Objective: To evaluate whether concurrent analysis of CSF samples from 2 collection sites (cerebellomedullary cistern [CMC] and lumbar subarachnoid space [LSS]) versus only 1 site could improve the diagnostic sensitivity of CSF analysis for dogs with suspected steroid-responsive meningitis arteritis (SRMA).
Animals: 111 client-owned dogs with SRMA diagnosed at 3 veterinary referral hospitals between 2011 and 2017.
Procedures: Only dogs with CSF collected from both sites (CMC and LSS) and with no previous history of corticosteroid administration were included. Medical record data and logistic regression were used to identify factors associated with having a total nucleated cell concentration (TNCC) within the reference interval in a CSF sample from 1 collection site.
Results: The TNCC was within the reference interval (TNCC < 5 cells/μL) in the CSF sample from 1 collection site for 8 of 111 (7%) dogs and was only slightly high (TNCC, 5 to 9 cells/μL) in the sample from 1 or both sites for 10 (11%) other dogs. For each of these 18 dogs, results for samples from 1 site were consistent with SRMA. The proportion of CSF samples that had a TNCC within the reference interval was comparable between sites. As age increased, so did the risk of having an unremarkable TNCC in the CSF sample from 1 site, albeit only slightly (OR, 1.08; 95% confidence interval, 1.01 to 1.16).
Conclusions and clinical relevance: CSF samples from both the CMC and LSS should be analyzed when evaluating dogs with suspected SRMA to improve the chance of detecting a high TNCC.