Read the French translation: Les Pièges des Radiographies du Thorax--Comment les Éviter?
There are a number of unapparent sources of trouble, or pitfalls that can make interpretation of the thoracic radiograph more difficult. These are often normal variations that are diagnosed as disease, or disease that is overlooked due to technical issues with the radiographic images. This lecture will discuss a variety of frequently seen normal variations that are confused with disease with numerous case examples in dogs and cats. In addition, the participant will learn practical approaches to performing thoracic radiographs and interpreting common radiographic diseases through case examples. Tunnel vision and how it can lead to misdiagnosis will also be discussed as a pitfall of radiographic interpretation.
Unfortunately, technical errors leading to poor radiographic quality are the most common cause of improper diagnosis. The most common errors include: not including the entire thorax in the collimated field, not centering, not taking on inspiration, rotation, respiratory and patient motion, under or over exposure and developing errors.
A very common problem that makes analysis of the pulmonary parenchyma difficult if not impossible. This is compounded when the animal is also obese. The resulting increased opacity of the lung appears like a generalized interstitial or bronchial pattern. The cardiac silhouette also appears larger during expiration. In small breed dogs, such as the Yorkshire Terrier or Dachshund, the increased soft tissue opacity in the caudal lung fields due to expiration, combined with a small thoracic shape and apparent large heart, can be misdiagnosed as cardiogenic pulmonary edema. In obese animals, obtaining good inspiration is difficult and diagnosing cardiogenic edema equally so. In obese dogs, the dorsoventral projection is crucial to visualizing the air-filled lung superimposed with the diaphragm. Here one can scrutinize the pulmonary vessels for their border definition and assess the lung for patchy unstructured interstitial or alveolar disease when trying to rule out decompensated left heart failure.
In any position, the dependent portion of the lung will quickly develop a decreased air volume which appears as an increased soft tissue opacity. It can also cause shifting of the heart to the side of decreased volume. In obese animals, right and left lateral as well as ventro- and dorsoventral radiographs can be used to allow the non-dependent, air-filled lung to be analyzed in order not to miss pathology due to positional atelectasis. Luckily, positional atelectasis is dynamic and turning the animal into a different position with the affected lung non-dependent, allows repeat radiographs to be performed to rule out this artifact.
Oblique VD or DV projections can simulate unilateral increased opacity or unilateral hyperlucency. The lungs will look more opaque in the lateral projections than in the orthogonal views. This is due to the atelectasis occurring in the dependent lung which has a decreased air content and increased opacity and creates a summation image of increased opacity over the lungs. If an increased opacity is suspected in the right or left lateral projections, the same increased opacity should be appreciated in the VD or DV projections if disease is actually present. Rotated lateral projections falsely project the trachea closer to the spine simulating cardiomegaly.
Obesity leads to a wide variety of pitfalls and often poses a diagnostic challenge. A widened cranial mediastinum can appear like a cranial mediastinal mass. In such cases, the position of the trachea should be observed for dorsal or right sided deviation, which is suggestive of a mass. A widened mediastinum due to fat can also simulate fluid accumulation. Obese animals also appear to have cardiomegaly. Fat surrounding the heart, however, can usually be differentiated from the soft tissue opacity of the heart. The surrounding fat appears as a hazy halo and is especially visible in the ventrodorsal projection. Intrathoracic fat often can be confused with pleural fluid and has been called "pseudo pleural effusion". In order to differentiate pleural fat bands from pleural fissure lines, the shape of the soft tissue band should be assessed. In fat animals, the "widened fissures" are wider centrally, while actual fluid lines are widest at their lateral aspect, close to the thoracic wall. In the fat cat, the tip of the caudal lung lobes can be displaced ventrally, away from the spine in the lateral view and be confused with pleural effusion. An important tip, is that intrathoracic fat does not change in position between VD and DV projections. Re-positioning the patient and comparing the radiographs are important steps to ruling out the presence of pleural fluid in obese animals. In the caudoventral mediastinum, there is a soft tissue band that can be identified on the ventrodorsal projections that can become quite large in diameter in obese animals and should not be confused with a mass or pleural fluid.
Extrathoracic / Extrapleural Structures
Nipples can appear as pulmonary masses. Faint, round, soft tissue nodules only visible in the ventrodorsal or dorsoventral projections, lateral to the cardiac silhouette should be located in the lateral projections. Nipples are easily visible in the lateral projection and comparisons can be made with the VD and DV. The best way to rule out the nipple as a source of a pulmonary nodule is to place barium on the nipples and re-radiograph the dog. We also use human nipple markers for mammography that come in convenient tape rolls and avoid the smearing of barium on the skin. Skin folds can simulate pleural fluid and pneumothorax. Any linear soft tissue opacity running the length of the patient should be traced along its entire length. Skin folds can always be traced to the front or hind limb flank. Furthermore, vessels can be detected lateral to skin fold lines whereas this is not the case with pneumothorax. Chondrodystrophic breeds have pleural indentations due to their costochondral deformities. In the VD or DV radiographs, this will appear as a soft tissue zone adjacent to the thoracic wall that simulates the appearance of pleural fluid. This can be recognized by the lack of pleural fissure lines. Large extrathoracic masses can lead to an increased lung opacity due to summation effect. If lateral and VD radiographs are performed, detection of the mass is generally not an issue and the cause of the increased opacity in one of the projections quickly becomes apparent.
Sedation, Anesthesia and Fluid Therapy
The heart can appear larger and the lungs more vascular and opaque when under sedation or general anesthesia. Also, positional atelectasis is often present, creating a mediastinal shift and increased opacity and decreased volume of the lobes lateral to the heart. Animals that have undergone bronchoalveolar lavage that receive thoracic radiographs following the procedure may have alveolar patterns and lung lobe consolidation in the lobes that were dependent during the procedure. Small cats and small dogs that have received an intravenous fluid overload can develop cardiomegaly, pleural fluid and increased lung opacity that can mimic heart failure.
Head and diaphragm positioning can cause kinking that simulates a cranial mediastinal or tracheobronchial mass due to lymphadenopathy. The trachea is a flexible tube that is attached at the larynx and carina and can bend in the middle. Excessive flexion or extension of the head can lead to a kink or undulating course of the trachea. Moreover, when the diaphragm is displaced cranially as with an upper airway obstruction, obesity and expiration, the trachea can also appear kinked. Careful attention to the VD or DV projection for lateral displacement is warranted. Also, if there is no evidence of an increased soft tissue mass or other space occupying lesion ventral to the trachea where it is kinked, then the head and diaphragmatic position should be scrutinized.
Enlarged tracheobronchial lymph nodes can create a large zone of soft tissue opacity at the carina. This can be confused with cardiogenic pulmonary edema if attention is not paid to the position of the stem bronchi. In instances of an enlarged left heart, the stem bronchi will be displaced dorsally. When the tracheobronchial lymph nodes are enlarged, the stem bronchi are displaced ventrally and the trachea will appear to have an undulating course. Furthermore, cardiomegaly is generally lacking when heart failure is not present.
Alveolar Pattern and Air Bronchograms
An air bronchogram is the hallmark sign of an alveolar pattern. Alveolar disease is due to the presence of cells or fluid. Not all patients with alveolar disease have air bronchograms and especially in cats. This is due to alveolar material not being extensive enough to surround an airway to create the air bronchogram. Air bronchograms are more easily seen in lobar disease versus multifocal alveolar disease. Intense, ill-defined soft tissue opacities that efface the borders of vessels should be considered alveolar even if air bronchograms are not present.
Normal lung can also be misinterpreted to have air bronchograms. The bronchus between two adjacent vessels can look suspiciously like an air-filled linear track similar to an air bronchogram. The key point is that the vessels are visible and well delineated.
Heart Size and Thoracic Conformation
Depending on the shape of the thorax, the cardiac silhouette can appear larger or smaller than expected. The cardiac silhouette in flat chested dogs, such as Dachshunds and Yorkshire Terriers appears to be larger, being greater than 2/3 of the height of the thorax in the lateral projection. Rotation of the thorax will accentuate this finding. In the same dog, the VD or DV projection will show the heart to be of normal size. This iterates the importance of orthogonal projections when assessing the size and shape of the heart.
Pulmonary osteomas are small (2-4mm) well-defined structures that are often confused with metastases. These are osseous metaplasia in the pulmonary parenchyma and are an incidental finding and are located in the interstitium of the lung. They are accentuated ventrally in the lateral radiograph. They are differentiated from actual metastases as these are soft tissue and require a minim size of 4-5 mm to be recognized radiographically. Due to the mineralization of the osteoma, they are easily detected even though of very small size.
Probably one of the worst perpetrators of misinterpretation is tunnel vision, i.e., advance knowledge of the details of the patient that affect how we assess the radiographic findings. For instance, pulmonary masses and nodules in middle aged or older animals do not always equal neoplasia. Pyogranulomatous diseases can lead to nodule formation in the lung that mimic metastatic infiltration and can occur at any age.