Robert M. Kirberger, BVS, MMedVet(Rad), DECVDI
Faculty of Veterinary Science, University of Pretoria, Onderstepoort, South Africa
Computed tomography (CT) is a non-invasive imaging technique that uses X-rays and computers to create cross-sectional images of structures. One advantage of CT over radiography is the ability to remove superimposition of overlying bone structures. Post-processing/manipulation allow transverse images to be reformatted in dorsal and sagittal planes, as well as 3-D images after the patient has left the CT suite (known as multiplanar reformatting or MPR, and volume rendering techniques or VRT). Because images are acquired digitally, they can be displayed using various windows (e.g., bone, mediastinal, lung and soft tissue windows, etc.) to best demonstrate the structure of interest. Tissue densities can be measured using numerical values called Hounsfield units (HU), which are units of density relative to water. The Hounsfield scale ranges from -1000 to +1000. Examples of differing tissue HU are: bone (+1000), mediastinal soft (+50), fat (-80 to -100), lungs (-713 to -846) and air (-1000). Virtual endoscopy is another specialized technique allowing real visualization of air-filled structures. The use of intravenous contrast media [CT angiography (CTA)] adds another dimension that allows greater visibility of cardiovascular structures and appreciation of perfusion of pathological structures. This presentation will give an overview of the indications for thoracic CT (excluding trauma) and the practical applications of taking a dog or cat for a thoracic CT examination.
Heavy sedation or short-acting anaesthesia is required for the patient to lie completely still during the scan, as movement will create undesirable artefacts. This applies particularly to the thorax and the abdomen where respiratory motion is most marked. Respiration can be stopped by means of muscle relaxants, but this requires an anaesthetist to be present. Alternatives are injection of a small dose of propofol, which will usually stop respiration for about a minute or to use a ventilator. Alternatively, the dog can be hyperventilated to create hypocarbia, which will stop respiration for 30+ seconds. Critically ill patients can be propped into position with restraining devices, or for cats and very small dogs a specially designed Vet mousetrapTM can be used and sedation is thus avoided. These scans may have to be repeated due to patient movement. It is advisable to scan the thorax from caudal to cranial, which means that if the dog does start breathing during the scan, it is likely to be towards the end of the scan when the cranial thorax is imaged. This will minimize the effect of respiratory movement as the diaphragm and caudal rib cage, which has the most respiratory movement, will have been scanned already. Anaesthetize the patient in the CT room and ensure you do this in sternal recumbency and keep the dog in this position when positioned in the gantry. A dog lying on its side for just 5 minutes will cause artefactual increased opacity of the dependent lung due to hypostasis, which will hamper lung interpretation.
If pleural effusion is present, a DV and VD scan can be made to move the free pleural fluid away from possible pathology that may be hidden by the fluid. If oesophageal or mediastinal pathology is present, place another endotracheal tube in the oesophagus and inflate the oesophagus with air just prior to scanning. This will enhance the visibility of oesophageal masses, as well as help define the origin of mediastinal masses.
It is important to know that CT results in very high radiation exposures of the patient and potentially of any person within the CT room due to scatter radiation. One human abdomen CT exam is equivalent to 500 radiographs. If you thus have to be in the CT room, ensure you follow the principles of the inverse square law and stand as far away from the gantry as possible. Also ensure that you wear at least a lead coat and thyroid shield. It is important to stand facing the gantry to avoid possible radiation of the side of your body. The least amount of scatter will be generated in the direction of the gantry table, so try to stand in line with the table.
CTA is frequently used to highlight arterial, venous and perfusion stages of organs or masses. For vascular studies, contrast should be injected with a pressure injector, which allows volumes of 5 ml/sec to be injected rapidly, resulting in optimal vessel opacification. For these studies, the biggest possible intravenous catheter must be placed to avoid the vein from blowing during the high-pressure injection. Perfusion studies are normally done 1–2 minutes after routine intravenous injection. Dosages are usually about 1–2 ml/kg with larger volumes used in smaller dogs. Adverse reactions to iodinated intravenous contrast media have been extensively reported and investigated in humans, but are rare in small animals. Patients at risk for anaphylactoid reactions or that have elevated creatinine levels and diabetic dogs are likely to have a greater risk of contrast-induced nephropathy. Dogs should as a routine be on a Ringer's drip for 2 hours before and after the contrast examination and patients at risk at least 12 hours before and after the procedure.
These include the osseous structures and overlying soft tissues, thoracic inlet and diaphragm. The greater sensitivity of CT to detect early bone changes compared to radiographs makes it an ideal modality to detect early osseous neoplasia, particularly of the ribs, and infection and particularly how much these involve the thoracic cavity and its associated structures. The extent of defects and organ displacement are easily defined in diaphragmatic hernias and ruptures. It is important to check the visible abdomen for pathology as well.
Tracheal and main stem bronchus pathology, such as collapse, pressure from adjacent masses and mucosal alterations are readily evaluated. The lungs are evaluated in a lung window allowing the detail of the blood vessels and bronchi to be seen. The greater sensitivity of CT allows bronchial disease, and in particular chronic idiopathic pulmonary fibrosis, which may not be seen on radiographs, be readily diagnosed as ground glass opacities, parenchymal bands, crazy paving appearance and bronchiectasis. Neoplastic pulmonary masses, which are often calcified in cats, can be defined for surgical planning. CT is much more sensitive than radiographs to detect pulmonary metastasis (radiographs miss 90% of nodules), as well as regional lymph node involvement. Pulmonary osteomata should not be confused with pulmonary metastatic disease. Radiographically visible hilar masses may originate from the lungs, mediastinal structures or heart base and CT can usually distinguish these, particularly with the use of contrast studies. Lung lobe torsion, which is often accompanied by pleural effusion, can be better defined with CT and is seen as an abnormal size or orientation of the lung lobe, vesicular pattern, abruptly ending bronchus and lack of contrast enhancement.
The mediastinum contains a variable amount of fat and a prominent thymus cranioventrally in young animals. Mediastinitis, possibly with associated lymphadenopathy, can be detected with abscessation seen as focal hypoattenuating structures with contrast enhancing walls. Inciting foreign bodies, such as kebab sticks, may also be seen. Mediastinal neoplasia may have numerous origins and the aetiology may often be implied from their location in the various mediastinal regions. Oesophageal pathology, such as foreign bodies, masses, Spirocerca lupi-associated nodules and their neoplastic transformation, can all be seen as well as part of the stomach within or adjacent to the terminal oesophagus in hernias and intussusceptions.
CT easily detects small-volume pneumothorax, which rises to the non-dependant area and CT may assist in finding the cause, such as bullae or foreign bodies. Effusion is detected in the dependant thoracic regions and generally has a HU of < 40. Haemorrhage is usually denser (HU > 50). CT lymphangiography can be performed to identify thoracic duct leakage. Pyothorax is often accompanied by pleural thickening, which enhances with contrast medium administration. Pleural masses, nodules or thickening may be due to neoplasia, mesothelial hypertrophy or bacterial infection. CTA studies are helpful to detect the margins of pleural pathology, particularly in the presence of pleural effusion. Migrating foreign bodies through the pleural cavity can also be detected.
CTA allows visualization of the atria and ventricles and thoracic vasculature. Incidental aortic bulb and coronary vessel, as well as Spirocerca lupi-associated aortic mineralization, are readily detected, as is aneurysm detection. It is advisable to perform CTA in vascular ring anomalies to detect less common causes, which will influence surgical planning. CTA is also useful in other congenital cardiac defects, such as patent ductus arteriosus, to define the extent of pathology and shunt direction. Pulmonary thromboembolism and Dirofilaria-associated changes are readily picked up with contrast studies as tortuous pulmonary arteries with filling defects.
1. Scwarz T and Saunders J (eds.) Veterinary Computed Tomography. Wiley Blackwell, Oxford, United Kingdom, 2011.
2. Nemanic S, London C, Wisner ER. Comparison of thoracic radiographs and single breath-hold helical CT for detection of pulmonary nodules in dogs with metastatic neoplasia. J Vet Intern Med. 2006;20:508–515.
3. Seiler G, Schwarz T, Vignoli M, Rodrigues D. Computed tomographic features of lung lobe torsion. Vet Radiol Ultrasound. 2008;49:504–508.
4. Van der Merwe LL, Kirberger RM, Clift S, et al. Spirocerca lupi infection in the dog: a review. Vet J. 2009;176:294–309.
5. Reetz JA, Buza EL, Krick EL. CT features of pleural masses and nodules. Vet Radiol Ultrasound. 2012;53:121–127.