RIG-I expression in perifascicular myofibers is a reliable biomarker of dermatomyositis

BACKGROUND: Dermatomyositis (DM) is inflammatory myopathy or myositis characterized by muscle weakness and skin manifestations. In the differential diagnosis of DM the evaluation of the muscle biopsy is of importance among other parameters. Perifascicular atrophy in the muscle biopsy is considered a hallmark of DM. However, perifascicular atrophy is not observed in all patients with DM and, conversely, perifascicular atrophy can be observed in other myositis such as antisynthetase syndrome (ASS), complicating DM diagnosis. Retinoic acid inducible-gene I (RIG-I), a receptor of innate immunity that promotes type I interferon, was observed in perifascicular areas in DM. We compared the value of RIG-I expression with perifascicular atrophy as a biomarker of DM. METHODS: We studied by immunohistochemical analysis the expression of RIG-I and the presence of perifascicular atrophy in 115 coded muscle biopsies: 44 patients with DM, 18 with myositis with overlap, 8 with ASS, 27 with non-DM inflammatory myopathy (16 with polymyositis, 6 with inclusion body myositis, 5 with immune-mediated necrotizing myopathy), 8 with muscular dystrophy (4 with dysferlinopathy, 4 with fascioscapulohumeral muscle dystrophy) and 10 healthy controls. RESULTS: We found RIG-I-positive fibers in 50% of DM samples vs 11% in non-DM samples (p < 0.001). Interestingly, RIG-I staining identified 32% of DM patients without perifascicular atrophy (p = 0.007). RIG-I sensitivity was higher than perifascicular atrophy (p < 0.001). No differences in specificity between perifascicular atrophy and RIG-I staining were found (92% vs 88%). RIG-I staining was more reproducible than perifascicular atrophy (κ coefficient 0.52 vs 0.37). CONCLUSIONS: The perifascicular pattern of RIG-I expression supports the diagnosis of DM. Of importance for clinical and therapeutic studies, the inclusion of RIG-I in the routine pathological staining of samples in inflammatory myopathy will allow us to gather more homogeneous subgroups of patients in terms of immunopathogenesis. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s13075-017-1383-0) contains supplementary material, which is available to authorized users.