A Personal 'Omics' Approach to Understand the Molecular Mechanism of Response to Biologic Therapy in Rheumatoid Arthritis

Abstract

Background: Anti-tumour necrosis factor (TNF) therapy has revolutionised the treatment of rheumatoid arthritis (RA). Nevertheless, a large number of patients fail to exhibit a complete response to these expensive drugs, which presents a challenge for responsible pharmacological spending and improving long term outcomes. To date, the quest for a biological biomarker or panel of biomarkers to predict and monitor treatment response has not yielded a replicable, clinically applicable test. Methods: Samples were selected from the Biologics in RA Genetics and Genomics Study Syndicate (BRAGGSS) cohort according to European League against Rheumatism (EULAR) criteria (50 good- and 20 non-responders to adalimumab; 37 good- and 18 non-responders to etanercept). Total RNA from baseline (pre-treatment) and 3-month samples were isolated from Tempus-stabilised whole blood and expression measured on the Affymetrix GeneChip Human Transcriptome Array (HTA) platform. Quality control and differential expression/splice analysis were assessed using appropriate Affymetrix and Bioconductor packages. A validation study was performed examining the extreme ends of adalimumab intermediate responders (n =24) using the OpenArray to measure gene expression in top hits from the initial HTA study. Results: In adalimumab good-responders, 198 transcripts were upregulated and 610 were downregulated at 3-months (fold-change > 1.2, false discovery rate (FDR) p-value < 0.05). No significant changes were observed in non-responders. In an equivalent study, 27 transcripts were upregulated and 395 transcripts were downregulated in etanercept good-responders at 3-months. Only six significant changes were observed in etanercept non-responders. Changes in adalimumab good-responders were most significantly enriched for genes related to 'B and T cell signalling in RA.' Modular analysis was performed to identify groups of co-regulated genes within the data that correlated with clinical outcomes. One of the modules most significantly associated with treatment timepoint in adalimumab good-responders was associated with osteoclast differentiation and leukocyte transendothelial migration. In etanercept good-responders, differential expression was most significantly associated with downstream TNF signalling. In total, expression changes at 97 unique genes were overlapping in good-responders to adalimumab and etanercept. The OpenArray study was able to validate 18 of the hits initially identified in adalimumab good-responders (fold-change > 1.2, FDR p-value < 0.05). Conclusions: Measurement of early biomarkers of response could be used for the prompt identification of non-responders and facilitate switching to an alternative therapy. This would reduce the impact of irreversible radiological damage and support efficient pharmacological spending. The signatures of response identified herein should be replicated in an alternative cohort and be investigated at an earlier timepoint of 4 weeks.

Date of Award	1 Aug 2018
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Anne Barton (Supervisor), Darren Plant (Supervisor) & Gisela Orozco (Supervisor)