Background Rheumatoid arthritis (RA) treatment may target B-cells using rituximab (RTX), a chimeric monoclonal antibody directed against the B-cell marker CD20. Transcriptome of PBMCs from RA patients before RTX has suggested that responders differs from non-responders on the expression of genes regulated by type 1 interferon (IFN) (1) suggesting a possible contribution of type 1 IFN pathway genes such as interferon regulatory factor 5 (IRF5), tyrosine kinase 2 (TYK2), signal transducer and activator of transcription 4 (STAT4), osteopontin (SPP1) and interferon regulatory factor 7 (IRF7).
Objectives To study 8 IRF5, TYK2, STAT4, SPP1 and IRF7 variants as candidate predictors of the clinical response to RTX in RA at 6 months defined by a good or moderate EULAR response.
Methods Patients from the SMART study who signed consent for genetic studies were included (2). The association between the EULAR response at 24 weeks and the polymorphism of the 8 chosen SNPs was analyzed by logistic regression. Response rates were compared across risk genotypes using the χ2 test or the Fisher exact test, when appropriate. Significant variables after univariate regression (P<0.15) were entered into a stepwise multivariate model adjusted to the sex, the age and the DAS28-CRP score.
Results Among the 115 included patients, 92 were women, mean age was 56 years, 93 patients were anti-CCP positive. At M6, 30 patients were considered as good responders (26.1%), 62 as moderate responders (53.9%) and 22 as non-responders (19.1%). Univariate analysis showed that IRF5 rs2004640 was significantly associated with a good/moderate EULAR response at week 24 (P=0.038) whereas IRF7 rs1131665 and SPP1 rs9138 were near statistical significance (P=0.092 and P=0.089, respectively). Since it was previously reported an association between TNFSF13B (the gene coding the BAFF cytokine) rs9514828 and RTX response in RA in the same cohort (2), this variant was also included in the multivariate analysis. The multivariate analysis with logistic regression analysis including IRF5, SPP1, IRF7 and TNFSF13B revealed that a genotypic combination of IRF5 rs2004640, SPP1 rs9138 and TNFSF13B rs9514828 was strongly associated with good/moderate response to RTX at M6: P=9.34x10–6, OR =11.37 [4.03–35.28] with a PPV of 91% and a NPV of 53% (Fig. 1).
Conclusions We found an association between a combination of type 1-interferon genes and response to RTX treatment in RA. To further validate the magnitude of the association detected an independent replication in a larger cohort is required. If validated, our model could be helpful before the decision of initiating RTX in RA.
Vosslamber S, et al. Ann Rheum Dis 2011.
Ruyssen-Witrand A, et al. Rheumatology 2013.
Acknowledgement Unrestricted grant from Roche-Chugai and SFR
Disclosure of Interest None declared