Longitudinal and Stability-Aware Analysis Reveals Treatment-Specific MicroRNA Response Signatures Following Immune-Reconstitution and B-Cell-Targeted Therapies in Multiple Sclerosis.

Disease-modifying therapies (DMT)s) for relapsing-remitting multiple sclerosis (RRMS) act through distinct immunological mechanisms, yet the within-patient molecular response programs associated with these therapies remain incompletely defined. Here, we reanalyzed publicly available peripheral blood mononuclear cell (PBMC) miRNA microarray data (GSE230064) using a longitudinal, robustness-focused framework to compare therapy-associated miRNA response patterns following cladribine versus ocrelizumab treatment. Baseline (t0) and 6-month post-treatment (t1) samples were paired within individuals and technical replicates consolidated prior to analysis, yielding a final paired cohort of four cladribine-treated and six ocrelizumab-treated patients. Within each treatment arm, we quantified per-patient Δ-miRNA (t1 - t0) values and prioritized therapy-associated response features using a multi-evidence framework integrating effect direction, magnitude, directional consistency across individuals, and leave-one-out sensitivity. Cladribine treatment was associated with a highly coordinated, directionally concordant upregulation of five miRNAs including hsa-miR-27a-3p, hsa-miR-27b-3p, hsa-miR-503-5p, hsa-miR-148a-3p, and hsa-miR-26a-5p, all exhibiting 100% directional stability across patients and mean Δ-expression values ranging from +0.77 to +1.38. These miRNAs target pathways relevant to MS pathophysiology, including Th17/Treg balance, Wnt-β-catenin signaling, macrophage polarization, and epigenetic immune regulation. In contrast, ocrelizumab elicited a more selective response pattern, with five miRNAs including hsa-miR-100-5p, hsa-miR-410-3p, hsa-miR-432-5p, hsa-miR-296-5p, and hsa-miR-485-3p showing moderate directional stability (83%) and greater inter-individual heterogeneity, consistent with the more targeted mechanism of CD20+ B-cell depletion. Notably, the two treatment-associated signatures were non-overlapping, with hsa-miR-27b-3p representing the only miRNA shared with prior cross-sectional analyses of this dataset. The identified ocrelizumab-associated miRNAs implicate pathways including mTOR/IGF1R signaling, NF-κB regulation, RNA editing, and mitochondrial biogenesis, several of which are dysregulated in progressive MS. Together, these findings demonstrate that cladribine and ocrelizumab induce distinct, treatment-specific miRNA response architectures that reflect their divergent immunological mechanisms. This work establishes a stability-aware analytic template for extracting reproducible longitudinal miRNA signals from small paired RRMS cohorts and provides a ranked set of biologically plausible candidate miRNAs for prospective validation and mechanistic investigation.