Radiographic interpretation paradigm: The small animal thorax (Proceedings)

Thoracic radiography is still the most common first line assessment for diseases and conditions of the thorax. With the advent of digital radiography, a new interest in diagnostic radiology has emerged. However, even though certain artifacts are not an issue (e.g., processing, exposure), problems with inadequately positioned patients still exist. If one does not quality control their radiographs, then there is no sense in even taking them as a poor set of radiographs can mislead a veterinary and result in inappropriate treatment and/or diagnosis for the pet. Thoracic radiography is a quick and easy test to perform and because of the air/soft tissue contrast, provides for an ideal radiographic subject for evaluation. However, interpretation of the thorax can be frustrating and thereby reduce the number of cases that are being imaged even in the face of clear indications for radiographs based on the clinical presentation of the pet. A systematic approach to interpretation is the key to success. This approach will include a step by step overview for the evaluation of all aspects of the thoracic radiograph. This approach will be reviewed in broad strokes during this hour, as all aspects cannot be covered.

Objectives of the Presentation

      1. Provide practitioners with a basic interpretation paradigm for the evaluation of the small animal thorax.

     2. Provide practitioners with the appropriate assessments for technical quality control of the thoracic radiographs.

     3. Provide practitioners with a step-by-step overview for assessing abnormalities of intrathoracic structures.

Key Etiologic and Pathophysiologic Points

     1. Technical factors including technique, phase of respiration and the positioning of the patient have to be taken into account when interpreting thoracic radiographs.

     2. Often, thoracic diseases are multicompartmental.

     3. Often, pulmonary patterns are mixed with disease being in transition or involving a variety of lung components.

     4. All radiographic abnormalities are described based on the standard roentgen sign approach. Ultimately, the abnormalities should be interpreted in light of the patient, physical examination and laboratory data. There will be incidental findings that may or may need to be pursued.

     5. Thoracic ultrasound is limited to peripheral masses (pleural, pulmonary or mediastinal) or generalized disease (pleural or pulmonary)

     6. Repeat thoracic radiographs should be thought of in terms of repeat blood work. You are evaluating a very small time period (1/120th of a sec or faster) of the disease process and all diseases are not static.

Key Clinical Diagnostic Points

     1. One should try to compartmentalize radiographic abnormalities into extrathoracic, pleural, pulmonary and mediastinal (including cardiac), recognizing that any disease can have multicompartmental components.

     2. Most radiographic changes are non-specific and the creation of a prioritized differential list that goes from general to as specific as possible should be written out in the medical record as a summary to the interpretation.

     3. Lesions within the down lung lobe are more difficult to see due to atelectasis of the normal lung adjacent to the abnormal lung.

     4. On a left lateral radiograph, the cardiac silhouette will rotate away from the sternum and should not be mistaken for a pneumothorax.

     5. Always practice radiation safety. Be sure that this is enforced with the techs taking the radiographs and there are penalties associated with violations of standard ALARA principles. Standard radiation safety measures are beyond the scope of this talk; however, remember several key factors: use high mA and low time stations for a given mAs; collimate, collimate, collimate, wear lead apron, thyroid shields and gloves (none of which protect from the primary beam), do not stand directly in front of the tube when making an exposure and use sedation and restraint devices as a first line prior to someone being in the room.

Key Therapeutic Points

     1. Follow-up radiographs should be obtained in rapid succession if warranted (within the same day or within 24 to 48 hours).

     2. Follow-up radiographs will often document the progression or regression of disease.

     3. In certain diseases, the radiographic resolution of abnormalities may lag behind the clinical response by the patient.

     4. Dogs and cats in congestive left sided heart failure usually clear within 24 hours if an adequate diuretic dose has been used.

Key Prognostic Points

     1. In cases of canine lymphoma, the presence of a cranial mediastinal mass has been shown to be a poor prognostic indicator of response to chemotherapy, length of remission and overall survival.

Discussion

One should always obtain at least two radiographs of the thorax; although, three radiographs (right lateral, left lateral and a VD or DV) have become the standard of care for veterinary patients. If one cannot get the VD or DV, then a right and left lateral are obtained. An attempt to make sure that all images are obtained on peak inspiration is critical. If you take a single right lateral expiratory radiograph of a Dachshund, you will always call their hearts big and interstitial edema; however, these animals are normal. A dog or cat in respiratory distress may only tolerate a DV radiograph to begin with and an initial assessment of pleural effusion, pneumothorax or severe pulmonary parenchymal disease can be made and appropriate supportive care initiated. However, once the patient is stabilized and after therapeutic pleurocentesis (if indicated), a complete radiographic study should be obtained.

When reviewing radiographs (whether using the old analog films and a view box or newer digital monitors), use a quiet and darkened environment. Review films with an additional practitioner as two pairs of eyes are always better than one. Be sure to write down all roentgen abnormalities and use orthogonal projections to confirm the presence of these changes. The basic radiographic opacities are air (black), fat, soft tissue, mineral (bone) and metal (white). The basic roentgen signs include: location, margin, number, opacity, shape and size. Every radiographic structure has expected (all 6) roentgen signs associated with them. Make sure you write down and list the abnormalities and then you can prioritize the changes accordingly from most to least (incidental) important. Be sure to always hang the images, whether on the view box or the monitor(s), the exact same way every time. For example, for a three-view thorax, when I am facing the view boxes, I will hang the left lateral on my left, the ventrodorsal in the middle and the right lateral on the right. This consistency will allow the occipital cortex to develop a "normal" pattern. Although this "pattern" recognition approach does not replace a systematic approach for evaluation of thoracic radiographs, it does allow for one to make an initial assessment (just do not stop there)

Several pitfalls in interpretation should be recognized. These include:

     • No systematic approach

     • Satisfaction of search based on preconceived bias prior to looking at the film

     • Satisfaction of search based on finding an abnormality and stopping the review of the films

     • Incomplete evaluation of an abnormality

     • Lack of recognition of an abnormality

     • Classification of an incidental finding as an significant abnormality

The largest problem when evaluating dogs is the large differences in chest size, shape and configuration based on the breed. This creates an infinite set of possibilities as to what normal really looks like. Several factors that should help in the quest for normal will be making sure the patient is positioned straight (thoracic limbs pulled cranially away from the cranial thorax) and catching the dog or cat on peak inspiration.

The normal thorax

Some basic rules of thumb can be applied to most dog breeds. The two breeds that will break all of the rules will be boxers and bulldogs. The widest point of the cardiac silhouette on the lateral view in dogs should be between 2.5 to 3.5 intercostal spaces. For a cat, on a lateral film, the widest point of the cardiac silhouette should be 2 to 3 intercostal spaces. On the right lateral radiograph the shape of the heart should be egg or oval shaped. On the VD/DV image (straightly positioned), the widest point of the heart should not cross 50% of the widest pleural to pleural width. In geriatric cats, the cardiac silhouette will tilt cranially and ventrally on the lateral radiograph (called a "lazy" heart) and there will be prominence to the aortic arch at the junction with the descending aorta (called a prominent aortic knob). In addition, on lateral radiographs, you can use the vertebral heart score to measure cardiac enlargement. This technique is rapid and simple to perform. You measure the apical to basilar length and measure the widest point of the cardiac silhouette. Then starting at the fourth vertebral body you count the total number of vertebral bodies (ignore the intervertebral disc spaces) that make up the width and length of the heart. Normal for the dog is 9.7 ± 0.5 vertebrae whereas for cats, the vertebral heart score or scale is 7.5 ± 0.3 vertebral bodies. Obese patients will lay down fat in the cranial mediastinum, the ventral caudal mediastinal reflection and in cats, ventral to the cardiac silhouette. The only opacity that contributes to the lungs includes the pulmonary vessels, the airways and the air within the alveoli. You should be able to trace the bronchi out to each of the lung lobes from the carina and termination of the caudal trachea into the principle right and left bronchus. The pulmonary artery and vein size should be matched at the same position whether viewing the lateral or the ventrodorsal images. On the lateral, the pulmonary artery is dorsal to the bronchus that is dorsal to the pulmonary vein and on the DV/VD; the pulmonary artery is lateral to the bronchus that is lateral to the pulmonary vein.

Systematic Evaluation

The entire right and left lateral and the VD/DV radiographs have to be reviewed. The easiest way to get into the habit is to hang a radiographic checklist (such as the one provided) next to the monitor or view box as a reminder until you get the system down. When in doubt or at the end of the evaluation always asks, "What have I missed?" We will divide the thorax into four compartments and evaluate each compartment starting with the outside of the radiograph and working our way inward.

The four compartments include: extrathoracic structures, the pleural space, the pulmonary parenchyma, and the mediastinum. Remember that diseases can impact multiple compartments; however, from a simplistic standpoint, if one can figure out which compartment is involved then one can move forward with looking at possible organs of origin and differentials for each compartment and then look for similar disease processes that could involve the affected compartments. When you list your prioritized differentials, you should think in terms of a "cone of certainty." This cone represents the confidence level you have that a disease (specific) is present. The higher one moves on the cone (scale of 1 to 5), the more confident you are as to a specific diagnosis. For example, there is a long list of differentials for an unstructured interstitial pulmonary pattern. But, if I include left sided cardiomegaly and pulmonary venous enlargement, the most likely cause of the pulmonary pattern is cardiogenic pulmonary edema.

The extrathoracic structures include all soft tissue and osseous structures that make up the caudal cervical region, the proximal thoracic limbs, the thoracic wall, the sternum and thoracic vertebrae, the diaphragm and cranial abdomen. Again, any abnormality should be described in terms of roentgen abnormalities.

Normally, there is nothing that is visualized in the pleural space. However, the areas where the pleural fissure lines are located should be evaluated as subtle pleural effusions can be a clue as to a focal disease process. On the left lateral radiograph, it is not uncommon to see some degree of pleural thickening between the right middle and the right caudal lung lobe superimposed over the caudal cardiac silhouette.

The basic questions to be answered regarding the pleural space include:      • Is there any pleural effusion?

     • Is there any air in the pleural space?

     • Are there any pleural mass(es)?

     • Is there an "extrapleural" sign?

     • What have I missed?

The pulmonary parenchyma has always been the "thorn in the side" for practitioners and students alike. The easiest way to evaluate the pulmonary parenchyma is based on inspiratory films. The first question asked is: Are the lung radiolucent or radiopaque (or mixed)? If the lung(s) are radiolucent then the most common reason is hypovolemia from what every cause followed by pulmonary thrombo-embolism. Causes of focal radiolucencies would include: pulmonary bulla, blebs, cavitated lesions (granulomas or tumors), hematocoeles or pneumatocoeles.

If the lungs are too radiopaque (white), then the first thing to consider is where? Describe the anatomic location of the abnormality including lung lobes involved and if partial lobar involvement, is the change peripheral, mid-zone or hilar? Next, is there a contralateral or ipsilateral mediastinal shift noted on the VD/DV image? Now we will approach the pulmonary patterns. We are going to work through the patterns from the easiest to identify to the most difficult (actually what we will be left with and thereby an easy diagnosis). The alveolar pattern is the easiest to diagnose. The components of an alveolar pattern include: uniform soft tissue opacity, the presence of air bronchograms, a lobar sign, border effacement with the heart or diaphragm and border effacement with the pulmonary vessels and outer serosal wall of the airways. The next pattern is the bronchial pulmonary pattern. In general, bronchial patterns are generalized and you are looking for thickened small airways that will create "rings and lines". The central airways will always be prominent and in older dogs can mineralize. This can be quite striking in appearance. Try to evaluate for the presence of small airways in the peripheral aspect of the lungs or the thin section of the lungs. Next, is the vascular pattern. There are three options for increased opacity. Increase in size of the pulmonary arteries (heartworm disease), increase in size of the pulmonary veins (left heart failure) or increase in size of both pulmonary arteries and veins (overcirculation from left to right congenital cardiac shunts, arteriovenous fistulas, heart failure in cats, or volume overload in renal failure patients). If none of the above fits, you are left with the treaded interstitial pattern. The interstitial pattern is either structured (miliary or nodular) or unstructured (hazy increase in background lung opacity with loss of normal vessel border definition). Lastly, assign some degree of severity to the pulmonary pattern (mild, moderate and severe). If you are arguing over a mild unstructured interstitial pulmonary pattern, forget it. You are not going to do anything about it anyway. So my final report may go something like this: "There is an alveolar pulmonary pattern that has a mid zone and peripheral distribution (cranioventral) within the right cranial and middle lung lobes, moderate in severity." Based on the anatomic location, bronchopneumonia or aspiration pneumonia become the top differentials, particularly with clinical signs of fever, labored breathing or a history of vomiting and/or regurgitation.

The caveats for lung patterns are:

     • The pulmonary patterns do not equate to histological diagnoses

     • The pulmonary patterns are often mixed for a given disease

     • The pulmonary pattern may represent a disease in transition (interstitial to alveolar)

     • Avoid jargon terms such as consolidation or infiltrate as there are no good definitions for these terms and there are no differential lists to look up for these terms

     • Avoid broncho-interstitial. Th is is not to say that a disease process (e.g., heartworm disease or PIE) cannot have both a bronchial and interstitial component, but again broncho-interstitial is not a category to look up in a table in a book. You will have to decide whether or not the predominant pattern is bronchial or interstitial

Finally, the mediastinum is to be evaluated. I divide the mediastinum into a cranial third, middle third and caudal third. Important structures in the cranial third include lymph nodes (sternal and cranial mediastinal), thymus, trachea and esophagus. Important structures within the middle mediastinum include the esophagus, tracheal termination (carina), tracheobronchial lymph nodes and the cardiac silhouette. Important structures in the caudal third include esophagus, aorta and caudal vena cava.

Summary

     • Technical quality impacts the ability to interpret thoracic radiographs

     • A systematic approach to film review is critical for complete assessment

     • Adequate time needs to be give to the review of thoracic radiographs

     • Multicompartmental disease can be confusing at first, but take each compartment separately and then tie them together

     • Pulmonary abnormalities should be described in terms of anatomic localization prior to a pulmonary pattern description

     • Interpretation of thoracic radiographs is fun!

Suggested Reading/References

Thrall, DE. Veterinary Diagnostic Radiology, 5th Edition. WB Saunders:2008.