Article Text

PDF
FRI0596 Rheumatoid Arthritis Risk Alleles, Hla-Drb1 Haplotypes, and Response To TNFI Therapy – Results from A Swedish Cohort Study
  1. X. Jiang1,
  2. J. Askling1,
  3. S. Saevarsdottir1,
  4. L. Padyukov1,
  5. S. Viatte2,
  6. T. Frisell1
  1. 1Karolinska University Hospital, Stockholm, Sweden
  2. 2Arthritis Research UK Epidemiology Unit, Center for Musculoskeletal Research, Institute of Inflammation and Repair, manchester, United Kingdom

Abstract

Background Tumour necrosis factor inhibitors (TNFi) have proven successful in suppressing both inflammation and joint damage in rheumatoid arthritis. Still, predictors of response are largely lacking, to increase our understanding of the disease-driving biology and response to treatment. Several pharmacogenetics studies, including five GWAS, have been conducted to identify genetic predictors to TNFi response, but few markers have reached genome-wide significance or withstood replications.

Objectives By taking a genetic risk score approach, combining information contained in individual RA risk markers, amino acids, and haplotypes, we may be able to better predict treatment response to TNFi.

Methods We linked recent onset RA patients from the Swedish EIRA study (genotyped with the Illumina Immunochip array) to prospectively recorded data on treatment and clinical characteristics in the nationwide Swedish Rheumatology Register (SRQ). The cohort was defined as those who started TNFi as first biological DMARD treatment 2000–2012, and had at least one return visit with recorded DAS28 within 2–8 months after TNFi start (N=867). Based on previous reports, we identified established RA genetic risk factors: 76 RA risk loci, 4 amino acid polymorphisms at HLA-DRB1 (position 11, 13, 71, 74), and calculated 16 haplotypes based on the amino acids. We calculated risk scores for the 76 SNPs, the 4 amino acids, and the haplotypes, by weighting SNP alleles/amino acid residues/copies of haplotype by their published log-odds. The association of RA risk scores and individual established genetic risk factors with DAS28-based EULAR response and a number of clinical disease activity measures at the evaluation visit closest to 5 months after initiation was estimated using logistic and linear regression models.

Results All risk scores were good predictors for RA, in particular for ACPA-positive RA. We found, however, no association for any of the risk scores, or SE, in achieving good/moderate EULAR response in RA overall, nor in any RA subset. When evaluating each of the 76 SNPs, individual amino acid residues and haplotypes separately, we found that the number of SNPs with significant associations to response for TNFi was not higher than expected by chance (5/76 SNPs had p<0.05 in ACPA-positive RA, 4/76 in ACPA-negative RA). Similarly, despite the borderline significant association for some residues in ACPA-positive RA (glycine at position 11, serine and tyrosine at position 13) and in ACPA-negative RA (proline at position 11, arginine at position 13), none survived correction for multiple testing. In addition, none of the haplotypes, neither those based on 4 amino acids (11/13/71/74) nor those based on 3 amino acids (11/71/74), were significantly associated with EULAR response. With regard to the changes in individual disease activity measures, no associations were identified except that the risk scores may explain a small proportion of variance in HAQ changes.

Conclusions Identified RA risks genes, amino acids, and haplotypes in our study do not predict response to TNFi therapies, neither individually nor when groups together into scores.

Disclosure of Interest None declared

Statistics from Altmetric.com

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.