Evaluation of different computed tomography techniques and myelography for the diagnosis of acute canine myelopathy

Forty-six dogs with either cervical (C1-C5 or C6-T2) or thoracolumbar (T3-L3) acute myelopathy underwent prospective conventional computed tomography (CT), angiographic CT, myelography, and CT myelography. Findings were confirmed at either surgery or necropsy. Seventy-eight percent of lesions were extradural, 11% were extradural with an intramedullary abnormality, 7% were intramedullary, 2% were intradural-extramedullary, and 2% had nerve root compression without spinal cord compression. Intervertebral disc herniation was the most frequent abnormality regardless of signalment or neurolocalization. Twenty-one of 23 Hansen type I disc extrusions but none of the Hansen type II disc protrusions were mineralized. Two chondrodystrophic dogs had acute myelopathy attributable to extradural hemorrhage and subarachnoid cyst. CT myelography had the highest interobserver agreement, was the most sensitive technique for identification of compression, demonstrating lesions in 8% of dogs interpreted as normal from myelography and enabling localization and lateralization in 8% of lesions incompletely localized on myelography due to concurrent spinal cord swelling. None of the imaging techniques evaluated permitted definitive diagnosis of spinal cord infarction or meningomyelitis but myelography and CT myelography did rule out a surgical lesion in those cases. While conventional CT was adequate for the diagnosis and localization of mineralized Hansen type I disc extrusions in chondrodystrophic breeds, if no lesion was identified, plegia was present due to concurrent extradural compression and spinal cord swelling, or the dog was nonchondrodystrophic, CT myelography was often necessary for correct diagnosis.