Diagnostic imaging forms a critical part of good cancer patient case management. Treatment options are defined by a number of factors but one of the most significant is the anatomical extent of disease. Imaging is also used for diagnosis of suspected cancer. Some neoplastic processes have distinctive features which are recognised by characteristic alterations to a diagnostic imaging study. For example, osteosarcoma of the distal radius is characterised by an osteolytic and osteoproductive lesion of the distal radial metaphysis in large and giant-breed dogs.

Some imaging modalities, most frequently ultrasonography but also computed tomography (CT), are used to direct biopsy procedures to ensure accurate placement of the biopsy device to ensure that the appropriate tissue is sampled and to protect neighbouring vital structures. Ultrasonography lends itself to this particularly well as a needle or other biopsy device can be observed in real time as it enters the target tissue.

Some tumours simply do not spread. Canine cutaneous histiocytoma is a good example. If a histiocytoma is diagnosed by fine-needle aspirate biopsy and cytology, no subsequent diagnostic imaging is required. The tumour can be cured by appropriate surgery.

Many tumours are capable of spread, however. For practitioners in normal first-opinion practice it is not reasonable to expect a full working knowledge of the biological behaviour of all tumour types. Therefore, if a tumour is diagnosed and a specific behaviour pattern is not known, it is wise to simply pursue reasonable diagnostic evaluations that would be expected to define the clinical stage of disease for any generic tumour.

Radiography lends itself very well to thoracic imaging to identify (radiographically detectable) pulmonary metastases. Thoracic radiography is therefore something that is very much a routine part of the evaluation of a patient with a known or suspected diagnosis of cancer. There are some cancerous conditions which have a very high likelihood of early pulmonary spread. If pulmonary metastasis is radiographically evident at the time of diagnosis, response to therapy is consistently poor. Therefore these tumours are perhaps best evaluated by thoracic radiography as soon as a diagnosis is made or even on the basis of a suspicion of a diagnosis of neoplasia. If both lateral projections are obtained and there is no indication of atelectasis, I am comfortable that a further orthogonal view is not required. If there is any doubt, a dorsoventral thoracic radiograph should be obtained.

Canine pulmonary metastases will typically be recognised as multiple soft tissue density nodules throughout the pulmonary parenchyma. Lesions are virtually impossible to differentiate from other soft tissue or fluid opacities if they are less than 5 mm in diameter. Typically multiple nodules of varying sizes are seen with diameters ranging from 5 mm to 2 cm.

The nodular pattern of radiographic change seen in canine pulmonary metastasis is rarely seen in cats. Instead a less nodular pattern is typically recognised. My explanation for this is simply that cats have much smaller lungs yet the radiographic resolution remains the same as it is in dogs. If cats undergo thoracic imaging by a higher resolution radiographic technique, CT imaging, the expected nodular interstitial pattern becomes very clear in patients with pulmonary metastasis.

Radiography is also an excellent investigative tool to narrow the list of differential diagnoses in cases of suspected neoplasia. Anatomical sites that lend themselves most well to radiographic diagnosis are the skeleton and the lungs. Importantly, these sites cannot be easily evaluated by the next most common imaging modality in veterinary practice, ultrasonography, because bone and air interfere with transfer of the ultrasound wave.

Lameness and bony swellings secondary to primary, or secondary metastatic, neoplasia are characterised by mixed patterns of bone destruction, osteolysis, and bone production, osteogenesis. It is not true to say that this combination of changes can only be explained by osteosarcoma. However, it is true to say that a lesion with the radiographic and clinical characteristics of osteosarcoma is almost certainly going to be an osteosarcoma.

Other sites which lend themselves well to investigation by radiography include the bladder, prostate and the sub-lumbar lymph nodes. Bulky sublumbar lymphadenopathy can be readily recognised by lateral abdominal radiography. This can be of great significance in the decision making for caudal abdominal and perineal tumours. Positive and negative contrast media can enhance the visibility of intravesical mass lesions. However, radiographic contrast studies of the lower urinary tract can be time-consuming procedures and anaesthesia is generally required. Furthermore, as veterinary surgeons become more familiar with ultrasonography, the value of radiography for evaluation of the lower urinary tract diminishes.

Ultrasonography and radiography provide different kinds of information about the abdomen. Abdominal radiography is quick. It provides information about the absolute and relative size of the viscera; it does not suffer significant loss of image quality with the presence of air or food in the bowel. Radiodense foreign bodies and mineral densities are readily identified. However, in the context of abdominal neoplasia, abdominal ultrasonography is proving increasingly beneficial. Ultrasonography allows the examination of the architecture of the abdominal organs; gross abnormalities are very readily identified as soon as the practitioner is experienced in identification of normal sonographic anatomy, though this is dependent on the individual in question seeking adequate training and having access to suitable equipment. So there is a learning curve associated with the useful application of abdominal ultrasound, but once gained, this experience is never lost. The further major advantage of ultrasonography is that it allows the acquisition of biopsy specimens under direct 'real-time' ultrasound guidance.

Abdominal and perineal tumours can readily spread to the liver and spleen. Perineal tumours are frequently found to have metastasised to the sublumbar lymph nodes. These organs are rarely satisfactorily imaged by radiography. For the identification of intra-abdominal metastasis, ultrasonography is therefore much more appropriate. However, the mere identification of alterations to the normal architecture of the abdominal organs does not prove the presence of neoplasia. Only cytology or histology can diagnose these structural changes and there are numerous non-neoplastic causes including metabolic hepatopathy, reactive lymphoid hyperplasia, haematoma and extramedullary haemopoiesis. The absence of nodular changes within the abdominal organs does not constitute utterly reliable proof of absence of neoplasia either. The whole liver can rarely be fully examined ultrasonographically due to obstruction by the rib cage. Diffuse infiltrative neoplasia such as lymphoma or metastatic mast cell neoplasia may not induce detectable sonographic changes. Knowledge of the cancer process under investigation allows the clinician to interpret the ultrasonographic findings and to determine whether further investigations, such as organ biopsy, are required.

CT and magnetic resonance imaging (MRI) scans provide different anatomical information from radiography and ultrasound. A CT scanner is effectively a more sophisticated version of an X-ray machine. X-ray data are captured digitally and then processed to create a three-dimensional simulation of the region scanned. Most CT data are viewed in the form of a series of imaginary slice images of a thickness defined by the radiographer. In this way much finer-resolution anatomical detail can be obtained compared to conventional radiography. In some cases, the potential risk to a patient associated with a surgical or medical procedure might be such that it would not be considered justifiable without maximally detailed images to exclude the presence of detectable metastasis. A significant benefit of CT imaging of the abdomen is that it presents the entire abdomen in an image format that can be viewed and reviewed at a later date. The spatial accuracy of CT imaging also makes anatomical referencing vastly superior. The principal negative aspects of CT imaging are its limited accessibility, the need for chemical restraint, higher radiation dose, increased cost and the fact that it is less useful than ultrasonography for image-guided biopsy.

MRI similarly produces data which are viewed in the form of a series of imaginary slice images. In medical oncology, the primary application of MRI is the detailed imaging of soft tissue tumours in the limbs and the trunk. MRI can allow determination of the detail of which tissue planes are affected by a tumour process so that a detailed surgical plan can be made to confidently excise the entire tumour field rather than simply cutting a measured distance away from the apparent edge of a tumour and hoping that will be enough.

As imaging technology advances, it becomes increasingly important for us to recognise that imaging studies cannot definitively diagnose neoplasia. Cancer is a diagnosis made by pathology evaluations, be they cytology or histology. Imaging studies can only yield data from which we infer that cancer is present and to what extent it is present. That said, imaging studies do provide the cornerstone by which we make cancer treatment decisions because they allow us to define the anatomical extent of neoplasia, and anatomical extent of neoplasia remains the most significant determinant of prognosis.