# Gene expression and molecular pathway analyses differentiate immunotherapy-induced myositis from spontaneous dermatomyositis

## TL;DR

This study compares gene activity in muscle tissue to distinguish between myositis caused by immunotherapy and a similar condition that occurs naturally.

## Contribution

The study identifies distinct gene expression patterns and molecular pathways that differentiate immunotherapy-induced myositis from spontaneous dermatomyositis.

## Key findings

- Immunotherapy-induced myositis shows upregulation of type-II IFN signaling, while spontaneous dermatomyositis shows type-I IFN signaling.
- Key genes like FOXP3 and CCL14 are differentially expressed, suggesting T cell activation in immunotherapy-induced myositis.
- Gene expression differences highlight distinct regulatory hubs and interactions in ICI-related myositis compared to spontaneous cases.

## Abstract

Immune checkpoint inhibitor therapy (ICI)-induced myositis (irMyositis) occurs in about 1% of patients treated with anti-PD1 or anti-CTLA-4 antibodies and can be debilitating or even fatal. We compared gene expression profiles from skeletal muscle biopsies between irMyositis patients, patients with spontaneous dermatomyositis (DM, comprising anti-Mi2-positive and anti-TIF1-γ-positive subtypes), and non-diseased controls (NDC). We used the NanoString nCounter PanCancer Immune Profiling Panel to perform differential gene expression (DGE) and pathway enrichment analyses. We identified 93 differentially expressed genes (DEGs) across conditions. Gene set enrichment analysis (GSEA) suggested activation of interferon gamma (type-II IFN) and interferon alpha/beta (type-I IFN) signaling in irMyositis and DM, respectively. For instance, type-II IFN was upregulated exclusively in irMyositis when compared to DM, which conversely showed upregulation of effector genes downstream type-I IFN. The observed fold-change of a subset of 33 genes drove the GSEA. We further characterized the DEGs using network interaction and expression correlation analyses. This allowed us to describe potential differences between regulatory hubs and cells involved in irMyositis susceptible to ICI effects. For example, the downregulation of FOXP3 we observed together with the upregulation of the chemokine CCL14 in irMyositis may have been a consequence of T cell activation upon ICI therapy. The gene expression correlation and putative molecular interactions set irMyositis apart from DM, particularly with respect to IFN response and DGE of interactors of ICI protein targets (CTLA4, PD-1, PD-L1). Our results suggest new avenues for understanding immunotherapy-related adverse events.

The online version contains supplementary material available at 10.1038/s41598-025-11944-5.