Longitudinal interferon signatures are associated with disease flare in dermatomyositis

Background Disease flares in dermatomyositis (DM) are unpredictable and often detected weeks to months after their onset. Interferon (IFN)-related gene signatures are implicated in DM, but their value for longitudinal flare prediction remains unclear. We hypothesized that longitudinal transcriptomic profiling using minimally invasive blood microsampling would detect IFN gene signatures preceding and during DM disease flares. Methods In this feasibility study, patients were included if they had inactive or minimally active disease and at least one of the following criteria: recent diagnosis within 2 years, a documented flare in the prior year, or ongoing tapering of immunosuppressive therapy. Patients self-assessed their disease activity weekly for 6 months using the International Myositis Assessment and Clinical Study group (IMACS) patient global assessment (PtGA; numerical rating scale, range 0-10). Disease flares, defined as a PtGA increase >2 for 2 consecutive weeks, triggered an in-person visit for confirmation by a physician. During the follow-up period, patients also self-collected weekly blood microsamples (125uL) that were shipped by courier to the study center. From those weekly blood microsamples, RNA sequencing was performed on 54 samples collected across 9 timepoints using an Illumina platform. Bioinformatic analyses included alignment, quality assessment, and gene-level quantification to evaluate differential gene expression and RNA processing changes during DM flares. Results Six DM patients were included in this study (2 anti-MDA5, 2 anti-Mi2, 1 anti-NXP2, 1 no myositis-specific antibodies). All were female with ages at study entry ranging from 46 to 72 years. Three of the six patients experienced a flare during their 6-month follow-up. The mean (SD) RNA integrity score for the weekly blood microsamples collected was 7,61 (0,64). IFN scores were significantly higher in flare-associated samples than in no flare samples (linear mixed-effects model with individual patient as a random intercept, P = 5.13 x 10E-4). Gene set enrichment analysis of hallmark gene sets identified enrichment of IFN-related gene sets in pre-flare and flare samples with IFN Alpha Response and IFN Gamma Response showing the highest positive normalized enrichment scores. Conclusion This study demonstrates the technical feasibility of combining home-based longitudinal blood microsampling with transcriptomic profiling in dermatomyositis. IFN gene score were higher in flare-associated than in no-flare samples, supporting further investigation of IFN-related signatures as candidate biomarkers for longitudinal disease monitoring.