Identification of a distinctive gene signature associated with disease activity in granulomatous myositis.

OBJECTIVES: Granulomatous myositis (GM) is defined by focal collections of activated macrophages that fuse to form multinucleated cells that aggregate into granulomas within skeletal muscle. This study aimed to elucidate the pathophysiology of GM by defining its specific transcriptomic profile. METHODS: Bulk RNA sequencing was performed on 722 muscle biopsies, including 38 from patients with GM, other myopathies, and healthy comparators. Spatial transcriptomics and immunohistochemistry assays were used to identify the location of specific transcripts and proteins within the tissue. RESULTS: We conducted a systematic analysis of the inflammatory pathways implicated in GM. This analysis revealed that GM muscle biopsies are characterized by elevated levels of IFNγ, IFNγ-inducible genes, and proinflammatory cytokine genes including IL1B, TNF, and TGFB1. Additionally, we identified a specific gene signature in GM, comprising 1293 genes that were significantly upregulated and 256 significantly downregulated. These genes were strongly correlated with transcriptomic markers of disease activity. Using this gene signature, support vector machine models accurately identified GM, achieving an area under the curve of 99.6% (99.0%-99.9%) and an accuracy of 98.6% (98.2%-98.9%). Among the identified genes, CHIT1 emerged as the most significantly upregulated gene, with its expression strongly correlating with disease severity. Furthermore, CHIT1 was detected at both the RNA and protein levels in granulomas and giant cells. CONCLUSIONS: GM is transcriptomically characterized by a strong IFNγ signature and overexpression of proinflammatory cytokines, including IL1B, TNF, and TGFB1. Additionally, it exhibits a unique transcriptomic profile, including CHIT1, which correlates with disease activity.