Wall CR, Cook CR, Cook JL.
Radiography, magnetic resonance imaging (MRI), and ultrasonography are commonly used for diagnosis of shoulder osteochondrosis and osteochondritis dissecans (OC/OCD) in dogs, however there is a lack of published information on the relative diagnostic sensitivities of these modalities. The purpose of this prospective study was to compare diagnostic sensitivities of these modalities for detecting shoulder OC/OCD in a group of dogs, using arthroscopy as the reference standard. Inclusion criteria were history and clinical findings consistent with osteochondrosis and/or osteochondritis dissecans involving at least one shoulder. With informed client consent, both shoulders for all included dogs were examined using standardized radiography, ultrasonography, MRI, and arthroscopy protocols. One of three veterinary surgeons recorded clinical and arthroscopic findings without knowledge of diagnostic imaging findings. One of two veterinary radiologists recorded diagnostic imaging findings without knowledge of clinical and arthroscopic findings. Eighteen client-owned dogs (n = 36 shoulders) met inclusion criteria. Diagnostic sensitivity, specificity, and accuracy (correct classification rate) values for detecting presence or absence of shoulder osteochondrosis/osteochondritis dissecans were as follows: radiography (88.5%, 90%, 88.9%), ultrasonography (92%, 60%, 82.6%), and MRI (96%, 88.9%, 94.4%). Odds of a correct diagnosis for MRI were 3.2 times more than ultrasonography and two times more than radiography. For MRI detection of lesions, the sagittal T2 or PD-FAT SAT sequences were considered to be most helpful. For radiographic detection of lesions, the additional supinated-mediolateral and pronated-mediolateral projections were considered to be most helpful. Findings from the current study support more evidence-based diagnostic imaging recommendations for dogs with clinically suspected shoulder osteochondrosis or osteochondritis dissecans.