Identification and validation of interferon-driven gene signature as a predictor of response to methotrexate in juvenile idiopathic arthritis

Abstract

Objectives: To identify and validate gene expression biomarkers of response to methotrexate (MTX) treatment in peripheral blood of children with juvenile idiopathic arthritis (JIA) measured before starting MTX treatment.

Methods: RNA sequencing was performed on sorted CD4+, CD8+, CD14+, and CD19+ cells, as well as peripheral blood mononuclear cells (PBMC) taken pre-treatment in a discovery cohort (n = 97) and 2 validation cohorts (n = 26 and n = 47, respectively) of patients with non-systemic JIA. Clinical data were recorded at baseline (timepoint 1) prior to treatment and 6 months (timepoint 2) of MTX treatment. Analysis tested the association of gene expression in specific cell types with treatment response using limma-voom, gene set enrichment analysis, and a novel 51-gene score against response to treatment. Parallel analysis, also using pre-treatment gene expression data, was performed in adult rheumatoid arthritis (RA) data (n = 240).

Results: In patients with JIA, the baseline expression of genes driven by interferon (IFN) alpha (type-I) or gamma (type-II) was associated with response to treatment at 6 months in 3 independent JIA cohorts. The direction of the association indicated that children with higher baseline expression of IFN-stimulated genes prior to MTX were more likely to be good responders. Comparison with adult RA indicated differences between PBMC and whole blood gene expression associations with response.

Conclusions: In children with JIA, a high IFN-driven gene signature is associated with a better response to MTX than those with a low IFN-driven gene signature. These data could pave the way to clinically validated tools to identify those most likely to require medications in addition to MTX to control inflammation.