Identification of genetic biomarkers of response to conventional and biologic therapies in rheumatoid arthritis

Around 30‐40% of rheumatoid arthritis (RA) patients fail to respond adequately to diseasemodifying anti‐rheumatic drugs (DMARDs). These individuals are at an elevated risk of irreversible joint damage resulting in functional impairment. Thus, there is an unmet need for markers that predict response to therapies so that treatment can be optimised before irreversible joint damage occurs. Genetic variation provides a potential pool of biomarkers and has been the subject of investigation across drugs, including methotrexate (MTX) and TNF antagonists. This doctoral work sought to identify single nucleotide polymorphisms (SNPs) that were statistically associated with response to DMARD therapy based on their association with disease susceptibility. Nominal association was found between HLA‐DRB1 shared epitope alleles and worse response. These susceptibility SNPs were investigated in the context of early joint damage in RA; associations were observed with alleles at the AFF3, C8orf13‐BLK, IL2RB and KIF5A‐PIP4K2C loci. Publicly available expression data from healthy primary B cells and monocytes was used to investigate cis acting expression quantitative trait loci (ciseQTL) over these regions. Multiple independent isoform specific cis‐eQTLs were identified with AFF3 transcripts. A candidate gene association study investigated SNPs across the genomic regions encoding proteins involved in MTX metabolism and transport. A non‐synonymous SNP, rs8923, in the MTHFS gene provides a putative causal candidate for further investigation. TNF antagonist biologics containing Fc regions are likely to bind IgG receptors expressed on cells of the immune system. Several genes encoding the low‐affinity IgG receptors (FcγR) are subject to copy number variation (CNV). An assay to measure CNV of FCGR2C was developed and applied in an association study of FCGR genetic variation with response to TNF antagonist therapy. Association was observed between FCGR2A alleles and response to biologic therapies in a drug class dependent manner. These data highlight the complexity involved in identifying treatment response biomarkers.