Comparison of radiography, computed tomography, and magnetic resonance imaging for evaluation of appendicular osteosarcoma in dogs

OBJECTIVE: To determine which imaging modality best determines the microscopic extent of primary appendicular osteosarcoma in amputated limbs in dogs. DESIGN: Case series. ANIMALS: 10 dogs with appendicular osteosarcoma. PROCEDURE: 10 dogs with appendicular osteosarcoma that did not receive neoadjuvent chemotherapy were treated by use of limb amputation. Amputated limbs were imaged by use of radiography, computed tomography (CT), and magnetic resonance imaging (MRI) and examined microscopically to determine longitudinal extent of neoplastic cell involvement and length of associated intramedullary fibrosis. Changes detected by use of the various imaging studies were compared with the actual tumor length determined microscopically. Data were analyzed to determine which imaging technique most closely predicted tumor length. RESULTS: Measurements obtained by use of craniocaudal radiographic views were most accurate at predicting tumor length but underestimated tumor length substantially in 1 limb and slightly in another limb. Measurements made by use of CT were most accurate at predicting tumor length when intramedullary fibrosis was taken into account but underestimated tumor length in 1 limb. Measurements made by use of MRI were least accurate but did not underestimate tumor length in any of the limbs. CONCLUSIONS AND CLINICAL RELEVANCE: Although radiography is used in diagnosis of osteosarcoma in dogs, additional imaging studies to confirm the extent of neoplasia prior to limb-sparing ostectomy may be beneficial. Underestimation of tumor length would be associated with higher incidence of incomplete excision and local tumor recurrence.