The Utility of Immune Profiling on Formalin Fixed Nervous Tissue to Diagnose Canine Meningoencephalitis of Unknown Etiology
Meningoencephalitides of Unknown Etiology (MUE), including subtypes Granulomatous Meningoencephalitis (GME), Necrotizing Meningoencephalitis (NME) and Necrotizing Leukoencephalitis (NLE), are canine idiopathic inflammatory neurologic diseases for which an immune mediated pathogenesis is suspected. Despite immunosuppressive therapy, the prognosis for all subtypes remains guarded. Similar clinical presentations limit the ante-mortem diagnosis of MUE to exclusion of infectious causes, and the subtyping of MUE to post-mortem histopathology. Immune profiling is a next generation sequencing technique which characterizes adaptive immune responses by analysis of lymphocyte antigen receptor repertoires. It was hypothesized that immune profiling may be adapted as a novel ante-mortem diagnostic tool for MUE, provided that GME, NME and NLE are associated with disease specific immune repertoires. Case collection also involved histopathologic review and clinical data (signalment, treatment, survival time, etc.) comparisons to the current MUE literature. Post-mortem formalin-fixed and paraffin-embedded (FFPE) nervous tissues from 309 dogs with inflammatory brain disease archived at the University of Guelph and University of California-Davis from 1996-2018 were re-evaluated histologically, and 94 MUE cases were identified and subtyped. Medical record clinical data was in agreement with the literature understanding of MUE. There was a predilection in signalment for small breed, spayed, female dogs over a range of ages, although large breeds and males were also affected. GME was the most commonly diagnosed subtype. Pugs and Yorkshire Terriers most commonly presented with their predisposed subtypes (NME and NLE respectively). Steroid monotherapy was the most common treatment, but extended survival was associated with combination therapy. Lymphocyte receptor repertoires could be sequenced from 80 of 94 cases. The majority of clonotypes were private (unique to each individual), however, a subset of lymphocyte receptor sequences were public (shared) in 3-4 individuals. Unfortunately, this degree of repertoire overlap was too low to correlate with histologic subtype or clinical data parameters. These results provide proof of principle for repertoire sequencing from FFPE nervous tissues, however larger sample numbers are needed to determine if repertoire signatures are of diagnostic or prognostic utility. If so, immune profiling may eventually facilitate the clinical diagnosis of canine inflammatory neurologic diseases, including MUE, improving response to therapy and outcome.