The diagnostic value of routine muscle sampling at autopsy: high prevalence and systematic under-detection of type 2...

Type 2 myofiber atrophy is a clinically significant manifestation of muscle wasting. Its detection at autopsy depends on recognition of fiber size variation on routine hematoxylin and eosin (H&E) staining - a potentially subtle finding that may be missed without neuropathological training. Its prevalence is therefore likely to be under-reported, particularly in autopsy populations which are enriched for chronic illness, malignancy, and therapeutic high dose steroids. We conducted a prospective diagnostic accuracy study using a hospital autopsy cohort signed out by a board-certified neuropathologist, in which myosin fast immunohistochemistry was ordered prospectively as a reference standard. A non-neuropathology-trained pathologist's retrospective autopsy cohort, in which IHC was not ordered, served as a comparison. Diagnostic accuracy measures were calculated for fiber size variation as a predictor of type 2 atrophy. Logistic regression was performed to identify clinical predictors of type 2 atrophy. Diagnostic histopathological findings were present in 58.0% of prospective cases and included type 2 atrophy, necrosis and metastatic calcification. Type 2 fiber atrophy was identified in 52.2%. Fiber size variation as a predictor of type 2 atrophy had a sensitivity of 90.2%, specificity of 89.3%, positive predictive value of 92.5%, and negative predictive value of 86.2%. The non-neuropathology-trained pathologist identified type 2 atrophy in 10.5% of cases. Longer hospital stay was significantly associated with type 2 atrophy (median 19.0 vs. 4.0 days; p=0.0004; OR=6.37). In a hospitalized autopsy population, fiber size variation on H&E is a highly accurate proxy for clinically significant type 2 fiber atrophy. Type 2 atrophy may be substantially underdiagnosed at routine hospital autopsy, with a five-fold detection gap between neuropathology-trained and autopsy pathologists. Universal muscle sampling and targeted H&E evaluation for fiber size variation can achieve efficient and cost-effective detection of important diagnoses in the autopsy setting.