Background/Purpose: Despite the revolutionary impact of TNF inhibitor (TNFi) therapy in rheumatoid arthritis (RA), up to 40% of patients fail to respond adequately. Whilst non-responder (NR) patients can be switched to alternative therapies at 3 months, many experience a delay in switching as some will subsequently respond to treatment by 6 months, which may be detrimental to patient outcomes. Ideally, blood-based biomarkers would be available to objectively predict and monitor response to TNFi but current measures, such as CRP levels, do not correlate very well with objective measures of synovitis. The aim of my work therefore, was to identify an expression signature to predict or monitor treatment response to adalimumab in a large sample cohort of patients about to start treatment.

Methods: 50 extreme EULAR good-responders (GR) and 20 extreme NR to adalimumab were selected from the Biologics in RA Genetics and Genomics Study Syndicate (BRAGGSS) cohort. Total RNA was extracted from whole blood using the MagMAX™ RNA isolation kit before (baseline) and following 3 months of therapy. RNA was amplified and converted into biotinylated sense-strand DNA using the Affymetrix WT PLUS kit and hybridized onto Affymetrix GeneChip® human transcriptome arrays. Quality control and differential expression analysis (both individual transcripts and splice variants) were assessed using the Affymetrix expression and transcriptome analysis console™ and appropriate Bioconductor packages. Weighted gene co-expression network analysis (WGCNA) was performed to identify co-regulated genes correlated with response. WGCNA alleviates the multiple testing issue inherent in microarray data analysis and could hold greater power for detecting clinically applicable biomarkers.

Results: In GR, there were 11 gene co-expression modules, which significantly changed over 3 months of treatment. The most significant module (p= 1e-05) was highly enriched for genes involved in macrophage function, specifically osteoclast differentiation, chemokine signaling and leukocyte transendothelial migration. Transcript significance for treatment time-point and module membership were highly correlated, suggesting genes which change over time are the most important genes within the module. The module was also highly correlated with individual DAS components: tender joint count, swollen joint count, patient global health and CRP. This immune based signature of response is consistent with findings at the individual transcript and splice-level, which showed significant changes in HLA, T-cell signalling and MMP genes implicated in RA pathogenesis. Specifically, upregulation of immune genes at 3 months could reflect migration of immune cells from the inflamed joint into the peripheral blood in a positive response to therapy. No significant changes were observed over time in NR.

Conclusion:  Identification of an immune blood-based signature of response early in the treatment time-course could aid timely therapeutic switching in NR. It could also offer a superior measure of ultrasound-determined synovitis than CRP alone. Subsequent work will include replication in an independent cohort and integration of genotype and serum microRNA data.

« Back to ACR/ARHP Annual Meeting 2016

ACR Meeting Abstracts - https://acrabstracts.org/abstract/identification-of-immune-gene-modules-in-good-responders-to-adalimumab-in-rheumatoid-arthritis/