Background Up to 40% of rheumatoid arthritis (RA) patients exhibit insufficient response to TNF inhibitor (TNFi) therapy, which has an adverse effect on long term outcomes. Without reliable biomarkers to direct treatment decisions, many non-responder (NR) patients experience a delay in switching to an alternative therapy. Ideally, blood-based biomarkers would be measured before treatment (baseline) and throughout treatment in order to select and monitor therapeutic response to maximise the chances of responding to the first biologic therapy.
Objectives To compare transcriptomic changes between patients administered etanercept and adalimumab therapy, and to identify biomarkers to predict or monitor response.
Methods From the Biologics in RA Genetics and Genomics Study Syndicate (BRAGGSS) cohort, 37 EULAR good-responders in clinical remission (GR) and 18 NR to etanercept, and 50 GR and 20 NR to adalimumab were selected. Total RNA was isolated from Tempus™-stabilised whole blood samples collected at baseline and following 3-months (3M) of therapy using the MagMAX™ RNA extraction kit. RNA was amplified and converted into biotinylated sense-strand DNA using the Affymetrix WT PLUS kit for hybridisation onto Affymetrix GeneChip® Human Transcriptome arrays. Quality control and differential expression/splice analysis were assessed using the Affymetrix Expression and Transcriptome Analysis Console™ and appropriate Bioconductor packages. Differential transcript expression was adjusted for baseline DAS, age, gender and concurrent DMARD therapy. Pathway analysis was performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) and Ingenuity Pathway Analysis (IPA) tools.
Results In adalimumab GR, 636 genes were downregulated and 253 upregulated at 3M (FDR p <0.05, fold-change >1.2). There was significant upregulation of immune cell components, most notably HLA genes including HLA-DRB1, other RA susceptibility genes (SLC2A4, PADI4 and CD28) and many B and T cell signalling genes. Etanercept GR exhibited a milder transcriptomic change overall, showing little overlap with adalimumab GR; 395 genes were downregulated and 27 upregulated at 3M (FDR p<0.05, fold-change >1.2). Downregulated genes included downstream TNF components such as mitogen activated protein (MAP) kinases, as well as genes involved in NOD-like receptor, Toll-like receptor and NF-κB signalling. Such significant changes were absent in NR to adalimumab and etanercept. Furthermore, alternative splice changes in RA-relevant genes such as MMP9 were apparent in adalimumab GR at 3M but not etanercept GR.
Conclusions The heterogeneity in the blood-based transcriptomic profiles of etanercept and adalimumab response observed herein suggests that different TNFi therapies function by alternative mechanisms that impact patient outcomes. It also calls into question the reliability of response studies that consider TNFi therapies as a homogenous group. The candidate biomarkers identified require replication in independent datasets but may provide early and objective response biomarkers to inform timely therapeutic switching in patients who are not responding to their current TNFi drug.
Disclosure of Interest None declared