knee Meniscus - basics
by David Rubin and Robin Smithuis
This article is based on a presentation by David Rubin and adapted for the Radiology Assistant by Robin Smithuis. Interactive cases are presented in the menubar to test your knowledge. by David Rubin and Robin Smithuis Both horns are triangular in shape and have very sharp points. The posterior horn is always larger than the anterior horn (figure). If this is not the case than the shape is abnormal, which can be a sign of a meniscal tear or a partial meniscectomy. The posterior root is immediately anterior to the posterior cruciate ligament. If it is missing on the sagittal images, then there is a meniscal root tear (figure). The anterior horn has an insertion on the tibia and a second portion that travels from medial to lateral to connect to the anterior horn of the lateral meniscus ( intermeniscal or transverse ligament). On sagittal images the posterior horn is higher in position than the anterior horn. Both horns are about the same size. The lateral meniscus posteriorly comes up higher over the tibial spine to insert near the posterior cruciate ligament. This upward position of the posterior horn may be the reason for the higher signal intensity of the posterior horn in all planes due to magic angle effect. The two most important criteria for meniscal tears are an abnormal shape of the meniscus and high signal intensityon PD-images unequivocally contacting the surface . It is a misunderstanding that menisci should be homogeneously low in signal intensity on proton-density images. The meniscus does not have to be black. Only when the high signal unequivocally reaches the surface of the meniscus you can make the diagnosis of a tear. If there is doubt whether the high signal touches the surface, look at all the adjacent images and if there still is doubt than do not diagnose a tear. If you have a questionmark in your head, say meniscus is normal. (figure) Shapes. There are 3 basic shapes of meniscal tears: longitudinal, horizontal and radial . Complex tears are a combination of these basic shapes. Displaced Tears Bucket-handle tear = displaced longitudinal tear. Flap tear = displaced horizontal tear. Parrot beak = displaced radial tear. Longitudinal tears Longitudinal tears parallel the long axis of the meniscus dividing the meniscus in an inner and outer part. So the distance between the tear and the outer margin of the meniscus is always the same (figure). The tear never touches the inner margin. Longitudinal tears follow the collagen bundles that parallel the contour of the meniscus. If a longitudinal tear has other components (horizontal or radial) than it is a complex tear violating the collagen bundles. This requires a higher energy trauma. Longitudinal tear (2) Bucket handle tear is a displaced longitudinal On coronal images bucket handle tears are easier to recognize. Normally there are only two structures in the intercondylar fossa: the anterior and posterior cruciate ligament. Any other structure in the intercondylar fossa is abnormal and a displaced meniscal fragment is the most likely possibility. More on Bucket handle tears in the article 'Knee Meniscus: special cases'. Longitudinal tear (3) Flipped meniscus is a form of bucket handle tear. There is a capsular detachment or peripheral tear of the meniscus, usually the posterior horn. The posterior horn flippes over onto the anterior horn. More on Flipped meniscus in the article 'Knee Meniscus: special cases'. Horizontal tears Horizontal tears divide the meniscus in a top and bottom part (pitta bread). If horizontal tears go all the way from the apex to the outer margin of the meniscus they may result in the formation of a meniscal cyst. The synovial fluid runs through the horizontal tear and accumulates periferally to the meniscus. The connection with the joint space is often lost, so they will not fill with contrast on MR-arthrography. The synovial fluid is absorbed and is replaced by a gelatinous substance. There are 3 criteria for the diagnosis of a meniscal cyst: 1. Horizontal tear. 2. Fluid accumulation bright on T2. 3. Flat lining against the outside margin of the meniscus. The diagnosis of a meniscal cyst is important to the surgeon because it takes one operation on the outside of the knee to remove the cyst and another operation on the inside for the meniscus. Radial tears Radial tears are perpendicular to the long axis of the meniscus. They violate the collagen bundles that parallel the long axis of the meniscus. These are high energy tears. They start at the inner margin and go either partial or all the way through the meniscus dividing the meniscus into a front and a back piece. Radial tears are difficult to recognize. You have to combine the findings on sagittal and coronal images to make the diagnosis. The following combination of findings is diagnostic: In one plane: triangle missing the tip and in the other plane: a disrupted bow tie. Small radial tears are difficult to diagnose. Sometimes the only sign is a disrupted bow tie. If you image a complete radial tear directly along the length of the tear you will see an absent or empty meniscus. These complete radial tears open up and give the impression that there is a part missing. However you will not find a displaced meniscal fragment. It is simply separation of the meniscal parts. More on empty meniscus sign in the article 'Knee Meniscus: special cases'. A meniscal root tear is a radial tear located at the meniscal root. Normally when you image the posterior cruciate ligament on sagittal images you should see a considerable posterior horn of the meniscus on that image or the image adjacent to it. If this is not the case it is an absent or empty meniscus-sign indicating a radial tear. More on meniscal root tears in the article 'Knee Meniscus: special cases'. Post-operative Menisci are harder to evaluate because the two most important criteria, i.e. abnormal signal and abnormal shape, do not work any more. Abnormal signal is not anymore a reliable sign of a tear, because if there has been a suture repair, this will heal with scar tissue, which also has high signal on PD-images (figure). However if there is also high signal on T2-weighted images than you can make the diagnosis of a tear as this is the result of synovial fluid leaking into a tear. This however is an uncommon finding. Abnormal shape can be the result of partial meniscectomy. So you need to know what procedure was performed during arthroscopy. Only when you can compare with prior postoperative images, you can say if an abnormal shape is a new finding indicative of a new tear. Sometimes differation between normal post-op findings and a re-tear is not possible on conventional MR-images. In these cases MR-arthrography with 40cc diluted Gadolinium helps to make the distinction because even small amounts of Gadolinium that leak into a tear are readily visible on fat saturated T1 images. Post-operative Meniscus 1 The case on the left shows a meniscus with an abnormal shape aswell as abnormal signal touching the surface on PD but not on T2W-images. This patient had a prior partial meniscectomy and a suture repair. On the basis of these imaging findings it's impossible to tell if this is a tear or normal postoperative finding. This patient had another operation for ACL reconstruction. They also looked at the meniscus and the meniscus was found to be normal i.e. no tear. Post-operative Meniscus 2 This patient had a suture repair for meniscal tear. There was a new injury. On the new MR impossible to determine if the old tear had healed. However a new tear is seen, so this case ia easy. On a MR-arthrogram there was very high signal intensity in the new tear comparable with the synovial fluid, but only moderate signal intensity at the healed old tear. So comparison with the old films was diagnostic for the new tear, while the arthrogram showed that the old tear has healed. Post-operative Meniscus 3 This patient also had a suture repair for meniscal tear. After a new injury the PD-images show high signal unequivocally reaching the surface of the meniscus (seen on the original films, but not clearly seen on the compressed image on the left. On this image it is not possible to tell if the tear has healed. So a MR-arthrogram was performed which showed that the tear has healed.David Rubin and Robin Smithuis Medial meniscus Lateral meniscus Criteria for tears Nomenclature of Meniscal Tears Longitudinal, horizontal and radial tears Meniscal root tearKnee Meniscus - basicsRadiology department of the Washington University School of Medicine, St. Louis, USA and the Rijnland hospital in Leiderdorp, the Netherlands
msk12
1
Knee Meniscus - special cases
by Robin Smithuis
In this article we will show some examples of special meniscal pathology in more detail. For the basics of meniscal pathology we advise you first to read the article 'Knee Meniscus - Part 1'. On most images you can click to get an enlarged view, but this does not work on the iPhone application. by Robin Smithuis Study the image on the left and try to determine what the problem is with this meniscus. Then continue with the next consecutive images of the same patient. Read the article 'Knee Meniscus - basics' for the basics Use the scrollbar to scroll through the images. Then continue reading.. As you already suspected by reading the title of this paragraph, this is a flipped meniscus. A flipped meniscus is a special form of bucket-handle tear. A flipped meniscus occurs when the ruptured fragment of the posterior horn is flipped anteriorly so the anterior horn of the meniscus appears to be enlarged. On the left another flipped meniscus. Now on the medial side. Part of the anterior horn is flipped posteriorly. Only a small part of the anterior horn is seen anteriorly. Most flipped menisci occur on the lateral side. The ACL prevents the meniscal fragment from completely migrating into the intercondylar notch. Illustration of the mechanism in a flipped meniscus. On a coronal image you will first see an enlarged bulky anterior horn. Posteriorly a very small posterior horn will be seen. See next case. On the left another case of a flipped lateral meniscus. Use the scrollbar on the right. Same case sagittal images. Notice how the ruptured part of the meniscus runs anteriorly through the intercondylar fossa (arrows)  On the left sagittal PD-images of a flipped meniscus. The whole posterior horn is flipped anteriorly resulting in an empty meniscus sign (arrow). First study the images on the left. Then continue reading.  Bucket handle tears are displaced vertical longitudinal tears. The displaced inner fragment resembles the handle of a bucket. The remaining larger peripheral portion of the meniscus resembles the bucket. These tears account for about 10% of all meniscal tears. The double posterior cruciate ligament (PCL) sign is a low-signal-intensity band that is parallel and anteroinferior to the PCL on sagittal MR images. It is a highly specific indicator of a bucket-handle meniscal tear (3). First study the image on the left and try to recognize the meniscal tear. These tears often go unnoticed. Then continue with the next images. A radial tear is present at the posterior root junction of the medial meniscus which extends through the entire thickness of the meniscus with a cleft of fluid tracking through the defect (red arrows). Meniscal root tears are often associated with extrusion of the meniscus beyond the margin of the tibial plateau. More than 3 mm meniscus extrusion is often associated with tears involving the meniscal root (6). In the case on the left there is a complete radial tear separating the posterior horn from its root (red arrows). There is also minimal extrusion of the meniscus (image 1/6). On the left another medial meniscal root tear. Notice how easily you can overlook these tears and think that the posterior horn is normal. On the left another typical case of a medial meniscal root tear. Notice the lateral discoid meniscus. When there is a complete radial tear, the two meniscal fragments can be completely separated. This can result in an empty meniscal space or empty meniscus sign (arrow). If you image a complete radial tear directly along the length of the tear you will see an absent or empty meniscus. These complete radial tears open up and give the impression that there is a part missing. However you will not find a displaced meniscal fragment. It is simply separation of the meniscal parts. On the left an illustration of a complete radial tear, which can result in an empty meniscus sign. On the left coronal PD-images of a patient with a complete radial tear resulting in an empty meniscus sign.  On the left consecutive coronal PD-images of a posterior horn of a meniscus that at first glance might give the impression that it is normal. Continue with the sagittal images. When you take a good look at the sagittal images, you will notice the empty meniscus sign, where normally the meniscal root attaches (red arrows). This means that we are dealing with a meniscal root tear. The meniscus posteriorly should come up over the tibial spine to insert near the posterior cruciate ligament. First scroll through the images on the left. Try to figure out what is going on with this meniscus. Then continue reading. At first impression this looks like a tear within a discoid meniscus. At closer look you will notice that the horizontal structure has a lower signal intensity than the meniscus and looks funny. This is a normal variant caused by a vacuum phenomenon. A vacuum phenomenon is caused by negative pressure within the joint due to the position of the patient, which results in the accumulation of nitrogengas. Continue with the radiograph of the same patient. Vacuum phenomenon on the lateral side. First scroll through the images on the left. Try to figure out what is going on with this meniscus. Then continue reading. The hypointense structure on the lateral side is a discoid meniscus (blue arrow). The structure on the medial side is again a vacuum phenomenon. On an adjacent slice this vacuum phenomenon is not seen any more. First scroll through the images on the left. Try to figure out what is going on with this meniscus. Then continue reading. There is a longitudinal tear in the periphery of the meniscus (red arrow). The outside one-third of the meniscus is called the 'red' zone, because it has a rich blood supply. The location of a meniscal tear is of importance because tears in this vascular portion of the meniscus are more likely to heal spontaneously than tears in the avascular portion or white zone of the meniscus. Sometimes extensive triangular or wedge-shaped high signal intensity can be encountered that does not reach te surface of the meniscus. This is sometimes referred to as meniscus within meniscus sign. Since this meniscal abnormality does not reach the meniscal surface, it does not fullfill the criteria for a meniscal tear. It was found that in half of these patients and symptoms warranting arthroscopic follow-up had meniscal tears (4). On the left another meniscus with diffuse high signal in the meniscal body. On other images (not shown) there was no evidence of a tear. Notice severe extrusion of the meniscus beyond the margin of the tibia plateau. Tears involving the meniscal root (central attachment) are also significantly related to the severity of meniscal extrusion, seen in 3% with minor extrusion and 42% with major extrusion. With meniscus extrusion, the meniscus is unable to resist hoop stresses and cannot shield the adjacent articular cartilage from excessive axial load. Over time, this can lead to symptomatic knee osteoarthritis. Tears of the posterior meniscal root can be easily missed because of inconsistent clinical symptoms and can be overlooked without thorough arthroscopic examination. Detection of meniscal extrusion is important not only because it is associated with underlying tear but also because meniscal extrusion itself is thought to be related to development of osteoarthritis. A Segond fracture is an avulsion of the lateral capsular ligament. The mechanism of injury is internal rotation and varus stress. On a radiograph it manifests as an elliptic bony fragment off the lateral proximal tibia (figure). A Segond fracture has a high association with a tear of the anterior cruciate ligament (75-100%) and injuries of the medial and lateral menisci (66-70%). On the radiograph you could easily miss the Segond fracture (red arrow). Notice that there is also an avulsion of the medial collateral ligament. Continue with the MR-images. On the left three consecutive coronal PD-images:  Continue with the sagittal images. A Segond fracture is almost pathognomonic for an anterior cruciate ligament tear, which was also demonstrated in this patient. On the left an AP-view of another patient. In association with a Segond fracture (red arrow), there is also an avulsion fracture of the anterior cruciate ligament (blue arrow). A meniscal cyst results from extrusion of synovial fluid through a peripherally extended horizontal meniscal tear. Medial meniscal cysts are most commonly located adjacent to the posterior horn and lateral meniscal cysts are most commonly located adjacent to the anterior horn or body. On the left a PD- and a T2-weighted image demonstrating a lateral meniscal cyst are adjacent to the anterior horn as a result of a complex tear. On the left three consecutive images of a small meniscal cyst (blue arrow) as a result of a horizontal tear (red arrow). On the left coronal PD-images without fatsat and with fatsat. A large meniscal cyst is seen in relation to a horizontal tear (red arrow). by Keith W. Harper, Clyde A. Helms, H. Stanley Lambert and Laurence D. Higgins. AJR 2005; 185:1429-1434 by DH Wright, AA De Smet and M Norris s. American Journal of Roentgenology, Vol 165, 621-625 by Marc A. Camacho November 2004 Radiology, 233, 503-504. by Thomas R. McCauley et al AJR 2002; 179:645-648 by Jeffrey M. Brody et al by Rosalia Costa et al AJR 2004; 183:17-23 by So Yeon Lee et al AJR 2008; 191:81-85 by Scot E. Campbell AJR 2001; 177:409-413Robin SmithuisKnee Meniscus - special casesRadiology department of the Rijnland hospital in Leiderdorp, the Netherlands
msk13
1
Knee - Non-Meniscal pathology
by David Rubin and Robin Smithuis
This article is based on a presentation given by David Rubin and adapted for the Radiology Assistant by Robin Smithuis. This review focusses on all the non-meniscal pathology of the knee. See the article entitled Knee MRI - meniscal pathology for the pathology of the meniscus. Interactive cases are presented in the menubar. by David Rubin and Robin Smithuis The ACL has interesting anatomy. It is an intra-articular structure, but it is extra-synovial. The synovium folds over the ligament. So at arthroscopy they look through the synovium. Sometimes when there is a tear ,the synovium layer is intact and only a hemorrhagic ACL is seen. The ACL is composed of 3-5 layers of fibers. Between the fibers there can be fat or synovium or sometimes a little bit of fluid. This explains why the ACL is not black on PD-images. Do not look at the ACL on PD-images because this may give a false impression of pathology. Only look at the ACL on T2W-images and even on these images the ACL does not have to be entirely black. Criteria for the normal ACL are: So on MR the primary signs of a tear are: discontinuity on T2, abnormal orientation or non-visualisation. Many secondary signs of tears have been described, but these are not helpfull, since we have to rely on direct visualisation of the ligament. Only bone bruises can be a helpfull secondary sign. Notice that on coronal and axial images fibers of the ACL are right next to the bone of the intercondylar notch (arrows). There should never be any fluid between these ACL-fibers and the bone of the lateral condyle ('empty notch sign'). Also notice that the PCL is also composed of many fibers. Anterior Cruciate Ligament (2). The case on the left shows a ligament that's too flat and we see disrupted fibers so there is abnormal orientation and discontinuity. Based on these images we cannot differentiate between complete tear, high grade partial tear or partial tear. MRI does not accurately differentiate between partial or complete ACL tear. But yes we can differentiate between high grade or low grade injury. A high grade injury is 'not able to see 50% of the fibers'. So if the othopaedic surgeons operate on a high grade injury, they will either find a totally torn ACL or a high grade partial tear, that needs to be repaired. On the other hand if most of the fibers appear to be intact on MR indicating a low grade ACL tear, they will find an intact or partially torn ACL, that is stable and doesn't need any treatment. Anterior Cruciate Ligament (3). Bone bruises appear in a very typical location indicating the dislocation, that was the cause of the ACL-tear. Anterior Cruciate Ligament (4) On X-rays an important indirect sign of an ACL-tear is a Segond fracture. Difficult to see on MR, but much more easy to see on radiographs. A Segond fracture is an avulsion fracture at the attachment of the lateral collateral band due to internal rotation and varus stress. In 75-100% there will also be a tear of the ACL. The unhappy triad or O'Donoghues syndrome is a different combination of injuries. The unhappy triad injury commonly occurs in contact sports such as football when the knee is hit from the outside. This causes an injury to three knee structures: Anterior Cruciate Ligament (5) Case on the left shows a torn ACL. Fibers have an abnormal orientation (too flat). Yet it is difficult to see if these are attached to the femur. The acute angulation in the ligament is due to fact that the ACL and PCL have scarred together (see below). Sometimes it is easier to see whether these fibers are attached in the coronal plane. Against the interior part of the lateral condyle there never should be fluid. If this is the case it is called the 'empty notch sign' indicating that the ACL is torn from it's attachment to the femur. Also in the axial plane there should be ligament next to the condyle. At a lower level we see the torn ACL attached to the posterior cruciate ligament. They have scarred together. This is a very common appearance of a chronic ACL tear. This scarring leads to the acute angulation of the ligament. Even though the ACL is connected to the PCL it is not strong enough and still needs reconstruction. Anterior Cruciate Ligament (6) Case on the left shows a non-visualisation of the ACL on a PD-image. But the lesson is 'do not look at ligaments on a PD-image'. If you want to judge the ACL-ligament look at the T2W-images. The T2W-images show fibers going all the way from the tibia to the femur with a normal orientation. So the ACL is intact. This is a case of mucoid degeneration. Normally between the ACL-fibers there can be synovium or fat. In normal aging that can change into gelatinous material. This has no effect on the strenght of the ACL. Anterior Cruciate Ligament (7) Another case of ACL Mucoid degeneration. Often this is associated with cyst-formation in the bone. You could call it ganglion cyst, but you could also call it normal because it has no clinical meaning. This is part of normal aging. Anterior Cruciate Ligament (8) Case on the left shows cyst seperate from ACL unlike mucoid degenaration. This is a ganglion cyst. Probably also a form of degeneration. The difference with Mucoid degeneration is that these cysts can be symptomatic. Sometimes these cysts will drained under ultrasound guidance. Be sure to use a very large needle, because it is very thick material. We use the same criteria for all the other ligaments in the body. The case on the left shows a high grade PCL tear. The superficial medial collateral ligament (MCL) extends from the medial epicondyle to insert not just near the joint but 7 cm below the joint space. At that point there are three landmarks: the inferomedial geniculate artery and paired veins (figure). The deep part of the MCL, even when it is normal, you may not be able to see. It is closely applied to the medial meniscus and the superficial MCL. Medial collateral ligament (2) The case on the left shows a Grade I sprain of the medial collateral ligament. Medial collateral ligament (3) The case on the left shows a Grade II sprain of the medial collateral ligament. Medial collateral ligament (4) The case on the left shows a superficial MCL that is torn from it's attachment on the tibia. Remember it should be attached 7 cm below the joint line. Deep MCL is also torn the ligament is absent. Normal anatomy Posterolateral corner contains seven or eight structures. Only three of them are important to us because they are visible on MR and because the surgeon might want to fix them. These structures are: 1. Fibular collateral ligament 2. Biceps femoris muscle and tendon. 3. Popliteal tendon The fibular collateral ligament together with the tendon of the biceps femoris form the letter V on sagittal images. They inserts on the fibulahead as the conjoined tendon. Posterolateral corner injury (2) On the left a football player, who was hit in the front part of the knee. The image on the far left shows a bone bruise anteromedially. So you suspect ligamentous injury on the contralateral side, which is the posterolateral corner. The next image shows a normal popliteus tendon but biceps femoris tendon is not attached to the fibula. On the left more images of the same patient located more anteriorly. The fibular collateral ligament has a normal proximal attachment but is not attached to the fibula. On a sagittal plane there is a gap between biceps femoris tendon and collateral ligament on one side and the fibular head on the other. These findings indicate a conjoined tendon rupture. Posterolateral corner injury (3) On the left PD-fatsat images after severe injury. There are bone bruises and many ligaments are ruptured. There is a posterolateral corner injury with proximal rupture of the fibular collateral ligament. There is also a rupture of the popliteus tendon because it is not attached proximally. There are about 12 named bursae and recesses in the knee. Some very common and others uncommon. These are synovial lined structures. The most common recess is the popliteal or Baker's cyst. The origin is between the semimembranosus and gastrocnemius tendon. On the left the typical imaging findings of prepatellar bursitis. An uncommon form of bursitis is the deep infrapatellar bursitis. Sometimes associated with Osgood-Schlatter. These bursae are all named by the structures next to them. So a bursitis of the bursa between the deep MCL and the superficial MCL is called a medial collateral ligament bursitis. Cysts, Bursae and Recesses (2) Adventitial bursae are bursae, that are formed in places where normally there is no bursa> The bursa is formed due to abnormal friction. Iliotibial Band Friction syndrome A common place for abnormal friction is between the iliotibial band and the lateral condyle in speedwalkers, bicyclists and sometimes runners. When a bursa is formed in this location it is called the 'Iliotibial Band Friction syndrome'. On the left a speedwalker with lateral knee pain. Between iliotibial band and the lateral condyle there should be fat, but in this case it is missing. Same patient. On axial images fluid within a bursa is seen between the iliotibial tract and the underlying femur. Sometimes fluid in this location has to be differentiated from joint fluid. You have to look at all the images. In this case the joint fluid stops at the red arrows. Remember that not everything that's bright on a T2W-image is fluid. You have to be suspicious, if there is something, that looks like a fluid collection, but it is not in a location, where there normally is a bursa, cyst or recess. Give Gadolineum to differentiate cystic from solid. Normal Extensor mechanism The extensor mechanism of the knee is composed of the quadriceps muscle and tendon, the patella and the patellar tendon. The quadriceps tendon is made of four tendons but comes in three layers on sagittal images. It has a broad attachment all the way from the front of the patella almost to the back. The tendons of the quadriceps aswell as the patellar tendon are homogeneous in signal but don't have to be black on PD-images. They have a sharp posterior demarcation. There should be no focal thickness. Quadriceps tendon tear The case on the left shows an abnormal quadriceps attachment. There is only one layer and the attachment does not go from the front of the patella to nearly the back. In such a case extra images higher up have to be made after repositioning of the coil to see what's going on up there. The missing part of the torn quadriceps tendon is retracted. The deep part is still intact Same patient, axial images. The torn quadriceps tendon is very thick indicating tendinopathy. Normal tendons do not tear, so always look for signs of pre-existing tendinopathy. Anywhere in the body, if you see a tendon that looks torn, but there is no pre-existing tendinopathy, think hard, if you really have the right diagnosis. An image below this level shows normal vastus intermedius muscle and tendon. Another example of a partially torn quadriceps tendon. If there is no continuity between the patella and the quadriceps tendons it is a complete tear. Jumper's Knee Jumper's knee is a spectrum from tendinopathy to tear. Just the same as with the quadriceps tendon or any other tendon the MR shows a spectrum from eccentric tendon thickening, indistinct posterior border, increased signal on T2W-images and finally fiber disruption. Patellar tendinopathy Patient on the left is a professional ballet dancer with pain underneath the knee cap. Patellar tendon proximally is too thick. Posterior border is indistinct. In patella a little bit of edema ( or bone bruise). If left untreated could end up like... Partial patellar tendon tear. Image on the right of a different patient. Complete Patellar tendon tear. Images on the left show no continuity between fibers and patella. The tendon is thickened. Patellar sleeve avulsion. In children we have a different situation. They don't develop tendinopathy. The case on the left shows images of a girl who had pain beneath the patella after doing gymnastics. Although the X-ray is normal there accually is a fracture through the cartilage part of the lower pole. On MR it looks just like the jumper's knee above. Only on coronal images the dark fractureline within the bright cartilage is visible. Usually these fractures are sutured. When these lesions are unrecognized they heal with ossification just below the patella. Normal anatomy The patellar cartilage is the thickest in the body. It should have smooth contours. The most important part of the medial retinaculum is the medial patellofemoral ligament which inserts all the way posteriorly just in front of the MCL. Case on the left is a female soccerplayer who twisted her knee. Four MR-images from caudal to cranial demonstrate all the imaging features of a patellar dislocation with rupture of the medial patellar femoral ligament . The patella was dislocated and the medial facet has bumped onto the lateral condyle. The patella has spontaneously reduced. Bone bruise may be complicated by cartilage fracture. Patellar dislocation (2) Patellar dislocation is a common condition, but clinically often unrecognized because the patella after the dislocation comes back in it's normal position. The patient comes with a swollen painfull knee which could be anything from ACL-, MCL- or meniscal tear to a fracture. So the MRI-findings are important in recognizing this condition. Patients who have loose bodies or continuing dislocation may undergo operation with retinaculum repair. In adults the bone marrow is largely composed of fat. Normal islands of red marrow may produce confusing images. Red bone marrow can be pronounced in young women, cigarette smoking, high altitude, hemoglobinopathy or for no reason at all. As long as the criteria on the left are fullfilled it is normal. Normal red marrow on the left. Restricted to the metaphysis and not into the epiphysis. Comes in islands. On T1 brighter than muscle. Abnormal Bone Marrow Case on the left shows abnormal bone marrow. On T1W the signal intensity is lower than muscle. On T2W-images the signal is very bright. The abnormal signal comes into the epiphysis. Another case with abnormal marrow. In this case the marrow is too dark on T1 and T2 due to iron deposition in the marrow after many blood transfusions in a patient with hemosiderosis. The most common marrow abnormality is Avascular Necrosis (AVN). Some people will say 'AVN, Osteochondrosis Dissecans and Stress fracture all look the same'. There is however a distinct difference. AVN has the following features: 1. Focal abnormality is subchondral and originates in the bone. 2. Normal cartilage (until it collapses). 3. Wedge shaped marrow edema due to bone infarction. The wedge-shaped pattern of bone marrow edema is just the same as any other infarction in the body i.e. liver infarction or kidney-infarction. On the left a different entity, but the patient had the same symptoms. Acute onset of medial pain. There is diffuse marrow edema on T2W-image. On T1W-image the focal abnormality is not directly subchondral. The abnormality on the T1 is more inside the edema. On the T1W-image a dark line is visible indicating a insufficiency fracture. This patient will get better with no weight bearing. On the left another patient with knee pain after trauma. There is some effusion but otherwise the x-rays are normal. In the same patient the MRI shows an obvious tibiaplateau fracture. The point is that any patient who is unable to bear weight in the hip, knee or ankle with normal X-rays needs another study. The diagnosis Osteochondritis Dissecans is usually made on X-rays. The question for MRI is whether it is stable or unstable. The case on the left is unstable for two reasons: - small cysts at the base of the lesion (red arrow) - even more important is fluid at the base of the lesion (blue arrow) Notice that this layer of fluid is different from AVN where the fluid is between the cartilage and the bone. Not helpfull for the discussion stable versus unstable OD are - bone marrow edema (could be stable or unstable) - break in the osteochondral surface. So the case on the left is unstable because there is fluid at the base of the lesion. The case on the left shows a OD with bone marrow edema and a break in the osteochondral surface. But since there is no fluid we cannot tell if this is stable or unstable. At operation the OD was found to be stable In those cases where you cannot tell whether the lesion is stable or unstable MR-arthrogram is helpfull. We look for Gadolineum tracking around the osteochondral lesion.David Rubin and Robin Smithuis Anterior Cruciate Ligament Posterior Cruciate Ligament Medial collateral ligament Posterolateral Corner injury Prepatellar bursitis Deep infrapatellar bursitis Normal and abnormal bone marrow Avascular Necrosis Insufficiency fracture Osteochondritis DissecansKnee - Non-Meniscal pathologyRadiology department of the Washington University School of Medicine, St. Louis, USA and the Rijnland hospital in Leiderdorp, the Netherlands
msk14
1
Muscle MR - non-traumatic changes
by Mini Pathria and Jennifer Bradshaw
This article is based on a presentation given by Mini Pathria and was adapted for the Radiology Assistant by Jennifer Bradshaw. In part I we discussed the MR features of various muscle injuries. In part II we will discuss non-traumatic muscle changes. When assessing muscle pathology try to decide which one of the four basic patterns of abnormality is present: Some of these conditions, such as polymyositis, require biopsy for appropriate therapy to be initiated. In other conditions, such as myositis ossificans, biopsy should be avoided because it may lead to an incorrect diagnosis of a neoplasm and thus to inappropriate therapy. Clues to the correct diagnosis and whether biopsy is necessary are often present on the MR images, especially when they are correlated with clinical features and the findings from other imaging modalities (1). The patient on the left had slipped on the ice in the hospital parking lot and torn the hamstrings. The hamstring tear was associated with sciatic neuritis, when the sciatic nerve became irritated by the hematoma. Muscle edema is the most common MR-pattern. It is hard to make a specific diagnosis based on the MR-findings alone. Be sure to get the right history because it usually provides the clue. The most common cause of muscle edema is trauma, which was discussed in 'Muscle MR - traumatic changes - Part I'. Inflammatory myopathy is a term that defines a group of muscle diseases involving inflammation of skeletal muscle and often the adjacent fascia, with elevated CPK. They are thought to be autoimmune disorders. When using the term inflammatory myopathy, one is actually considering three separate disease entities, namely dermatomyositis (DM), polymyositis (PM), and inclusion body myositis (IBM). On the left an example of inflammatory myopathy. Note increased signal of all the muscles, in all the compartments. This is edema. There is also some edema of the subcutaneous tissues. It is very unusual for a trauma, for example, to present with edema in all compartments. There are no fluid collections within the muscles, but notice the perifascial fluid collections. On the left a patient with myositis. Again we see that multiple compartments are affected, there is a vast amount of edema, skin involvement and perifascial fluid. It is non-specific but myositis could be suggested. Inflammatory myositis is generally bilaterally symmetrical. MR features of myositis include normal architecture on T1-weighted images, feathery edema with enhancement, skin reticulation and abnormalities NOT limited to specific compartmental or neural anatomy. On the left a patient with polymyositis (PM), one of the inflammatory myopathies. The large proximal muscles are involved, generally in a symmetric pattern. Generally speaking, not all muscles are involved, so MR can help locate the best area for biopsy. Sometimes whole body MR is used for diagnosis and follow-up of polymyositis after steroid therapy has been initiated. Induction body myositis, one of the inflammatory myopathies, is a more recently recognized form of myositis of unknown cause. It is the most common acquired myopathy in patients > 50 years and makes up about a quarter (16-28%) of all inflammatory myopathies, although inflammation is not a prominent feature in this disease. It is an indolent disease and there are no skin changes. The muscles that tend to be involved are the deltoid, quadriceps (see next example), finger flexors and ankle dorsiflexors. Patients present with muscle weakness, the disease owes its name to the histological finding of vacuoles and filamentous inclusions. Although the findings are non-specific, it is worth considering this diagnosis in a older patient with abnormalities of the above mentioned muscles. On the left a patient with inclusion body myositis. Notice the symmetrical involvement of the quadriceps and the lack of edema in the surrounding tissues. Patients with underlying collagen vascular diseases can develop myositis, such as rheumatoid arthritis, systemic lupus erythematosus (SLE), mixed connective tissue disease and Sjogren syndrome. For example, as in this patient with SLE, it can be very focal (coronal image, right leg, adductor loge) or nodular. A type of nodular myositis is focal proliferative myositis, it is the least common form. This can be seen in association not only with collagen vascular diseases but also lymphoma. The lesion caused by myositis is sometimes indistinguishable from lymphoma itself, and biopsy is necessary to make the diagnosis. On the left another patient with focal nodular myositis which looks like any other mass on T1-weighted, T2-weighted, and post contrast. With a history of lymphoma you could suggest focal nodular myositis, but there is nothing definitive about the images. The relationship of myositis to underlying malignancy remains controversial, and the frequency of this association is not well established. 2 types of cancer show a relationship with myositis: ovarian cancer and Non-Hodgkin lymphoma (shown on the left a patient with, strangely enough, metastatic thyroid cancer). Myositis can precede malignancy (as in a paraneoplastic syndrome), which is not to say that routine screening for malignancy is called for in patients presenting with myositis. Myositis due to radiation can be seen many years after the therapy. It seems to be a vascular problem which doesn't disappear after cessation of the therapy. The history is usually the clue, but also you may see a band like appearance where the radiation changes in the muscles stop, corresponding with the radiation field. On the left a well-known example of inflammatory myopathy which has an endocrine etiology: Graves disease, otherwise known as thyroid eye disease. The inflammation of the eye muscles and orbital fat with subsequent volume increase leads to proptosis. Same patient, coronal T2-weighted image. Several drugs can induce myositis and in the author's practice the most frequent culprit seems to be a lipid-lowering statin. In a significant number of patients statins induce muscle pain and myositis, the dosage then needs to be decreased or the drug needs to be discontinued. On the left an example, note the inflammatory changes in the large muscles around the buttocks. After discontinuation of the drug, the muscle pain will disappear in about 2 weeks, the MRI however will still show abnormalities until roughly a month later. The best time to do a follow-up MR is about 6 weeks after stopping the drug. This was an older patient with elevated cholesterol who was put on Lipitor. The patient developed muscle aches and pains, CPK was mildly elevated. The changes are rather subtle, we see perifascial fluid collections, around the edges of the muscle (the epimysium). Also there are minimal skin changes. Antiretroviral drugs (used in HIV positive patients) can also cause myositis because they interfere with the mitochondrial repair mechanism. The abnormalities are very similar to those in myositis induced by lipid-lowering statins. Again, the patients present with weakness and pain, the changes are centrally located in the larger muscles. The differential in HIV positive patients is relatively long (autoimmune, HIV wasting syndrome with type II muscle fiber atrophy, denervation and infection). It is obviously important to be able to rule out infection in these patients. One way to differentiate between the two is to look for symmetry. Zidovudine (AZT) myopathy, or HIV myositis, is symmetrical. Infection is usually unilateral or at least asymmetrical. Muscle infection or myositis without abscess or necrosis may produce edema as the sole abnormality on MR images. The MR images and clinical history may suggest the presence of such an infection. Bacterial myositis frequently progresses to abscess formation and thus often has a masslike appearance on MR images. Viral infection does not progress to abscess formation. Muscle infection can be due to: Important groups at risk for muscle infection are diabetics, immuno-compromised patients, patients with a penetrating wound (including ulcers that cause infection to spread deep or skin infections). The hallmark of muscle infection is fluid collections present inside the muscle (figure). On the left T2-weighted, T2FS, and post contrast sagittal images of the knee. On the T2-weighted image we see a posterior mass with inflammatory changes in the surrounding muscles. The T2-weighted image with fatsat shows an ill-defined fluid collection and the inflammatory changes in the muscle are clearer. So far this is a non-specific finding. It could be a tumor but tumors tend not to have so much inflammatory change around them. Lack of central enhancement combined with surrounding inflammatory changes and a suggestive history will help to make the diagnosis of pyomyositis. Same patient, T1-weighted image post Gadolineum with fatsat. Pyomyositis (2) On the left another example of pyomyositis, much more extensive, in a patient with AIDS who had a loculated abscess. Note the thick enhancing walls. On the left another case of pyomyositis. Note the asymmetry with enlargement of the left extremity. There is subcutaneous, fascial, and muscular inflammation. Generally speaking muscle infection is a clinical diagnosis, but MR can help determine the depth and extent of the disease. MR also is helpful to locate fluid collections or abscess formation, which can then be aspirated for culture. Pyomyositis (3) On the left bilateral psoas abscesses as a complication of osteomyelitis of the spine in a patient with TB. Necrotizing fasciitis is a rare infection of the deeper layers of skin and subcutaneous tissues, easily spreading across the fascial plane within the subcutaneous tissue. Group A streptococcus is the most frequent pathogen found in necrotizing fasciitis. These bacteria are sometimes called 'flesh-eating bacteria', which is a misnomer as the bacteria do not actually eat the tissue. They cause the destruction of skin and muscle by releasing toxins, which include streptococcal pyogenic exotoxins. On the left an example of another inflammatory disease: Sarcoidosis. Sarcoid is confusing on MR, because you will see changes in the skin and a peculiar nodular enhancement pattern in the muscle. 1-2% of patients with active sarcoid will have muscle involvement and there are always skin changes on the overlying skin. The enhancement pattern is also referred to as the 'Stars and Stripes' pattern, mostly because of the stripes on the long axis of the muscle. The fat present in a muscle will be either intramuscular or intermuscular. There obviously is a wide range of amount of fat present. Far left an example of an olympic gymnast with 6% body fat. Next to it an example of most of us mortals, the so called couch potato, with much more intra- and inter-muscular fat. The image on the far left demonstrates fatty infiltration of muscle in Charcot-Marie-Tooth disease. Charcot-Marie-Tooth disease is also known as hereditary motor and sensory neuropathy or peroneal muscular atrophy. It affects both motor and sensory peripheral nerves. The image next to it demonstrates atrophy of the supraspinatus muscle. The supraspinatus muscle is small and there is peritendinous atrophy. This is the pattern that you will see when the tendon is torn and there is disuse. In the acute phase of denervation, the MR is normal. In the early subacute phase after one week, there will be uniform muscle edema and paradoxical hypertrophy (the muscle looks swollen). In the late subacute phase after 3 weeks, there will be mixed edema and atrophy. The chronic phase is characterized by only atrophy. On the left a patient with cervical root avulsion. The para-spinal musculature shows a mixture of edema and atrophy, 3 weeks post-trauma. On the left a patient with peroneal nerve entrapment by a ganglion, leading to atrophy of the peroneus longus, peroneus brevis and the anterior compartment. Note the increased signal intensity of the anterior muscles, due to edema, meaning early atrophy. On the left an example of atrophy in a patient with a resection arthroplasty of the right hip (Girdlestone procedure). There is atrophy of the thigh musculature and gluteal muscles due to disuse. There is a decrease in size of the muscles and there is fatty replacement in a peritendinous pattern. On the left a patient with a ganglion (blue arrow) in the shoulder. Stop and think for a moment about what causes ganglions in the shoulder. There is a subtle increase in signal in the infraspinatus muscle (yellow arrow), due to edema. Remember that the early subacute stage of denervation presents with edema. The nerve is being impinged by the ganglion. This patient had a labral tear (not pictured), which can lead to ganglion formation. It is important to make this diagnosis while the muscle is not yet atrophic. If the ganglion is aspirated or removed on time, the nerve function will be restored before the muscle becomes atrophic. On the left a different patient with a paralabral ganglion (not pictured). There is atrophy of both the supraspinatus and infraspinatus muscle. There is volume loss and edema, without any focal fluid colllections. On the left a case in which the obturator nerve was clipped at surgery. This is the chronic phase where there is volume loss and fatty replacement of the muscle. On the left another patient with a tear of the posterior labrum and a paralabral cyst. Study the images and try to determine if there is any atrophy. Then continue with the T1-weighted image. This case illustrates how important it is to include a T1-weighted image to your exam! If you were to have only the T2-weighted image with fat sat, you would miss the atrophy of the supraspinatus muscle, which has been entirely replaced by fat. The chronic phase of denervation is characterized by only atrophy. On the left images of a patient with an Erb's palsy. On the left images of a patient with chronic denervation due to polio. On the left a patient with edema of both the deltoid muscle and the supraspinatus. This is an important observation since these muscles are innervated by different nerves. The deltoid muscle is innervated by the axillary nerve and supraspinatus muscle by the suprascapular nerve. This means that the lesion is more proximal. This patient had Parsonage-Turner Syndrome, also known as brachial neuritis. This is an inflammatory disorder characterized by severe pain, followed by weakness due to denervation. Note also edema of the infraspinatus muscle. Muscular dystrophy is a less common cause of atrophy. It is diagnosed clinically usually in children, with patients experiencing symmetrical muscle weakness, pseudohypertrophy of the calves and difficulty in standing up. The muscle initially is edematous and then rapidly becomes atrophic. There are many types of muscular dystrophy. The most common type is Duchenne. On the left an example of adult onset muscular dystrophy. There is subtle high signal intensity of the quadriceps muscle bilaterally due to edema, with a feathery appearance. Most of the adductor muscle is normal. Note the lack of skin edema. This is an important finding to be able to differentiate from inflammatory myositis, in which the skin is also edematous. The imaging findings correspond to the acute stage of muscular dystrophy. In a chronic setting, the muscle will be atrophied with fatty replacement. On the left T1- and T2-weighted images of the thigh muscles. Notice that there is an obvious difference between the signal intensities of the quadriceps muscle and the posterior muscles. On the T1-weighted image only the posterior muscles contain normal fat. On the T2-weighted image there is edema of the quadriceps, which is a sign of early muscular dystrophy. Although the exact diagnosis cannot be made by MR, it can be helpful in suggesting a location for biopsy to determine the type of muscle degenerative disease. On the left an example of chronic muscular dystrophy. The muscle has been entirely replaced by fat. When the muscle loses its nerve supply it becomes atrophic. Dysfunction at any level of the motor unit can lead to denervation with muscle atrophy as a result. Accessory muscles may present as an asymptomatic painless mass or with symptoms of nerve entrapment or vascular entrapment. On the left an example of an accessory soleus, on the medial side of the ankle, which caused compression of the tibial nerve (i.e. tarsal tunnel syndrome). In order to diagnose this entity, you must be really familiar with the anatomy of the area being studied. Patients with accessory muscles will usually present with a painless mass, which will be marked by the lab technician. Initially, the MR appears normal. However, be aware that there are 3 questions that you must consider in these cases: On the left a wrist examination showing an accessory forearm flexor muscle. On the left an example of an accessory muscle at the dorsum of the wrist (T1- and T2-weighted). Under the marker is a well-defined mass, iso-intense to normal muscle. It is a muscle at mid-carpal level, with normal signal. Normally, at this level, there is no muscle on the extensor side of the wrist, just tendon. This is an accessory extensor digitorum manus brevis. A recent article in Radiographics lists the accessory muscles in the human body (Sookur PA et al. Accessory Muscles: Anatomy, Symptoms, and Radiologic Evaluation. Radiographics 2008;28:481-499). On the left an example of a common accessory muscle: the accessory soleus. Normally the soleus muscle inserts almost entirely onto the achilles tendon with a very small soleal tendon passing anteriorly to the achilles tendon. In about 1-2% of the population however, the soleus comes down and inserts directly onto the calcaneus. This will present as a palpable mass and is usually, but not always, bilateral. On the left a low lying soleus muscle, but it did not have a separate tendinous insertion on the calcaneus. On the left another example of an accessory muscle which lies medial to the flexor hallucis longus (middle) and peroneus brevis (lateral). This is a common area for accessory muscles, and there are a lot of different muscles that can be found here (to differentiate you need to determine where the tendon inserts). The reason that this is presented for imaging is because it compresses the adjacent neurovascular bundle leading to atrophy of the muscles of the foot or tarsal tunnel syndrome with tingling of the sole of their foot or weakness. On the left an elbow, the medial side is on the left. Note that there is a muscle directly behind the ulnar nerve, which in a normal situation whould not be present. This is an accessory anconeus epitrochlearis, found in 10% of the population. It is a common cause of ulnar neuritis, due to compression, with pain and tingling of the ulnar side of the hand, and sometimes atrophy of the hypothenar and thenar musculature. Always look carefully at the nerve when you have encountered this muscle. David A. May et al October 2000 RadioGraphics, 20, S295-S315. by J. M. Mellado et al. AJR 2004; 182:1289-1294Mini Pathria and Jennifer Bradshaw Inflammatory myopathy Polymyositis Inclusion body myositis Myositis in collagen vascular disease Radiation myositis Graves disease Drug induced myositis HIV myositis Myositis due to infection Pyomyositis Necrotizing fasciitis Sarcoidosis Atrophy patterns Denervation and Peripheral nerve entrapment Muscular dystrophyMuscle MR - non-traumatic changesDepartment of Radiology of the University of California School of Medicine, San Diego, USA and the Medical Centre Alkmaar, the Netherlands
msk15
1
Muscle MR - traumatic changes
by Mini Pathria and Jennifer Bradshaw
This article is based on a presentation given by Mini Pathria and was adapted for the Radiology Assistant by Jennifer Bradshaw. In part I we will discuss MR features of various muscle injuries. In part II we will discuss non-traumatic muscle changes. Dr. Pathria is a Professor of Clinical Radiology at the University of California, San Diego . Dr. Pathria's specific areas of interest include musculoskeletal trauma, emergency radiology, and musculoskeletal MR imaging. She is the author of the book MRI of the Musculoskeletal System. When looking at muscle on MR there are a few rules to keep in mind: When looking at muscle pathology try to decide which one of the four basic patterns of abnormality is present: On the left an example of a lipoma creating a mass effect in the anterior compartment of the calf. The most common muscle injury is muscle strain (1). It is an injury to the musculotendinous junction. Typical for muscle strain is edema centered along the musculotendinous junction (1). More severe muscle strains contain fluid collections such as hematomas and may contain grossly interrupted muscle fibers and thus may show masslike features on MR images in addition to muscle edema. Muscle contusions are caused by a direct blow. MR images reveal interstitial hemorrhage as well as edema at the injured site . More severe contusions may contain hematomas and thus reveal a masslike lesion in addition to the edema. Muscle strain is an injury to the musculotendinous junction. The tendinous junction is where the muscle fibers meet the tendon, and the shape of it varies in different muscles. In many muscles, the tendon extends deeply into the muscle creating a long musculo-tendinous junction (figure). This area is especially important in a trauma setting, because it is often involved. The epimysium is the fibrous tissue that lies at the edge of the muscle. It becomes the muscle sheath that fuses with the tendon. This is also an important area to consider because when there is a tear in the muscle, fluid tends to leak out and collect in the epimysium. The pattern of edema in muscle strain will depend on the architecture and shape of the musculotendinous junction involved. The image on the left nicely demonstrates the feather-like arrangement of the musculotendinous junction in an atrophic muscle. The image on the left demonstrates edema surrounding the musculotendinous junction in a feather-like arrangement. This is a complete tear to the rectus femoris (arrows). The architecture can be very confusing. For example on the left the rectus femoris, which can show a variety of edema patterns depending on where (anatomically) the injury took place. The blue arrow demonstrates the tendon of the indirect head, which comes from the hip, it has a vertical orientation on this axial image. Along the posterior portion of the muscle (yellow arrows), there is a flat area of tendon originating from the knee. When a muscle has different orientations of the tendons it means that there are different patterns of edema possible depending on the tendon injured. Therefore this is a pattern of edema corresponding to an injury arising from the knee. Muscle Strain (2) Muscle strain is an acute injury. The history is usually very concise. The muscles that are most prone to strain are the long fusiform muscles that cross 2 articulations. Most at risk are the hamstrings, rectus femoris and medial gastrocnemius. Strain involving the upper extremity is slightly less common and then usually involves the biceps brachii. The hallmark of strain is a lot of edema around the myotendinous juntion because that is where the tearing takes place. There is edema around the tendon and sometimes the tendon itself will show signal changes. There are 2 patterns found with muscle strain. By far the most common is the myotendinous junction pattern, which occurs roughly 97% of the time. Depending on the severity of the strain, there might also be fluid collections. The remaining 3% will show an epimysial strain pattern, with the abnormalities found at the periphery of the muscle. Muscle Strain (3) On the left an example of a weight-lifter with an epimysial strain pattern. The tendon tears at the myotendinous junction, and the fluid leaks around the edge of the muscle showing an epimysial appearance. Do not confuse this with a degenerative or impingement-type tear. On the left a strain with partial tear of the subscapularis. The subscapularis is a convergent muscle (like for example the pectoral muscle) with multiple tendons. Edema will have a multipennate distribution pattern, as the edema tracks in different directions along the multiple tendons. On the far left a complete tear of the indirect head of the rectus femoris (yellow arrow). This tear originates in the hip. The image next to it, which was also shown above, shows a completely different finding. There is edema surrounding the distal tendon that is connected to the patella. The other tendon is completely normal (blue arrow). Clinically the severity of a muscle injury is graded from 1-3. Trying to grade a muscle injury by the signal intensity is tricky. Chronic injuries can show mild signal changes and yet still be high grade injuries according to the clinical classification. On the left an example of a tear in the left pectoral muscle. On T1-weighted image there is a gap in the muscle with a small amount of fat filling it up. The gradient echo demonstrates focal fluid accumulation and some increased signal within the muscle. It does not look that severe. However, when asked to fully contract the pectoral muscles there is an obvious asymmetry due to a complete tear in the muscle (blue arrow). This clinically is a grade 3 injury, with a complete loss of function of the muscle. On the left a low grade injury of the flexor hallucis longus. There is normal muscle architecture on the T1-weighted image, and increased signal on T2-weighted image and STIR. The injury will heal completely within a couple of weeks. This example shows edema with an epimysial pattern which is common around the flexor hallucis. When grading acute injuries, look not only at the architecture but also at the length of the muscle. Studies have shown that the length of the muscle strain is the best predictor of the duration of disability, with longer lesions requiring more time to recover. (Reference article by Dr. Connell DA et al, AJR 2004, 83: 975-984). So the longer the abnormality, the longer it takes for the injury to heal. On the left a patient with 2 grades of injury to the gastrocnemius. There is a low grade injury to the lateral head. On the left the same patient. There is also a moderate grade injury to the medial head. Note the fluid accumulations around the muscle head. The more fluid, the higher the grade. On the left an injured rectus femoris muscle. The images demonstrate a moderate grade injury, with architectural distortion and a fluid collection (arrow). Notice the edema at the bipennate musculotendinous junction. On the left an example of a high grade injury. There is a complete tear of the tendon or myotendinous junction of the pectoral muscle. This patient will lose all function in this muscle. On the left a complete tear of the left hamstring at the musculotendinous junction. Tthe tendons are avulsed and there is fluid accumulation. On the left a different patient with also a complete hamstring rupture. There is an epimysial pattern of edema and sciatic nerve irritation. A hamstring syndrome may occur. This is a painful condition caused by post-traumatic scar formation around the sciatic nerve (arrow). On the left images of a patient who had a prior muscle strain. There are typical chronic changes such as focal tendon thickening (blue arrow) and peritendinous muscle atrophy (yellow arrow). On the left images of a patient who had an injury to the long head of the biceps femoris muscle. There are typical chronic changes such as focal tendon thickening (blue arrow) and severe muscle atrophy. Muscle contusions are caused by a direct blow. The MR findings in contusion are similar to strain but without the typical myotendinous junction localization seen in the latter. Typically, there is also skin edema and sometimes, bone contusion. MR images reveal edema at the injured site, frequently due to interstitial hemorrhage as well as edema. More severe contusions may contain hematomas and thus reveal a masslike lesion in addition to edema. It is mostly seen in the superficial muscles. On the left images of a patient who has a mass-like swelling of the fore-foot. The findings are nonspecific, but the history 'Slammed car door on foot' was specific. Hemorrhage can present as a discrete hematoma or as parenchymal hemorrhage. In the case on the left there is a mixed pattern. The signal intensity of a hematoma on T1W- and T2W-images depends on the stage of the hematoma (Table). On the left images of a patient who fell on a slippery floor. There is a hyperacute hematoma. Low signal intensity on T1W and high signal on T2W. On the left an acute hematoma. It is isointense or hypointense to muscle on T1. The hypointensity in T2WI in the acute period is due to the high concentration of intracellular deoxyhemoglobin. On the left images of two different patients with an early subacute hematoma. On the far left a T1-weighted image. The hyperintensity at the periphery of the hematoma is due to methemoglobin which is seen after 2-7 days and can persist for months. The image on the right shows the same hyperintensity on a T2-weighted image. On the left a late subacute hematoma with layering. On the left images of two different patients with a chronic hematoma in the calf. On the left a T1-weighted image. Notice the dark rim of hemosiderin surrounding the hematoma. On the right a T2-weighted image of a similar case. Notice that the hemosiderin is also dark on T2-weighted image. On the left a chronic hematoma known as Morel-Lavall?e lesion. A Morel-Lavall?e lesion is the result of separation of the skin and subcutis from the fascia, producing a cavity that is filled with fluid and debris. These lesions are found around the thigh and have a well-defined oval or fusiform shape. They are usually partially or wholly encapsulated. A hematoma can look like a tumor and vice versa. On the left a metastasis of a renal cell carcinoma. When in doubt use gadolinium to see if the abnormality enhances. Continue with the post-Gad image. The majority of the lesion enhances, making hematoma unlikely. In the centre there is no enhancement as a result of necrosis. Hematomas can show some enhancement, but only at edge. Severe blunt trauma causing an intra-muscular hematoma may result in delayed ossification in the soft tissues known as myositis ossificans. Myositis ossificans has a variable appearance depending on the maturity: On MRI myositis ossificans can be difficult to differentiate from osteosarcoma. On X-rays and CT soft tissue ossification not attached to bone is seen. On the left another case of myositis ossificans with bone formation. Compartment syndrome is a limb-threatening and life-threatening condition observed when perfusion pressure falls below tissue pressure in a closed anatomic space. A fasciotomy procedure with incision in the skin and the muscle fascia is necessary to release the pressure and regain normal function of the capillaries. Compartment syndrome progresses to rhabdomyolysis if untreated. Necrosis of tissue may begin at interstitial pressure as low as 30 mm. In the lower leg there are four compartments: the anterior, deep and superficial posterior compartment and a small lateral compartment. On the left T1W-images of a patient one month post trauma. On the post-Gadolinium image the necrosis in the anterior and lateral compartment is seen. The posterior compartment is normal. On the left a T2W-image of a patient with a chronic lateral compartment syndrome. On the left a compartment syndrome in the upper leg which progressed to rhabdomyolysis. Rhabdomyolysis is a dissolution of skeletal muscles that causes extravasation of toxic intracellular contents from the myocytes into the circulatory system and can lead to kidney failure. Calcific myonecrosis is a rare post- traumatic entity characterized by latent formation of a dystrophic calcified mass occurring almost exclusively in the lower limb. In calcific myonecrosis an entire single muscle is replaced by a fusiform mass with central liquefaction and peripheral calcification. They can present as enlarging soft tissue masses with clinical features that suggest an enlarging soft-tissue neoplasm or infection. On the left a patient who met up with the wrong end of a knife. The man was caught by his wife while cheating on her with another woman and he was rewarded with a stab into the groin. This resulted in a laceration of his right pectineus muscle. MR imaging is usually not required for laceration, since these patients usually go directly to the ER or OR for surgical exploration, but this case nicely demonstrates the atrophied muscle and the scar tissue. Delayed onset muscle soreness (DOMS) develops 1-2 days following exercise and resolves in 1-2 weeks (for example after the first days on the ski slopes). DOMS is a type of overuse injury that does not become symptomatic until hours or days after the overuse episode, in contrast with a muscle strain or contusion, which usually is immediately painful. The MR findings show diffuse muscle edema that does not localize to the myotendinous junction and can persist for weeks. On the left a patient who had gone for a run for the first time in quite a while. The muscle is irritated as illustrated by edema in the gastrocnemius (arrows). Because there is a delay in symptoms, patients are not always aware of when or how the injury was actually caused. On the left a navy recruit with delayed onset muscle soreness after weight-lifting. Note the swollen edematous brachialis muscle. These abnormalities can last for weeks. A fascial tear presents as a mass, the signal is usually normal (rather like an accessory muscle). The muscle herniates through the fascial defect, protruding upon muscle contraction. It is an intermittent mass and can be missed on MR if it is only visible during contraction. A fascial tear is a typical sports injury and most commonly involves the calf (figure). This type of muscle injury is well evaluated with ultrasound, because it is an dynamic examination. David A. May et al October 2000 RadioGraphics, 20, S295-S315. Connell DA et al, AJR 2004 183:975-984 by J. M. Mellado et al. AJR 2004; 182:1289-1294 by Helena M. O. Dwyer et al, AJR 2006; 187:W67-W76 by Dennis L. Janzen et al AJR 1993;160:1072-1074Mini Pathria and Jennifer Bradshaw Muscle injury Grading muscle strain Low grade muscle strain Moderate grade muscle strain High grade muscle strain Chronic changes of muscle strain Calcific myonecrosisMuscle MR - traumatic changesDepartment of Radiology of the University of California School of Medicine, San Diego, USA and the Medical Centre Alkmaar, the Netherlands
msk16
1
Shoulder MR - Anatomy
by Robin Smithuis and Henk Jan van der Woude
The glenohumearal joint has a greater range of motion than any other joint in the body. The small size of the glenoid fossa and the relative laxity of the joint capsule renders the joint relatively unstable and prone to subluxation and dislocation. MR is the best imaging modality to examen patients with shoulder pain and instability. In Shoulder MR-Part I we will focus on the normal anatomy and the many anatomical variants that may simulate pathology. In part II we will discuss shoulder instability. In part III we will focus on impingement and rotator cuff tears. The glenohumeral joint has the following supporting structures: Anterior graphic of the shoulder. The tendon of the subscapularis muscle attaches both to the lesser tubercle aswell as to the greater tubercle giving support to the long head of the biceps in the bicipital groove. Dislocation of the long head of the biceps will inevitably result in rupture of part of the subscapularis tendon. The rotator cuff is made of the tendons of subscapularis, supraspinatus, infraspinatus and teres minor muscle. Posterior graphic of the shoulder. The supraspinatus, infraspinatus and teres minor muscles and tendons are shown. They all attach to the greater tuberosity. The rotator cuff muscles and tendons act to stabilize the shoulderjoint during movements. Without the rotator cuff, the humeral head would ride up partially out of the glenoid fossa, lessening the efficiency of the deltoid muscle. Large tears of the rotator cuff may allow the humeral head to migrate upwards resulting in a high riding humeral head.  The supraspinatus tendon is the most important structure of the rotator cuff and subject to tendinopathy and tears. Tears of the supraspinatus tendon are best seen on coronal oblique and ABER-series. In many cases the axis of the supraspinatus tendon (arrowheads) is rotated more anteriorly compared to the axis of the muscle (yellow arrow). When you plan the coronal oblique series, it is best to focus on the axis of the supraspinatus tendon.   Labral tears The abduction external rotation (ABER) view is excellent for assessing the anteroinferior labrum at the 3-6 o'clock position, where most labral tears are located. In the ABER position the inferior glenohumeral ligament is stretched resulting in tension on the anteroinferior labrum, allowing intra-articular contrast to get between the labral tear and the glenoid. Rotator cuff tears The ABER view is also very useful for both partial- and full-thickness tears of the rotator cuff. The abduction and external rotation of the arm releases tension on the cuff relative to the normal coronal view obtained with the arm in adduction. As a result, subtle articular-sided partial thickness tears will not lie apposed to the adjacent intact fibers of the remaining rotator cuff nor be effaced against the humeral head, and intra-articular contrast can enhance visualization of the tear (3). Images in the ABER position are obtained in an axial way 45 degrees off the coronal plane (figure). In that position the 3-6 o'clock region is imaged perpendicular. Notice red arrow indicating a small Perthes-lesion, which was not seen on the standard axial views.  There are many labral variants. These normal variants are all located in the 11-3 o'clock position. It is important to recognise these variants, because they can mimick a SLAP tear. These normal variants will usually not mimick a Bankart-lesion, since it is located at the 3-6 o'clock position, where these normal variants do not occur. However labral tears may originate at the 3-6 o'clock position and subsequently extend superiorly. There are 3 types of attachment of the superior labrum at the 12 o'clock position where the biceps tendon inserts. In type I there is no recess between the glenoid cartilage and the labrum. In type II there is a small recess. In type III there is a large sublabral recess. This sublabral recess can be difficult to distinguish from a SLAP-tear or a sublabral foramen. These images illustrate the differences between an sublabral recess and a SLAP-tear. A recess more than 3-5 mm is always abnormal and should be regarded as a SLAP-tear. The image shows the typical findings of a sublabral recess. A sublabral foramen or sublabral hole is an unattached anterosuperior labrum at the 1-3 o'clock position. It is seen in 11% of individuals. On a MR-arthtrogram a sublabral foramen should not be confused with a sublabral recess or SLAP-tear, which are also located in this region. A sublabral recess however is located at the site of the attachment of the biceps tendon at 12 o'clock and does not extend to the 1-3 o?lock position. A SLAP tear may extend to the 1-3 o'?lock position, but the attachment of the biceps tendon to the superior labrum should always be involved. Scroll through the images and notice the unattached labrum at the 12-3 o'clock position at the site of the sublabral foramen. Notice the smooth borders unlike the margins of a SLAP-tear. A Buford complex is a congenital labral variant. The anterosuperior labrum is absent in the 1-3 o'clock position and the middle glenohumeral ligament is usually thickened. It is present in approximately 1.5% of individuals. On these axial images a Buford complex can be identified. The anterior labrum is absent in the 1-3 o'clock position and there is a thickened middle GHL. The thickened middle GHL should not be confused with a displaced labrum. It should always be possible to trace the middle GHL upwards to the glenoid rim and downwards to the humerus. Failure of one of the acromial ossification centers to fuse will result in an os acromiale. It is present in 5% of the population. Usually it is an incidental finding and regarded as a normal variant. The os acromiale may cause impingement because if it is unstable, it may be pulled inferiorly during abduction by the deltoid, which attaches here. On MR an os acromiale is best seen on superior axial images. An os acromiale must be mentioned in the report, because in patients who are considered for subacromial decompression, the removal of the acromion distal to the synchondrosis may further destabilize the synchondrosis and allow for even greater mobility of the os acromiale after surgery and worsening of the impingement (4). The axial MR-images show an os acromiale with degenerative changes, i.e. subchondral cysts and osteophytes (arrow). by Jaideep J. Iyengar, MD; Keith R. Burnett, MD; Wesley M. Nottage, MD ORTHOPEDICS August 2010;33(8):562. by Schreinemachers SA, van der Hulst VP, Willems WJ, Bipat S, van der Woude HJ. Skeletal Radiol. 2009; 38(10):967-975. by Herold T, Bachthaler M, Hamer OW, et al. Radiology. 2006; 240(1):152-160. MRI of the shoulder second edition by Michael Zlatkin.Robin Smithuis and Henk Jan van der Woude Axial anatomy and checklist Axis of supraspinous tendon Coronal anatomy and checklist Sagittal anatomy and checklist ABER view ABER - anatomy Sublabral recess Sublabral Foramen Buford complexShoulder MR - AnatomyRadiology department of the Rijnland hospital, Leiderdorp and the Onze Lieve Vrouwe Gasthuis, Amsterdam, the Netherlands
msk17
1
Shoulder MR - Bankart lesions
by Robin Smithuis and Henk Jan van der Woude
A Bankart lesion is an injury of the anterior glenoid labrum due to anterior shoulder dislocation. These labral tears make the shoulder unstable and susceptible to repeated dislocations. In this article we will focus on:  A Clockwise approach to the labrum is the easiest way to diagnose labral tears and to differentiate them from normal labral variants. There are two types of labral tears: SLAP tears and Bankart lesions. SLAP is an acronym that stands for 'Superior Labral tear from Anterior to Posterior'. SLAP tears start at the 12 o'clock position where the biceps anchor is located, which tears the labrum off the glenoid. SLAP tears typically extend from the 10 to the 2 o'clock position, but can extend more posteriorly or anteriorly and even extend into the biceps tendon. Bankart lesions are typically located in the 3-6 o'clock position because that's where the humeral head dislocates. There are many labral variants that may simulate a labral tear. They also have a typical location. They are not in the 3-6 o'clock position, which makes it easy to differentiate them from a Bankart tear. A Bankart tear can extend to the 1-3 o'clock position, but then there should also be a tear in the 3-6 o'clock position. Labral variants however may mimick a SLAP tear. The shoulder is a very mobile and therefore unstable joint. It is the most dislocated joint in the body. The humeral head is almost always displaced anteriorly, inferiorly and medially below the coracoid process (95% of cases). Motion to superior is limited by the acromion, coracoid process and rotator cuff (figure). Motion in a posterior direction is limited by the posterior rim of the glenoid which is in an anteverted position. The dislocation of the humeral head to antero-inferior causes damage to the antero-inferior rim of the glenoid in the 3 - 6 o'clock position (marked in red). Especially in younger patients this results in a Bankart fracture or a Bankart lesion which is a tear of the anteroinferior labrum. This results in instability and recurrent dislocations. Dye to these recurrent dislocations significant bone loss and erosion of the anterior glenoid rim may occur, which will further increase the instability. The images show a subtle Bankart fracture (arrows). This is a post-reduction view. Notice the very large fracure of the glenoid rim with displacement. On the coronal-oblique and sagittal reconstruction the displaced fracment is seen in the 3-6 o'clock position. There is also a large Hill-Sachs impression fracture. 3D-reconstruction of a large bony Bankart in the 2 - 6 o'clock position. On MR a Hill-Sachs defect is seen at or above the level of the coracoid process. Hill-Sachs is a posterolateral depression of the humeral head. It is above or at the level of the coracoid in the first 18 mm of the proximal humeral head. It is seen in 75-100% of patients with anterior instability. It is chondral or osteochondral. MRI is 94% accurate. The physiologic groove in the humerus or cysts and erosions at the attachment site of the infraspinatus tendon can simulate a Hill-Sachs, but usually this is not a diagnostic problem (figure). Posterior dislocations are uncommon and easily missed, because there is less displacement compared to the anterior dislocation. On the AP-view the head looks strange due to the internal rotation. On the transscapular-Y view the humeral head is displaced posteriorly. Sometimes the displacement is difficult to appreciate, especially when the transscapular-Y view is slightly rotated. Sometimes an axillary view can be of help, but when in doubt go to CT. Images of another patient with a posterior dislocation. On the transscapular-Y view the humeral head is displaced posteriorly. Notice the distance between the humeral head and the glenoid on the AP-view, which is abnormally wide. Posterior dislocations are uncommon and not as obvious on the X-rays as an anterior dislocation. Approximately half of the posterior shoulder dislocations go undiagnosed on initial presentation, because of a low level of clinical suspicion and insufficient imaging. Posterior dislocations account for 2-4% of all shoulder dislocations. Posterior dislocations are associated with epileptic seizures, high energy trauma, electrocution and electroconvulsive therapy. This case is a posterior dislocation-fracture. The MR-images are of a patient who had undergone both an anterior aswell as a posterior dislocation. This resulted in both a Hill-Sachs impression fracture on the posterior aspect of the humeral head (blue arrow) and an impression fracture on the anterior aspect as a result of posterior dislocation (red arrow). This was an incidental finding on a chest-film. There is a superior dislocation of the humeral head. This is probably the result of a very large long-standing rotator cuff tear with progressive cranialisation of the humeral head and erosion of the acromion. Bankart-lesions and variants like Perthes and ALPSA are injuries to the anteroinferior labrum. These injuries are always located in the 3-6 o'clock position because they are caused by an anterior-inferior dislocation. The only exception to this rule is the reverse Bankart, which is the result of a posterior dislocation and injury to the inferoposterior labrum. Bankart tears may extend to superior, but this is uncommon.  Bankart lesions are labral tears without an osseus fragment. MR arthrography or arthroscopy are optimal to diagnose Bankart or Bankart-like lesions. There is a detachment of the anteroinferior labrum (3-6 o'clock) with complete tearing of the anterior scapular periosteum. The arrow points to the disrupted periosteum. On MR-athrography the labrum is missing on the anterior glenoid and the labral fragment is displaced anteriorly (arrow). Bankart lesions with an osseus fragment are common findings in patients with an anterior dislocation and are frequently seen on the x-rays or CT-scan. On MR-arthrography it may be difficult to depict the osseus fragment. On CT it is easy to appreciate the osseus fragment of the anterior glenoid (arrow). Scroll through the images. There is an osseus Bankart lesion (curved red arrow). The tear extends to superior (black arrows). There is also a Hill-Sachs defect (red arrow). Sagittal MR-arthrogram demonstrates the superior extension of the Bankart tear. Here another patient with an osseus Bankart seen on four consecutive images of a MR arthrogram in ABER-view. Notice the abnormal contour of the anterior glenoid and the avulsed anterior rim (arrow) CT-images in another patient show a reversed osseus Bankart in a patient with posterior dislocation. Axial MR-arthrogram of a reverse Bankart. Another example of a reverse Bankart. Notice the detatched labrum at the 6-9 o'clock position on the sagittal MR-arthrogram. A Perthes lesion is a labroligamentous avulsion like a Bankart, but with a medially stripped intact periosteum. On images of the shoulder with the arm in a neutral position, the torn labrum may be held in its normal anatomic position by the intact scapular periosteum, which thereby prevents contrast media from entering the tear. This means that MR-arthrography with the arm in the neutral position may fail to detect the labral tear. In the ABER position however there is tension on the antero-inferior labrum by the stretched anterior band of the inferior glenohumeral ligament and you have more chance to detect the tear. The arrow points to the intact periosteum. The images in ABER-position demonstrate a detached anterior labrum. The image on the right is rotated 90? anti-clockwise. Sometimes this makes it easier to understand the anatomy. Images of a MR-arthrogram. The image on the left shows an absent anterosuperior labrum, which is called a Buford complex. The image on the right shows a cartilage defect in the 4 o'clock position. It is not clear whether the labrum is normal. Continue with the images in ABER-position. In the ABER-position it is obvious that there is a Perthes lesion (black arrow). Due to the ABER-position the anterior band of the inferior GHL creates tension on the anteroinferior labrum and contrast fills the tear. The red arrow points to the absent labrum - Buford complex. An ALPSA-lesion is an Anterior Labral Periosteal Sleeve Avulsion. The anterior labrum is absent on the glenoid rim. The arrow points to the medially displaced labroligamentous complex. Images of a patient with an ALPSA-lesion. Notice the medially displaced labrum. Images of another patient with an ALPSA-lesion. The ABER-view shows an absent antero-inferior labrum. The coronal images shows the medially displaced labrum (red arrow). This is a difficult case. First scroll through the images and try to find out what is going on. Then continue reading. First notice the Hill-Sachs defect indicating a prior anterior dislocation (blue arrow). Now you know that you have to look for a Bankart or variant. Next notice the high signal at 12 o' clock (red arrows). On coronal images you want to make sure whether this is a variant like a labral recess or labral foramen or whether this is a SLAP. Notice how this high signal continues posteriorly, which means that it is a SLAP-lesion. The yellow arrow points to the anterior glenoid rim. The anterior labrum is absent at the 1-3 o 'clock position This is a Buford complex, which is a normal variant. The structure anterior to the glenoid is not a thorn labrum, but the middle glenohumeral ligament. Notice extention of the SLAP-tear further to posterior (red arrow). Finally there is a medially displaced inferoanterior labrum at the 3-6 o 'clock position, i.e. an ALPSA-lesion (black arrow). A GLAD-lesion is a GlenoLabral Articular Disruption. It represents a patial tear of the anteroinferior labrum with adjacent cartilage damage. The arrow points to the cartilage defect. The images show a partial tear of the anteroinferior labrum with adjacent cartilage damage at the 4-6 o 'clock position (arrows). Scroll through the images. There is a Bankart lesion with extension into the cartilage, i.e a GLAD-lesion (red arrows). HAGL is a Humeral Avulsion of the inferior Glenohumeral Ligament. There is discontinuity of the IGHL attachment on the humerus with leakage of contrast. Another patient with an avulsion of the inferior glenohumeral ligament from the humeral insertion. by Asgar M. Saleem, Joong K. Lee, Leon M. Novak AJR 2008; 191:1024-1030 by Glenn A. Tung et al AJR June 2000 vol. 174 no. 6 1707-1715 by Michel De Maeseneer et al October 2000 RadioGraphics, 20, S67-S81.Robin Smithuis and Henk Jan van der Woude Clockwise approach to labral pathology Anterior dislocation Bankart fracture Hill-Sachs Posterior dislocation Bankart lesion Osseus Bankart Reverse Bankart Perthes lesion ALPSA GLADShoulder MR - Bankart lesionsRadiology department of the Rijnland hospital, Leiderdorp and the Onze Lieve Vrouwe Gasthuis, Amsterdam, the Netherlands
msk18
1
Stress fractures
by Ferco Berger, Milko de Jonge, Robin Smithuis and Mario Maas
One of the most common injuries in sports is the stress fracture. In this review we will discuss: A stress fracture is an overuse injury. Bone is constantly attempting to remodel and repair itself, especially when extraordinary stress is applied. When enough stress is placed on the bone, it causes an imbalance between osteoclastic and osteblastic activity and a stress fracture may appear. Muscle fatigue can also play a role in the occurrence of stress fractures. For every mile a runner runs, more than 110 tons of force must be absorbed by the legs. Bones are not made to withstand so much energy on their own and the muscles act as shock absorbers. As muscles become tired and stop absorbing, all forces are transferred to the bones. Stress fractures usually occur after a recent change in training regimen has taken place. Especially professional or recreational athletes and militairy recruits are subject to change in training intensity (increased), type of training or training circumstances (new shoes, other training surface etc.) and thus at increased risk of developing a stress fracture. However, sedentary people may also develop stress fractures if suddenly an active lifestyle is adopted. Insidious onset of pain and swelling over the affected region is the most important complaint, initially during the activity. With ongoing exposure, pain will last after the training, eventually causing the athlete to stop exercising. Finally pain is experienced at rest. Stress fractures are most common in the weight-bearing bones of the lower extremity, especially the lower leg and the foot (Figure). Radiographs have a sensitivity of 15-35% for detecting stress fractures on initial examinations, increasing to 30-70% at follow up due to more overt bone reaction. Therefore, radiologists should not be comforted by negative radiographs and should initiate further state of the art imaging. Radiographs are however mandatory in order to show overt fractures and to rule out other diseases, like infections or tumours. On the left a 42-year old female who walks long distances and has been experiencing forefoot pain for a month. On the initial radiograph no fracture is seen. After 4 weeks, a follow up radiograph clearly marks callus formation at the site of the stress fracture. On the left a 28-year old female with recent onset of pain over a region of the 2nd metatarsal bone. At presentation, the radiograph was negative for fracture of the second metatarsal bone. An MRI STIR (Short TI Inversion Recovery)sequence showed a high signal intensity of bone marrow and the surrounding soft tissue, indicating bone marrow edema as a result of a stress fracture. Stress fractures radiographically show the following signs: MRI has surpassed bone scintigraphy as the imaging tool for stress fractures, showing equal sensitivity (100%) but a higher specificity (85%), probably by giving better anatomical detail and more precisely depicting the tissues involved. STIR (short tau inversion recovery), T1-weighted (T1WI) and T2-weighted images (T2WI) are used for characterization and grading. Grading is based on signs seen at MRI: On the left a 22-year old female, a professional athlete with a recent onset of forefoot pain, persisting after training. At presentation MRI showed a high signal on the STIR- and a low signal on T1WI (i.e. grade 3 stress fracture). On the left a 27-year old soccer player in the highest league of amateur football. He suffered from midfoot pain with a recent increase in complaints. T1WI shows a definite fracture line in the navicular bone, indicating a grade 4 stress fracture. Corresponding CT shows a fracture line and sclerosis on the axial images and coronal reconstructions. There are two types of stress fractures of the femoral neck: On the left we see a compression fracture of the femoral neck. The radiograph is normal, but MR depicts the fracture and bone marrow edema (i.e.grade 4). A radiograph made one month later shows evolvement to complete fracture. Although this is a low-risk fracture, the follow-up radiographs at 3 and 13 months did show poor healing tendency. The tibia is the most common location of stress fractures (more than 50%). On the left a 42-year old man with pain in his left knee. The pain had started gradually during a 10-mile running competition. The initial x-ray was reported as normal, but a T2-weigthed gradient echo of the knee shows bone marrow edema in the proximal tibia indicating the presence of a stress fracture. In retrospect, the sclerotic line on the x-ray also indicates the stress-fracture. On the left a 24-year old runner with pain in his lower leg since four months. Initially the pain was only present during running, but finally it was present even in rest. The x-ray was initially reported as normal. A bone-scan (not shown) showed a focal increase of activity. A CT-scan was performed for further differentiation and revealed a vertically oriented fissure at the insertion of the flexor digitorum longus muscle. The patient was treated with six weeks of rest, followed by a gradual increase in training-activity. On the left a 50-year old male, who led a sendentary life. He participated in a 10-mile walking contest without any training beforehand. Gradually pain developed in the lower leg and in the end he was unable to walk any further. The x-rays show a stress fracture of the lower tibia. Doing too much too soon is a common cause of stress fractures. On the left a 25-year old professional soccer player with complaints of the ankle. Evident marrow abnormalities on coronal STIR sequence MRI was seen, but there was doubt about the presence of a fracture line. At 11 months follow-up a clear fracture line is visualized by CT. On the left the postoperative radiograph with screws and lower leg cast at 12 months. It shows a just discernable fracture line at the typical location: the junction of the tibial plafond and inner vertical line of the medial malleolus Fibular fractures account for 10% of stress fractures. Stress fractures of the fibula typically occur in the distal one-third. On the left an athlete with pain just above both ankles, more pronounced on the left than on the right. Radiographs made at presentation were unremarkable. Bone scintigraphy 2 weeks later shows stress fractures of the distal fibula on both sides. The radiograph at 6 weeks follow-up (not shown) confirmed bilateral stress fractures with healing tendencies. The navicular bone is the most common site for stress fractures of the tarsus. On the left a 16-year old male athlete with a high weekly mileage. He complained of a recent onset of midfoot pain during training, lasting for several hours afterwards. There is high signal intensity in the navicular bone on the sagittal STIR-image. On the axial T1WI there is low signal intensity, but no definite fracture line. The metatarsal bones are common sites for stress fracures (25% of stress fractures). On the left a 15-year old female with no history of trauma. Recent onset of lateral forefoot pain with walking. The radiograph taken at presentation is unremarkable. Follow-up at 3 weeks shows complete fracture of the distal shaft of the 4th metatarsal with overt periosteal reaction On the left a 39-year old female with forefoot pain which began during a biking holiday. The radiograph at presentation is normal. At 1 and 3 months follow-up, clear healing tendencies can be seen, indicating the presence of a stress fracture Sesamoid bones are uncommon sites for stress fractures. On the left a 14-year old male soccer player with persistent plantar forefoot pain. Stress fracture of the medial sesamoid of the great toe is indicated by a high signal intensity on an MR sagittal STIR-sequence at presentation. A CT performed at presentation shows sclerosis of the medial sesamoid and confirms the diagnosis of stress fracture. Stress fractures can be divided into high and low risk stress fractures according to their likelihood of uncomplicated healing with conservative therapy. High Risk fracture sites: Low Risk fracture sites: Fredericson M, Bergman AG, Hoffman KL, Dillingham MS. Tibial stress reaction in runners. Correlation of clinical symptoms and scintigraphy with a new magnetic resonance imaging grading system. Am J Sports Med 1995; 23:472-481 Arendt EA, Griffiths HJ. The use of MR imaging in the assessment and clinical management of stress reactions of bone in high-performance athletes. Clin Sports Med 1997; 16:291-306 Three previously healthy persons with a stress fracture. by J.L.Bron, G.B.van Solinge, A.R.J.Langeveld, T.U.Jiya en P.I.J.M.Wuisman Ned Tijdschr Geneeskd. 2007;151:621-6 Stress fractures in the lower extremity. The importance of increasing awareness amongst radiologists. Berger, FH, de Jonge, MC and Maas, M. European Journal of Radiology 62 (2007), 16-26.Ferco Berger, Milko de Jonge, Robin Smithuis and Mario Maas Location Radiography MRI Tibia Fibula Tarsal bones Metatarsal bones Sesamoid bonesStress fracturesFrom the Radiology Department of the Academical Medical Centre, Amsterdam and the Rijnland Hospital, Leiderdorp, the Netherlands
msk19
1
US-guided injection of joints
by James Collins, Robin Smithuis and Matthieu Rutten
This article describes the application of Ultrasound guidance for diagnostic and therapeutic joint injections. Ultrasound is a valuable alternative to procedures performed either blind or under fluoroscopic or CT guidance. Anterior approach In the anterior approach the patient is lying supine with the extended arm externally rotated (figure). The transducer is placed ventrally parallel to the long axis of the subscapular tendon. The grey line on the side of the transducer indicates the long axis. Local anaesthetics are not needed if needles are used with a diameter of 21-gauge or thinner. For joint aspirations one may need to use a larger bore needle due to high viscosity of the aspirate. In such cases local anaesthetics are indicated. To facilitate injection of medication or contrast, one may use a connection tube in between the needle and the syringe, the latter being held and managed by an assistant. A 22-gauge, 50mm needle is used connected to a syringe containing the contrast media, held by an assistant who upon proper needle position injects 15-20 mL of the contrast medium. The needle is advanced perpendicular to the medial edge of the humeral head, penetrating the subscapular tendon. If one hits the cartilage of the humeral head, the needle should be pulled back 1 or 2 mm, slightly angled by about 15◦ and then advanced tangentially to the head into the joint with the bevel of the needle facing into the joint (figure). No resistance to injection should be felt and one should see the contrast flow freely into the joint and if present into the subscapular recess. A. The needle is in the intra-articular position with the tip underneath the infraspinatus tendon (ISP) and posterior labrum (L) and bordering the hyaline cartilage (asterisks) of the humeral head. B. Corresponding cadaver section showing the optimal needle track (white line). C. Sonogram after injection of 15 mL contrast. The correct intra-articular position of the needle can be visualized real-time during injection, but is also confirmed by the 'comma'-like configuration of the posterior labrum (arrowheads), which is lifted by the intraarticularly injected fluid. The subacromial-subdeltoid bursa is a synovial lined space, which contains no observable or only a minimal amount of fluid. The bursa consists of two bursal leaves. The outer and inner leaves are fused with the deltoid muscle fascia and rotator cuff, respectively. The bursal leaves can easily glide over each other, thus facilitating the range of movement of the shoulder. Blind subacromial injection of drugs into the subacromial bursa is a frequently performed therapy by general practitioners and orthopedic surgeons. The incidence of injections that miss the subacromial bursa range from 12% to 70%. For injection of the elbow the patient is supine with the arm in 90◦ flexion, raised and resting on a cushion. The joint space between the radial head and the capitulum is easily palpated. The hand is pronated or may be turned into the thumb up position, which is necessary to open the joint maximally. The transducer is placed over the joint visualizing the joint space. The needle (22 gauge, 30 mm) is directed at a slight craniocaudal angle on the dorsolateral side of the joint toward the joint space aiming toward the articular surface of the radial head. When seen to have entered the joint and upon feeling the cartilage of the radial head, the needle is slightly pulled back to ensure that the bevel is free from the cartilage and facing into the joint. 5-8 mL of contrast media is injected. No resistance to injection should be felt. The patient is supine with the wrist resting flexed over a 45◦ sponge or a rolled-up towel. In some cases it may be helpful to hold the wrist in ulnar deviation in order to open the joint space even more. The space between the radius and the scaphoid is identified on ultrasound. A 23-25-gauge, 30 mm needle is advanced under ultrasound guidance into the joint directed toward the articular surface of the radius until one feels contact with the radius. After ensuring that the tip of the needle is free from the radial cartilage 2-4 mL contrast is injected. The bevel of the needle is facing toward the joint space and the contrast is seen to flow into the joint. A linear array transducer is axially positioned dorsally over the distal radius and ulna. Along the short axis of the transducer a 23-25- gauge, 30 mm needle is inserted being directed from proximally to distally in a caudal direction. A total amount of 0.5-1 mL is injected according to rising pressure during injection. Physicians and specialists routinely perform intra-articular punctures and injections on small wrist and finger joints to relieve joint effusion or to inject drugs. The failure rate and frequency of occurrence of peri-articular injections are high: 15% - 32%, especially with the joints of the little finger and the DIP joints. Unintended peri-articular drug injection moreover may affect the surrounding ligaments or tendons, leading to serious complications. A dorsal approach using a 23-25-gauge needle is preferable. Although high-frequency linear array transducers with frequencies from 18 to 12 MHz are often used for scanning the superficial soft tissue structures of the wrist and hand, a small footprint transducer may allow better access to the small peripheral joints. Generally, 0.5-1 mL of contrast material is instilled after confir- mation of correct needle placement. The sacroiliac joint has been implicated as a source of low back and lower extremity pain, which is thought to be caused by sacroiliitis. Treatment consists of intra-articular injection of corticosteroids. Diagnostic injections or blocks are frequently performed, to distinguish between the probable causes of low back pain, because in 15-25% this is generated by the SI joint. Upper level SI joint injection The axially orientated transducer is moved from the level of the fifth lumbar vertebra caudally, depicting the dorsal surface of the sacrum with the median and lateral sacral crest, the gluteal surface of the ilium, and the first posterior sacral foramen. The needle is inserted along the short axis of the transducer into the hypoechoic cleft located between the surface of the sacrum and the contour of the ilium. Angulations of needle insertion are adjusted to the orientation of the hypoechoic cleft of the SI joint, which presents cranially a more medial to lateral orientation, and caudally a more vertical orientation. Lower level SI joint injection From the upper level the transducer is moved downward by delineation of the median and lateral sacral crest, at the dorsal surface of the sacrum and the gluteal surface of the ilium until the second posterior sacral foramen is visualized. As with the upper level, the needle is inserted into the hypoechoic cleft between the sacrum and ilium. The patient is placed supine. The leg is held in slight endorotation and abduction thereby reducing tension on the capsular structures and moving the iliopsoas tendon and bursa medially out of the intended needle path. Preferably a 5-3.5 MHz curved array transducer is used, which provides the necessary penetration depth. Usually a 21 gauge needle with a length of 9 cm is used for the average adult. In smaller adults or children a 23-gauge, 5 cm hypodermic needle might be used. Local anaesthetic may be injected prior to the main injection but this entails extra manipulation as well as non-contrast fluid (anaesthetic) in the joint leaving less room for the contrast in the limited joint space as well as possibly 2 punctures. The needle may inadvertently be withdrawn from the joint after anaesthetic injection during the switch to connect the syringe with contrast. This could be avoided by using a three-way connector between the two syringes containing the anaesthetic and the injection fluid (contrast or medication). The needle is advanced at a caudo-cranial angle along the long axis of the transducer aiming for the anterior recess near the junc- tion of the femoral neck with the femoral head (Fig. B and C). The bevel of the needle should be facing toward the joint. When the needle makes contact with the femoral head-neck junction it is slightly retracted. If one sees that it is within the joint capsule, 10-15 mL contrast or medication is injected and one sees the anterior recess swell with fluid confirming the intra-articular positioning. Indications for CT or MR arthrography of the knee are evaluation of the post-operative meniscus, query intra-articular bodies, evaluation of the stability of osteochondral lesions and evaluation of articular cartilage. One may also be requested to inject medication such as corticosteroids and/or a local anaesthetic. For injection we do not use ultrasound guidance but use the standard "blind" procedure introducing the needle (21-gauge, 50 mm) behind the patella using a lateral midpatellar approach. The patella is lateralized and the needle introduced from the mid lateral side aiming toward the centre of the patella indicated by the left forefinger. The needle is introduced horizontally aiming posterior to the centre of the patella until one makes contact with the lateral patellar facet or the lateral femoral condyle and when felt to be in the joint 40 mL contrast media is injected. Prior to CT or MR one can choose to apply a tight bandage above the patella thereby forcing contrast from the suprapatellar recess into the joint space proper. CT or MR arthrography may be used to query ligamentous, osteochondral or chondral injury, eval- uation for free bodies or query stability of ostechondral lesions. For injection of the ankle (tibiotalar joint) the patient is supine with the foot in slight plantar flexion. The medial side of the tibiotalar joint is investigated anteriorly with ultrasound to deter- mine a suitable place for injection, at the same time checking for any excessive joint fluid. We use a small curved array 8 MHz transducer but if preferred one can use an 18-12 MHz linear array transducer. The long axis of the probe is held in a sagittal plane. The needle, usually 22-gauge (length: 30 mm), is introduced in line with the long imaging axis of the transducer on the medial side of the anterior joint space, medial to the anterior tibial ligament, avoiding ligaments and vessels. One should identify the talar dome and the overhanging anterior tibial lip. The needle is angled caudo-cranially into the joint under the ventral lip of the distal tibia aiming for the articular surface of the distal tibia. Contact is felt and once again one ensures that the needle tip is free from the tibial cartilage and that the bevel is facing into the joint. 8-10 ml of contrast is injected into the tibiotalar joint and one sees the anterior capsule swells up with the fluid. There should be no resistance to injection or pain experienced by the patient. The subtalar or talocalcaneal joint is composed of 3 facets: a broad posterior facet representing the primary articulating surface, a medially located middle facet in which the sustentaculum tali articulates with the medial process of the talus, and an anterior facet. Subtalar arthrography may be performed via an anterolat- eral, posterolateral or posteromedial approach. 2-4 ml of contrast material is injected into the posterior subtalar joint. The sinus tarsi is a cone-shaped cavity that courses in a postero- medial to anterolateral direction. It is located in the lateral aspect of the foot between the neck of the talus and the anterosuperior surface of the calcaneus. The tarsal sinus continues medially as the tarsal canal, which is a funnel-shaped space between the talus and the calcaneus. It contains fat, an arterial anastomosis, joint cap- sules, nerve endings, and five ligamentous structures-the medial, intermediate, and lateral roots of the inferior extensor retinaculum; the cervical ligament; and the interosseous talocalcaneal ligament (figure). This space can be the cause of foot pain in the sinus tarsi syn- drome. The first step in treatment is infiltration of the sinus tarsi with a mixture of Depomedrol and local anaesthetic (Lidocaine). This can be challenging for the surgeon in a non-guided approach but is reasonably easily and accurately achieved with ultrasound guidance. US-guided injection of the sinus tarsi at the right-hand side with a lateral approach. The transducer is held in a coronal oblique plane. The needle is introduced along the long axis of the transducer. The sinus tarsi can easily be visualized using ultrasound. The patient turns onto the contralateral side laying the foot to be treated with its medial surface against the table top, the lateral side of the foot being uppermost. The transducer is held in a coronaloblique plane with regards to the foot. The sinus tarsi is identified as a triangular space between the anterior process of the calcaneus and the talar neck. The tip of the needle (arrow head) is seen within the cone shaped sinus tarsi, which is bordered by the talus (T) and calcaneus (C). Depending on the degree of inflammation there may be hyperemia of the space and there may be intervening vessels visible, which one wishes to avoid. This is relatively easy, especially with colour doppler by Collins JM, Smithuis R, Rutten MJ. Eur J Radiol. 2012 Oct;81(10):2759-70James Collins, Robin Smithuis and Matthieu Rutten Glenohumeral joint Posterior approach Subacromial bursa Radiocarpal joint Distal radioulnar joint (DRUJ) Carpal, carpometacarpal and interphalangeal joints Tibiotalar joint Posterior subtalar joint Sinus tarsiUS-guided injection of jointsDepartment of Radiology of the Medical Center, Leeuwarden, the Rijnland Hospital, Leiderdorp and the Jeroen Bosch Hospital, ‘s-Hertogenbosch, The Netherlands
msk20
1
Wrist - Carpal instability
by Louis A. Gilula and Ileana Chesaru
This article is based on a presentation given by Louis Gilula and adapted for the Radiology Assistant by Ileana Chesaru. First a systematic analysis of the wrist is presented to look for carpal instability and fracture dislocation. Secondly cases are presented as examples in the chapter systematic review and diagnosis. When you analyse the wrist to look for possible carpal instability and fracture dislocation, you should ask yourself the following questions: - Is there good positioning of the patient. This is essential to be able to make statements about improper alignment or abnormal axes of carpal bones. - Is there normal alignment between the carpal bones. The carpals should be parallel when profiled. Any overlap indicates abnormal tilting, dislocation or fracture. - Is there any disruption of the three carpal archs. Disruption indicates ligament tear or a fracture. - What is the shape and axis of the carpal bones. Give special attention to lunate, scaphoid and capitate. An abnormal shape indicates abnormal tilt with or without dislocation. Answering these questions will help you find clues to carpal instability, dislocation and fractures. Positioning PA view should be taken with the wrist and elbow at shoulder height. Only in this position, the radius and the ulna are parallel. Moving the arm down makes the radius cross the ulna and become relatively shorter. So it will be impossible to make any statements on the length of the ulna (plus or minus variant) Lateral view is taken with the elbow adducted to the side. Shoulder, elbow and wrist are again in one plane. This positioning will make the lateral view exactly perpendicular to the PA view. PA view A correctly positioned PA view will show the extensor carpi ulnaris groove radial to the midportion of the ulnar styloid. The PA and lateral view are equally important and thus should both be studied carefully. The PA view usually shows what is wrong and the lateral view shows in what direction the bones move. Sometimes an oblique view will also be obtained, especially if you want to look at the trapezium-trapezoid joint in profile. Lateral view Only on a good positioned lateral view one can see the volar edges of respectively scaphoid, pisiform and capitate separately and lined up as shown on the left. Lunate is the semilunar bone that fits in the distal radius. Looking through that, one can see the convexity of the scaphoid. Distally from the scaphoid is the trapezium. The angular shaped bone visible dorsally is the triquetrum. The square bone that bridges the proximal and distal half of the wrist is the pisiform. Capitate is the rounded bone fitting inside the distal lunate. Distally between the metacarpals, one can make out the hook of the hamate. Improper positioning may result in the same view of the ulna on both the PA and lateral view as shown in the case on the left. Nowhere in the body you would accept two views giving you the same image of a bone. Oblique view An oblique view is not routinely performed. It is however the only view showing the trapezio-trapezoidal joint. The joint spaces of the wrist have a width of 2 mm or less. Only the radiocarpal joint is slightly wider. The carpometacarpal joints are slightly narrower than the midcarpal joints. The capitolunate joint is considered the baseline joint width to which other joint spaces can be compared. One should make sure to look at all of them: the radiocarpal, the proximal intercarpal, the midcarpal, the distal intercarpal and the carpometacarpal joint spaces. Carpal joints should be symmetrical. Furthermore, when viewed in profile (tangentially), the cortical margins of the bones constituting that joint should be parallel. Bone edges that are not viewed in profile do not display this parallelism, e.g. the distal portion of the scaphoid that articulates with the capitate. Studying this parallelism is easier when regarding the carpal bones as pieces of a jigsaw puzzle that all fit together, as opposed to tracing carpal bones by their outer cortical margins producing the outlines (figure). When one bone is not paralleling the others, that is out of place. If the rest of the bones still parallel each other, they have stayed together. The picture on the left shows abnormal overlapping of the lunate with the capitate, hamate and triquetrum. We also see the medial profile surface of the scaphoid, but nothing paralleling it. There is also abnormal widening of the radiolunate space. The other joints are nicely parallel and symmetric. This leads to the conclusion that the lunate is displaced while the other bones have stayed together. The next step is looking at the three carpal arcs: smooth curves joining the surfaces of the carpal bones as shown on the left. The first arc is a smooth curve outlining the proximal convexities of the scaphoid, lunate and triquetrum. The second arc traces the distal concave surfaces of the same bones, and the third arc follows the main proximal curvatures of the capitate and hamate. Distruption of carpal arcs An arc is disrupted if it cannot be traced smoothly. A break in one of the arcs indicates a fracture or the disruption of a ligament leading to a subluxation or dislocation. On the left one can note the disruption of arc I at the lunotriquetral joint. A disruption of the second carpal arc at the scapholunate joint and the lunotriquetral joint is seen on the left. Although there is a gap in the first arc, it can still be traced by a smooth curve so arc I is considered intact. Disruption of the third carpal arc is shown in the next case on the left. There is an abnormal step off at the capitohamate joint. The lunate has a trapezoidal shape, as the sides converge from the proximal surface to the distal surface, which are grossly parallel. If the lunate is tilted, it becomes triangular in shape. Awareness of this fact prevents thinking the lunate might be dislocated based only on its appearance, that in fact changes with its position. So it may be dislocated with tilting or just be tilted. Common dislocations of the wrist are the lunate and perilunate dislocations. The key to differentiation between both is what is centered over the radius. If the capitate is centered over the radius and the lunate is tilted out, it is a lunate dislocation. If however the lunate centers over the distal radius and the capitate is dorsal, we are dealing with a perilunate dislocation (figure). The scaphoid shape changes with movement of the wrist. In ulnar deviation or extension the scaphoid elongates to fill the space between the radial styloid and the base of the thumb (the trapezium). Both with radial deviation aswell as flexion of the wrist the space between the radial styloid and trapezium is reduced. As scaphoid fills this space it will foreshorten and tilt towards the palm. This will give scaphoid a signet ring appearance (figure). Drawing the longitudinal axes of some of the carpal bones on a lateral radiograph and measuring the angles between them is a good method of determining the wrist bones? spatial relationship. The three most important axes are those through the scaphoid, the lunate and the capitate, drawn on the lateral radiograph. The true axis of the scaphoid is the line through the midpoints of its proximal and distal poles. Since the midpoint of the proximal pole is often difficult to appreciate, an almost parallel line can be used that is traced along the most ventral points of the proximal and distal poles of the bone (figure). The axis of the lunate runs through the midpoints of the convex proximal and concave distal joint surfaces and can best be drawn by finding the perpendicular to a line joining the distal palmar and dorsal borders of the bone as demonstrated on the left. Scapholunate angle Normal: 30 - 60? Questionably abnormal: 60 - 80? Abnormal: > 80? This indicates instability of the wrist. The capitate axis joins the midportion of the proximal convexity of the third metacarpal and that of the proximal surface of the capitate. Capitolunate angle Normal: Abnormal: > 30?.This indicates instability of the wrist. DISI is short for dorsal intercalated segmental instability. The intercalated segment is the proximal carpal row identified by the lunate. The term 'intercalated segment' refers to it being the part in between the proximal segment of the wrist consisting of the radius and the ulna and the distal segment, represented by the distal carpal row and the metacarpals. So all this means is that in DISI or dorsiflexion instability the lunate is angulated dorsally. If you think lunate is tilted, measure the scapholunate angle ( 30-60?is normal, 60-80?is questionably abnormal, >80? is abnormal) and the capitolunate angle ( In the figure on the left the scapholunate angle is measured: it is 105 degrees. As mentioned before this angle is considered abnormal if greater then 80 degrees. Volar intercalated segmental instability or palmar flexion instability is when the lunate is tilted palmarly too much. While most DISI is abnormal, in many cases VISI is a normal variant, especially if the wrist is very lax. In the next cases we advise you to first look at the images on the left and give a full description of the radiographs. Look for symmetry, parallelism, and the shape and axis of the carpal bones. Find out if there are any fractures and then try to make the diagnosis Then read the text on the right to see if you're right. Case 1 Systematic interpretation of the case on the left shows us the following: 1. On the PA-view all the carpal bones parallel each other except for the lunate. 2. The carpal arcs I and II are disrupted at the LT and SL joints. 3. Triangular shaped lunate So by just looking at the PA view we can make the diagnosis of lunate dislocation. Case 2 Analysis: 1. No parallelism at the TL joint since there is overlapping of the triquetrum and the lunate. 2. Also overlapping of the hamate and the lunate. 3. There is parallelism between radius, lunate, proximal pole of scaphoid and proximal pole of capitate. So these bones form a unit. 4. Also parallelism between triquetrum, hamate, distal pole of capitate, trapezium and distal pole of scaphoid. 5. Fracture of capitate and scaphoid So these findings indicate that this is a transscaphoid, transcapitate perilunate fracture-dislocation. On the left you see the same case with a line indicating the fracture-dislocation line. Same case with additional oblique and lateral view showing the dorsal dislocation. Case 3 Analysis: 1. Fracture of scaphoid and ulnar styloid process. 2. Broken arcs I and II at LT joint. 3. Some parallelism between lunate and proximal pole of scaphoid with the radius. 4. Scaphoid is foreshortened so it is tilted and has moved towards the palm. 5. All the other carpals show parallelism exept for lunate, the proximal pole of scaphoid and the radius. Although this probably is a perilunate dislocation, based on the PA-view alone it is very difficult to say if this is a lunate or perilunate dislocation. The triangular shape of the lunate could be the result of just tilting or dislocation with tilting. Same case with the lateral view also shown. Now we see that there definitely is a perilunate dislocation. So the triangular shape of the lunate is the result of just tilting. On the lateral view a fracture of the volar tip of lunate is seen. So this patient is at risk for recurrent dislocation. Case 4 Analysis: The case on the left shows severe arthrosis at the STT joint and CMC1 joint with subluxation. Carpal arcs are normal and there is normal paralelism.The scaphoid is elongated which means it is dorsally tilted. On the lateral view we can see that the lunate is also tilted dorsally. The proximal carpal row has moved as a unit, so there is no dissociation. Final diagnosis: non-dissociated DISI with arthrosis and subluxation of STT joints. Case 5 Analysis: 1. Loss of parallelism at LT joint resulting in broken arc I and II. 2. Lunate and scaphoid are parallel to each other but not to the other carpals. 3. Scaphoid is foreshortened due to palmar tilting. 4. Lunate is parallel to scaphoid. So the triangular shape must be the result of palmar tilting. 5. The proximal carpal row is not a unit since arc I is broken. Final diagnosis: VISI with dissociation at the LT joint. Case 6 Analysis: 1. Widened and narrowed joints, but there is normal parallelism , so there is no dislocation. 2. Scapholunate dissociation with widening of the SL joint and foreshortening of the scaphoid due to palmar tilt. 3. Arthrosis of the Radioscaphoid and Capitolunate joint due to the abnormal movements of scaphoid and lunate. 4. Dorsal tilt of lunate indicating DISI. This condition is known as SLAC. In this case it is post-traumatic due to the SL-ligament tear. SLAC (scapholunate advanced collapse) refers to a specific pattern of osteoarthritis and subluxation which results from untreated chronic scapholunate dissociation or from chronic scaphoid non-union The degenerative changes occur in areas of abnormal loading, which is the radial-scaphoid joint, followed by degeneration in the unstable lunatocapitate joint, as capitate subluxates dorsally on lunate. On the left another case of SLAC. This case is due to CPPD. Characteristics of CPPD with SLAC are: - Decreased size of proximal scaphoid due to erosion and resorption. - Scaphoid fossa erosionLouis A. Gilula and Ileana Chesaru Lunate shape Lunate vs. perilunate dislocation Scaphoid shape Scaphoid axis Lunate axis Capitate axis DISI or dorsiflexion instability VISI or volarflexion instabilityWrist - Carpal instabilityMallinckrodt Institute of Radiology Washington University St. Louis, Missouri, USA and the Westeinde hospital the Hague the Netherlands.
msk21
1
Wrist - Fractures
by Robin Smithuis
Fractures of the distal radius account for one-sixth of all fractures seen in the emergency department. The radiologist must possess an understanding of the factors that alter clinical decision making and patient treatment. In this review we will discuss: PA view should be taken with the wrist and elbow at shoulder height. This means that the wrist, elbow and shoulder are all in the transverse plane, perpendicular to the x-ray beam. Only in this position, the radius and the ulna are parallel. Lowering the arm makes the radius cross the ulna and become relatively shorter resulting in improper measurement of the length of the radius. Lateral view is taken with the elbow adducted to the side. Shoulder, elbow and wrist are again in one plane, i.e. the sagittal plane. This positioning will make the lateral view exactly perpendicular to the PA view. On a correctly positioned PA view the extensor carpi ulnaris tendon groove (arrow) can be seen. The extensor carpi ulnaris tendon groove should be at the level of or radial to the base of the ulnar styloid. A true lateral view is defined by the relationship between the pisiforme, capitate and scaphoid bones. On a standard lateral view, the palmar cortex of the pisiform bone should overlie the central third of the interval between the palmar cortices of the distal scaphoid pole and the capitate head. Apparent volar tilt of the surface of the distal radius, as measured on the lateral view, increases with supination and decreases with pronation of the wrist (5). A change of 10? rotation between two consecutive control lateral radiographs is not uncommon during clinical follow-up and results in 5? change in apparent tilt. CT should be performed if conventional radiographs provide insufficient detail about radiocarpal articular step-off and gap displacement. On the left a patient with a communitive intraarticular fracture of the distal radius with displacement of the volar rim of the radius together with the carpus (i.e. a volar Barton's). There is an axial CT image with 3D-, coronal and sagittal reconstructiosn. On the left sagittal reconstructions of 1mm axial CT slices. Scroll through the images and notice how well CT demonstrates the fracture components and the displacement. Magnetic resonance (MR) imaging is of benefit when concomitant injuries of ligaments and triangular fibrocartilage complex (TFCC) are suspected or if a fracture is suspected but not demonstrated on routine radiographs. On the left a fracture of the ulnar styloid process not visible on standard radiography, but clearly demonstrated with MR. Radial length or height Radial length is measured on the PA radiograph as the distance between one line perpendicular to the long axis of the radius passing through the distal tip of the radial styloid. A second line intersects distal articular surface of ulnar head. This measurement averages 10-13 mm. Radial inclination or angle Radial inclination represents the angle between one line connecting the radial styloid tip and the ulnar aspect of the distal radius and a second line perpendicular to the longitudinal axis of the radius. The radial inclination ranges between 21? and 25?. Loss of radial inclination will increase the load across the lunate.  Radial tilt Radial tilt is measured on a lateral radiograph. The radial tilt represents the angle between a line along the distal radial articular surface and the line perpendicular to the longitudinal axis of the radius at the joint margin. The normal volar tilt averages 11? and has a range of 2?-20?. There are many ways to describe distal radial fractures and there are several classification systems. In clinical practice however frequently eponyms like Colles' and Barton's are used. When these epomyms are used, an accurate description of the fracture characteristics should always be included in the report(5). In addition it should also be noted if there is osteoporosis or additional findings such as ligamentous injuries. We will discuss the following: One of the most important characteristics is whether a fracture is extraarticular or intraarticular. Extraarticular fractures are usually less complicated, unless they are comminutive. Intraarticular fractures either involve the radiocarpal joint, distal radioulnar joint, or both. Always mention whether the fracture is transverse (good prognosis), oblique or comminuted (multifragmented). When a fracture is oblique or when it is comminuted with crossing of the mid axial line, it can be unstable. On the left a patient with an extraarticular distal radius fracture. Notice the oblique course on the lateral view. Fractures with this configuration frequently show loss of reduction at follow up and need surgical treatment. On the left a sagittal reconstruction of an oblique intraarticular fracture of the distal radius. This is a volar Barton's type fracture. Even in a cast the volar fragment will show progressive displacement at follow-up and a volar buttress plate is needed to hold the volar rim in place. Fractures are either displaced or nondisplaced. A fracture with an offset of 2 mm or more in any plane or 2 mmm offset involving the articular surface is considered displaced. Displacement ican be dorsal, volar, radial or proximal. Axial shortening, radial inclination and radio-ulnar displacement can be measured on the routine posterior/anterior film. Dorsal tilt and dorsal or palmar displacement can be measured on the routine lateral X-ray. Fragment displacement and rotation may be further determined using CT. Instability is defined as a high risk of secondary displacement after initial adequate reduction. Radiographic signs that favor instability are displacement and an oblique or comminuted configuration (as mentioned above). These signs are listed in the table on the left. Although the initial x-ray after reduction may look good, always look for loss of reduction at follow up. Articular incongruity is the most important factor in the development of posttraumatic osteoarthritis of the wrist. Assessment of a wrist fracture must also include a description of the distal ulna and distal radioulnar joint (9). The distal ulna articulates with the sigmoid notch of the radius. Type I: stable Avulsion fractures of the tip of the ulnar styloid and stable fractures of the ulnar neck have a good prognosis. Following reduction of the radius the DRUJ is congruent and stable. Extraarticular unstable fractures however, require plate fixation. Type II: unstable There is subluxation or dislocation of the ulnar head as a result of avulsion of the base of the ulnar styloid or tear of the TFCC and/or capsular ligaments. The subluxation has to be reduced with closed or operative treatment to avoid chronic instability and arthosis. Type III: potentially unstable Intraarticular fractures of the sigmoid notch and intraarticular fractures of the ulnar head are potentially unstable because the incongruity of the DRUJ. Subluxation is possible. A Colles' fracture is a fracture of the distal metaphysis of the radius with dorsal angulation and displacement leading to a 'silver fork deformity'. Colles fractures are seen more frequently with advancing age and in women with osteoporosis. In many cases a Colles' fracture is an extraarticular, uncomplicated and stable fracture, but it can be intraarticular. So look for signs of instability in all Colles' fractures, especially: On the left a detailed AP view of the same patient as above. In addition to the dorsal angulation seen on the lateral view, notice the following: Just calling this fracture a Colles' fracture would be insufficient. All the characteristics have to be mentioned in the radiology report to convey the full extent of the injury, possible complications and treatment. Smith's fractures occur in younger patients and are the result of high energy trauma on the volar flexed wrist. Volar comminution and intraarticular extension are more common. On the left an extraarticular Smith's fracture with palmar and radial angulation and displacement. There is also an avulsion of the ulnar styloid process. Volar-type Barton's is a fracture-dislocation of the volar rim of the radius. This type is the most common. Dorsal-type Barton's is a fracture-dislocation of the dorsal rim of the radius. Dislocation of the radiocarpal joint is the hallmark of Barton's fractures. These are shear type fractures of the distal articular surface of the radius with translation of the distal radial fragment and the carpus. These fractures have a great tendency for redislocation and malunion. They usually require operative treatment. On the left a volar-type Barton's fracture. The radiographic findings are the following: On the left a dorsal-type Barton's fracture. The radiographic findings are the following: A die-punch fracture is a depression fracture of the lunate fossa of the distal radius. It is the result of a transverse load through the lunate. The radiographic findings can be very subtle. In many cases there is also a subtle proximal displacement of lunate, seen as a break in carpal arc I. (see the article Wrist - Carpal instability). On the left a typical die-punch fracture. The blue arrow indicates the depressed fragment of the lunate fossa. Notice the articular step-off. The yellow arrow indicates a subtle fracture of the radial styloid process. There is no disruption of carpal arc I. Notice that you can easily overlook such a fracture. On the left two 3D-reconstructions of the same fracture as above. An isolated fracture of the radial styloid process is also called a Hutchinson's or chauffeur's fracture. Displacement of the fragment is uncommon. There can be associated injury to the scapholunate ligament. In most cases a fracture of the radial styloid process is part of a comminutive intraarticular fracture. Ulnar styloid process fracture An ulnar styloid process fracture is usually associated with radial fractures and rarely isolated. An isolated fracture of the tip is clinically insignificant. Displaced fractures of the base are usually associated with TFC tears and can be associated with instability of the distal radioulnar joint (DRUJ). On the left a subtle fracture of the tip of the ulnar styloid process (blue arrow) in a patient with a volar Barton's fracture. Notice the depression of the volar rim. Torus fractures, or buckle fractures, are extremely common injuries in children. Because children have softer bones, one side of the bone may buckle. The word torus is derived from the Latin word 'Tori' meaning swelling or protuberance. These injuries tend to heal much more quickly than the similar greenstick fractures. These are partial fractures, since only one part of the bone is broken and the other side is bent. The name is derived from an analogy of breaking a young, fresh tree branch. Most often the greenstick fracture must be bent back into the proper position. Greenstick fractures can take a long time to heal because they tend to occur in the middle, more slowly growing parts of bone. These are usually Salter Harris type II epiphysiolysis fractures. Restorage of the anatomical situation is necessary to prevent growth disturbances. Redislocation is common after closed reduction. In many cases they need percutaneous pinning. The M?ller AO-classification is adapted by the Orthopaedic Trauma Association. In reference (6) a link is provided to download the illustrations of the M?ller AO Classification of Fractures. A = extra-articular fracture B = partial articular fracture C = complete articular fracture of radius This classification is popular, since it addresses the mechanism of injury and the consequent treatment options. Instability is defined as a high risk of secondary displacement after initial adequate reduction. Associated traumatic lesions are ligamentous rupture, nerve compression and compartment syndrome. Type 1 : Bending fracture Type 2 : Shearing fracture Type 3 : Compression fracture Type 4 : Avulsion fracture Type 5 : Combined fractures The treatment decision of a distal radius fracture is complex and depends on the type of the fracture, the age and activeness of the patient and the quality of the bone. At one extreme a stable, undisplaced extra-articular fracture has an excellent prognosis. At the other an unstable, displaced intra-articular fracture is difficult to treat and has a poor prognosis. If the alignment of the bones is not acceptable, they need to be reduced by closed or open reduction. Many authors suggest that distal radial fractures be reduced anatomically, but the real question is 'what is acceptable and what is not?'. Based on basic science and clincal studies some of the recommendations of the International Distal Radius Fracture Study Group are presented in the table on the left, although these recommendations are still the subject of ongoing debate (5). The initial treatment for most radius fractures is closed reduction and plaster immobilization. A displaced fracture is reduced under regional or general anaesthetic. First the arm is placed under traction to unlock the fragments. The deformity is then reduced with appropriate closed reduction, depending on the type of deformity. A splint or cast is placed in such a way that the risk of re-displacement is minimized. X-rays are taken to ensure that the reduction was successful. The cast is usually maintained for about 6 weeks. Guidelines for non-acceptable reduction are (8): On the left a control radiograph made after reduction. The reduction was unsuccessful, because there is a dorsal tilt > 10?, loss of inclination and radial shortening. Although in most cases closed reduction is attempted, surgical intervention is required when there is failure to obtain or maintain closed reduction. 40% of distal radial fractures are considered to be unstable and require surgical fixation. Many techniques of fixation are now available, including percutaneous pinning, intramedullary pinning, external fixation, and internal fixation with customized implants, including the Distal Volar Radius (DVR) system. Surgical fixation allows almost immediate mobility. Ultimately less stiffness and greater function is possible. On the left a post-operative image of a Salter-Harris II fracture, which is held in place with two pins after closed reduction. On the left a patient with a die-punch fracture, nicely shown on an oblique radiograph. The fracture fragment of the lunate fossa was replaced and fixated with a screw. Volar buttress plate Barton's fractures are rarely successfully treated with closed reduction due to the shearing nature of the injury. A volar buttress plate is the treatment of choice. Comminution or osteoporotic bone make external fixation the preferred surgical treatment option. On the left an intraarticular fracture of the distal radius with shortening of the radius. The ulna abutts the lunate. External fixation was used to lengthen the radius. On the left a patient with a dorsal Barton's fracture (shown before). After closed reduction the position of the dorsal rim is better, but this still is an unstable situation. Volar plates were used with screws to lock the dorsal rim. The volar approach was chosen, because this is an easier approach. Final result after one of the plates has been removed. Non-union is uncommon in distal radial fractures, since there is excellent vascularisation of this region. Malunion however is a common complication and is related to radial shortening, angulation and incongruity of the articular surface. This results in malfunction and early osteoarthritis. More than 2 mm incongruity of articular surface is the most important factor in the development of posttraumatic osteoarthritis of the wrist. On the left a patient with malunion. The radial shortening results in the ulna abutting the lunate. Notice the loss of radiocarpal joint space indicating osteoarthritis. Closed reduction is frequently unsuccessful when the fracture has an oblique course or when the fracture is comminutive. On the left a patient with an intraarticular fracture with dorsal tilt (i.e. intraarticular Colles' fracture). On the lateral radiograph at presentation there is an extreme dorsal tilt. After closed reduction and at follow up after one week, there is an acceptable tilt. Finally at 6 weeks follow-up, there is malunion with extreme dorsal tilt, radial shortening and loss of inclination. The ulna abutts the lunate. The final result will be malfunction, radiocarpal and distal radioulnar osteoarthritis. Risks specific to cast treatment relate to the potential for compression of the swollen arm causing compartment syndrome or carpal tunnel syndrome. Reflex sympathetic dystrophy and median nerve injury are uncommon complications. Complications associated with plating include tendon irritation or rupture and the need for plate removal. On the left another patient with malunion and osteoarthritis. There is also scapholunate dissociation as a result of associated ligamentous rupture with volar tilt of lunate indicating volar flexion instability (VISI). On the left another patient after unsuccessful treatment. There is loss of radial inclination and radial shortening, dorsal tilt and an articular step-off. by Charles A. Goldfarb, MD, Yuming Yin, MD, Louis A. Gilula, MD, Andrew J. Fisher, MD and Martin I. Boyer, MD Radiology. 2001;219:11-28. by Kevin C. Chung et al The Journal of Bone and Joint Surgery (American). 2006;88:2687-2694. by Marco Zanetti, MD, Louis A. Gilula, MD, Hilaire A. C. Jacob, PhD and Juerg Hodler, MD Radiology. 2001;220:594-600 A Report by the IFSSH BONE AND JOINT COMMITTEE Download the illustrations as packages: Radius/Ulna: ZIP File (351 KB) A PROSPECTIVE, RANDOMISED STUDY OF IMMOBILISATION IN A CAST VERSUS SUPPLEMENTARY PERCUTANEOUS PINNING by T. Azzopardi et al. J Bone Joint Surg Br 2005; 87-B: 837-840 by David L. Nelson, MD of the International distal radius fracture study group by Diego Fernandez, Jesse Jupiter Springer, New York, Second Edition, 2002, ISBN 0-387-95195-4.Robin Smithuis Positioning Measurements Location Configuration Displacement Instability Ulna and Distal radioulnar joint (DRUJ) Colles' fracture Smith's fracture Barton's fracture Die-punch fracture Chauffeur's fracture Torus fracture Green stick fracture Epiphysiolysis fracture M?ller AO-classification Fernandez Classification Indications for Reduction in Distal Radius Fractures Closed Reduction Surgical treatment MalunionWrist - FracturesRadiology department of the Rijnland Hospital in Leiderdorp, the Netherlands
neuro1
1
Brain Anatomy
by Robin Smithuis
Scroll through the images on the left.  On the left a coronal view of the segments of the middle cerebral artery. The anterior commissure is a bundle of white fibers that connects the two cerebral hemispheres across the middle line. At this level frequently perivascular CSF-spaces of Virchow-Robin are seen. At this level the basal ganglia are seen. The two dark lines medially of the thalamus are the internal cerebral veins. On the left a coronal illustration of the anatomy of the pituitary gland and the surrounding structures. On the left a coronal illustration of the area of the hippocampus. Neuroimaging Primer - Harvard Medical School lecture notes: Introduction to Neuroimaging by Keith Johnson and Alex Becker Developed by Jeffrey E. Zapawa and Anthony L. Alcantara, MDRobin Smithuis Circle of Willis Anterior commissure Thalamic level Pituitary gland HippocampusBrain AnatomyRadiology department, Rijnland Hospital Leiderdorp, the Netherlands.
neuro2
1
Brain Ischemia - Imaging in Acute Stroke
by Majda Thurnher
This review is based on a presentation given by Majda Thurnher and was adapted for the Radiology Assistant by Robin Smithuis. We will discuss the following subjects: The goal of imaging in a patient with acute stroke is: In this way we can select patients who are candidates for thrombolytic therapy. CT has the advantage of being available 24 hours a day and is the gold standard for hemorrhage. Hemorrhage on MR images can be quite confusing. On CT 60% of infarcts are seen within 3-6 hrs and virtually all are seen in 24 hours. The overall sensitivity of CT to diagnose stroke is 64% and the specificity is 85%. In the table on the left the early CT-signs of cerebral infarction are listed. The reason we see ischemia on CT is that in ischemia cytotoxic edema develops as a result of failure of the ion-pumps. These fail due to an inadequate supply of ATP. An increase of brain water content by 1% will result in a CT attenuation decrease of 2.5 HU. On the left a patient with hypoattenuating brain tissue in the right hemisphere. The diagnosis is infarction, because of the location (vascular territory of the middle cerebral artery (MCA) and because of the involvement of gray and white matter, which is also very typical for infarction. Hypoattenuation on CT is highly specific for irreversible ischemic brain damage if it is detected within first 6 hours (1). Patients who present with symptoms of stroke and who demonstrate hypodensity on CT within first six hours were proven to have larger infarct volumes, more severe symptoms, less favorable clinical courses and they even have a higher risk of hemorrhage. Therefore whenever you see hypodensity in a patient with stroke this means bad news. No hypodensity on CT is a good sign. Obscuration of the lentiform nucleus, also called blurred basal ganglia, is an important sign of infarction. It is seen in middle cerebral artery infarction and is one of the earliest and most frequently seen signs (2). The basal ganglia are almost always involved in MCA-infarction. This refers to hypodensity and swelling of the insular cortex. It is a very indicative and subtle early CT-sign of infarction in the territory of the middle cerebral artery. This region is very sensitive to ischemia because it is the furthest removed from collateral flow. It has to be differentiated from herpes encephalitis. This is a result of thrombus or embolus in the MCA. On the left a patient with a dense MCA sign. On CT-angiography occlusion of the MCA is visible. 15% of MCA infarcts are initially hemorrhagic. Hemorrhage is most easily detected with CT, but it can also be visualized with gradient echo MR-sequences. Once you have diagnosed the infarction, you want to know which vessel is involved by performing a CTA. First look at the images on the left and try to detect the abnormality. Then continue reading. The findings in this case are very subtle. There is some hypodensity in the insular cortex on the right, which is the area we always look at first. In this case it is suggestive for infarction, but sometimes in older patients with leukencephalopathy it can be very difficult. A CTA was performed (see next images). Now we feel very comfortable with the diagnosis of MCA infarction. CT Perfusion (CTP) With CT and MR-diffusion we can get a good impression of the area that is infarcted, but we cannot preclude a large ischemic penumbra (tissue at risk). With perfusion studies we monitor the first pass of an iodinated contrast agent bolus through the cerebral vasculature. Perfusion will tell us which area is at risk. Approximately 26% of patients will require a perfusion study to come to the proper diagnosis. The limitation of CT-perfusion is the limited coverage. Studies were performed to compare CT with MRI to see how much time it took to perform all the CT studies that were necessary to come to a diagnosis. It was demonstrated that Plain CT, CTP and CTA can provide comprehensive diagnostic information in less than 15 minutes, provided that you have a good team. In the case on the left first a non-enhanced CT was performed. If there is hemorrhage, then no further studies are necessary. In this case the CT was normal and a CTP was performed, which demonstrated a perfusion defect. A CTA was subsequently performed and a dissection of the left internal carotid was demonstrated. On PD/T2WI and FLAIR infarction is seen as high SI. These sequences detect 80% of infarctions before 24 hours. They may be negative up to 2-4 hours post-ictus! On the left T2WI and FLAIR demonstrating hyperintensity in the territory of the middle cerebral artery. Notice the involvement of the lentiform nucleus and insular cortex. High signal on conventional MR-sequences is comparable to hypodensity on CT. It is the result of irreversible injury with cell death. So hyperintensity means BAD news: dead brain. DWI is the most sensitive sequence for stroke imaging. DWI is sensitive to restriction of Brownian motion of extracellular water due to imbalance caused by cytotoxic edema. Normally water protons have the ability to diffuse extracellularly and loose signal. High intensity on DWI indicates restriction of the ability of water protons to diffuse extracellularly. First look at the images on the left and try to detect the abnormality. Then continue reading. The findings in this case are very subtle. There is some hypodensity and swelling in the left frontal region with effacement of sulci compared with the contralateral side. You probably only notice these findings because this is an article about stroke and you would normally read this as 'no infarction'. Now continue with the DWI images of this patient. When we look at the DWI-images it is very easy and you don't have to be an expert radiologist to notice the infarction. This is why DWI is called 'the stroke sequence'. When we compare the findings on T2WI and DWI in time we will notice the following: Pseudo-normalization of DWI This occurs between 10-15 days. The case on the left shows a normal DWI. On T2WI there is may be some subtle hyperintensity in the right occipital lobe in the vascular territory of the posterior cerebral artery. The T1WI after the administration of Gadolinium shows gyral enhancement indicating infarction. First it was thought that everything that is bright on DWI is dead tissue. However now there are some papers suggesting that probably some of it may be potentially reversible damage. If you compare the DWI images in the acute phase with the T2WI in the chronic phase, you will notice that the affected brain volume in DWI is larger compared to the final infarcted area (respectively 62cc and 17cc). Perfusion with MR is comparable to perfusion CT. A compact bolus of Gd-DTPA is delivered through a power injector. Multiple echo-planar images are made with a high temporal resolution. T2* gradient sequences are used to maximize the susceptibility signal changes. The area with abnormal perfusion can be dead tissue or tissue at risk. Combining the diffusion and perfusion images helps us to define the tissue at risk, i.e. the penumbra. On the left we first have a diffusion image indicating the area with irreversible changes (dead issue). In the middle there is a large area with hypoperfusion. On the right the diffusion-perfusion mismatch is indicated in blue. This is the tissue at risk. This is the brain tissue that maybe can be saved with therapy. On the left a patient with sudden onset of neurological symptoms. MR was performed 1 hour after onset of symptoms. First look at the images on the left and try to detect the abnormality. Then continue reading. These images are normal and we have to continue with DWI. See next images. On the DWI there is a large area with restricted diffusion in the territory of the right middle cerebral artery. Notice also the involvement of the basal ganglia. There is a perfect match with the perfusion images, so this patient should not undergo any form of thrombolytic therapy. On the left another MCA infarction. It is clearly visible on CT (i.e. irreversible changes). There is a match of DWI and Perfusion, so no therapy. On the left another case. The DWI and ADC map is shown. Continue for the perfusion images Now we can see that there is a severe mismatch. Almost the whole left cerebral hemisphere is at risk due to hypoperfusion. This patient is an ideal candidate for therapy. by R von Kummer et al. Radiology 1997, Vol 205, 327-333, by N Tomura et al Radiology 1988, Vol 168, 463-467 by Ashok Srinivasan et al RadioGraphics 2006;26:S75-S95Majda Thurnher Hypo attenuating brain tissue Obscuration of the lentiform nucleus Insular Ribbon sign Dense MCA sign Hemorrhagic infarcts Diffusion Weighted Imaging (DWI) Perfusion MR ImagingBrain Ischemia - Imaging in Acute StrokeDepartment of Radiology, Medical University of Vienna
neuro3
1
Brain Ischemia - Vascular territories
by Robin Smithuis
Knowledge of the vascular territories is important, because it enables you to recognize infarctions in arterial territories, in watershed regions and also venous infarctions. It also helps you to differentiate infarction from other pathology. On most images you can click to get an enlarged view, but this does not work on the iPhone application.  On the left a detail to illustrate the vascular supply to the basal ganglia. On the left CT-images of a left-sided PICA-infarction. Notice the posterior extention. The infarction was the result of a dissection (blue arrow). On the left MR-images of a left-sided PICA-infarction. In unilateral infarcts there is always a sharp delineation in the midline because the superior vermian branches do not cross the midline, but have a sagittal course. This sharp delineation may not be evident until the late phase of infarction. In the early phase, edema may cross the midline and create diagnostic difficulties. Infarctions at pontine level are usually paramedian and sharply defined because the branches of the basilar arery have a sagittal course and do not cross the midline. Bilateral infarcts are rarely observed because these patients do not survive long enough to be studied, but sometimes small bilateral infarcts can be seen. On the left MR-image of a cerebellar infarction in the region of the superior cerebellar artery and also in the brainstem in the territory of the PCA. Notice the limitation to the midline. Anterior cerebral artery: The territory of the anterior choroidal artery encompasses part of the hippocampus, the posterior limb of the internal capsule and extends upwards to an area lateral to the posterior part of the cella media. The whole area is rarely involved in AChA infarcts. On the left an uncommon infarction in the hippocampal region. Part of the territory of the anterior choroidal artery and the PCA are involved. The MCA has cortical branches and deep penetrating branches, which are called the lateral lenticulo-striate arteries. The territory of the lateral lenticulo-striate perforating arteries of the MCA is indicated with a different color from the rest of the territory of the MCA because it is a well-defined area supplied by penetrating branches, which may be involved or spared in infarcts separately from the main cortical territory of the MCA. On the left a T2W-image of a patient with an infarction in the territory of the middle cerebral artery (MCA). Notice that the lateral lenticulo-striate perforating arteries of the MCA are also involved (orange arrow). Medial lenticulostriate arteries Branches of the A1-segment of the anterior cerebral artery. They supply the anterior inferior parts of the basal nuclei and the anterior limb of the internal capsule. Lateral lenticulostriate arteries Branches of the horizontal M1-segment of the middle cerebral artery. They supply the superior part of the head and the body of the caudate nucleus, most of the globus pallidus and putamen and the posterior limb of the internal capsule. On the left images of a hemorrhagic infarction in the area of the deep perforating lenticulostriate branches of the MCA. On the left enhanced CT-images of a patient with an infarction in the territory of the middle cerebral artery (MCA). There is extensive gyral enhancement (luxury perfusion). Sometimes this luxury perfusion may lead to confusion with tumoral enhancement. Deep or proximal PCA strokes cause ischemia in the thalamus and/or midbrain, as well as in the cortex. Superficial or distal PCA infarctions involve only cortical structures (4). On the left a patient with acute vision loss in the right half of the visual field. The CT demonstrates an infarction in the contralateral visual cortex, i.e left occipital lobe. Only about 5% of ischemic strokes involve the PCA or its branches (3). On the left CT-images of a patient with a PCA-infarction. Notice the loss of gray/white matter differentiation in the regio of the left occipital lobe. Variations in perfusion territories in the brain can be visualized with selective arterial spin-labeling (9). The ability to visualize these perfusion territories is important in specific patient groups with cerebrovascular disease, such as acute stroke, large artery steno-occlusive disease, and arteriovenous malformation, as it provides valuable hemodynamic information. On the left the time-of-flight MR angiography-images of brain-feeding arteries showing the planning of the selective slabs for perfusion territory imaging of the left and right internal carotid artery and the vertebrobasilar artery. On the left a patient with a lacunar infarction on the left with normal perfusion territories. On the left a patient with a watershed infarct in the left hemisphere and also a cortical infarction in the left frontal lobe (arrow). Notice that there is a variation in the brain perfusion since the left frontal lobe is supplied by the right internal carotid artery. On the left another variation in the brain perfusion in a patient with multiple infarctions as demonstrated on the diffusion images. There is a small cortical infarction in the left occipital lobe which happens to be perfused by the left internal carotid artery (arrow). Notice that there is no contribution by the vertebrobasilar arteries. Watershed infarcts occur at the border zones between major cerebral arterial territories as a result of hypoperfusion. There are two patterns of border zone infarcts: On the left three consecutive CT-images of a patient with an occlusion of the right internal carotid artery. The hypoperfusion in the right hemisphere resulted in multiple internal border zone infarctions. This pattern of deep watershed infarction is quite common and should urge you to examine the carotids. On the left images of a patient who has small infarctions in the right hemisphere in the deep borderzone (blue arrowheads) and also in the cortical borderzone between the MCA- and PCA-territory (yellow arrows). There is abnormal signal in the right carotid (red arrow) as a result of occlusion. In patients with abnormalities that may indicate borderzone infarcts, always study the images of the carotid artery to look for abnormal signal. On the left another example of small infarctions in the deep borderzone and in the cortical borderzone between the MCA- and PCA-territory in the left hemisphere. On the left an example of infarctions in the deep borderzone and in the cortical borderzone between the ACA- and MCA-territory. The abnormal signal intensity in the right carotid is the result of an occlusion. This combination of findings is so common, that once you know the pattern, you will see it many times. Lacunar infarcts are small infarcts in the deeper parts of the brain (basal ganglia, thalamus, white matter) and in the brain stem. Lacunar infarcts are caused by occlusion of a single deep penetrating artery. Lacunar infarcts account for 25% of all ischemic strokes. Atherosclerosis is the most common cause of lacunar infarcts followed by emboli. 25% of patients with clinical and radiologically defined lacunes had a potential cardiac cause for their strokes. On the left a T2W- and FLAIR image of a lacunar infarct in the left thalamus. On the FLAIR image the infarct is hardly seen. There is only a small area of subtle hyperintensity. Lacunes may be confused with other empty spaces, such as enlarged perivascular Virchow-Robin spaces (VRS). The VRS are extensions of the subarachnoid space that accompany vessels entering the brain parenchyma. Widening of VRS often first occurs around penetrating arteries in the substantia perforata and can be seen on transverse MRI slices around the anterior commisure, even in young subjects (5). On the left CT- and MR-images at the level of the anterior commisure (blue arrows). On the CT there is a hypodense area in the right hemisphere, which follows the signal intensity of CSF on T2W- and FLAIR-images, which is typical for widened VRS. PRES is short for Posterior Reversible Encephalopathy Syndrome. It is also known as reversible posterior Leukoencephalopathy syndrome [RPLS]. It classically consists of potentially reversible vasogenic edema in the posterior circulation territories, but anterior circulation structures can also be involved (6). Many causes have been described including hypertension, eclampsia and preeclampsia, immunosuppressive medications such as cyclosporine. The mechanism is not entirely understood but is thought to be related to a hyperperfusion state, with blood-brain-barrier breakthrough, extravasation of fluid potentially containing blood or macromolecules, and resulting cortical or subcortical edema. The typical imaging findings of PRES are most apparent as hyperintensity on FLAIR images in the parietooccipital and posterior frontal cortical and subcortical white matter; less commonly, the brainstem, basal ganglia, and cerebellum are involved. On the left images of a patient with reversible neurological symptoms. The abnormalities are seen both in the posterior circulation as well as in the basal ganglia. Continue. Four days later most of the abnormalities have disappeared. There is great variation in the territories of venous drainage. The illustrations on the left should be regarded as a rough guide. Cerebral venous thrombosis results from occlusion of a venous sinus and/or cortical vein and usually is caused by a partial thrombus or an extrinsic compression that subsequently progresses to complete occlusion (7). Dehydration, pregnancy, a hypercoagulable state and adjacent infection (eg, mastoiditis) are predisposing factors. Cerebral venous thrombosis is an elusive diagnosis because of its nonspecific presentation. It often presents with hemorrhagic infarction in areas atypical for arterial vascular distribution. Imaging plays a key role in the diagnosis. On the far left a MRA with non-visualization of the left transverse sinus. Since the venous anatomy is variable, this can be due to absence of the transverse sinus or thrombosis. The T1W-image on the right clearly demonstrates, that there is a transverse sinus on the left, so the MRA findings are due to thrombosis. Continue with next images. On the left the CT nicely demonstrates the dense thrombosed transverse sinus (yellow arrow). The FLAIR image demonstrates the venous infarction in the temporal lobe. Thrombosis of deep cerebral veins The clinical presentation of thrombosis of the deep cerebral venous system are severe dysfunction of the diencephalon, reflected by coma and disturbances of eye movements and pupillary reflexes. Usually this results in a poor outcome. However, partial syndromes without a decrease in the level of consciousness or brainstem signs exist, which may lead to initial misdiagnoses. Deep cerebral venous system thrombosis is an underdiagnosed condition when symptoms are mild and should be suspected if the patient is a young woman, if the lesions are within the basal ganglia or thalamus and especially if they are bilateral. On the left images of a patient with deep cerebral vein thrombosis. Notice the bilateral infarctions in the basal ganglia. Continue. There is absence of flow void in the internal cerebral veins, sinus rectus and right transverse sinus (blue arrows). On the MRA the right transverse sinus is not visualized. The vascular territories of the carotid and vertebrobasilar systems. Diagrams based on CT studies of infarcts. by Savoiardo M. Ital J Neurol Sci. 1986 Aug;7(4):405-9. by Savoiardo M, Bracchi M, Passerini A, Visciani A. AJNR Am J Neuroradiol. 1987 Mar-Apr;8(2):199-209. by Colin P. Derdeyn et al Radiology. 2001;220:195-201 by Michael D Hill in eMedicine by Frederik Barkhof Journal of Neurology Neurosurgery and Psychiatry 2004;75:1516-1517 by Alexander M. McKinney et al. AJR 2007; 189:904-912 in eMedicine by Mahesh R Patel by Walter M. van den Bergh NEUROLOGY 2005;65:192-196 by Peter Jan van Laar, Jeroen van der Grond and Jeroen Hendrikse February 2008 Radiology, 246, 354-364.Robin Smithuis PICA SCA ACA Anterior choroidal artery Middle cerebral artery Lenticulostriate arteries Posterior cerebral artery (PCA) Cerebral venous thrombosisBrain Ischemia - Vascular territoriesRadiology department of the Rijnland Hospital in Leiderdorp, the Netherlands
neuro4
1
Brain Tumor - Systematic Approach
by Robin Smithuis and Walter Montanera
This review is based on a presentation given by Walter Montanera and was adapted for the Radiology Assistant by Robin Smithuis. In this review a systematic approach for the analysis of a possible brain tumor is described. By clicking on one of the subjects in the list on the left, you will go directly to this item. by Robin Smithuis and Walter Montanera When we analyze a potential brain tumor, there are many questions that need to be answered. Since different tumors occur in different age groups we first of all need to know the age of the patient. Next we need to know where the lesion is located - is it intra- or extra-axial and in what anatomical compartment does it lie? Is it located in the sellar or pontocerebellar region for example? Is it a solitary mass or is there multi-focal disease? On CT and MR we look for tissue characteristics like calcifications, fat, cystic components, contrast enhancement and signal intensity on T1WI, T2WI and DWI. Most brain tumors are of low signal intensity on T1WI and high on T2WI. Therefore high signal intensity on T1WI or low signal on T2WI can be an important clue to the diagnosis. Finally we have to consider the possibility that we are dealing with a lesion that simulates a tumor - like an abscess, MS-plaque, vascular malformation, aneurysm or an infarct with luxury perfusion. Roughly one-third of CNS tumors are metastatic lesions, one third are gliomas and one-third is of non-glial origin. Glioma is a non-specific term indicating that the tumor originates from glial cells like astrocytes, oligodendrocytes, ependymal and choroid plexus cells. Astrocytoma is the most common glioma and can be subdivided into the low-grade pilocytic type, the intermediate anaplastic type and the high grade malignant glioblastoma multiforme (GBM). GBM is the most common type (50% of all astrocytomas). The non-glial cel tumors are a large heterogenous group of tumors of which meningioma is the most common. The age of the patient is an important factor for the differential diagnosis. Specific tumors occur under the age of 2, like choroid plexus papillomas, anaplastic astrocytomas and teratomas. In the first decade medulloblastomas, astrocytomas, ependymomas, craniopharyngeomas and gliomas are most common, while metastases are very rare. When they do occur at this age, metastases of a neuroblastoma are the most frequent. In adults about 50% of all CNS lesions are metastases. Other common tumors in adults are astrocytomas, glioblastoma multiforme, meningiomas, oligodendrogliomas, pituitary adenomas and schwannomas. Astrocytomas occur at any age, but glioblastoma multiforme is mostly seen in older people. Although cancer is rare in children, brain tumors are the most common type of childhood cancer after leukemia and lymphoma. Most of the tumors in children are located infratentorially. The most common supra- and infratentorial tumors are listed in the table on the left. The most common tumors in adults are listed in the table on the left. Note that metastases are by far the most common. It is important to realize that 50% of metastases are solitary. Particularly in the posterior fossa, metastases should be in the top 3 of the differential diagnostic list. Hemangioblastoma is an uncommon tumor, but it is the most common primary intra-axial tumor in the adult. Supratentorially, metastases are also the most common tumors, followed by gliomas. When we study an intracranial mass, the first thing we want to know is whether the mass lies in- or outside of the brain. If it is outside the brain or extra-axial, then the lesion is not actually a brain tumor, but derived from the lining of the brain or surrounding structures. Eighty percent of these extra-axial lesions will be either a meningioma or a schwannoma. On the other hand, in an adult an intra-axial tumor will be a metastasis or astrocytoma in 75% of cases. The T2W-images show a schwannoma located in the cerebellopontine angle (CPA). This case nicely demonstrates the typical signs of an extra-axial tumor. There is a CSF cleft (yellow arrow). The subarachnoid vessels that run on the surface of the brain are displaced by the lesion (blue arrow). There is gray matter between the lesion and the white matter (curved red arrow). The subarachnoid space is widened because growth of an extra-axial lesion tends to push away the brain. All these signs indicate that this is a typical extra-axial tumor. In the region of the CPA 90% of the extra-axial tumors are schwannomas. Another sign of an extra-axial origin is a broad dural base or a dural tail of enhancement as is typically seen in meningiomas. This may also occur in other extra-axial tumors, but it is less common. Another sign of an extra-axial origin are bony changes. Bony changes are seen in bone tumors like chordomas, chondrosarcomas and metastases. They can also be secondary, as is seen in meningiomas and other tumors. On the left an example of a meningioma with a broad dural base and a dural tail of enhancement. There is hyperostosis in the adjacent bone and the lesion enhances homogeneously. Extra-axial tumors are not derived from brain tissue and do not have a blood-brain-barrier, so most of them enhance homogeneously. Intra- vs Extra-axial (2) The differentiation between intra-axial versus extra-axial is usually straight forward, but sometimes it can be very difficult and imaging in multiple planes may be necessary. The tumor in the case on the left was thought to be a falcine meningioma, i.e. extra-axial and was presented for surgery. This lesion surely has the appearance of a meningioma: these tumors can be hypointense on T2 due to a fibrocollageneous matrix or calcifications and frequently produce reactive edema in the adjacent white matter of the brain. However, there is gray matter on the anteromedial and posteromedial side of the lesion (red arrow). This indicates that the lesion is intra-axial. If the lesion was extra-axial the gray matter should have been pushed away. This proved to be a melanoma metastasis. Local tumor spread (1) Astrocytomas spread along the white matter tracts and do not respect the bounderies of the lobes. Because of this infiltrative growth, in many cases the tumor is actually larger than can be depicted with MR. Ependymomas of the fourth ventricle in children tend to extend through the foramen of Magendie to the cisterna magna and through the lateral foramina of Luschka to the cerebellopontine angle (figure). Oligodendrogliomas typically show extension to the cortex. Subarachnoid seeding Some tumors show subarachnoid seeding and form tumoral nodules along the brain and spinal cord. This is seen in PNET, ependymomas, GBMs, lymphomas, oligodendrogliomas and choroid plexus papillomas. Primitive neuroectodermal tumours (PNET) form a rare group of tumors, which develop from primitive or undifferentiated nerve cells. These include medulloblastomas and pineoblastomas. One of the most important roles of imaging is to assess the extent of a tumor. This is shown in the case on the left in a patient who presented with multiple cranial nerve abnormalities. On the images we see an extra-axial tumor in the region of the left cavernous sinus. There is homogeneous enhancement with a broad dural tail. This is typical for a meningioma. Only by studying all the images we do appreciate that the actual extent of the tumor is greater than expected. The tumor is situated in the pterygopalatine fossa and extends into the orbit. It also spreads anteriorly into the middle cranial fossa Local tumor spread (2) Another important consideration is the effect on the surrounding structures. Primary brain tumors are derived from brain cells and often have less mass effect for their size than you would expect, due to their infiltrative growth. This is not the case with metastases and extra-axial tumors like meningiomas or schwannomas, which have more mass effect due to their expansive growth. On the left is an image of a diffusely infiltrating intra-axial tumor occupying most of the right hemisphere with only a minimal mass effect. This is typical for the infiltrative growth seen in primary brain tumors. There is no enhancement so this would probably be a low-grade astrocytoma. The ability of tumors to cross the midline limits the differential diagnosis.  Multiple tumors in the brain usually indicate metastatic disease (figure). Primary brain tumors are typically seen in a single region, but some brain tumors like lymphomas, multicentric glioblastomas and gliomatosis cerebri can be multifocal. Some tumors can be multifocal as a result of seeding metastases: this can occur in medulloblastomas (PNET-MB), ependymomas, GBMs and oligodendrogliomas. Meningiomas and schwannomas can be multiple, especially in neurofibromatosis type II. Multiple brain tumors can be seen in phacomatoses: Many non-tumorous diseases like small vessel disease, infections (septic emboli, abscesses) or demyelinating diseases like MS can also present as multifocal disease. Most intra-axial tumors are located in the white matter. Some tumors, however, spread to or are located in the gray matter. The differential diagnosis for these cortical based tumors includes oligodendroglioma, ganglioglioma and Dysembryoplastic Neuroepithial Tumor (DNET). A DNET is a rare benign neoplasm, usually in a cortical and temporal location. Patients with a cortically based tumor usually present with complex seizures. On the left a 45-year-old female with a stable seizure disorder (complex-partial) for 15 years. There is a non-enhancing, cortically based tumor. This is a ganglioglioma. The differential diagnosis includes DNET and pilocytic astrocytoma. These cortically based tumors have to be differentiated from non-tumorous lesions like cerebritis, herpes simplex encephalitis, infarction and post-ictal changes. On the left are images of a 52-year-old female who, over the period of one year, complained of headache and neck pain. There is a recent onset of tonic-clonic seizures. The CT shows a mass with calcifications, which extends all the way to the cortex. Although this is a large tumor there is only limited mass effect on surrounding structures, which indicates that this is an infiltrating tumor. The most likely diagnosis is oligodendroglioma. The differential diagnosis includes a malignant astrocytoma or a glioblastoma. Fat has a low density on CT (- 100HU). On MR, fat has a high signal intensity on both T1- and T2WI. On sequences with fat suppression fat can be differentiated from high signal caused by subacute hematoma, melanin, slow flow etc. When you see high signal on T1WI always look for chemical shift artefact, as this indicates the presence of fat. The chemical shift artefact occurs as alternating bands of high and low signal on the boundaries of a lesion and is seen only in the frequency encoding direction. Fat within a tumor is seen in lipomas, dermoid cysts and teratomas. On the left a patient with the classical findings of a ruptured dermoid cyst. Some tumors can have a high density on CT. This is typically seen in lymphoma, colloid cyst and PNET-MB (medulloblastoma). Calcification Calcification is seen in many CNS tumors (Table). When we think of a calcified intra-axial tumor, we think oligodendroglioma since these tumors nearly always have calcifications. However an intraaxial calcified tumor in the brain is more likely to be an astrocytoma than a oligodendrogliomas, since astrocytomas, although less frequently calcified, are far more common. A pineocytoma itself does not calcify, but instead it 'explodes' the calcifications of the pineal gland. On the left is an image of a calcified mass in the suprasellar region, causing obstructive hydrocephalus. This location in the suprasellar region and the calcification are typical for a craniopharyngioma. Craniopharyngiomas are slow growing, extra-axial, squamous epithelial, calcified, cystic tumors arising from remnants of Rathke's cleft. They are located the (supra)sellar region and primarily seen in children with a small second peak incidence in older adults. On the left are images of a tumor with a small calcification. . The calcification is not appreciated on the MR images, but is easily seen on CT. The calcification and the extension of the tumor to the cortex are very typical for an oligodendroglioma. An astrocytoma should be in the differential. On the left are images of a patient with progressive visual loss. On the coronal and sagittal TW1I there is a large mass centered around the sella with a broad dural base. There is extension into the sella. This patient was booked for decompression. Only after the CT was performed, was it appreciated how densely calcified this tumor is. It would be impossible to operate this tumor and preserve the patient's vision. Cystic versus Solid There are many cystic lesions that can simulate a CNS tumor. These include epidermoid, dermoid, arachnoid, neuroenteric and neuroglial cysts. Even enlarged perivascular spaces of Virchow Robin can simulate a tumor. In order to determine whether a lesion is a cyst or cystic mass look for the following characteristics: An arachnoid cyst is isointense to CSF on all sequences. Tumor necrosis may sometimes look like a cyst, but it is never completely isointense to CSF. On the far left a craniopharyngioma with an enhancing rim surrounding the cystic component. In the middle a neuroenteric cyst with the contents of which have the same signal intensity as CSF. On the right a glioblastoma multiforme (GBM) with a central cystic component. The enhancement in GBM is usually more irregular. Most tumors have a low or intermediate signal intensity on T1WI. Exceptions to this rule can indicate a specific type of tumor. On the left is a list of causes for T1-shortening. Calcifications are mostly dark on T1WI, but depending on the matrix of the calcifications they can sometimes be bright on T1. Especially on gradient echo images slow flow can be seen as bright signal on T1WI and should not be confused with enhancement. This is particularly pronounced on gradient echo images. If you only do an enhanced scan, remember that high signal is not always enhancement. On the left are some images of tumors with high signal intensities on T1WI. On the far left images of a patient who presented with apoplexy. The high signal is due to hemorrhage in a pituitary macroadenoma. The patient in the middle has a glioblastoma multiforme, which caused a hemorrhage in the splenium of the corpus callosum. On the right is a patient with a metastasis of a melanoma. The high signal intensity is due to the melanin content. Most tumors will be bright on T2WI due to a high water content. When tumors have a low water content they are very dense and hypercellular and the cells have a high nuclear-cytoplasmasmic ratio. These tumors will be dark on T2WI. The classic examples are CNS lymphoma and PNET (also hyperdense on CT). Calcifications are mostly dark on T2WI. The differential diagnosis of calcified tumors was discussed above. Paramagnetic effects cause a signal drop and are seen in tumors that contain hemosiderin. Proteinaceous material can be dark on T2 depending on the content of the protein itself. A classic example of this is the colloid cyst. Flow voids are also dark on T2 and indicate the presence of vessels or flow within a lesion. This is seen in tumors that contain a lot of vessels like hemangioblastomas, but also in non-tumorous lesions like vascular malformations. On the left some examples of tumors with a low signal intensity on T2WI.  Normally water protons have the ability to diffuse extracellularly and loose signal. High intensity on DWI indicates restriction of the ability of water protons to diffuse extracellularly. Restricted diffusion is seen in abscesses, epidermoid cysts and acute infarction (due to cytotoxic edema). In cerebral abscesses the diffusion is probably restricted due to the viscosity of pus, resulting in a high signal on DWI. In most tumors there is no restricted diffusion - even in necrotic or cystic components. This results in a normal, low signal on DWI. Perfusion imaging can play an important role in determining the malignancy grade of a CNS tumor. Perfusion depends on the vascularity of a tumor and is not dependent on the breakdown of the blood-brain barrier. The amount of perfusion shows a better correlation with the grade of malignancy of a tumor than the amount of contrast enhancement. Blood brain barrier The brain has a unique triple layered blood-brain barrier (BBB) with tight endothelial junctions in order to maintain a consistent internal milieu. Contrast will not leak into the brain unless this barrier is damaged. Enhancement is seen when a CNS tumor destroys the BBB. Extra-axial tumors such as meningiomas and schwannomas are not derived from brain cells and do not have a blood-brain barrier. Therefore they will enhance. There is also no blood-brain barrier in the pituitary, pineal and choroid plexus regions. Some non-tumoral lesions enhance because they can also break down the BBB and may simulate a brain tumor. These lesions include like infections, demyelinating diseases (MS) and infarctions. Contrast enhancement cannot visualize the full extent of a tumor in cases of infiltrating tumors, like gliomas. The reason for this is that tumor cells blend with the normal brain parenchyma where the blood brain barrier is still intact. Tumor cells can be found beyond the enhancing margins of the tumor and beyond any MR signal alteration - even beyond the area of edema. On the left is an image of a 42 y/o male with mild head trauma. On the T2WI there is a lesion in the left temporal lobe, found incidentally. There was no enhancement and the DWI was normal. During follow-up there was a slight increase in size. This was diagnosed as a low-grade astrocytoma. It is not possible to resect such a lesion, since the infiltrating tumors cells are within the normal-appearing brain tissue. In gliomas - like astrocytomas, oligodendrogliomas and glioblastoma multiforme - enhancement usually indicates a higher degree of malignancy. Therefore when during the follow up of a low-grade glioma the tumor starts to enhance, it is a sign of malignant transformation.. Gangliogliomas and pilocytic astrocytomas are the exceptions to this rule: they are low-grade tumors, but they enhance vividly. As discussed above, it recently has been shown that tumor angiogenesis as shown by perfusion MR correlates better with tumor grade than enhancement after the administration of intravenous contrast. The amount of enhancement depends on the amount of contrast that is delivered to the interstitium. In general, the longer we wait, the better the interstitial enhancement will be. The optimal timing is about 30 minutes and it is better to give contrast at the start of the examination and to do the enhanced T1WI at the end. No enhancement is seen in: On the left is an image of an intra-axial tumor in an adult. It is centered in the temporal lobe and involves the cortex. Although there is massive infiltrative growth involving a large part of the right cerebral hemisphere, there is only minimal mass effect. There is no enhancement. These features are typical for a low-grade astrocytoma. Homogeneous enhancement can be seen in: Patchy enhancement can be seen in: On the left is an example of a glioblastoma multiforme (GBM). The enhancement indicates that this is a high-grade tumor, but only parts of it enhance. Notice that there is also a cystic component with ring enhancement. The tumor cells probably extend beyond the area of edema as seen on the FLAIR image. This is because gliomas grow infiltratively into normal brain - initially without any MR changes. Patchy enhancement (2) On the left are images of a tumor located in the right hemisphere. Although is a large tumor, the mass-effect is limited. This indicates that there is marked infiltrative growth, a characteristic typical for gliomas. Notice the heterogeneity on both T2WI and FLAIR. There is patchy enhancement. All these findings are typical for a GBM. Virtually no other tumor behaves in this way. Ring enhancement Ring enhancement is seen in metastases and high-grade gliomas. It is also seen in non-tumorous lesions like abscesses, some MS-plaques and sometimes in an old hematomas. On the left three different ring enhancing lesions. Conspicuity of tumors with contrast The case on the left demonstrates the value of Gadolinium in the conspicuity of tumors. This is a patient with Neurofibromatosis II. After the administration of contrast the two meningiomas and the schwannoma are easily seen. Leptomeningeal metastases are usually not seen without the administration of intravenous contrast. The case on the left demonstrates the abnormal enhancement along the brainstem, along the folia of the cerebellum (yellow arrow) and along the fifth intracranial nerve (blue arrow) in a patient with leptomeningeal metastases. Common skull base tumors are listed in the table on the left. These tumors either arise from extracranial structures like the sinuses (sinonasal carcinoma), or from the skull base itself (chordoma, chondrosarcoma, fibrous dysplasia). Chordoma is usually located in the midline, while chondrasarcoma usually arises off the midline. On the left a midline tumor arising from the clivus. This is the typical presentation of a chordoma. The differential diagnosis would include a metastasis and a chondrosarcoma. On the left another skull base tumor located off midline. This is a typical presentation for a chondrosarcoma. The differential diagnosis would include a metastasis and a paraganglioma. Chondrosarcomas can be located in the midline and chordomas are sometimes located off midline but those cases are exceptional. On the left an example of a Skull Base Paraganglioma. On the left CT images of a 58-year-old male with a gradual onset of right facial pain and numbness and a recent onset of double vision. First study the images, than continue. There is an enhancing mass anterior to the skull base and also in the region of the right cavernous sinus. In the bone window setting there is sclerosis of the skull base, particularly in the region of the clivus. Continue with the MR images. On the left enhanced sagittal and coronal T1WI. The most striking finding is the black clivus due to the sclerosis. A normal clivus is bright on T1WI as a result of the fatty bone marrow. There is an enhancing mass anterior to the clivus. On the coronal images we see the enhancement extending through the foramen ovale to the right of the cavernous sinus. The diagnosis is a nasopharyngeal squamous cell carcinoma with intracranial extension. The differential diagnosis would include: skull base metastasis, lymphoma, chronic infection and even a meningioma - although this would be an unusual way for a meningioma to spread. On the left is a list of common sellar and suprasellar tumors. In this region it is important to keep the possibility of an aneurysm in the differential diagnosis. On the left are images of a mass in the suprasellar cistern. On the NECT we can see that it contains calcium. On the T1WI there is a hyperintense area that shows no enhancement (i.e. cystic). There are other components that show enhancement. The tumor is complicated by a hydrocephalus. These findings are very specific for a craniopharyngeoma. On the left NECT and enhanced CT-images of a 33-year-old female with severe headache (worse in the a.m.), reduction in visual acuity and visual fields and papilledema. Continue with the MR images. Notice the normal inferiorly displaced pituitary gland. This means it is not a macroadenoma. The diagnosis is again a craniopharyngioma. The differential diagnosis would include an astrocytoma and a meningioma. Common CP Angle Tumors are listed in the table on the left. On the left a 52-year-old male with hearing loss on the right. The images show an unusual cystic mass with enhancing septations. There is also some enhancement within the internal acoustic canal. Based on the images the most likely diagnosis would be a cystic schwannoma, but this happened to be an uncommon, cystic presentation of a meningioma. Common pineal region tumors are listed in the table on the left. On the left a tumor located in the pineal region. Based on these images the differential diagnosis would include: This happened to be a meningioma. On the left are typical images of a ruptured pineal region dermoid. On the left images of a 12 y/o male with upward gaze paralysis. There is a tumor located in the pineal region. The tumor contains calcifications. There is homogeneous enhancement, which is common for a tumor in the pineal region (discussed above). Based on the age of the patient, the location and the tumor characteristics, this is most likely a germinoma. Common intraventricular Tumors are listed in the table on the left. On the left a tumor located in the 3rd ventricle. The tumor contains calcifications. The diagnosis is a giant cell astrocytoma. In children tumors in the 4th ventricle are very common. Astrocytomas are the most common followed by medulloblastomas (or PNET-MB), ependymomas and brainstem gliomas with a dorsal exophytic component. In adults tumors in the 4th ventricle are uncommon. Metastases are most frequently seen, followed by hemangioblastomas, choroid plexus papillomas and dermoid and epidermoid cysts. Many non-tumorous lesions can mimic a brain tumor. Abscesses can mimic metastases. Multiple sclerosis can present with a mass-like lesion with enhancement, also known as tumefactive multiple sclerosis.. In the parasellar region one should always consider the possibility of a aneurysm. Infections and vascular lesions can also mimic a CNS tumor. by James Smirniotopoulos by Namik Erdag et al. AJR 2001; 176:1319-1326 Diagnostic Neuroradiology by Anne G. Osborn Mosby 1994Robin Smithuis and Walter Montanera Incidence of CNS tumors Age distribution Intra- versus Extraaxial Midline crossing Multifocal disease Cortical based tumors Fat - Calcification - Cyst - High density High on T1 Low on T2 Diffusion weighted imaging Perfusion Imaging Skull base Sella/suprasellar Cerebello-pontine angle Pineal region Intraventricular 4th ventricleBrain Tumor - Systematic ApproachRadiology Department of the Rijnland hospital, Leiderdorp, the Netherlands and the Division of Neuroradiology of the St. Michael's Hospital, University of Toronto, Canada
neuro5
1
Cerebral Venous Thrombosis
by Barbara Simons, Geert Lycklama a Nijeholt and Robin Smithuis
Cerebral venous thrombosis is an important cause of stroke especially in children and young adults. It is more common than previously thought and frequently missed on initial imaging. It is a difficult diagnosis because of its nonspecific clinical presentation and subtle imaging findings. In this article we will focus on: Cerebral venous thrombosis is located in descending order in the following venous structures: Clinically patients with cerebral venous thrombosis present with variable symptoms ranging from headache to seizure and coma in severe cases. In neonates shock and dehydration is a common cause of venous thrombosis. In older children it is often local infection, such as mastoiditis, or coagulopathy. In adults, coagulopathies is the cause in 70% and infection is the cause in 10% of cases. In women, oral contraceptive use and pregnancy are strong risk factors. Venous thrombosis has a nonspecific presentation and therefore it is important to recognize subtle imaging findings and indirect signs that may indicate the presence of thrombosis. Although these findings are often present on initial scans, they are frequently detected only in retrospect. Clinically patients with venous thrombosis often present with seizures, which is not a symptom in patients with an arterial infarction. On a routine non-enhanced MR or CT you should think of the possibility of venous thrombosis when you see: Direct visualization of a clot in the cerebral veins on a non enhanced CT scan is known as the dense clot sign. It is seen in only one third of cases. Normally veins are slightly denser than brain tissue and in some cases it is difficult to say whether the vein is normal or too dense (see pitfalls). In these cases a contrast enhanced scan is necessary to solve this problem. Dense clot sign (2) Visualization of a thrombosed cortical vein that is seen as a linear or cord-like density, is also known as the cord sign. Another term that is frequently used, is the dense vessel sign. Dense clot sign (3) On the left images of a patient with a hemorrhagic infarction in the temporal lobe (red arrow). Notice the dense transverse sinus due to thrombosis (blue arrows). The empty delta sign is a finding that is seen on a contrast enhanced CT (CECT) and was first described in thrombosis of the superior sagittal sinus. The sign consists of a triangular area of enhancement with a relatively low-attenuating center, which is the thrombosed sinus. The likely explanation is enhancement of the rich dural venous collateral circulation surrounding the thrombosed sinus, producing the central region of low attenuation. In early thrombosis the empty delta sign may be absent and you will have to rely on non-visualization of the thrombosed vein on the CECT. The sign may be absent after two months due to recanalization within the thrombus. Empty delta sign (2) On the left a case of thrombosis of the right transverse sinus and the left transverse and sigmoid sinus (arrows). There is enhancement surrounding the thrombosed hypoattenuating veins. On spin-echo images patent cerebral veins usually will demonstrate low signal intensity due to flow void. Flow voids are best seen on T2-weighted and FLAIR images, but can sometimes also be seen on T1-weighted images. A thrombus will manifest as absence of flow void. Although this is not a completely reliable sign, it is often one of the first things, that make you think of the possibility of venous thrombosis. The next step has to be a contrast enhanced study. On the left a T2-weighted image with normal flow void in the right sigmoid sinus and jugular vein (blue arrow). On the left there is abnormal high signal as a result of thrombosis (red arrow). Absence of normal flow void on MR (2) The images on the left show abnormal high signal on the T1-weighted images due to thrombosis. The thrombosis extends from the deep cerebral veins and straight sinus to the transverse and sigmoid sinus on the right. Notice the normal flow void in the left transverse sinus on the right lower image. Absence of normal flow void on MR-images can be very helpful in detecting venous thrombosis, but there are some pitfalls as we will discuss later. Slow flow can occur in veins and cause T1 hyperintensity. The other sign that can help you in making the diagnosis of unsuspected venous thrombosis is venous infarction. Venous thrombosis leads to a high venous pressure which first results in vasogenic edema in the white matter of the affected area. When the proces continues it may lead to infarction and development of cytotoxic edema next to the vasogenic edema. This is unlike in an arterial infarction in which there is only cytotoxic edema and no vasogenic edema. Due to the high venous pressure hemorrhage is seen more frequently in venous infarction compared to arterial infarction. Since we are not that familiar with venous infarctions, we often think of them as infarctions in an atypical location or in a non-arterial distribution. However venous infarctions do have a typical distribution, as shown on the left. Since many veins are midline structures, venous infarcts are often bilateral. This is seen in thrombosis of the superior sagittal sinus, straight sinus and the internal cerebral veins. Venous infarction (2) - Superior sagittal sinus thrombosis The most frequently thrombosed venous structure is the superior sagittal sinus. Infarction is seen in 75% of cases. The abnormalities are parasagittal and frequently bilateral. Hemorrhage is seen in 60% of the cases. On the left bilateral parasagittal edema and subte hemorrhage in a patient with thrombosis of the superior sagittal sinus. On the left reconstructed sagittal CT-images in a patient with bilateral parasagittal hemorrhage due to thrombosis of the superior sagittal sinus. The red arrow on the contrast enhanced image indicates the filling defect caused by the thrombus. Venous infarcts (3) - vein of Labbe Another typical venous infarction is due to thrombosis of the vein of Labb?. On the left images demonstrating hypodensity in the white matter and less pronounced in the gray matter of the left temporal lobe. There is a broad differential diagnosis including arterial infarction, infection, tumor etc. Notice that there is some linear density within the infarcted area. This is due to hemorrhage. In the differential diagnosis we also should include a venous infarct in the territory of the vein of Labbe. The subtle density in the area of the left transverse sinus (arrow) is the key to the diagnosis. This is a direct sign of thrombosis and the next step is a CECT, which confirmed the diagnosis (not shown). On the left images of a patient with hemorrhage in the temporal lobe. When the hemorrhagic component of the infarction is large, it may look like any other intracerebral hematoma with surrounding vasogenic edema. The clue to the diagnosis in this case is seen on the contrast enhanced image, which nicely demonstrates the filling defect in the sigmoid sinus (blue arrow). On the left a similar case on MR. There is a combination of vasogenic edema (red arrow), cytotoxic edema and hemorrhage (blue arrow). These findings and the location in the temporal lobe, should make you think of venous infarction due to thrombosis of the vein of Labb?. The next examination should be a contrast enhanced MR or CT to prove the diagnosis. Venous infarction (4) - Deep cerebral veins On the far left a FLAIR image demonstrating high signal in the left thalamus. When you look closely and you may have to enlarge the image to appreciate this, there is also high signal in the basal ganglia on the right. These bilateral findings should raise the suspicion of deep cerebral venous thrombosis. A sagittal CT reconstruction demonstrates a filling defect in the straight sinus and the vein of Galen (arrows). On the left a young patient with bilateral abnormalities in the region of the basal ganglia. Based on the imaging findings there is a broad differential including small vessel disease, demyelinisation, intoxication and metabolic disorders. Continue with the T1-weighted images in this patient. Notice the abnormal high signal in the internal cerebral veins and straight sinus on the T1-weighted images, where there should be a low signal due to flow void. This was unlike the low signal in other sinuses. The diagnosis is bilateral infarctions in the basal ganglia due to deep cerebral venous thrombosis. Venous infarction (5) - Edema In some cases of venous thrombosis the imaging findings can resolve completely. On the left a patient with a subcortical area of high signal intensity. The first impression was that this could be a low grade glioma. On a follow up scan the abnormalities had resolved completely. In retrospect a dense vessel sign was seen in one of the cortical veins and the diagnosis of venous thrombosis was made. The high signal intensity can be attributed to vasogenic edema due to the high venous pressure that resulted from the thrombosis. CT-venography is a simple and straight forward technique to demonstrate venous thrombosis. In the early stage there is non-enhancement of the thrombosed vein and in a later stage there is non-enhancement of the thrombus with surrounding enhancement known as empty delta sign, as discussed before. Unlike MR, CT-venography virtually has no pitfalls. The only thing that you don't want to do, is to scan too early, i.e. before the veins enhance or too late, i.e. when the contrast is gone. Some advocate to do a scan like a CT-arteriography and just add 5-10 seconds delay. To be on the safe side we advocate 45-50 seconds delay after the start of contrast injection. We use at least 70 cc of contrast. On the left some images of a CT-venography demonstrating thrombosis in many sinuses. On the left images of a patient with an infarction in the area of the vein of Labb?. On the non-enhanced images you can appreciate the dense thrombus within the transverse sinus and the hemorrhage in the infarcted area. On the enhanced images a filling defect can be seen in the transverse sinus. You can scroll through the images. The MR-techniques that are used for the diagnosis of cerebral venous thrombosis are: Time-of-flight (TOF), phase-contrast angiography (PCA) and contrast-enhanced MR-venography: When you use MIP-projections, always look at the source images. On the left a lateral and oblique MIP image from a normal contrast-enhanced MR venography. Notice the prominent vein of Trolard (red arrow) and vein of Labbe (blue arrow). Every MR techniques has its own pitfalls as we will discuss in a moment. Contrast-enhanced MR venography has the disadvantage that you need to give contrast, but has less pitfalls. Angiography is only performed in severe cases, when an intervention is planned. On the left images of a patient with venous thrombosis, who was unconsious and did not respond to anticoagulant therapy. There is thrombosis of the superior sagittal sinus (red arrow), straight sinus (blue arrow) and transverse and sigmoid sinus (yellow arrow). Continue with the video of the thrombectomy. On the left a video of the thrombectomy. Arachnoid granulations are small protrusions of the arachnoid through the dura mater. They protrude into the venous sinuses and may mimic filling defects caused by thrombus. Usually these granulations are easily to differentiate from thrombosis. The dense triangle sign can be mimicked in infants by the combination of the hypointensity of the unmyelinated brain and the physiologic polycythemia resultig in high density of the blood in the sagittal sinus. A pseudodelta sign can also be seen in patients with hyperattenuating acute subarachnoid hemorrhage around the sinus or subdural empyema or in patients with a posterior parafalcine interhemispheric hematoma. In these cases, administration of contrast material should opacify the sinus, obliterating the lucent center of the pseudodelta. Normally veins are slightly denser than brain tissue and in some cases it is difficult to say whether it is normal or too dense. In these cases a contrast enhanced scan is necessary to solve this problem. On the left an image of a thrombosed transverse sinus and next to it a normal transverse sinus. On the left three images of a patient with venous thrombosis in the superior sagittal sinus. On the far left we see a dense vessel sign on the unenhanced CT. In the middle an image made 25 seconds after the start of the contrast injection. There is arterial enhancement and it looks as if the superior sagittal sinus enhances, but in fact what we see is the shine through of the dense thrombus. Only on the image on the right, which was made 45 seconds after contrast injection there is an empty delta sign, which proves the presence of a thrombus in the sinus. Usually there is no problem in differentiating a hematoma from a thrombosed sinus. On the left a patient with a peripheral intracerebral hematoma, that on first impression simulates a thrombosed transverse sinus. On the left a patient with an subdural hematoma, that in the region of the superior sagittal sinus results in a pseudo empty delta sign. By scrolling through the data set, it was obvious that it was an extention of the hematoma. Hypoplasia and aplasia of the right or left transverse sinus is a common finding. It can easily be mistaken for sinus thrombosis, because on the MRA one of the transverse sinuses is missing. When you suspect, that there is a hypoplastic transverse sinus, then you should look at the size of the jugular foramen. On the left images of a patient with hypoplasia of the left transverse sinus. Notice the size difference of the jugular foramen. On the left a transverse MIP of phase-contrast images. To differentiate whether there is a hypoplastic transverse sinus or thrombosed sinus, you need to look at the source images. On the source image on the right you can see that there is no hypoplasia (blue arrow). In this case there thrombosis of the left transverse sinus. On the left another case that demonstrates that you cannot fully rely on phase contrast imaging. The signal in the vein depends on the velocity of the flowing blood and the velocity encoding by the technician. On the far left a patient with non visualization of the left transverse sinus. This could be hypoplasia, venous thrombosis or slow flow. On the contrast enhanced T1-weighted image it is obvious that the sinus fills with contrast and is patent. Normally when there is low signal in a vein, it is attributed to flow void and a sign of patency of the vein. However at some stage of the thrombus there is intracellular deoxyhemoglobin, which is dark on T2 and mimics flow void. On the left there is a thrombosed right transverse sinus with a delta sign on the contrast enhanced image. The sinus has a low signal intensity on the T2-weighted image as a result of the intracellular deoxyhemoglobin. On the contrast enhanced T1-weighted image it is obvious that the sinus is filled with thrombus. On the contrast enhanced T1 images on the left there is an area of low signal intensity within the enhancing transverse sinus. This could easily been mistaken for a central thrombus within the sinus. This however is the result of flow void. Continue with the phase contrast images. On the phase contrast images it is obvious that the transverse sinus is patent. We can conclude that MRI has many false positives and negatives in the diagnosis of venous thrombosis. Contrast enhanced MR-venography is the most reliable MR technique. CT-venography is even more reliable, because it is easy and less sensitive to pitfalls. Pitfalls in TOF imaging are: Chronic dural sinus thrombosis can lead to dural arteriovenous fistula formation and to increased CSF pressure. A DAVF or dural arteriovenous fistula is an abnormal connection between dural arteries, which are branches of the external carotid with the venous sinuses. Sinus thrombosis is seen in many patients with a dural arteriovenous fistula, but the pathogenesis is still unclear (10). There are two possible mechanisms: (a) thrombophlebitis of the dural sinus may induce a dural fistula and (b) in the course of a dural fistula flow reversal may lead to thrombosis. Current classifications of DAVF focus mainly on the presence of leptomeningeal reflux related to cerebral venous hypertension leading to cerebral venous infarction or hemorrhage. On the left DSA images of a patient with a DAVF. Notice the direct communication between the branches of the external carotid artery and the transverse sinus (blue arrow). Continue with the T2-weighted images. On the left T2-images during the follow up. In april 2008 there were no abnormalities. In january 2009 there are signs of intracranial hypertension like CSF surrounding the optic nerve and CSF within the stalk of the hypophysis. In some patients dural sinus thrombosis may, even after recanalisation, lead to persisting disturbances in venous circulation. This may lead to raised intracranial CSF pressure as assessed by lumbar puncture. Clinically, these patients complain of headaches and they may have vision disturbances due to papil edema. On MRI, one may see increased CSF around the optic nerve and an empty sella. Apparently in some patients a residual stenosis persists. On the left a T2-weighted image demonstrating papil edema and an empty sella. Continue with the sagittal T1-weighted image. On the left a sagittal T1-weighted image demonstrating the empty sella (arrow). On the left an illustration of the territories of the venous drainage. There is great variation in these territories and the illustration should be regarded as a rough guide. by James L. Leach et al October 2006 RadioGraphics, 26, S19-S41 in eMedicine by Mahesh R Patel by J. Linn et al American Journal of Neuroradiology 28:946-952, May 2007 by Mathieu H. Rodallec et al October 2006 RadioGraphics, 26, S5-S18. by Emil J. Y. Lee September 2002 Radiology, 224, 788-789. by Colin S. Poon et al AJR 2007; 189:S64-S75 by J van Gijn JRSM Volume 93, Number 5 Pp. 230-233 by Phua Hwee Tang et al Ann Acad Med Singapore 2008;37:397-401 by N. Khandelwal et al AJR 2006; 187:1637-1643 by L K Tsai et al J Neurol Neurosurg Psychiatry 2004;75:1639-1641Barbara Simons, Geert Lycklama a Nijeholt and Robin Smithuis Dense clot sign Empty delta sign Absence of normal flow void on MR Venous infarction CT-venography MR-venography DSA Arachnoid Granulations Pseudodelta sign Wrong bolus timing Hematoma simulating venous thrombosis Hypoplastic transverse sinus Low signal intensity in thrombus Flow void on contrast-enhanced MR DAVF Thrombosis and increased CSF pressureCerebral Venous ThrombosisRadiology department of the Medical Centre Haaglanden in the Hague and the Rijnland hospital in Leiderdorp, the Netherlands
neuro6
1
Dementia: role of MRI
by Frederik Barkhof, Marieke Hazewinkel, Maja Binnewijzend and Robin Smithuis
This review is based on a presentation given by Frederik Barkhof at the Neuroradiology teaching course for the Dutch Radiology Society and was adapted for the Radiology Assistant by Robin Smithuis. First publication: 1-3-2007. Updated version: 9-1-2012. This presentation will focus on the role of MRI in the diagnosis of dementia and related diseases. We will discuss the following subjects: The role of neuroimaging in dementia nowadays extends beyond its traditional role of excluding neurosurgical lesions. Radiological findings may support the diagnosis of specific neurodegenerative disorders and sometimes radiological findings are necessary to confirm the diagnosis. It is a challenge for neuroimaging to contribute to the early diagnosis of neurodegenerative diseases such as Alzheimer's disease. Early diagnosis includes recognition of pre-dementia conditions, such as mild cognitive impairment (MCI). In addition, early diagnosis allows early treatment using currently available therapies or new therapies in the future. Neuroimaging may also be used to assess disease progression and is adopted in current trials investigating MCI and AD. The coronal image shows the hippocampus, the main structure involved in many forms of dementia. Coronal-oblique T1-weighted images are used for the assessment of medial temporal lobe and hippocampal atrophy. They are obtained in a plane orthogonal to the long axis of the hippocampus; this plane is orientated parallel to the brainstem. These should be thin-section images and are ideally obtained by reformatting a sagittal 3D T1 sequence through the entire brain. Additional sagittal reconstructions will enable the assessment of midline structures as well as parietal atrophy, which may be involved in certain neurodegenerative disorders. FLAIR images are used to assess global cortical atrophy (GCA), vascular white matter hyperintensities and infarctions. T2-weighted images are used to assess infarctions, in particular lacunar infarctions in the thalamus and basal ganglia, which can be missed on FLAIR images. T2*-weighted images are necessary to detect microbleeds in amyloid angiopathy. These images can also depict calcifications and iron deposition. DWI should be considered as a supplemental sequence in young patients or in rapidly progressive neurodegenerative disorders (DD - vasculitis, CJD). CT can be useful when contraindications prevent MRI or when the only reason for imaging is to rule out surgically treatable causes of cognitive decline. In the transverse plane the scan angle should be parallel to the long axis of the temporal lobe. Use of multi-detector CT will enable coronally reformatted images to be reconstructed perpendicular to the long axis of the temporal lobe for optimal vizualisation of the hippocampus. Use of multi-detector CT will enable coronally reformatted images to be reconstructed perpendicular to the long axis of the temporal lobe for optimal vizualisation of the hippocampus. An MR-study of a patient suspected of having dementia must be assessed in a standardized way. First of all, treatable diseases like subdural hematomas, tumors and hydrocephalus need to be excluded. Next we should look for signs of specific dementias such as: So when we study the MR images we should score in a systematic way for global atrophy, focal atrophy and for vascular disease (i.e. infarcts, white matter lesions, lacunes). When we study the MR images we must systematically score for global atrophy, focal atrophy and for vascular disease (i.e. infarcts, white matter lesions, lacunes). This standardized assessment of the MR findings in a patient suspected of having a cognitive disorder includes: GCA scale is the mean score for cortical atrophy throughout the complete cerebrum: Cortical atrophy is best scored on FLAIR images. In some neurodegenerative disorders the atrophy is asymmetric and occurs in specific regions. A radiological report should mention any regional atrophy or asymmetry. When assessing atrophy in different regions keep in mind that cranially, the central sulcus lies more posteriorly than you would expect (figure). The MTA-score should be rated on coronal T1-weighted images at a consistent slice position. Select a slice through the corpus of the hippocampus, at the level of the anterior pons. > 75 years : MTA-score 3 or more is abnormal (i.e. 2 can still be normal at this age) Data from a study with 222 controls and patients with various forms of dementia in which this visual rating scale was used to assess temporal lobe atrophy suggest that sensitivities and specificities of 85% can be obtained for patients with AD. The score is based on a visual rating of the width of the choroid fissure, the width of the temporal horn, and the height of the hippocampal formation. < 75 years: score 2 or more is abnormal. > 75 years: score 3 or more is abnormal. Here you can scroll through the images for examples of MTA score 0-4. < 75 years: score 2 or more is abnormal. > 75 years: score 3 or more is abnormal. A high MTA-score is very sensitive for the diagnosis of Alzheimer disease and is present in the vast majority of patients with AD, while in controls a positive score is almost always absent (table on the left). Therefore it is a good test to discern controls from patients with AD. This test is not completely specific for AD however, as MTA can also be found in other forms of dementias (7). On the other hand if a patient with mild cognitive impairment (MCI) a possible 'prodromal state of AD' has a negative MTA-score, it is very unlikely that this patient will develop AD (high sensitivity yields high negative predictive value), except in very young subjects, in whom a more posterior pattern of atrophy can be observed in AD. If there is a strong suspicion of Alzheimer's disease, it can be useful to repeat the examination to see if there is any progress of the (medial temporal lobe) atrophy. The images show a follow-up examination at 18 and 36 months in a patient who was at risk for familial AD, demonstrating progression of the disease. An alternative approach would be to perform a SPECT- or PET-scan to look for changes in perfusion/metabolism of the temporo-parietal cortex, as these changes precede the development of atrophy. On MR, white matter hyperintensities (WMH) and lacunes - both of which are frequently observed in the elderly - are generally viewed as evidence of small vessel disease. The Fazekas-scale provides an overall impression of the presence of WMH in the entire brain. It is best scored on transverse FLAIR or T2-weighted images. Score: Fazekas 1 is considered normal in the elderly. Fazekas 2 and 3 are pathologic, but may be seen in normally functioning individuals. They are however, at high risk for disability. In 600 normally functioning elderly people the Fazekas score predicted disability within one year (table). In the Fazekas 3 group 25% was disabled within one year (10). Three year follow-up shows that severe white matter changes independently and strongly predict rapid global functional decline (17). The findings in a normally aging brain can overlap with findings in dementia. As implicated earlier, there may be some degree of atrophy, though mainly of the white matter with increasing prominence of the perivascular (Virchow-Robin) spaces and non-specific fronto-parietal sulcal widening. There may also be some degree of medial temporal lobe atrophy. A MTA-score of 2 for individuals older than 75 years of age may be normal. As the brain ages, there is an increasing deposition of iron in specific areas of the brain: mainly the basal ganglia, nucleus ruber and pars reticluaris of the substantia nigra. There also may develop a rim of high signal intensity on T2W and FLAIR images around the ventricles, known as caps and bands (figure). A limited amount of white matter hyperintensities may also occur in the normally ageing brain (Fazekas grade 1). Lacunes are always pathological. Strategic infarctions are infarctions in areas that are crucial for normal cognitive functioning of the brain. These areas are summarized in the table. Strategic infarctions are best seen on transverse FAIR and T2W sequences. The images show bilateral thalamic infarctions - lesions often associated with cognitive dysfunction. Study the images of two different patients. Then continue reading. The image on the far left shows an infarct in the vascular territory of the Posterior Cerebral Artery (PCA), with involvement of the inferior medial temporal lobe which includes the hippocampus. This is a strategic infarction, since it is in the dominant hemisphere, it will result in cognitive dysfunction. The image next to it is a transverse FLAIR image showing another infarct in the PCA-territory, with involvement of the temporo-occipital association area. This is another example of a strategic infarction that can result in cognitive dysfunction. In addition to medial temporal lobe atrophy, parietal atrophy also has a positive predictive value in the diagnosis of AD. Atrophy of the precuneus is particularly characteristic of AD (15). This is particularly the case in young patients with AD (presenile AD), who may have normal MTA-scores. The Koedam scale rates parietal atrophy - assessed in sagittal, coronal and axial planes. In these planes, widening of the posterior cingulate and parieto-occipital sulci as well as parietal atrophy (including the precuneus) is rated (Table). Koedam scale grade 0-1 Sagittal T1-, axial FLAIR- and coronal T1-weighted images illustrating the Koedam scale of posterior atrophy. When different scores are obtained in different orientations, the highest score must be considered (16). Koedam scale grade 2-3 Sagittal T1-, axial FLAIR- and coronal T1-weighted images illustrating the Koedam scale of posterior atrophy. The yellow arrows point to extreme widening of the posterior cingulate en parieto-occipital sulci in a patient with grade 3 posterior atrophy. In addition to clinical findings, CSF and MRI, PET-imaging is useful in diagnosing AD. In AD FDG-PET can show hypometabolism in the temporoparietal regions and/or the posterior cingulum. This may help differentiate AD from FTD, which shows frontal hypometabolism on FDG-PET. The images show FDG-PET and axial FLAIR images of a normal subject and of patients with AD and FTD. FDG-PET (top row) and axial FLAIR images of a normal subject and of AD and FTD patients. In AD there is a decreased metabolism of the parietal lobes (yellow arrows), whereas in FTD, there is frontal hypometablism (red arrows). The prevalence of specific forms of dementia is age-dependent. In patients In patients > 65 years there are more cases of senile AD and vascular dementia. In many older patients with manifest AD there is co-existing vascular disease, which contributes to the demented state. AD accounts for 50%-70% of all cases of dementia in the elderly population. Age is a strong risk factor, with the disease affecting approximately 8% of individuals over the age of 65 and 30% over the age of 85 years. The progression of AD is gradual and the average patient lives 10 years after the onset of symptoms. With the increasing percentage of elderly in the population, the prevalence of AD is expected to triple over the next 50 years. In end-stage AD there is widespread atrophy, which is no different from other end-stage dementias. In imaging we therefore have to try to identify AD in an earlier stage and we have to concentrate on the hippocampus and the medial temporal lobe, because that is where AD starts. The role of MRI in the diagnostic process of AD is twofold: Study the image, then continue reading. The findings are consistent with the diagnosis of end stage AD, because there is: It is not specific for AD however, since severe GCA occurs in other end-stage disorders as well Presenile AD ( Although there usually is some mild hippocampal atrophy, the most striking finding is parietal atrophy with atrophy of the posterior cingulum and the precuneus; the hippocampus can be normal. Mild cognitive impairment is a relatively recent term used to describe people who have some problems with their memory, but do not actually have dementia, since dementia is defined as having problems in two or more cognitive domains. Some of these patients will be in the early stages of Alzheimer's disease or another dementia, so it is important to identify them. Finding MTA is a strong risk-factor for progression to dementia. Vascular dementia (VaD) is thought to be the second most common cause of dementia after Alzheimer's disease. It can sometimes be distinguished from AD by a more sudden onset and association with vascular risk factors. VaD can be characterized by its stepwise deterioration with periods of stability followed by sudden decline in cognitive function. Most patients, however, have small vessel disease, which is typified by a more gradual and subtle pattern of deterioration. Control of vascular risk factors is the treatment of choice, but cholinesterase inhibitors (drugs that are being used in AD) are also increasingly being used to treat vascular dementia. The images show a patient with a strategic PCA infarction involving the hippocampus. This type of infarct can result in sudden dementia if located in the dominant hemisphere. It will usually not result in dementia if it occurs in the non-dominant hemisphere. In most patients with VaD there is diffuse white matter disease with large confluent lesions (Fazekas 3). In some of these patients the ventricles may be dilated due to global atrophy and some will also have medial temporal lobe atrophy. The images are of a patient who had VaD, but the medial temporal lobe was normal. Cognitive dysfunction in VaD can be the result of (2): There is an increasing awareness for the importance of small vessel disease as a predictor of cognitive decline and dementia. Moreover, it seems to amplify the effects of pathologic changes of Alzheimer's disease. On the left we see a patient who was diagnosed as having VaD. White matter disease is seen as severe WMH (hypointense on T1) in the periventricular regions. In addition to these vascular changes, there is also MTA. Presumably this patient has both VaD and AD, a finding seen in many elderly patients. These findings should be described separately as it may have therapeutic consequences. The problem however is, that white matter hyperintensities and lacunes are also frequently observed in non-demented elderly and at some level can be regarded as normal findings in aging. To overcome this problem the NINDS-AIREN International Work Group has formulated criteria for the history and physical, radiological, (see above) and pathological examination to classify patients as having possible, probable and definite VaD. However considerable interobserver variability exists for the assessment of the radiological part of these NINDS-AIREN criteria and some level of training is mandatory (2). The medial nuclei of the thalamus play an important role in memory and learning. A large unilateral infarction or bilateral infarctions in this region can cause dementia. You have to pay special attention to these areas to find these small infarctions. On FLAIR images you will easily miss these infarctions, because they can be isointense to the surrounding structures (8). A high resolution T2WI is needed to detect these thalamic infarcts. FLAIR in the infratentorial region and in the spinal cord is of limited value as it suppresses not only the signal of water, but also pathology with a long T1-relaxation time. This phenomenon can also be seen in the detection of Multiple Sclerosis, where FLAIR is of limited value in the infratentorial region and of no use in the spinal cord. Dementia may be the clinical presentation in CAA, a condition in which ?-amyloid is deposited in the vessel walls of the brain. The result is hemorrhage, usually microhemorrhages, but also subarachnoid hemorrhage or lobar hematomas may occur. On MR, the T2* sequence will show multiple microhemorrhages, typically in a peripheral location (as opposed to hypertensive microhemorrhages, which are usually more centrally located, e.g. in the basal ganglia and thalami). In addition, FLAIR will reveal moderate to sever white matter hyperintensities (Fazekas grade 2 or 3) T2* images in a patient with CAA show multiple peripherally located microbleeds. FLAIR images of the same patient show Fazekas 2 white matter hyprintensities. T2* images in a patient with CAA microbleeds. T2* images demonstrate multiple lobar microbleeds in a patient with CAA. FTLD, formerly called Pick's disease, is a progressive dementia, that accounts for 5-10% of cases of dementia., and occurs relatively more frequently in presenile subjects FTLD is clinically characterized by behavioral and language disturbances that may precede or overshadow memory deficits. There is currently no treatment for this condition. Imaging plays an important role in the diagnosis as the findings are easy to recognize. Radiological findings are pronounced atrophy of frontal and / or temporal lobes. In some forms of FTLD the atrophy might be strikingly asymmetric, e.g. in Semantic Dementia, a disease subtype with progressive aphasia and left-sided temporal lobe degeneration. The images are of a patient with progressive aphasia. The most prominent finding is the striking asymmetric atrophy of the temporal lobe on the left side with not only atrophy of the hippocampus, but also the temporal poles. The atrophy has resulted in gyri that appear as sharp as knives ('knife blade atrophy'). There is also some increased signal intensity seen on the FLAIR image, probably due to gliosis. These findings are pathognomonic for the diagnosis of FTLD. Patients with left-sided temporal atrophy are usually clinically obvious. Right-sided atrophy is usually not as easily recognized as these patients only present with subtle disturbances in recognizing faces. Dementia with Lewy bodies is responsible for approximately 25% of dementias and belongs to the atypical Parkinson syndromes together with progressive supranuclear palsy (PSP) and multi-system atrophy (MSA). The clinical manifestations can be similar to that of AD or dementia associated with Parkinson's disease. Patients typically present with one of three symptom complexes: detailed visual hallucinations, Parkinson-like symptoms and fluctuations in alertness and attention. Pathologically, the disease is characterized by the presence of Lewy bodies in various regions of the hippocampal complex, subcortical nuclei and neocortex with a variable number of diffuse amyloid plaques. Cholinesterase inhibitors are currently the treatment of choice for this condition. The role of imaging is limited in Lewy body dementia. Usually the MR of the brain is normal, including the hippocampus. This finding is important as it enables us to differentiate this disease from Alzheimer';s disease, the main differential diagnosis. Nuclear imaging can be used to demonstrate an abnormal dopaminergic system (so-called DaTscan) PSP is also one of the atypical parkinsonian syndromes. In PSP there is pronounced atrophy of the midbrain (mesencephalon), which accounts for the typical upward gaze paralysis. Normally the upper border of the midbrain is convex. The atrophy of the midbrain in PSP results in a concave upper border of the midbrain with the typical 'humming bird sign' (figure). MSA is also one of the atypical parkinsonian syndromes. MSA is a rare neurological disorder characterized by a combination of parkinsonism, cerebellar and pyramidal signs, and autonomic dysfunction. MSA can be classified as MSA-C, MSA-P or MSA-A. In MSA-C (formerly known as sporadic olivopontocerebellar atrophy or sOPCA) the cerebellar symptoms predominate, whereas in MSA-P the parkinsonian symptoms dominate (MSA-P was formerly known as striatonigral degeneration). MSA-A is the form in which autonomic dysfunction predominates and is the new term for what was formerly known as Shy-Drager syndrome. Usually there is pronounced cerebellar atrophy and severe atrophy of the pons. In MSA-P: low T2 SI dorsolateral putamen and slit-like increased SI lateral to putamen on T2. In contrast to PSP, we don't see the humming bird sign, because the midbrain has a normal convex upper border. The so-called 'hot cross bun sign', which is a result of pontine hyperintensity, is typical for MSA-C. CJD is a very rare and incurable neurodegenerative disease, caused by a unique type of infectious agent called a prion. The first symptom of CJD is rapidly progressive dementia, leading to memory loss, personality changes and hallucinations. The disease is characterized by spongiform changes in the cortical and subcortical gray matter, with loss of neurons and replacement by gliosis. The abnormalities can sometimes be detected on FLAIR, but are most conspicuous on DWI sequences, affecting either the striatum, the neo-cortex, or a combination of both. New variant CJD New variant of CJD is also known as the 'mad cow disease' (12). It is a disease fortunately hardly encountered anymore. In this variant the changes are seen in the posterior part of the thalamus, called the pulvinar. CBD is a rare entity which may present with cognitive dysfunction, usually in combination with Parkinson-like symptoms. The so-called 'Alien-hand' syndrome is a typical manifestation. MRI shows asymmetric parietal cortical atrophy, sometimes with associated hyperintensity of the white matter on T2W images. Axial FLAIR image shows striking asymmetric cortical parietal atrophy in a patient with CBD. Huntington disease is a hereditary neurodegenerative disease (autosomal dominant trait, but often de novo mutations), and can present with early onset dementia as well as choreoathetosis and psychosis. Imaging shows characteristic atrophy of the caudate nucleus and subsequent enlargement of the frontal horns of the lateral ventricles. CADASIL is another hereditary disease which may present with a progressive cognitive dysfunction. Other presenting symptoms include migraines, stroke-like episodes and behavioral disturbances. It affects the small vessels of the brain. Confluent white matter hyperintesities in the frontal and especially anterior temporal lobes in combination with (lacunar) infarcts and microbleeds are seen on imaging. The FLAIR images show classic findings in CADASIL - confluent white matter hyperintensities with lacunar infarcts and involvement of the anterior temporal lobes. Long term sequelae of traumatic brain injury such as cerebral contusions and diffuse axonal injury (DAI) may include cognitive impairment. Frontobasal/temporal parenchymal loss or T2* black dots typical for DAI in a patient with a history of trauma must therefore be taken into consideration when assessing MR images for dementia. The FLAIR images show classic post-traumatic tissue loss with gliosis in both frontal lobes, the left occipital lobe and right temporal lobe. In the book on the left you can find more information about the role of MR in dementia (9). View Philip Scheltens' presentation at the 8th International Conference on Alzheimer's Disease and Related Disorders van Straaten EC, Scheltens P, Knol DL et al. Stroke 2003; 34: 1907-1912. L. O. Wahlund, MD, PhD; F. Barkhof, MD, PhD; F. Fazekas, MD et al on behalf of the European Task Force on Age-Related White Matter Changes. Stroke. 2001;32:1318 Wiesje M. van der Flier, PhD et al Stroke. 2005;36:2116. Jeffrey R. Petrella, MD, R. Edward Coleman, MD and P. Murali Doraiswamy, MD State of the Art article in Radiology 2003;226:315-336. G B Frisoni, P h Scheltens, S Galluzzi, F M Nobili et al. Journal of Neurology Neurosurgery and Psychiatry 2003;74:1371-1381 Barber R et al. Neurology 1999;52:1153 Ant?nio J. Bastos Leite, MD et al. Stroke. 2004;35:415 by Jaap Valk, Frederik Barkhof, Philip Scheltens. Inzitari D, Simoni M, Pracucci G, Poggesi A, Basile AM, Chabriat H, Erkinjuntti T, Fazekas F, Ferro JM, Hennerici M, Langhorne P, O'Brien J, Barkhof F, Visser MC, Wahlund LO, Waldemar G, Wallin A, Pantoni L; LADIS Study Group. Arch Intern Med. 2007 Jan 8;167(1):81-8 Webpath: the Internet Pathology Laboratory for Medical Education of the Florida State University College of Medicine by Collie DA, Sellar RJ, Zeidler M, Colchester AC, Knight R, Will RG. Clin Radiol. 2001 Sep;56(9):726-39. Clinical Neuroradiology, Gasser M. Hathout, 2009 Pp103-114 Diagnostic Imaging: Brain, 2nd edtion, Osborne, Salzman, Barcovich et al. Pp I (2) 36-39, I(4) 76-79, I (9) 76-79, I (10) 70-121 by Lehmann M, Koedam EL, Barnes J, Bartlett JW, Ryan NS, Pijnenburg YA, Barkhof F, Wattjes MP, Scheltens P, Fox NC. Source Dementia Research Centre, UCL Institute of Neurology, Queen Square, London, UK. Neurobiol Aging. 2011 May 17. by Koedam EL, Lehmann M, van der Flier WM, Scheltens P, Pijnenburg YA, Fox N, Barkhof F, Wattjes MP Eur Radiol. 2011 Dec;21(12):2618-25. by Inzitari D, Pracucci G, Poggesi A, Carlucci G, Barkhof F, Chabriat H, Erkinjuntti T, Fazekas F, Ferro JM, Hennerici M, Langhorne P, O'Brien J, Scheltens P, Visser MC, Wahlund LO, Waldemar G, Wallin A, Pantoni L; LADIS Study Group. BMJ. 2009 Jul 6;339:b2477.Frederik Barkhof, Marieke Hazewinkel, Maja Binnewijzend and Robin Smithuis CT protocol GCA-scale for Global Cortical Atrophy MTA-scale for Medial Temporal lobe Atrophy Fazekas scale for WM lesions Normal ageing Strategic infarctions Koedam score for Parietal Atrophy FDG-PET Alzheimers Disease Presenile AD Mild Cognitive Impairment (MCI) Vascular Dementia (VaD) Strategic infarcts and small vessel disease Cerebral Amyloid Angiopathy (CAA) Frontotemporal Lobar Degeneration (FTLD ) Dementia with Lewy bodies Progressive supranuclear palsy (PSP) Multi System Atrophy (MSA) Creutzfeldt-Jakob disease (CJD) Corticobasal Degeneration (CBD) Huntington Disease Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencehalopathy (CADASIL) Traumatic Brain Injury (TBI)Dementia: role of MRIAlzheimer Centre and Image Analysis Centre, Vrije Universiteit Medical Center, Amsterdam and the Rijnland Hospital, Leiderdorp, The Netherlands
neuro7
1
Disc Nomenclature
by Robin Smithuis
Disc is normal in development and there are no signs of disease, trauma or aging. Anular tears are also called anular fissures and are separations between anular fibers, avulsion of fibers from their vertebral body insertions, or breaks through fibers involving one or many layers of the anular lamellae. The terms 'tear' or 'fissure' does not imply that the lesion is consequent to trauma. In case of a traumatic event the term 'rupture' is appropriate. Disc Herniation is displacement of disc material beyond the limits of the intervertebral disc space. A herniated disc can be contained (covered by outer anulus fibrosus) or uncontained. Focal herniation is a herniated disc less than 90? of the disc circumference. Broadbased herniation is a herniated disc in between 90?-180? of the disc circumference. Bulging Disc is the presence of disc tissue 'circumferentially' (180?-360?) beyond the edges of the ring apophyses and is not considered a form of herniation. Protrusion indicates that the distance between the edges of the disc herniation is less than the distance between the edges of the base. Extrusion is present when the distance between the edges of the disc material is greater the distance at the base. Migration indicates displacement of disc material away from the site of extrusion, regardless of whether sequestrated or not. Sequestration is used to indicate that the displaced disc material has lost completely any continuity with the parent disc Central or medial (orange). Since the PLL (posterior longitudinal ligament) is at its thickest in this region, the disc usually herniates slightly to the left or right of this central zone. Paramedian or lateral recess (blue). Because the PLL is not as thick in this region, this is the number one region for disc herniations to occur in. Foraminal or subarticular (red). It is rare for a disc to herniate into the intervertebral foramen. Only 5% to 10% of all disc herniation occur here or farther out. When herniations do occur in this zone, they are often very troublesome for the patient. This is because a super-delicate neural structure called the 'Dorsal Root Ganglion' (DRG) lives in this zone resulting in severe pain, sciatica and nerve cell damage. Extraforaminal or lateral (green). Disc herniations in this region are uncommon. Website on disc pathology with excellent teaching and illustrations.Robin Smithuis Normal Disc Anular tear - Disc herniation Focal herniation - Broad based hernia - Bulging Disc Disc Protrusion - Extrusion Migration - Sequestration Axial localisation of herniated discsDisc NomenclatureRadiology department of the Rijnland Hospital, Leiderdorp, the Netherlands
neuro9
1
Role of MRI in Epilepsy
by Laurens De Cocker, Felice D'Arco and Philippe Demaerel and Robin Smithuis 
In many patients with epilepsy antiepileptic drug treatment is unable to control the seizures. Using a dedicated MRI-protocol, it is possible to detect an epileptogenic lesion in 80 percent of these patients. Resection of these lesions can lead to seizure freedom in many patients. We will discuss the MRI protocol and the typical findings in the most common epilepsy-associated diseases. The illustration summarizes the most common causes of seizures in patients with medically uncontrollable epilepsy. Some of these lesions are readily identifiable. Meso temporal sclerosis and focal cortical dysplasia are the most common causes and can only be depicted with a dedicated protocol. The table also summarizes epileptogenic lesions that are detected in patients with uncontrollable seizures. Mesial temporal sclerosis is the most common cause of intractable epilepsy. In medication refractory epilepsia the most common location of the epilectogenic lesion is temporal lobe (60%), frontal lobe (20%) and parietal lobe (10%), periventricular (5%) and occipital (5%). Seizures are common. About 4 percent of all people will have at least one seizure during their lifetime. In patients with a first ever seizure imaging will mostly show no brain-abnormalities, because the seizure is provoked by fever, drugs, dehydration or sleep deprivation. The term epilepsy is used, when there are recurrent unprovoked seizures. About 60 percent of patients with epilepsy can be controlled with antiepileptic drugs. Most patients with uncontrollable seizures have complex partial seizures. Partial seizures - also called focal seizures - are seizures which affect only a part of the brain at onset. They usually start in the temporal lobe. In simple partial seizures the person remains conscious. A simple partial seizure can be a precursor to a larger seizure and then it is called an aura. A complex partial seizure affects a larger part of the hemisphere and the person may lose consciousness. If a partial seizure spreads from one hemisphere to the other this will give rise to a secondarily generalised seizure. The person will become unconscious and may have a tonic clonic seizure. The table shows a dedicated epilepsy protocol. Some will also use Inversion Recovery and not use contrast on a routine base. T1WI Superior for cortical thickness and the interface between grey and white matter. On T1WI look for grey matter occuring in an aberrant location as in gray matter heterotopia. FLAIR Look very carefully for cortical and subcortical hyperintensities on the FLAIR, which can be very subtle. Since FLAIR may show false-positive results due to artefacts, the abnormalities should be confirmed on T2WI. T2* or SWI Helpful when searching for haemoglobin breakdown products as in posttraumatic changes and cavernomas, or to look for calcifications in tuberous sclerosis, Sturge-Weber, cavernomas and gangliogliomas. Mesial temporal sclerosis (MTS) is a specific pattern of hippocampal neuronal loss accompanied by gliosis and atrophy. The etiology is unknown, but there is a relationship between MTS and prolonged febrile seizures earlier in life, complicated delivery and developmental processes. In 15% of patients another developmetal abnormality can be found, mostly focal cortical dysplasia. This is called dual pathology. MTS is the most common cause of partial complex epilepsy in adults and is also the most common etiology in young adult patients undergoing surgery. Surgical removal of visible MRI changes associated with unilateral mesial temporal sclerosis leads to seizure freedom in up to 80% of cases. Coronal T2W and FLAIR images are the most sensitive for detecting MTS. On axial slices mesial temporal sclerosis is commonly overlooked. Bilateral mesial temporal sclerosis is difficult to detect due to the lack of comparison with the unaffected contralateral hippocampus. The coronal T2WI and FLAIR images show right-sided mesial temporal sclerosis. Notice the volume loss, which indicates atrophy and causes secondary enlargement of the temporal horn of the lateral ventricle. The high signal in the hippocamous reflects gliosis. Mesial temporal sclerosis may occur in association with other pathology, especially focal cortical dysplasia. This is called dual pathology. The images show mesial temporal sclerosis with a hyperintense and shrunken hippocampus (red arrows), and secondary enlargement of the left temporal horn of the left laterale ventricle. Also notice associated subcortical hyperintensity in the left temporal lobe indicating focal cortical dysplasia. 35-year-old patient with refractory temporal lobe epilepsy. MR shows subtle hyperintensity of the left hippocampus on the axial FLAIR (blue arrow) and atrophy of the left hippocampus on coronal images (yellow arrow). The patient was succesfully treated with amygdalo-hippocampectomy on the left. Hippocampal hyperintensity on T2WI or FLAIR images with volume loss is diagnostic for mesial temporal sclerosis in the appropriate clinical setting. Hippocampal hyperintensity without volume loss is seen in: Status epilepticus The imaging findings in status epilepticus can mimick mesotemporal sclerosis. In status epilepticus a hyperintense hippocampus can be seen, but there is swelling and no atrophy. Axial FLAIR, axial DWI and coronal T2WI demonstrate a hyperintense hippocampus with a slightly compressed temporal horn of the lateral ventricle consistent with hippocampal edema. DWI shows diffusion restriction due to cytotoxic edema in the acute stage of the status epilepticus. DNET mimicking mesial temporal sclerosis Axial T2WI shows hyperintense, but enlarged hippocampus with a bubbly appearance. This is typical for a DNET or dysembryoplastic neuroepithelial tumor, which we will discuss in a moment. The coronal contrast-enhanced T1WI shows an enlarged hippocampus without uptake of contrast medium. key findings Focal cortical dysplasia is a congenital abnormality where the neurons fail to migrate in the proper formation in utero. MRI findings may be very subtle or may even be negative, therefore a high index of suspicion is mandatory! The most common findings are cortical or subcortical hyperintensities especially seen on FLAIR-images. These are often found at the bottom of a deep sulcus. Another finding is a blurred interface between grey and white matter, because the white matter looks a little bit like gray matter because it contains neurons that did not reach the cortex. The images show typical focal cortical dysplasia. There is cortical thickening and blurring of the grey/white matter junction on T1WI (left). The FLAIR image on the right shows the subcortical hyperintensity. The images demonstrate cortical and subcortical signal abnormalities on T2WI and FLAIR in the left temporal lobe indicating focal cortical dysplasia. Notice associated T2/FLAIR hyperintense and shrunken hippocampus as a result of mesial temporal sclerosis, i.e. dual pathology. Another case of focal cortical dysplasia. Notice the hypoplastic left temporal lobe with cortical thickening (arrow) and atrophy of the white matter. Axial T1WI, T2WI and FLAIR-images of a 15 year old boy with epilepsy. Notice thickening and hyperintensity of the cortex of the left superior frontal gyrus. The FLAIR-images also show high signal in the subcortical white matter. These findings are typical for focal cortical dysplasia. Sometimes the hyperintensity is seen extending from the subcortical area to the margin of the ventricle. This is called the transmantle sign. This finding represents the arrested neuronal migration. Images of a 27-year-old male with refractory occipital lobe epilepsy. Coronal FLAIR and axial T2WI show T2-hyperintense cortical thickening and high signal in cortex and subcortical region. Notice subcortical hyperintensity extending to the right ventricle indicating transmantle sign (blue arrow). Transmantle sign seen in another patient with focal cortical dysplasia. Cortical and glial scars usually result from meningitis or birth injury. Ulegyria is a specific type of scar. It is defined as cerebral cortex scarring due to perinatal ischemia. Ulegyria typically affects full term infants. In these infants there is greater perfusion to the apex of the gyri than to the cortex at the depth of the sulci. The resulting pattern is that of a shrunken cortex in which the deep portions of the gyri are more shrunken than the superficial portions, leaving pedunculated gyri on long stalks with a mushroom appearance. Ulegyria must be differentiated from microgyria. MR will shows tissue loss and gliosis underneath a shrunken cortex. The shrunken cortex is best appreciated on a 3D-T1WI because of its high resolution and the superior delineation of the cortex, while FLAIR will show the hyperintensity associated with the gliosis. Therefore always use the FLAIR-sequence to search for hyperintensities in an epileptic patient and subsequently correlate these findings with the cerebral cortex in the affected area on high resolution T1WI. Cavernoma is also known as cavernous malformation or cavernous angioma. It is a benign low flow vascular malformation with a tendency to bleed. 75 percent occur as solitary sporadic lesions and 10-30 percent occur as multiple lesions. Cavernomas consist of locules of variable size that contain blood products in different stages of evolution which produces a popcorn appearance. A complete hemosiderin rim surrounds the lesion, but not when there is a recent bleeding. Unenhanced CT may show a hyperdense nodule or calcification, but in 50% of cases cavernomas will be occult on CT. T2WI and T2* gradient echo show multiple cavernomas. Notice the popcorn appearance with peripheral rim of hemosiderin on the T2WI. The lesions are almost completely black on the gradient echo due to blooming artefacts. T2* and susceptibility weighted imaging (SWI) markedly increase the sensitivity of MRI to detect small cavernomas. The five black dots in the left cerebral hemisphere on the T2* are also cavernomas and are not visible on the T2WI. Cavernomas are associated with developmental venous anomalies (DVA's). The unenhanced CT shows a small calcification in the right lentiform nucleus. Enhanced CT shows a venous anomaly draining the cavernoma into the right internal cerebral vein. Coronal T2WI shows the venous anomaly as a curvilinear flow void. Cavernoma in the postcentral gyrus on T1WI, T2WI and SWI. Notice popcorn appeance and blooming artefact. Same patient. Notice the hemosiderin coating of the precentral gyrus consistent with superficial siderosis due to prior hemorrhage of the cavernoma (red arrowheads). In patients with multiple small black dots the differential diagnosis is: Diffuse axonal injury (DAI) A 46 year old biker presented with seizures after being hit by a car. CT-image shows only minimal subarachnoidal hemorrhage (arrow). MRI was performed several weeks after the injury because of a change in personality. T2*-images show multiple hemosiderin depositions at the interface between grey and white matter, consistent with diffuse axonal injury (DAI). Notice that the location of the microbleeds is different from the peripheral located CAA-bleeds. All brain tumors may present with epilepsy, but there are some typically epilepsy associated tumors. These tumours share the following characteristics: key findings Ganglioglioma is the most common tumor associated with temporal lobe epilepsy. Calcification is common in ganglioglioma and is an important distinguishing factor from DNET and pleomorphic xanthoastrocytoma. Ganglioglioma in a young child. Note large cyst with enhancement of mural solid tissue. Small cystic ganglioglioma with a small enhancing nodule. key findings DNET in typical cases present as a bubbly mass which expands the affected gyri. The bubbly cystic appearance is seen as small cyst-like intratumoral structures that are very hyperintense on T2WI. DNET in an 11-year old boy presenting with refractory partial seizures. The tumor shows a characteristic bubby appearance and there is subtle scalopping of the skull. key findings Pleomorphic xanthoastrocytoma (PXA) is a rare cause of temporal lobe epilepsy. Peritumoral edema may be seen in PXA, while it is not a feature of either ganglioglioma or DNET. Thickening and enhancement of the adjacent leptomeninges is highly characteristic but is not always present. When meningeal involvement is not present, than a pleiomorphic xanthoastrocytoma is indistinguishable from a ganglioglioma. key findings Hypothalamic hamartoma is also known as diencephalic or tuber cinereum hamartoma. It represents nonneoplastic congenital grey matter heterotopia in the region of tuber cinereum of the hypothalamus. It is seen in infants presenting with seizures and precocious puberty. key findings T2WI shows right hemimegalencephaly. Notice the asymmetric skull and slightly enlarged lateral ventricle. Hemimegalencephaly is the only condition in which an increase in parenchymal volume is associated with an increase in ipsilateral ventricular volume. Hemimegalencephaly is a rare disease characterized by hamartomatous growth of one cerebral hemisphere or part of it. Patients present with early seizures, macrocrania and severe developmental delay with contralateral hemiparesis. The thickened cortex may show a wide spectrum of abnormalities, such as lissencephaly, pachygyria or polymicrogyria. In the late stage, the involved hemisphere may atrophy due to constant seizure acitivity. Most of the affected children die in the first years of life because of status epilepticus. CT and T2WI in a patient with a right hemimegalencephaly. There is dysplastic thick cortex and ventricular dilatation on the affected side. 9-y-old girl with refractory nocturnal epilepsy. MRI shows overgrowth of the left cerebral hemisphere. T1WI shows heterotopic gray matter lining the left lateral ventricle (blue arrow). In hemimegalencephaly it is important to exclude contralateral abnormalities, as these form a contraindication to hemispherectomy. key findings Rasmussen's encephalitis is a progressive hemispheric atrophy of unknown origin. Patient develop an increasing frequency of seizures and progressive hemiplegia. Notice that, opposed to hemimegalencephaly, the smaller hemisphere is the site of abnormality, and the lateral ventricle is larger in the smaller hemisphere. key findings in the brain Tuberous sclerosis or Bourneville's disease is an inherited condition characterized by the presence of hamartomas in many organs including angiomyolipoma of the kidney, cardiac rhabdomyoma and cortical and subependymal tubers in the brain. Some patients have lymphangioleiomatosis, a cystic lung disease seen in women. The classic clinical triad is focal epilepsy, adenoma sebaceum and mental retardation (mnemonic: fits, zits and nitwits). The cortical hamartomas are called tubers and are similar to cortical dysplasia. Subependymal nodules are small lesions protruding into the lateral ventricles. Sometimes they are calcified. Seizure surgery in TSC is contemplated if a particular tuber can be implicated in seizure activity, or if a subependymal giant cell astrocytomas obstructs the foramen of Monro causing hydrocephalus. CT of a patient with Tuberous Sclerosis shows multiple cortical and subcortical calcifications. CT and MRI in a patient with Tuberous Sclerosis. There are multiple cortcal and subependymal nodules. The CT shows that most of the lesions are calcified. Subependymal giant cell astrocytoma (SEGA) This is a tumor that develops from a subependymal nodule near the foramen of Monro. They have a poor prognosis because they lead to obstruction of CSF flow. They are characterized by marked enhancement and their typical location. Axial T2WI and T1WI-CE show a giant cell astrocytoma at the level of the left foramen of Monro causing obstructive hydrocephalus. Also notice tuber on the left. Sagittal T1WI post contrast shows a giant cell astrocytoma in the right foramen of Monro. key findings Sturge-Weber is also called encephalotrigeminal angiomatosis. It is a vascular malformation with capillary venous angiomas in the face (port-wine stain), choroid of the eye and leptomeninges. Venous occlusion and ischemia lead to angiomatosis with cortical calcium deposition and atrophy Clinical features are seizures, hemiparesis, anopsia, mental retardation and port-wine stain. The MR-images show leptomeningeal angiomatosis which is mainly localized in the occipital lobes. Venous stasis and calcifications are best seen on the SWI. MRI in patients with Sturge-Weber can show: Coronal MR-images of a patient with Sturge-Weber show leptomeningeal enhancement in the right posterior hemispere. CT in a patient with Sturge-Weber shows huge cortical and subcortical tram-track calcifications involving the left posterior hemispere. 4-year-old boy with Sturge-Weber syndrome. Notice atrophy of the left posterior cerebral hemisphere with leptomeningeal enhancement and thickening. In Sturge-Weber a vascular malformation of the choroid of the eye is seen. These patients present with buphthalmos (enlarged eye) due to increased intraocular pressure and hemianopsia. Eye abnormalities in a 4-year-old boy with Sturge-Weber syndrome. Notice FLAIR-hyperintensity (red arrow) and excessive enhancement of the wall of the left globe (blue arrow) consistent with a diffuse choroidal hemangioma. key findings Polymicrogyria is a malformation due to an alteration of the cortical development in the late stage of neuronal migration. The deeper layers of the cortex form multiple small gyri with derangement of the normal lamination and sulcation. The T1W-images show a comparison between normal lamination and sulcation on the left and polymicrogyria on the right (arrow). Heterotopic Grey Matter results from an arrested migration of normal neurons along the radial path between the ventricular walls (ependyma) and the subcortical regions. There are two types of heterotopia: subependymal and subcortical. The most common clinical presentation is intractable seizures. Heterotopia present as nodular foci of grey matter intensity on all sequences. They do not enhance. Images of a typical subependymal heterotopia. Another case of heterotopia with typical subcortical nodules (arrows). Schizencephaly is a cleft in the brain that connects the lateral ventricle to the subarachnoid space. The cleft is lined by polymicrogyric gray matter. Open-lip schizencephaly is characterized by separation of the cleft walls. Closed-lip schizencephaly is characterized by cleft walls in apposition to each other. Patients have seizures and hemiparesis, which is proportional to the size of the cleft and are more common in the open-lip type. First study the images and then continue reading. This patient has a bilateral schizencephaly. There is an open-lip type on the right and a closed-lip type on the left (red arrow). Notice the track of grey matter in the left hemisphere on the axial image. The differential diagnosis of schizencaphaly is porencephaly, which is also a cleft, but it is not lined by grey matter. by Abdel Razek AA et al. AJNR. 2009 Jan;30(1):4-11 by Barkovich J et al. Amirsys 2007 by Barkovich AJ. Neuroradiology 2010 52:479-487 by Bien CG, et al Brain 2002; 125:1751-1759. by Bien CG et al Brain 128(pt 3):454-71,2005 by Chiapparini L, et al Neuroradiology 2003; 45:171-183. by Chinchure S et al Neurol India 2010 May-Jun,58(3):361-70 by Demaerel P JBR-BTR 2008 Nov-Dec;91(6):254-7 by Flores-Sarnat L J Child Neurol 2002; 17:373-384 by Hanefeld F, Kruse B, Holzbach U, Christen HJ, Merboldt KD, Hanicke W, Frahm J. J Magn Reson Imaging 2008 aug,28(2):300-7 by Kim SJ et al. Pediatr Neurol 27(4):282-8,2002 by Maria BL, et al J Child Neurol 1998; 13:606-618. by Martin N, et al Neuroradiology 1990; 31:492-497 by Montenegro MA et al Arch Neurol 2002; 59:1147-1153 by Radhakrishnqn R et al Journ Clin Imag Sci 2011; 1(2):1-11 by Urbach H et al JNR 2004 Jun-Jul;25(6):916-26 by Tortori-Donati P, Rossi A Springer 2005 by Woermann FG, Vollmar C Epilepsy Behav 2009 May;15(1):40-9Laurens De Cocker, Felice D'Arco and Philippe Demaerel and Robin Smithuis Common causes of Epilepsy Seizures and Epilepsy MRI epilepsy protocol Differential of hippocampal hyperintensity Transmantle sign Differential diagnosis of microbleeds Ganglioglioma DNET Pleomorphic xanthoastrocytoma Hypothalamic hamartoma Diffuse choroidal hemangiomaRole of MRI in Epilepsy
neuro10
1
Sella Turcica and Parasellar Region
by Walter Kucharczyk and Marieke Hazewinkel
This review is based on a presentation given by Walter Kucharczyka and was adapted for the Radiology Assistant by Marieke Hazewinkel. In this review a systematic anatomic approach to differential diagnosis of a sellar or parasellar mass is described. By clicking on one of the subjects in the list on the left, you will go directly to this item. If you have printing problems with the margins of the document, you may have to adjust the margins in the page set up of your internet browser, which you will find in the top left of the menu bar. In order to analyze a sellar or parasellar mass on MRI we use the following anatomic approach: Pituitary gland On a coronal section through the brain the reference structure is the pituitary gland which lies in the sella turcica. It is usually larger in females than in males - in females the superior border tends to be convex, whereas in males it is usually concave. The most common abnormalities that arise in the pituitary gland are pituitary adenoma, Rathke's cleft cyst and craniopharyngioma. Pituitary stalk The next structure to identify is the pituitary stalk. This is a vertically oriented structure which connects the pituitary gland to the brain. It is thinner at the bottom and thicker at the top. Embryologically, it is also derived from Rathke's cleft epithelium and therefore the pathologies, which can arise in the pituitary gland can also arise in the stalk. There are a few unusual things to be considered in children, such as germinomas and eosinophilic granulomas. In adults metastases and occasionally lymphoma can arise in the pituitary stalk. Optic chiasm Another major structure in the suprasellar cistern is the optic chiasm. It is an extension of the brain and looks like the number 8 lying on its side. It is glial tissue - therefore the most common tumors to originate here are gliomas. In the US and Europe another frequent pathology in this region is demyelinating disease - particularly multiple sclerosis. This can also be associated with some swelling of the optic chiasm. Hypothalamus Further cephalad lies the base of the brain, which at this location is the hypothalamus. Anatomically the hypothalamus forms the lateral walls and floor of the third ventricle. The most common pathologies to arise here are gliomas - in children hamartomas, germinomas and eosinophilic granuloma. Carotid artery A very important structure in this area is the internal carotid artery. It runs a complex anatomic course as it passes through the skull base shaped like an S on lateral views. It passes through the cavernous sinus. The segment cranial to this is known as the supracavernous segment. This bifurcates into the anterior cerebral artery, which passes cranially to the optic chiasm, and the middle cerebral artery, which runs laterally. Aneurysms and ectasias are pathologies that can arise here. One must also be aware of congenital variations in the course of the internal carotid Sometimes it is very medially positioned and can actually lie in the midline. Cavernous sinus The cavernous sinus is a paired complex of venous channels. In the lateral wall of the sinus run nerve III (oculomotorius), IV (trochlearis), V1 and V2 (trigeminus). The sixth cranial nerve (abducens) runs more medially and is located caudal to the carotid artery. The most common pathologies occurring in the cavernous sinus include schwannomas arising from the cranial nerves and inflammation, which can lead to thrombosis. This is known as cavernous sinus thrombophlebitis. Carotid-cavernous fistulas are fistulous communications between the carotid artery and the veins of the cavernous sinus. Meninges The meninges cover the cavernous sinus. They are thicker laterally and superiorly than medially and inferiorly. The most common tumor to arise from the meninges is of course the meningioma. Dural metastasis is the second most common tumor to arise here. Also inflammatory pathologies occur in the basal meninges - the most common infection being tuberculous meningitis. Of the non-infectious inflammatory pathologies sarcoidosis is the commonest. Sphenoid sinus Inferior to the pituitary gland lies the sphenoid sinus. This structure contains air and is lined by mucosa and bone. Posterior to the sphenoid sinus lies the clivus (not shown on this coronal section through the brain). Pathology that arises in this area includes carcinomas arising from the mucosa of the sphenoid sinus - squamous cell carcinoma and adenoid cystic carcinoma are the most common. Chordomas arise in the clivus and chondrosarcomas and osteosarcomas also occur in this area. Metastases can occur anywhere. Bacterial or fungal inflammatory processes in the sphenoid sinus can spread intracranially via the cavernous sinus. By definition, pituitary microadenomas are less than 10 mm in diameter and are located in the pituitary gland. These images show a classic case: on T1 a lesion about 3-4 mm in diameter, slightly hypointense compared to normal pituitary tissue, located in the pituitary gland. On T2, the lesion is slightly hyperintense. The differential diagnosis: pituitary microadenoma or Rathke's cleft cyst (the two can be indistinguishable). The sensitivity of an unenhanced MRI scan for detecting pituitary microadenomas is about 70%. It is not always necessary to give intravenous contrast for detecting pituitary microadenomas as patients with a negative scan generally receive the same symptomatic treatment as patients with a microadenoma (usually these patients are women with symptoms of hyperprolactinemia). The purpose of the scan is to rule out any large lesions. In possible surgical candidates (for example patients with failed medical therapy or pituitary disease not amenable to medical therapy such as Cushing's disease) it is necessary to give contrast to localize the lesion as accurately as possible. On an unenhanced scan, approximately 70% of all pituitary microadenomas can be detected. If you give gadolinium, you can reduce the false-negative rate from 30% to 15%. As mentioned earlier, this usually does not affect patient management. Coronal T1 and T2-weighted images and T1-weighted images before and after gadolinium. In this patient the lesion in the pituitary gland is only detectable after the administration of intravenous contrast. The differential diagnosis: pituitary microadenoma or Rathke's cleft cyst. By definition, pituitary macroadenomas are adenomas over 10mm in size. They tend to be soft, solid lesions, often with areas of necrosis or hemorrhage as they get bigger. As they grow, they first expand the sella turcica and then grow upwards. In this example of a pituitary macroadenoma there is suprasellar extension with elevation and compression of the optic chiasm. Because they are soft tumors, they usually indent at the diaphragma sellae, giving them a 'snowman' configuration. This is one feature that can help distinguish between a pituitary macroadenoma and a meningioma. Another feature which can help differentiate them is enlargement of the sella turcica - this generally only occurs with pituitary macroadenomas that originate in the sella. On the left another example of a pituitary macroadenoma. The lesion starts in the sella, which is enlarged, and extends into the suprasellar cistern. Note the classic 'snowman' configuration caused by constriction by the diaphragma sellae. Notice the blood-fluid level, indicating hemorrhage. The usefulness of observing the inclination of the diaphragmatic leaflets was referred to earlier. On the T2-weighted images on the right you can see that the leaflets are displaced upwards by this macroadenoma which started in the sella and is growing upwards. A lesion originating above the sella and growing downwards would push the leaflets in the other direction (this can be seen with meningiomas for example). Usually the diagnosis of a macroadenoma is straightforward. Sometimes a meningioma can give a similar appearance. On the left an example of a meningioma. Note there is no diaphragmatic constriction and there is uniform enhancement after the administration of intravenous gadolinium which is typical of meningioma. These images are of a transsphenoidal resection of a pituitary macroadenoma. After the bony floor of the sella turcica has been removed, the dura is incised with a cruciate incision. Because the pressure above the dura is larger than the pressure below, the macroadenoma then delivers itself into the sphenoid sinus. Intra-operative MRI was performed in an experimental setting to determine whether the neurosurgeon had successfully removed all of the tumor. Because using this surgical approach means a limited field-of-view, it is important to know beforehand what it is you are operating on. As we will see there are lesions you do not want to operate using this approach! Another common pathway of extension is laterally into the cavernous sinus. It is not always possible to tell if there is cavernous sinus invasion, but there are three signs to look out for: -Is there more than 50% encirclement of the carotid artery? Note: meningiomas tend to constrict the carotid artery, macroadenomas do not. -Is there lateral displacement of the lateral wall of the cavernous sinus compared to the opposite side? -Is there an increased amount of tissue interposed between the carotid artery and the lateral wall of the cavernous sinus? At medical school they teach you that a rare manifestation of a common lesion is more likely than a rare abnormality. Since pituitary adenomas are the most common lesions of the skull base, it is prudent to always include them in the differential diagnosis if you can not identify a normal pituitary gland when confronted with a mass in this region. This patient presented with nasal obstruction. She went to an ENT specialist who saw a large endonasal mass and she was referred to the neurosurgeon for planned major skull base resection. The neurosurgeon had seen something similar before, and checked her prolactin-level. This was 4000 (25 or less is normal). Endonasal biopsy revealed prolactinoma. After treatment with bromocriptine the mass shrunk down and no surgery was necessary. Rathke's cleft cyst is the second of three pathologies derived from Rathke's cleft epithelium. The cyst is fluid-filled and has very thin walls with a thickness of only one or two cell layers. This is illustrated by the microscopic image. These walls can contain cells which secrete fluid, allowing the cyst to grow and compress adjacent structures. Rathke's cleft cysts can occur either in or above the sella turcica. On the images above there is a normal pituitary gland, a normal optic chiasm and a normal carotid artery on each side. The pituitary stalk is not identifiable, however, due to a round mass in this area. The mass has a high signal intensity on the unenhanced T1-images. Now the only two things that are this bright on unenhanced T1-weighted images are either fluid (blood or proteinacious fluid) or fat. Solid masses are not this bright. Therefore it is most likely a cystic structure originating from the pituitary stalk, probably a Rathke's cleft cyst. A cystic craniopharyngioma is also in the differential diagnosis. These images illustrate the importance of unenhanced T1 images. They allow you to appreciate that the abnormality is located in the pituitary stalk alone. If you were only presented with images after the administration of intravenous contrast, you might think the pituitary gland was abnormal as well. These T1, T2 and T1-weighted images after gadolinium demonstrate another Rathke's cleft cyst located in the pituitary gland. Unlike the normal pituitary tissue and pituitary stalk it does not enhance after the administration of intravenous contrast. The normal pituitary tissue is compressed and displaced far to the left. It is important to recognize this as it could be mistaken for an enhancing component of the cystic mass. In general, all extra-axial masses , i.e. masses outside of the brain like the pituitary gland and stalk, will enhance because they do not have a blood-brain barrier. If you have a non-enhancing extra-axial mass, there are three possibilities: Craniopharyngioma is the third of the three pathologies derived from Rathke's cleft epithelium. Technically these are benign tumors, but unlike Rathke's cleft cysts, they have thick walls and are locally invasive. Macroscopically, it is a complex mass with multiple nodules at the base of the brain, sinuating along the fissures. Often, it can not be completely resected. The picture on the right shows a thick-walled cyst as part of the craniopharyngioma. In over 50% of cases craniopharyngiomas have a pathognomonic appearance. On these unenhanced and enhanced T1-weighted sagittal images, a compressed pituitary gland can be identified. There is a large intrasellar and suprasellar mass with cystic and enhancing components as well as calcifications. These findings in a child are virtually pathognomonic for craniopharyngioma (perhaps with only a dermoid in the differential diagnosis). Coronal images of the same mass. And axial images. Unenhanced CT shows the calcifications more clearly. After intravenous contrast the total extent of the lesion and its cystic components are much less evident. The most common intracranial tumor in adults is the meningioma with 20% of occurring at the skull base. This is an autopsy specimen with the brain removed, showing a meningioma sitting on the diaphragma sellae. Meningiomas are almost always solid lesions, sometimes with a cyst on the edge. They can lift up the arachnoid a little bit and enhance uniformly as a general rule. On the top-left unenhanced and enhanced CT-images, the main differential diagnosis of the enhancing mass would include meningioma, pituitary adenoma and an aneurysm. The post-constrast MR-image on the top-right rules out an aneurysm as a possible diagnosis (no flow void), but on axial images a pituitary adenoma and meningioma are still difficult to differentiate. Notice the spread of the lesion along the meninges. The epicentre of the lesion is above the sella. On the coronal images (T1 and T1-postcontrast), a compressed pituitary gland can be identified at the bottom of the sella turcica. Above it lies a large mass, partially intrasellar and partially suprasellar. Although the diaphragma sellae can not be identified on these images, it is probably a suprasellar mass growing downwards. When pituitary macroadenomas get this size they usually have areas of hemorrhage or necrosis - in mengiomas this is less often the case. This is an important case to keep in mind. This patient is a woman in her late forties, who presented to her family doctor with galactorrhea. The family doctor did a number of tests, including a determination of her prolactin level. This was about 150 (25 or less is normal). Thinking the patient had a pituitary adenoma, the family doctor ordered this CT scan. It is easy to get tunnel vision when reporting on a scan like this as a radiologist when the clinical information includes hyperprolactinemia and galactorrhea. Of course your first thought is a pituitary adenoma. If you look at the location of the lesion however (partially in the sella turcica and partially in the cavernous sinus), there are other possibilities, including a meningioma or an aneurysm. The radiologist reported this as a pituitary adenoma, and the patient was treated with bromocriptine. The bromocriptine had no effect, and the patient went to a neurosurgeon for a surgical opinion. The neurosurgeon ordered this MRI. The lesion partly in the right cavernous sinus and partly in the sella turcica is predominantly black on this T1-weighted image. In general there are three things that are black on MRI: air, bone and rapid blood flow. In this case it is black due to rapid blood flow in a carotid aneurysm. This is the corresponding angiogram. Obviously, this is not a lesion to be operated on transsphenoidally! Why did the aneurysm cause hyperprolactinemia and galactorrhea in this patient? It was caused by compression of the pituitary stalk. The pituitary stalk connects the hypothalamus to the pituitary gland and hormones produced in the hypothalamus are transported to the anterior lobe of the pituitary gland via portal veins running along the stalk. Most of these hormones stimulate the production of other hormones in the pituitary gland (such as TRH, GnRH, GHRH and CRH), but the release of dopamine inhibits the production of prolactin by the anterior lobe of the pituitary. Therefore when the stalk is compressed by a mass or is transected, the level of prolactin rises while all the other hormone levels decrease. This is known as the 'Stalk Section Effect'. It is the reason why masses other than adenomas can cause hyperprolactinemia. This is also why an unenhanced MRI scan suffices in a patient with hyperprolactinemia: it is not the size of the microadenoma, but ruling out other pathology that matters. On the left the T1-weighted image of a thrombosed aneurysm with high signal intensity on the unenhanced scan. It originates in the intracavernous segment of the right internal carotid artery. On the right the T2-weighted images: the thrombosed aneurysm has a dark rim. This is an example of a partially thrombosed aneurysm in the suprasellar cistern. The patent lumen is black on these T1-weighted images. It is surrounded by clot of different ages arranged in layers reaching from the lumen to the wall. It resembles an onion cut in half. On the left an autopsy specimen. You can see that this patient suffered a massive intraventricular and subarachnoid hemorrhage. The layers of bloodclot are very nicely reflected in the MR images. One of the most difficult differential diagnoses on CT is aneurysm versus meningioma. In this patient there is a large mass on the right hand side, possibly originating from the meninges or cavernous sinus. On CT it is impossible to tell whether this mass is an aneurysm or a meningioma. This is an MRI of the same patient. The mass is predominantly black and there is a large flow artefact running in the phase-encoding direction. These findings correspond to rapid blood flow, and the mass must therefore be an aneurysm. Angiogram of the same patient. It demonstrates that the flow in the aneurysm is not laminar, but that it swirls, gradually filling the lumen with contrast. Hamartomas are masses of dysplastic tissue found almost exclusively in young children. One of the most common locations is the floor of the third ventricle. This is a pathology specimen showing a small nodule hanging in the suprasellar cistern. They are benign lesions, but patients do succumb to them because of the bad location. These are CT images of a hamartoma suspended from the floor of the third ventricle. It does not enhance after the administration of intravenous contrast. MR images of a similar small nodule suspended from the floor of the third ventricle. The best images to see hamartomas on are enhanced sagittal T1-weighted MR images. Here you can see the non-enhancing hamartoma attached to the tuber cinereum between the pituitary stalk and mamillary body. There really is no differential diagnosis. Gliomas can occur in any part of the brain and the optic chiasm is a common location, particularly in patients with neurofibromatosis type 1. This enhanced CT shows an example of an optic nerve glioma in a patient with neurofibromatosis. There is a suprasellar mass which is indistinguishable from the optic chiasm. Further forward at the level of the orbits the optic nerve is abnormal on both sides. These consecutive coronal MR-images show the mass at the optic chiasm and the swollen optic nerves. On these axial images you can see the optic nerves and chiasm enhance after the administration of intravenous gadolinium. The way a patient is normally positioned, slices through the nerves themselves are not obtained. These slices can be used to make oblique images along the axis of the nerves. With these images as a result. Note the enhancement of the nerve after intravenous contrast with sparing of the meninges. Approximately 25% of optic nerve gliomas do not enhance, so a lack of enhancement should not prevent you from making the diagnosis. This is another example of a right-sided optic nerve glioma with enhancement after gadolinium. Note the normal pituitary gland and stalk. The following case concerns a 9-year-old male with a history of headache, nausea and vomiting. Sagittal T1 images before and after intravenous contrast show a mass in the midline, on the floor of the third ventricle. The mass enhances after gadolinium. Continue with next images. T2- and T1 weighted sagittal images of the same patient show a similar mass in the epiphysial area. This is a germinoma - an intracranial germ cell tumor that occurs primarily in children and adolescents. These are typical localisations. These lesions crawl along the floor of the 3rd ventricle. Chondromas are the most common lesions of the clivus, also a favored location for metastases and chondrosarcomas. This patient has a normal pituitary gland. Posterior to this is a large, fungating mass positioned at the level of the clivus. The CT shows some calcifications in this area. The differential diagnosis for this mass would be chondroma or chondrosarcoma. Chordomas tend to occur in the midline, whereas chondrosarcomas tend to occur off the midline. The patient on the left is a patient with lung cancer who presented with a sixth cranial nerve palsy. The abnormality is in the clivus, which should have a high signal intensity on this sagittal T1-weighted image (as in the image on the left). A low signal intensity means the normal fatty marrow has been replaced by some other tissue. In this case by tumor metastasis. Also lymphomas, myelomas or diffuse bone abnormalities can give this appearance. Therefore always take a minute to look at the clivus. So again in order to analyse a sellar or parasellar mass on MRI we use the following anatomic approach:Walter Kucharczyk and Marieke Hazewinkel Aneurysm vs MeningiomaSella Turcica and Parasellar RegionRadiology department of the University of Toronto, Canada and the Radiology department the Medical Centre Alkmaar, the Netherlands
neuro11
1
Spine - Cervical injury
by Adam Flanders
This review is based on a presentation given by Adam Flanders and adapted for the Radiology Assistant by Robin Smithuis. Approximately 3 % of patients who present to the emergency department as the result of a motor vehicle accident or fall have a major injury to the cervical spine. 10-20% patients with head injury also have a cervical spine injury. Up to 17% of patients have a missed or delayed diagnosis of cervical spine injury, with a risk of permanent neurologic deficit after missed injury of 29%. Most cervical spine fractures occur predominantly at two levels. One third of injuries occur at the level of C2, and one half of injuries occur at the level of C6 or C7. In this overview we will discuss the most common cervical spine injuries. You can click on some of the images to get a larger image. The most common fracture mechanism in cervical injuries is hyperflexion.   Unstable fractures: There are two types of injury to the spinal cord: There is a strong correlation between the length of the spinal cord edema and the clinical outcome. The most important factor however is whether there is hemorrhage, since hemorrhagic spinal cord injury has an extremely poor outcome. The chart on the left is showing the motor recovery rate for patients with edema alone (in blue) versus edema plus cord hemorrhage (in red). The motor recovery rate is for the legs only. Spinal cord syndromes (2): On the left images of spinal cord injury after a stab wound with a screwdriver. This resulted in a Brown-Sequard syndrome due to hemisection of the spinal cord. Hyperflexion sprain injuries are injuries to the soft tissues of the spine without fracture. On x-rays this can only be suspected when there is angulation or translation MR will demonstrate subtle injuries to the soft tissues. On the left images of a patient who has been in a car accident and complained of neck pain. The x-rays were normal and there were no neurological symptoms. First study the images on the left. Then continue reading. The findings are: In this patient we can conclude that there was mild hyperflexion strain and we do not know if a special treatment is required, since these were isolated MR-findings without evidence of fracture or abnormal positioning. There is controversy regarding the meaning of soft tissue abnormalities detected only on MRI. Signal changes do not necessarily equate with structural failure. These findings still require better validation. In trauma centres up to 25% of all patients with neck injury have signal abnormalities on MR and the significance is indeterminate. Hyperflexion sprain (2) On the left images of a 44 year old female, who sustained a fall on the ice. She subsequently had a second fall the following morning, where after she had complete loss of motor and sensation. On physical examination there was lower extremity paraparesis with some upper extremity weakness on the right. Central cord injury was proposed initially. The radiographs were normal. First study the images on the left. Then continue reading. The findings are: These CT-findings are very subtle and do not seem to match the neurological problem. In such a case MRI is the next step. First we show you a coronal and axial CT with also a soft tissue window-setting. There is high density material at the back of the disc space, which is very suggestive for a traumatic disc herniation. A epidural hematoma should be in the differential, but this finding was limited to just the area of the disc space, unlike a hematoma. Continue with the MR. Hyperflexion sprain (3) The MRI explains the neurological status of this patient. First study the images on the left. Then continue reading. The MR-findings are: Continue with the axial image. The axial image shows the spinal cord injury and in addition to it there is absence of flow void in the right vertebral artery. This indicates thrombosis as a result of dissection. In conclusion we can say that this patient had no fracture, but a severe hyperflexion sprain with acute disc herniation, non-hemorrhagic spinal cord injury and vertebral thrombosis. The MRA confirms the occlusion of the right vertebral artery. Unilateral interfacet dislocation is due to a hyperflexion injury with rotation. The superior facet on one side slides over the inferior facet and becomes locked. This results in an anterior subluxation of the upper vertebral body of about 25% of the AP diameter of the body. Simple unilateral facet dislocation is a stable injury. 30% of patients have an associated neurologic defect. MRI plays an important role in the diagnosis in order to see if there is disc extrusion leading to cord compression. On the left images of a 20 year old male who had a rollover motor vehicle accident. First study the images on the left. Then continue reading. The radiographic findings are: The CT confirms the unilateral dislocation. The contralateral facetjoint is only distracted. Unilateral interfacetal dislocation (2) On the axial view the left facet joint is normal and the configuration has similarities with the hamburger. On the right side the classic 'inverted hamburger sign' is seen. Continue with the MR-images First study the MR-images. Then continue reading. The MRI-findings are: Bilateral interfacetal dislocation (BID) is the result of extreme hyperflection. There is anterior dislocation of the articular masses with disruption of the posterior ligament complex, posterior longitudinal ligament, the disc and usually also the anterior longitudinal ligament. When the dislocation is complete, the dislocated vertebra is anteriorly displaced one-half of the AP diameter of the vertebral body. Because of its extensive soft tissue damage and dislocated facet joints, BID is unstable and is associated with a high incidence of cord damage. First study the images on the left. Then continue reading. The findings are: On the left CT-images of the same patient, which confirm the bilateral dislocation. Near one of the facets there is a small fleck of bone, but there is no major fracture, so this is basically just a hyperflexion soft tissue injury. On the axial images the inverted hamburger sign is seen on both sides. Bilateral interfacetal dislocation (2) On the left you can scroll through the 3D-reconstructions. Bilateral interfacetal dislocation (3) First study the MR-images for additional findings. Then continue reading. The MRI-findings are: Continue with the axial image. Notice on the axial image that the cord injury is located in the grey matter, which is more sensitive to damage. In order to regain normal alignment, progressive weights are used to lengthen the spine until reduction is achieved. Scroll through the images on the left. Notice, that with 60 pounds the facets start to move, but it finally takes about 110 pounds before the neck is reduced. Because someone is holding on to the neck while more weight is added, an actual 'clunk' can be felt in the neck indicating that reduction is achieved. Conti ue with the MR-images after reduction. Bilateral interfacetal dislocation (5) On the left images of a 15-year old, who was injuried during wrestling. There is 50% anteroposition of C3 on C4 as a result of bilateral interfacetal dislocation. There is complete disruption of the posterior complex. This boy had severe neurologic deficit. Bilateral interfacetal dislocation (6) On the left another bilateral interfacetal dislocation with complete transsection of the cord. This is a very uncommon finding, since the spinal cord is very plastic. This fracture is the result of a combination of flexion and compression, which is usually the result of a motor vehicle accident. The teardrop fragment comes from the anteroinferior aspect of the vertebral body. The larger posterior part of the vertebral body is displaced backward into the spinal canal. On x-rays the facet joints and interspinous distances are usually widened and the disk space may be narrowed. 70% of patients have neurologic deficit. It is an unstable fracture associated with complete disruption of ligaments and anterior cord syndrome. On the left images of a 21 year old male who sustained a diving injury, striking his head in a swimming pool. He had immediate onset of upper and lower extremity weakness. First study the images. Then continue reading. The findings are: Some would just call this a severe hyperflexion injury, but this entity is better known as a 'flexion tear drop' fracture. Look for additional findings on the CT-images and then continue reading. The findings are: Continue with the MR-images. Flexion tear drop fracture (2) Look for additional findings on the MR-images and then continue reading. The findings are: Flexion tear drop fracture (3) On the left images of a similar case. There is a C5 flexion teardrop fracture. Notice that the lower cervical area is not visualised well and additional imaging is required. The CT-images demonstrate the extreme axial loading. The tear drop fragment is displaced anteriorly and the larger part of the vertebral body is displaced posteriorly compressing the spinal cord. Continue with the MR-images. The MR-images demonstrate the spinal injury. It is a hemorrhagic injury, which has a poor outcome. Also notice the posterior ligamentous injury as a result of the hyperflexion with rupture of the ligamentum flavum and CSF leakage. Notice the central location of the spinal cord injury. The Hangman' s fracture is the most common cervical spine fracture. Classically it is an extension-fracture as the hangman puts the knot under the chin to produce maximal extension-force. That is why we discuss the hangman' s fracture in the chapter on hyperextension injuries. In some situations however it can also be the result of extreme flexion. The hangman's fracture is common in diving accidents. Although considered an unstable fracture, it seldom is associated with spinal injury, since the anteroposterior diameter of the spinal canal is greatest at this level, and the fractured pedicles allow decompression. When associated with unilateral or bilateral facet dislocation at the level of C2, this type of hangman's fracture is unstable and has a high rate of neurologic complications. Classification of Hangman' s fractures On the left images of a restrained passenger in a vehicle going about 55 miles per hour . She ran into a tree at about 9 p.m. the previous night with questionable loss of consciousness. She had cervical tenderness to palpation, but was alert and had no neurologic abnormalities on examination. First study the images on the left. Then continue reading. The findings are: Continue with the CT-images. The CT-images confirm the fracture-lines of the hangman's fracture. They run through the pars interarticularis resulting in a traumatic spondylolysis. In this case there was no neurologic deficit, because the spinal canal is widened at the level of the fracture. On the left images of another patient with a type I Hangman' s fracture. There is a hair-line fracture and there is no displacement. On the left images of a 90-year-old male who tripped and fell on his back and the back of his head. He had immediate quadriparesis after the event with no loss of consciousness. First study the images on the left. Then continue reading. The findings are: Continue with the CT-images. On CT we also see that there has been a hyperextension injury. The small black dots in the disc space are the result of a vacuum phenomena. The negative pressure resulted in a vacuum phenomena in the injured disc space. There is also some hyperdensity at the back of C5C6, which could be a herniated disc or just preexisting disc degeneration. In such a patient with spondylosis which has led to narrowing of the canal, a low velocity injury can lead to spinal cord injury. Continue with the MR-images. The MR shows a subtle increase in signal intensity of the spinal cord. Most of the time these patiets get a central cord injury. There is only injury to the central part of the cord and these patients have disproportioned weakness of their arms and normal strength in their legs. These injuries can be devastating, although it is uncommon that they are hemorrhagic. Hyperextension injury (3) On the left two other examples of this hyperextension injury. It is easy to find the injured disc, since it is the one with the high signal (arrows). Notice the prevertebral soft tissue swelling in the case on the right. As with flexion teardrop fracture, extension teardrop fracture also manifests with a displaced anteroinferior bony fragment (6). This fracture occurs when the anterior longitudinal ligament pulls a bony fragment away from the inferior aspect of the vertebra because of the sudden hyperextension. The fragment is a true avulsion, in contrast to the flexion teardrop fracture in which the fragment is produced by compression. This type of fracture is commonly seen in diving accidents and tends to occur at lower cervical levels. It also may be associated with the central cord syndrome due to buckling of the ligamenta flava into spinal canal during the hyperextension phase of injury. This injury is stable in flexion but highly unstable in extension. On the left images of a 70 year old female who fell down ten steps striking her head resulting in a subgaleal hematoma with possible loss of consciousness. There was no neurologic deficit. Notice the anteroinferior bony fragment of C2. Continue with the CT images. The CT confirms the displaced anteroinferior bony fragment. This fragment is a true avulsion, in contrast to the flexion teardrop fracture in which the fragment is produced by compression of the anterior vertebral aspect due to hyperflexion. Continue with the MR images. The MR also confirms that this is not a flexion injury, since the soft tissue injury is located anteriorly. Extension-teardrop fracture (2) On the left another extension-teardrop fracture of C2. This fracture is caused by a compressive downward force that is transmitted evenly through the occipital condyles to the superior articular surfaces of the lateral masses of C1. This process displaces the masses laterally and causes fractures of the anterior and posterior arches, along with possible disruption of the transverse ligament. Radiographically the fracture is characterized by bilateral lateral displacement of the articular masses of C1. Odontoid or dens-fractures are very common. They are seen in elderly, but also frequently in children due to the relatively large head-to-spine ratio. Classification On the left the most common type of odontoid fracture, which is type II through the base of the odontoid. These type II fractures have a tendency to nonunion, which occurs in 64%. On the left another type II odontoid fracture. Sometimes these fracture-lines can be difficult to see. There are fracture mimics like lucent lines as a result of overprojection or a prominent mach line (figure). Odontoid fracture (2) On the left images of a 26-year-old unrestrained passenger in a MVC who was ejected from the automobile. He had multiple injuries including subdural hematoma, hemothorax, epidural cord bleed, and a T-spine fracture, left L3 transverse process fracture as well as a left clavicle fracture. There was no neurologic deficit at physiacl examination. First study the images, then continue reading. The findings are: Look at the CT-images and then continue reading. The CT confirm the x-ray findings and shows two additional findings: The MR demonstrates: On the left transverse MR-images at the level of the cervical spine and the thoracic spine. Notice that at the thoracic level, there is also a epidural fluid collection, but it is located posteriorly. This resuted from the T-spine fracture. Continue with the axial images. Odontoid fracture (3) On the left coronal CT-images of another type III odontoid fracture. Odontoid fracture (4) On the left images of a type II unstable odontoid fracture. On the left an unstable type III odontoid fracture. Atlanto-occipital dislocation is an uncommon injury characterized by complete disruption of all ligaments between occiput and atlas with subluxation or complete dislocation of the occipitoatlantal factes. Anterior translation of the skull on the vertebral column is the most common presentation. Death usually occurs immediately from stretching of the brainstem, which causes respiratory arrest. It is reported to occur in 31% of fatal MVAs. It is more common in children due to the larger head. Up to 50% of atlanto-occipital dislocations are overlooked initially, which can have catastrophic results, since cervical traction can be fatal. Power's ratio is sometimes used as a measurement for the relationship between skull base and spine, but Harris' measurement for AO-dissociation has a greater sensitivity (figure). On the left images of an unrestrained passenger, who was ejected from a vehicle and found confused and combative at the scene. He was intubated and taken to a hospital, where he was found to be quadriplegic. On the scout view the abnormal relationship between skull and cervical spine is seen. The axial CT-image demonstrates blood surrounding the brainstem. On the images on the left notice the abnormal relationships of the basion, opisthion and the tip of the dens and the posterior arch of the atlas. The subarachnoid space is hyperdense due to the hemorrhage (arrow). On the left the MR-images. Notice the prevertebral hemorrhage and the compression on the cord. The NEXUS criteria state that a patient with suspected c-spine injury can be cleared providing the following:  The lateral view is the most useful view. Approximately 85-90% of spinal injuries are evident on this view. Systematic approach: The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. by Francis Denis. Spine 1983, volume 8, number 8 - 817 in current diagnosis & treatment - Orthopedics - fourth edition by C. Argenson, F. de Peretti, A. Ghabris, P. Eude, J. Lovet, I. Hovorka in Wheeles' Textbook of Orthopaedics by Jerome R. Hoffman et al NEJM Volume 343:94-99 July 13, 2000 Number 2 by Moira DavenportAdam Flanders Flexion injuries Extension injuries Axial compression injuries Stability Spinal cord injury Hyperflexion Sprain Unilateral interfacet dislocation Bilateral Interfacetal Dislocation --> Reduction under fluoroscopy Flexion tear drop fracture Hangman' s fracture Hyperextension with superimposed spondylosis Extension teardrop fracture Jefferson fracture Odontoid fracture Atlanto-occipital dislocationSpine - Cervical injuryDepartment of Radiology and Regional Spinal Cord Injury Center of the Delaware Valley, Thomas Jefferson University Hospital, Philadelphia
neuro12
1
Spine - Myelopathy
by Majda M. Thurnher and Robin Smithuis
In this article we will focus on spinal cord diseases that are characterised by high signal within the cord on T2WI. The most common causes are inflammatory and demyelinating disorders like Multiple Sclerosis, Neuromyelitis Optica, Acute Disseminating Encephalomyelitis and Transverse myelitis. We will discuss the differential diagnosis including tumors, inflammatory and vascular disorders.  If we exclude myelopathy due to cord compression as seen in trauma, degeneration and metastatic disease, which is usually not a diagnostic dilemma, then the most common diseases of the spinal cord are demyelinating diseases. MS is by far the most common demyelinating disease. The diagnosis can only be made if there is dissemination over time and space. Many patients who are diagnosed as having acute disseminating encephalomyelitis (ADEM) or Transverse Myelitis, may have recurrent disease and later turn out to have MS (red arrow). Whenever there is an abnormality in the spinal cord, we need a systematic approach to analyse the findings. Clinical findings can be helpful but can be quite similar in most spinal cord disorders. On MR look for the following:  Transverse images are very helpful in the differential diagnosis. You need high resolution images. Look for how much is involved (both halves or not), which part is involved and what is the form of the involvement.  Brain abnormalities In many cases of myelopathy there will also be brain abnormalities and these can be a diagnostic clue to the diagnosis. key facts: MS is the most common demyelinating disease and there is overlap between these diseases. NMO was first thought to be a form of MS, but is now considered to be a distinct form. ADEM can relapse and progress to MS. The partial form of transverse myelitis Here we have images of a typical case. Many times the clinical history is very helpful like in this case. This 24-year old patient had visual disturbances on one eye followed by weakness and sensory disturbances of the lower and upper extremities a couple of years later. Now she presents with sensory disturbances of both lower extremities. So we already think MS. In the cord there are some well-defined lesions, but also some ill-defined foggy lesions. The transverse image shows the dorsal location and the typical triangular shape. Continue with the contrast-enhanced images On the contrast-enhanced images there is no enhancement. Active MS-lesions in the spine may enhance, but it is not that common as we see in active lesions in the brain. Whenever spinal lesions are encountered, it can be helpful to image the brain aswell. Now sometimes the patient is only sceduled for MRI of the spine and you don't have time to do a full brain examinations. In those cases consider to do only a sagittal FLAIR. Continue with the images of the brain. The MRI of the brain shows periventricular lesions and a lesion in the corpus callosum. These locations are very specific for MS. In another patient there are non-specific lesions in the cord. Based on the examination of the spine alone, we have a broad differential diagnosis. However when we examine the brain, it becomes obvious that we are dealing with MS. Continue with the images of the brain. In this case the findings in the brain are very helpful. The location of the lesions is very typical: pons, periventricular and subcortical. Now what can we expect in the spinal cord of patients with MS. It is like in the brain. Frequently the lesions are focal like we see on the left image. Less commonly there are diffuse abnormalities and then we have a tough differential diagnosis which will include TM and NMO. In long standing cases atrophy will be seen. One third of MS patients will have spinal symptoms. One third of patients have isolated spinal MS without any findings in the brain. However pathologic studies have shown that 95% of MS patients have spinal cord lesions, whether they have spinal symptoms or not. On transverse images MS lesions typically have a round or triangular shape and are located posteriorly or laterally. So can we exclude MS if a lesion is located anteriorly? Unfortunately not. MS is the great mimicker and can also be located anteriorly like in this patient who has a lesion in the typical location (blue arrow) but also a lesion ventrally in the cord (red arrow). This is uncommon, but you can not exclude MS. When MS lesions are active, they can enhance, but enhancement is not as common as in the brain. The enhancement patterns are non-specific. You can see ring enhancement, intense and less-intense enhancement. The less intense or vague enhancement is the most common pattern. These are images of a patient with longstanding MS and acute exacerbation. There is enhancement in the active lesions. Continue with the images of the cervical spine. Also in the spinal cord there are multiple lesions. On the transverse image a typical triangular shaped dorsal lesion is seen. Continue with the contrast-enhanced images. Also in the spine there are multiple enhancing MS-lesions. MS usually presents with focal abnormalities. Diffuse abnormalities that can look like transverse myelitis or extensive astrocytoma are sometimes seen. This pattern is more common in primary progressive and secondary progressive MS. Some say that spinal cord atrophy is specific for primary progressive MS (PPMS). The atrophy correlates very well with the clinical disability. It is more prominent in the upper part of the spinal cord. Duration of the disease is the most important determinant of cord atrophy. key facts: On the left images of a child who presented with unilateral neuritis optica. Images of the brain were otherwise normal. Continue with MRI of the spine. Patients who have one episode of optic neuritis or myelitis and who test positive for NMO-IgG are at high risk for developing the full spectrum of NMO. One month later this child presented with acute transverse myelopathy, i.e. bilateral symptoms. The images show abnormal signal in the spinal cord with swelling and some enhancement. An astrocytoma could very well present with these images, but given the history of an optic neuritis and the acute myelopathy, we do not think of a tumor. This proved to be NMO and the Ig-test for NMO was positive. In the original description of Devic's disease the optic neuritis and the myelopathy were simultaneously, but now we know that this is not always the case. Brain lesions in NMO Previously it was thought that in NMO the brain was spared, but now we know, that brain lesions do occur. They are often distinct from those seen in MS. In Asia 60-80% of patients with NMO have brain abnormalities. In Europe only 25-40%. The location of the brain lesions in NMO is only around the ventricles. The reason why these brain lesions are located around the ventricles is the following: The NMO IgG auto-antibodies are directed against Aquaporin-4 water-channels. So just like sodium- and potassium channels in the cells, there are also water-channels. The highest concentration of these Aquaporin-4 water-channels is seen around the ventricles. The images show abnormal signal around the third and frontal horns of the lateral ventricles. Things become even more complicated, since it is also possible to have large lesions in the corpus callosum of patients with NMO as was described by Nakamura (6). So in any CNS disease with optic nerve and spinal cord involvement it is goos to do the test for NMO-IgG. key facts: On the left images of a teenage child with a typical history: This clinical history is typical for ADEM. Usually the brain is also involved. 30% of cases has spinal involvement. The imaging findings in this case are also typical. There is swelling and cord involvement like in TM and no enhancement. Continue with the images of the brain. First look at the images of the brain and decide what is different from MS-lesions. What is typical for ADEM and uncommon for MS is: The follow up MR shows that the cord has returned to normal again. On the left another case of ADEM. ADEM is typically seen in children. Again there is diffuse involvement of the spinal cord without enhancement and there is involvement of the brain. Another case of ADEM. Notice the typical involvement of the pons and basal ganglia. Continue with follow up scan. On follow up scan almost complete normalisation. key facts: The sagittal image shows a large segment of hyperintensity on T2WI. The transverse image shows that most of the cord is involved. These images are of a 31 year old male with headache, voiding disturbances, urinary retention, sensory level C3. The CSF analysis revealed 400/3 cells (meaning no infection) and a slightly higher protein level. The images show a long segment myelopathy with full transverse involvement. There is no swelling and no enhancement. It does not look like MS or tumor, so we are thinking ATM - acute transverse myelitis. Transverse myelitis may occur in isolation or in the setting of another illness. When it occurs without apparent underlying cause, it is referred to as idiopathic. Idiopathic transverse myelitis is assumed to be the result of abnormal activation of the immune system against the spinal cord. The table provides a list of illnesses associated with TM (7). Patients with an acute short segment TM (or APTM) are at risk of developing MS if there is one of the following: In children with TM the expression is usually more severe. 60-90% are unable to move their legs. However the outcome is usually better than in adults and in 30-50% there is complete recovery. Typical for TM is that on the initial MR the abnormalities are usually extensive and less or completely resolved on follow up scans. Longitudinal case series of TM reveal that approximately 1/3 of patients recover with little to no sequelae, 1/3 are left with moderate degree of permanent disability, and 1/3 have severe disabilities. Here images of a typical case of TM. There is multisegment high signal on STIR and T2WI with some swelling. Most of the cord in the transverse diameter is involved. There is no enhancement, which is usually the case in TM. Sometimes there is some patchy enhancement. When there is enhancement, it can be difficult to differentiate TM from an astrocytoma. On the left images of a 60 year old male with an astrocytoma. He presented with pain in the thoracic region and sensory disturbances in the left lower extremity followed by left hemiparesis. There is multisegment high signal on T2WI with some swelling, just like we have seen in cases of TM. On CE-T1WI there is a region of enhancement. The region of enhancement is more tumor-like, but the differentiation is difficult. As we have just seen, the major differential of the spinal cord diseases that we have discussed so far is an astrocytoma. Astrocytoma is a diffusely infiltrating tumor, that is not mass-like. Usually there is some patchy enhancement. On the left an astrocytoma in a 66 year old patient who presented with progressive sensory complaints. Biopsy revealed astrocytoma. Continue with the follow up. The patient was not operable and a follow up scan shows progressive disease. The other two common spinal cord tumors are ependymoma and hemangioblastoma and they just look like a tumor. They present as enhancing masses and will not cause a differential problem. All other cord tumors are uncommon. The images are of a patient with neurofibromatosis who has multiple ependymomas. They present as multiple enhancing masses. Spinal cord ischemia is typically seen as a complication of aortic aneurysm surgery or stenting. The images are of a patient who developed a paraparesis after stenting of an aortic aneurysm. Notice the high signal ventrally in the chord, which is typical for arterial infarction. On transverse images a typical snake-eye appearance can be seen. Aortic aneurysm stenting is the most common cause of spinal chord infarction. The diffusion images show restricted diffusion. The images are of a child with headache, fever, hyperalgesia and numbness of the left side of the body. There is also abducens nerve palsy. The images are non-specific with multiple focal lesions and probably the first choise would be MS. The differential diagnosis would include inflammation, infection and metastases. In such cases always perform a transverse image of the spine to look for the exact location and perform a MRI of the brain. Continue. The lesions are located dorsally and one of the lesions is enhancing. Now if this was infection or metastases it would be strange that not all lesions enhance. MS is still on our list. Continue with the MR of the brain. On the CE-T1WI only one lesion shows enhancement. The location of the lesions and the enhancement could very well fit to the diagnosis of MS, but this proved to be vasculitis. Vasculitis can be idiopathic, but is also seen in SLE, Sj?gren and Behcet. Normally you think of vasculitis as a disease of the vessels in the brain, but all vasculitis can be seen in the spine as well. It produces MS-like images. The most common vascular malformation of the spinal cord is the dural AV-fistula. It consist of an abnormal connection between the artery and the veins , which can lead to increased venous pressure and predisposes the cord to ischemia and less commonly to hemorrhage. AVF's are mostly seen in the elderly population and are believed to be the result of trauma. An accurate diagnosis is important because these lesions may represent a reversible cause of myelopathy. Notice the high signal in the lower thoracic cord and the surrounding dilated vessels on the T2WI. On the enhanced T1WI there is subtle enhancement. Another case with myelopathy and dilated veins as a result of an AVF. Another patient with myelopathy and dilated vessels surrounding the cord. Notice the hypointense areas on the T2WI which represents hemorrhage. Another AVF with myelopathy and dilated vessels. Although beyond the scope of this article, the most common cause of myelopathy is cord compression as seen in trauma, metastatic disease and epidural hemorrhage. This patient has a fracture with posterior displacement. There is myelopathy due to traumatic cord compression. Another case of cord compression in a patient who was treated with anticoagulantia. There is myelopathy as a result of compression by a dorsally located epidural hemorrhage. The most common cause of cord compression is metastatic disease. Notice the abnormal signal in the vertebral body as a result of a metastasis which extends into the vertebral canal. by Kerr et al Neurology. 2002 Aug 27;59(4):499-505 by Andrea Rossi Neuroimaging Clinics of North America, Volume 18, Issue 1, Pages 149-161 by Anu Jacob, M.D., and Brian G. Weinshenker, M.D., F.R.C.P. by Dean M Wingerchuk, Vanda A Lennon, Claudia F Lucchinetti, Sean J Pittock, Brian G Weinshenker The Lancet Neurology, 2007, 6:805-15 by Nakamura M, Misu T, Fujihara K, Miyazawa I, Nakashima I, Takahashi T, Watanabe S, Itoyama Y. Mult Scler. 2009 Jun;15(6):695-700 by Joanne Lynn, M.D.Majda M. Thurnher and Robin Smithuis Differential diagnosis Systematic approach Short or long segment involvement Transverse involvement Diseases associated with Transverse Myelitis Astrocytoma Arterial infarction Vasculitis Spinal AVFSpine - MyelopathyDepartment of Radiology of the Medical University of Vienna, Austria and Rijnland hospital in Leiderdorp, the Netherlands
neuro13
1
Spine - Thoracolumbar injury
by Adam Flanders
This review is based on a presentation given by Adam Flanders and adapted for the Radiology Assistant by Robin Smithuis. In this overview we will discuss the most common injuries of the thoracolumbar spine. You can click on some of the images to get an enlarged view. In the thoracolumbar spine there are three biomechanical regions. The upper thoracic region (T1-T8) is rigid due to the ribcage which provides stability. The transition zone T9-L2 is the transition between the rigid and kyphotic upper thoracic part and the flexible lordotic lumbar spine. This is where most injuries occur. Finally we have the L3-Sacrum zone which is flexible and this is the region where axial loading injuries occur. In the upper thoracic spine the center of gravity is anterior to the spine. Axial loading will result in compressive forces anteriorly and tensile forces posteriorly. This will result in flexion-type of injuries. In the lumbar spine due to the lordosis, the center of gravity is posteriorly. Flexion type of injuries will straigthen the lumbar spine and result in axial loading. In this area we will see many burstfractures. On the left the three column model of Denis. This model is used to predict the soft tissue injury from bone injury. Spinal stability is dependent on at least two intact columns. When two of the three columns are disrupted, it will allow abnormal segmental motion, i.e. instability. So a simple anterior wedge fracture or just sprain of the posterior ligaments is a stable injury. However a wedge fracture with rupture of the interspinous ligaments is unstable, because the anterior and the posterior column are disrupted. A burst fracture is always unstable because at least the anterior and middle column are disrupted. Criteria to predict soft-tissue injury from bony injury are: On the left images of a 31 year old male. He was working on a roof, fell approximately 5 meters landing on his feet. He complained of pain in left lower extremity and lower back. First study the images, then continue reading. On the x-ray there is a hyperflexion injury of L1 with involvement of the anterior column and possible involvement of the middle column. The sagittal reconstructions of the CT demonstrate that the posterior part of the vertebral body is of normal height, but there is some involvement of the posterior part of the vertebral body. There is debate on how to treat these patients and if there is any role of MRI in these cases. If you are aggressive you could call this a two column injury, which would require stabilizing surgery. If you are conservative you could call this an injury with only minor involvement of the middle column. On the left a coronal reconstruction and an axial image at the level of the fracture. Continue with the MR. The MR images show bone marrow edema in the involved vertebral body, but no additional soft tissue injury. Based on the fact that the MR did not show any additional findings, this patient was treated as having a single column injury. Consultation with orthopedic surgery recommended conservative management with a TLSO brace. Nowadays there is a tendency to treat these thoracolumbar injuries conservatively, even if there is slight involvement of the middle column. The role of MRI in these cases is not clear yet. On the left a fracture of the calcaneus and a lumbar spine fracture. This is called a 'jumpers fracture' or a 'lover's fracture', because it is usely seen in people jumping out of a window to escape from the police or a jealous husband. In this case it is clear that we are looking at an unstable fracture, because this is a burst fracture. Both the anterior and the middle column are disrupted. In addition there is edema in the posterior soft tissues indicating that there is also involvement of the posterior column. Notice also the marrow edema in the adjacent bodies due to the severe axial loading. On the left images of a 21-year-old female who presented after sustaining a seatbelt type injury. She had an exploratory laparotomy for repair of a ruptured duodenum. There was no neurologic deficit. First study the images, then continue reading. What we see is a classic example of a chance fracture, which is a three column injury with a horizontal orientation of the fracture. Continue with the CT-images. What is unique about the Chance fracture is the horizontal orientation, which is nicely demonstrated on the sagittal reconstructions on the left. Continue with the coronal reconstructions. Also on the coronal reconstructions we can see the horizontal orientation of the fracture. What type of circumstances results in a fracture of this type? The classic mechanism of this injury is a lap-belt injury. If you don't have an addtional shoulder belt, the body will fold over. Chance fracture (2) On the left another example of a Chance fracture. Chance fracture (3) On the left a Chance variant. This is a pure ligamentous injury, which is analogous to bilateral interfacet dislocation, which is also a pure ligamentous injury. There is rupture of the interspinous ligament, dislocation of the facet joints and a horizontal rupture of the disc. Pure ligamentous and combined osseous / ligamentous variants have an increased risk of instability compared to the osseus type. Always look for a split of the posterior elements, disc widening or widening of the spinous processes and facets. by Clare J. Groves et al. Radiology 2005;236:601 by Georges Y. El-Khoury in the VirtualAdam Flanders Biomechanics StabilitySpine - Thoracolumbar injuryDepartment of Radiology and Regional Spinal Cord Injury Center of the Delaware Valley, Thomas Jefferson University Hospital, Philadelphia
ped2
1
Diagnostic Imaging in Child Abuse
by Simon Robben
Child abuse is a relatively common problem in our society. In the U.S it is estimated that 4 million children a year are abused in some manner. At least two thousand children die as a result of this abuse. This overview focusses on the role of diagnostic imaging in depicting the findings that are specific for child abuse. Awareness of the radiologist is essential in finding these skeletal and CNS injuries in order to document child abuse, to stop further abuse and to protect siblings. by Simon Robben Battered child syndrome, shaken infant syndrome, stress-related infant abuse and non accidental trauma are all terms to describe the complex of non-accidental injuries in infants and young children as a result of abuse. The term shaken infant syndrome probably best describes the classic pattern of injuries. The child is held around the chest and violently shaken back and forth. This causes the extremities and the head to flail back and forth in a whiplash movement. Intracranial injury occurs as a result of severe angular acceleration, deceleration and direct impact as the head strikes a solid object. The chest is compressed resulting in rib fractures. Arms and legs move about in a whiplash movement resulting in the typical 'corner' or 'bucket-handle'-fractures in the metaphyseal region. The ability to identify child abuse constitutes an important concern to those involved in the medical care of children. Studies show that at least 10% of children under 5 years old who are brought to the emergency room with alleged accidents have actually suffered nonaccidental trauma. Since as many as 65% of all abuse cases are initially seen in the emergency room, the first step in correctly identifying abuse is to train hospital staff members to recognize abuse indicators. The wide range of findings, which can mimic other disease processes, results in misdiagnosis of many cases of inflicted head trauma. Jenny and colleagues reported that 31% of confirmed abusive head trauma cases were missed on initial presentation and many infants sustained additional injury because of the delay in diagnosis. The radiologist can be the first to suggest the diagnosis on the basis of CT studies performed to evaluate for seizures or other neurologic symptoms or on X-rays performed for other reasons. A high degree of suspicion, inability to explain the degree of injury or a reported mechanism of injury, that is inconsistent with the physical findings should alert the radiologist to possible inflicted injury. A protocol for imaging in suspected abuse should be present to provide high quality radiographs. The future safety of a child with the shaking infant syndrome rests on the radiologist's ability to recognize these characteristic features. When we look at X-rays at the emergency department, we have to realize, that the forces needed to break a bone in an infant or young child are enormous. Any fracture in this age group indicates a major traumatic event, not just a fall from a low height. Fractures with a high specificity for child abuse are listed in the table on the left. The classical metaphyseal corner or bucket handle fracture is virtually pathognomonic for abuse. Rib fractures are very common and highly specific for abuse in young children less than 2 year. Fractures of the acromion, sternum and spinous processes are so rare in other conditions, that this affords them a high specificity for abuse. Occipital impression and other skull fractures occur when the head strikes a solid object. The corner fracture was first described by Caffey who noted these peculiar fractures in children with subdural hematomas. When a small piece of bone is avulsed due to shearing forces on the fragile growth plate it is seen as the typical corner fracture. These fractures are often subtle, and the likelihood of detection is directly related to the quality of the radiologic studies. It is for this reason that skeletal surveys in cases of suspected infant abuse must be performed with utmost attention to the quality of the radiographs. Bucket handle fractures These fractures are essentially the same as corner fractures. The avulsed bone fragment is larger and seen 'en face' as a disc or bucket handle. These corner and bucket handle fractures are most common in the tibia, distal femora and proximal humeri. They are frequently bilateral. In violent shaking the child is held very tightly around the chest and squeezed while being shaken. This compresses the ribs front to back and tends to break them next to their attachment to vertebrae and laterally where they are being literally almost folded in half. Therefore, lateral and posterior rib fractures are highly specific for abuse. CPR is rarely, if ever, a cause of such fractures. These rib fractures in abused children may be found incidentally on chest X-rays performed for other reasons, such as evaluation for pneumonia. Rib fractures are very common and highly specific for abuse. In 31 children who died as a result of child abuse, there was an extremely high incidence of ribfractures and metaphyseal fractures. Rib fractures pose difficulties similar to those of metaphyseal injuries in that they are easily overlooked on radiographs. These fractures usually are not evident on radiographs in the acute stage, as little displacement occurs. They are identified in the healing stage as a result of associated callus. Skull fractures are common child abuse injuries, but they are also common in accidental trauma. Patterns of skull fracture that suggest child abuse are: - Multiple 'eggshell' fractures - Occipital impression fractures - Fractures crossing sutures The infant's skull is very resistent to trauma, so any fracture that is inconsistent with the history should raise the question of non-accidental injury. Diaphyseal fractures are non-specific as they do occur in both accidental and non-accidental injury. However in these cases the age of the child and the history become very important. A fall out of a bed will usually not produce a diaphyseal fracture. In order to break a femur you have to fold it with enormous power. Spiral fractures are a result of twisting forces which are uncommon in accidents in young children, but more common in adults. So a simple fall does not produce a spiral fracture in a child. Callus in diaphyseal fractures generally forms no earlier than 5 days after a fracture, but will usually form by 14 days. Thus, fractures without visible callus may be up to 14 days old, and fractures which demonstrate a little bit of callus are at least 5 days old. Large amounts of callus indicate at least 2 weeks old. These signs can be used to correlate with the history. For instance a child that fell out of bed the day before cannot have a fracture with callus formation. Metaphyseal fractures do not typically heal with callus as there is no periosteal disruption, so dating of metaphyseal fractures is difficult. CNS injury related to nonaccidental injury is a leading cause of morbidity and mortality in infants and children. Some state that eighty percent of deaths of children under 2 years of age result from nonaccidental head trauma. A baby's neck muscles are very weak and its head is large and heavy in proportion to the rest of its body. The infant brain is poorly myelinated and is surrounded by larger subarachnoid spaces than the brain in older children and adults. When a baby is shaken, the neck snaps back and forth, much like in a whiplash injury, causing the brain to hit the front and back of the skull. This can damage the brain and cause it to bruise, bleed and swell. Imaging studies of the head may show subdural or subarachnoid bleeding, diffuse axonal injury and associated cerebral edema or older injuries such as subdural effusions. The case on the left shows an ultrasound examination that demonstrates an old and a new subdural hematoma in an abused child. Subdural hematomas arise from disruption of delicate bridging veins extending from the cortex to the dural sinuses. Although bleeding can occur at any site, the tendency is for blood to extend into the posterior interhemispheric fissure. MR examination is even more sensitive in detecting subdural hematomas. The case on the left shows chronic bilateral subdural hematomas and new subdural hematomas in the right frontal and posterior interhemispheric region. The bright signal is a result of methemoglobin indicating subacute hematoma ( about one week old). Visceral injury Visceral injury is seen at autopsy of young infants, but it is rarely documented radiologically in living victims less than 1 year of age. It is estimated that 2-10% of all abdominal injury results from child abuse. The mean age of these children is about 2 years, which is older than the cases we have discussed before. It is more common in boys than girls. The mortality rate is 50% due to 'patients and doctors delay'. These children are brought to the hospital days after the injury, when a perforation already has resulted in peritonitis and sepsis. The history given by the abusers usually does not correlate with the symptoms, which makes these cases very difficult to evaluate for the clinician. The most common non-accidental abdominal injuries are: - visceral perforation or hematoma - liver- and pancreatic laceration - adrenal bleeding Suprisingly the most common abdominal accidental injuries, which are laceration or subcapsular bleeding of the spleen and the kidney, are unusual in these children. The figure on the left shows a case of pancreatic laceration in child abuse. The figure on the left shows a case of liver laceration in child abuse. These abdominal injuries are non specific and could also be attributed to accidental injury. However in most of these cases of child abuse, there is a history that does not correlate well with the injuries, that are found. So you have to look for other more specific skeletal injuries in these children. Retinal hemorrhage Retinal hemorrhage is seen in nearly all cases of infant abuse in which shaking is documented. Cervical spine compression Cervical spine compression results as shaking or impact injury damages the spinal cord. Infants are vulnerable to spinal cord injury because of their large head and weak underdeveloped paraspinous and neck musculature. Spinal cord injury may be difficult to document. These infants may exhibit apnea or vasomotor collapse similar to spinal shock. Radiographic skeletal survey is necessary in all children less than 2 years old suspected of abuse. It consists of individual AP X-rays of chest, skull (also lateral) and extremities. In children 12 months or younger, also perform a lateral thoracolumbar spine film. Head CT scan shoud be performed on all suspected abuse victims 1 year of age or younger and in all children with neurological symptoms. Expert attention to technique and detail is necessary to get quality radiographs that show some of the very subtle injuries of abuse. Do not perform a 'babygram'. Remember that these are the radiographs that will go to court. Repeated skeletal imaging in 7-10 days may provide evidence of a healing injury, that was inapparent on the initial study. Nuclear bone scan is usually not necessary. Perform this if there are equivocal findings on the skeletal survey or if there is a high clinical suspicion of skeletal injury but the skeletal survey is normal. Plain X-rays of the skeleton in the areas of abnormality identified at bone scan, are still needed to evaluate for the exact nature of the abnormality. Accidental injury Accidental subdural hemorrhages have been reported in infants after motor vehicle collisions or falls involving substantial angular deceleration. In cases of accidental head injury, the history is clear and consistent, the infant's symptoms reflect the forces described, and no unexplained skeletal injuries are identified. Birth trauma resulting from high birth weight and traumatic delivery has been postulated as a cause of rib fractures in infants, but this is extremely rare (figure). Rib fractures associated with accidental trauma are rare and require significant force to produce such as direct chest wall trauma from motor vehicle crashes, because the elastic and more flexible chest wall of infants allows for greater compression without injury. Cardiopulmonary resuscitation also has been implicated as a cause for rib fractures. Many critically ill children receive CPR and have no evidence of rib fractures, however, including children with osteogenesis imperfecta. More important, CPR does not cause posterior rib fractures. Coagulopathies A variety of coagulopathies is associated with intracranial hemorrhage in infants, including hemophilia and hypoprothrombinemia caused by vitamin K deficiency. These disorders are suggested by the clinical history, physical findings, and laboratory tests. Osteogenesis imperfecta Osteogenesis imperfecta is a rare inherited disorder of connective tissue. Other skeletal findings in these patients are generalized osteoporosis, wormian bones, bowing and angulation of healed fractures and progressive scoliosis. In addition to fractures, suggestive findings include blue sclerae, hearing impairment, dentinogenesis imperfecta, hypermobility of the joints, bruising and short stature. Subdural hemorrhage is a rare complication of the disease. Menke's disease Menke's disease is a very uncommon inborm error of metabolism. In these patient's small metaphyseal hooks can be seen that resemble corner fractures. Metaphyseal dysplasia A case of metaphyseal dysplasia is shown on the left. In these children the form of the metaphysis is irregular resembling an old corner fracture. Caffey's disease This is a rare disease of unknown etiology. These children have extreme periosteal reactions. Am. J. Roentgenol. Kleinman 155 (4): 703. Review article by Paul K. Kleinman The metaphyseal lesion in abused infants: a radiologic-histopathologic study PK Kleinman, SC Marks, and B Blackbourne, Am. J. Roentgenol., May 1986; 146: 895 - 905.Simon Robben Corner fracture Rib fractures Skull fractures Diaphyseal fractures Fracture healing Subdural hematomaDiagnostic Imaging in Child AbuseRadiology Departement of the Maastricht University Hospital in the Netherlands
ped4
1
Neonatal Brain US
by Erik Beek and Floris Groenendaal
Cranial sonography (US) is the most widely used neuroimaging procedure in premature infants. US helps in assessing the neurologic status of the child, since clinical examination and symptoms are often nonspecific. It gives information about immediate and long term prognosis. by Erik Beek and Floris Groenendaal Ultrasound is a fast and bedside examination which makes it ideal for premature infants. Try to get all the information you can. Do not limit yourself to only one transducer or only one acustic window (figure). Generally the large fontanel is used as acoustic window. The small fontanel however is a good window to the occipital lobes. This can be usefull in patients with borderline hyperechogenicity in these areas. Disadvantages of US are: PVL is also known as Hypoxic-Ischemic Encephalopathy (HIE) of the preterm. It is a white matter disease that affects the periventricular zones. In prematures this white matter zone is a watershed zone between deep and superficial vessels. Until recently ischemia was thought to be the single cause of PVL, but probably other causes (infection, vasculitis) play an additional role. PVL presents as areas of increased periventricular echogenicity. Normally the echogenicity of the periventricular white matter should be less than the echogenicity of the choroid plexus. PVL occurs most commonly in premature infants born at less than 33 weeks gestation (38% PVL) and less than 1500 g birth weight (45% PVL). Detection of PVL is important because a significant percentage of surviving premature infants with PVL develop cerebral palsy, intellectual impairment or visual disturbances. More than 50% of infants with PVL or grade III hemmorrhage develop cerebral palsy. PVL is graded according to the signs as listed in the Table on the left. Regular sonographic examination is mandatory as cysts in PVL can develop as long as 4 weeks after birth (especially in prematures Cranial ultrasonographic findings may be normal in patients who go on to develop clinical and delayed imaging findings of PVL. A good protocol is US-examination at least once a week until discharge ?nd at the age of 40 weeks. PVL grade 1 PVL is diagnosed as grade 1 if there are areas of increased periventricular echogenicity without any cyst formation persisting for more than 7 days. Increased periventricular echogenicity is however a nonspecific finding that must be differentiated from the normal periventricular halo or normal hyperechoic 'blush' posterosuperior to the ventricular trigones. Suspect PVL if the echogenicity is asymmetric, coarse, globular or more hyperechoic than the choroid plexus. The abnormal periventricular echotexture of PVL usually disappears at 2-3 weeks. PVL can be differentiated from hemorrhages because PVL lacks mass effect. PVL grade 2 The images on the left demonstrate a PVL grade 2 with small periventricular cysts. The echogenicity has resolved at the time of cyst formation. 2% of the preterm neonates born before 32 weeks develop cystic PVL. The severity of PVL is related to the size and distribution of these cysts. Cystic PVL has been identified on cranial ultrasounds on the first day of life, indicating that the adverse event was at least 2 weeks prenatal rather than perinatal or postnatal. US is highly reliable in the detection of cystic WM injury (PVL grade II or more), but has significant limitations in the demonstration of noncystic WM injury (PVL grade I). This deficiency of neonatal cranial US is important, because noncystic WM injury is considerably more common than cystic WM injury. PVL grade 3 PVL is diagnosed as grade 3 if there are areas of increased periventricular echogenicity, that develop into extensive periventricular cysts in the occipital and fronto-parietal region. PVL grade 4 PVL is diagnosed as grade 4 if there are areas of increased periventricular echogenicity in the deep white matter developing into extensive subcortical cysts. PVL grade 4 is seen mostly in fullterm neonates as opposed to PVL grade 1-3, which is a disease of the preterm neonate. Flaring persisting beyond the first week of life is by definition PVL garde 1. The term flaring is used to describe the slightly echogenic periventricular zones, that are seen in many premature infants in the first week of life. During this first week it is not sure if this is a normal variant or a sign of PVL grade 1. Flaring persisting beyond the first week of life is by definition PVL grade 1. Follow up is needed to differentiate flaring from PVL grade I. The case on the left shows a premature infant with flaring. At follow up no cyst formation was found and after the first week a normal periventricular white matter was seen. Germinal matrix hemorrhage (GMH) is also known as periventricular hemorrhage or preterm caudothalamic hemorrhage. These germinal matrix hemorrhages occur in the highly vascular but also stress sensitive germinal matrix, which is located in the caudothalamic groove. This is the subependymal region between the caudate nucleus and thalamus. Most infants are asymptomatic or demonstrate subtle signs that are easily overlooked. These hemorhages are subsequently found on surveillance sonography. Grade 1 intracranial hemorrhage On the left an intracranial hemorrhage confined to the caudothalamic groove. It is staged as grade 1 hemorrhage. In the acute phase these bleedings are hyperechoic, changing to iso- and hypo-echoic with time. Grade 2 intracranial hemorrhage On the left a grade 2 intracranial hemorrhage. On the coronal image only the cavum septi pellucidi is seen. Both lateral ventricles are filled with blood, but there is no ventricular dilatation. On the left the same patient after 3 days. The ventricles are dilated and clot formation is seen. Secondary hydrocephalus occuring several days after a grade 2 bleed should not be mislabeled as grade 3 hemorrhage. Grade 3 intracranial hemorrhage On the left a grade 3 intracranial hemorrhage filling the left lateral ventricle. Also note the wedge shaped hyperechoic area on the laterosuperior side of the ventricle. This represents a small venous infarction. Same patient as above. Two weeks later the venous infarction has developed into a hypoechoic area with cyst formation. Grade 4 intracranial hemorrhage Originally these grade 4 hemorrhages were thought to result from subependymal bleeding into the adjacent brain. Today however most regard these grade 4 hemorrhages to be venous hemorrhagic infartions, which are the result of compression of the outflow of the veins by the subependymal hemorrhage. On the left a grade 4 hemorrhage. There is a large subependymal bleeding but also a large area with increased echogenicity in the brain parenchyma lateral to the ventricle. This is probably the result of a venous infarct. These venous infarctions resolve with cyst formation. These cysts can merge with the lateral ventricle, finally resulting into a porencephalic cyst. On the left a different patient with a grade 4 hemorrhage at a later stage with extensive cyst formation. Grade 1 and 2 bleeds generally have a good prognosis. Grade 3 and 4 bleeds have variable long-term deficits, but outcome in grade 3 hemorrhages is usually good when no parenchymal injury has occurred. Hydrocephalus is a common complication and many infants require ventriculoperitoneal shunting. The mechanisms by which hydrocephalus develop include:  Common variants are listed in the Table on the left. Well known variants are the cavum of the septum pellucidum and the cavum vergae. The more premature the baby, the more frequently these cavities are present. They can persist until adulthood. A less frequently seen variant is the cavum of the velum interpositum. This presents as a cyst-like structure in the region of the tectum. It's shape is compared to a helmet. It can easily be confused with a subarachnoid cyst or a cyst of the pineal gland. In postnatal US these cysts of the chorio?d plexus are often incidental findings without clinical consequences. Chorio?d plexus cysts (CPC) are however of importance for obstetricians. At prenatal US these cysts can be predictive of trisomy 18. About half of babies with Trisomy 18 show a CPC on ultrasound, but nearly all of these babies will also have other abnormalities on the ultrasound, especially in the heart, hand, and feet. An exeption must be made for cysts that arise close to the foramen of Monro. Although these cysts often disappear spontaneously, follow up US is necessary to ensure disappearance. Some may produce symptoms of raised intracranial pressure due to obstruction to the cerebrospinal fluid (CSF) flow. Benign macrocrania is also known as extraventricular obstructive hydrocephalus. This is seen in children between 6 months and 2 years. The head circumference is above the 97th percentile. After the age of 2 years the head size normalizes. Often the mother or father of the child had large heads at that age. The cause is not known. Most state that it is a normal condition, although some state that these children have a slight developmental delay. When children with a large head are presented for US, examine the superficial subarachnoid space and the ventricles. Normal subarachnoid space measures The ventricles are often slightly enlarged. Thes prominent subarachnoid space and ventricular system in these children should not be interpreted as cerebral atrophia, as in atrophia there is a small head circumpherence. Mineralizing vasculopathy can be seen in the thalamostriatal and lenticulostriatal arteries and is caused by calcification of the arterial wall. A wide range of perinatal, acquired, and nonspecific clinical conditions may result in this sonographic finding. In the Wilhelmina Children's Hospital these children used to be tested for TORCH-infection, but currently the only test that is done is a urine-test for CMV. Germinolytic cysts Are located at the caudothalamic groove. They are tear shaped. There are no signs of intracerebral hemorrhage and these children have no neurological sequelae. The etiology is not known. Pseudocyst These are also called coarctation of the lateral ventricle. They are often bilaterally and have no neurological sequelae If cysts are seen around the lateral ventricles, it is important to determine their position in regard to the upper part of the lateral ventricle (figure). Measurement of the ventricular system should be done in an easy reproducible sonographic plane. Use a coronal section through the lateral ventricles slightly posterior to the foramen of Monro. You will see 3 echogenic dots representing the choroid plexus in the lateral ventricles and in the roof of the third ventricle. Furthermore you should see a symmetrical image of the Sylvian fissure on both sides and the hippocampus (green and orange arrows). Up to 40 weeks of gestational age the Levene-index should be used and after 40 weeks the ventricular index. The Levene index is the absolute distance between the falx and the lateral wall of the anterior horn in the coronal plane at the level of the third ventricle. This is performed for the left and right side. These measurements can be compared to the reference curve and are quite usefull for further follow-up. After 40 weeks the ventricular index or frontal horn ratio should be used, i.e. the ratio of the distance between the lateral sides of the ventricles and the biparietal diameter. When using this ratio you have to realise, that when the ventricular system widens, the frontal horns tend to enlarge in the craniocaudal direction more than in the left to right dimension. Real-time ultrasound was used to make exact measurements from the lateral wall of the body of the lateral ventricle to the falx (the ventricular index) in 273 infants of varying gestational ages (5). The measurement performed in an axial plane through the temporoparietal bone correlated closely with an actual measurement made in coronal plane in 50 infants. A cross-sectional centile chart was drawn up of the normal range for this measurement from 27 to 42 weeks' postmenstrual age. A further chart showing the rate of change of the ventricular index allowed growth of the ventricles to be assessed in a longitudinal manner. Use of these charts permits early detection of hydrocephalus or dilated ventricles secondary to cerebral atrophy. A more realiable indicator of widening of the ventricular system would be an area- or volume-measurement. This however is more time consuming. So although ventricular index has shortcommings it is still the most commonly used. In general practise, studying the images by eye is reliable, provided, that standard planes are used. An Atlas of Neonatal Brain Sonography by Paul Govaert, Gent University Hospital and Linda S. de Vries, Wilhelmina Kinderziekenhuis. with contributions of Frank van Bel, Erik Beek, Dirk Voet, An Bael, Linde Goossens Ariadne M. Roelants-van Rijn, Floris Groenendaal, F. J. A. Beek, Paula Eken, Ingrid C. van Haastert, Linda S. de Vries. in eMedicine, Author: Terence Zach, MD, in eMedicine, Author: David J Annibale, MD Levene MI., Measurement of the growth of the lateral ventricles in preterm infants with real-time ultrasound. Arch Dis Child. 1981 Dec;56(12):900-4.Erik Beek and Floris Groenendaal Grading PVL Cavum septi pellucidi, cavum vergae and cavum of the velum interpositum Chorio?d plexus cyst Benign macrocrania Mineralizing vasculopathy Cysts Levene index Ventricular indexNeonatal Brain USDepartment of Radiology and Neonatology of the Wilhelmina Children's Hospital and the University Medical Centre of Utrecht, the Netherlands