Background More than 40 loci have been definitively associated with RA susceptibility. However, they only explain a small fraction of RA heritability. Interactions between polymorphisms could explain part of the remaining heritability. Multiple interactions have been reported but only the shared epitope (SE) X PTPN22 interaction has been replicated convincingly. Two recent studies deserve attention because of their quality including replication in a second sample collection. One of them has identified interactions between PTPN22 and 7 SNPs1. The other showed interaction between SE and the null allele of GSTM1 in the anti-citrullinated peptide (ACPA) positive patients2.
Methods Samples of 1744 patients with RA and 1650 healthy controls of Spanish ancestry were studied. Polymorphisms were genotyped by single base extension; SE genotypes of 725 patients were available from previous studies. Interaction analysis was done with three methods that included the originally reported: assessment of the interaction term in logistic regression models, evaluation of the relative fit to the data of nested logistic regression models as implemented in LRASSOC3, and additivity deviation following Hosmer & Lemesshow4 as implemented in R.
Results Genotypes of one of the SNPs, rs4695888, failed quality control. Call rate for the other 8 polymorphisms was 99.9 %. All of them were in HWE. Their frequencies were similar in RA patients and controls, except for PTPN22. None of the interactions between PTPN22 and the 6 SNPs from Briggs et al.1 was replicated as a significant interaction term, the originally reported finding, or with any of the other two methods. Neither was replicated the interaction between GSTM1 and SE as a deviation from additivity in ACPA+ patients2 or with any of the other two methods.
Conclusions None of the interactions tested was replicated in spite of sufficient power and assessment with three different assays. These negative results indicate that we still do not know whether interactions are a significant contribution to RA susceptibility or not.
Briggs FBS, et al. Genes Immun. 2010; 11: 199–208.
Mikuls TR, et al. Arthritis Res Ther. 2010; 12:R213.
North BV, et al. Human Heredity 2005; 59: 79-87
Hosmer DW, Lemeshow S. Epidemiology. 1992; 3:452-6
Acknowledgements Funding was provided by grants PI11/01048, PI12/01909 and by RETICS Program, RD08/0075 (RIER) of the Instituto de Salud Carlos III (Spain) that are partially financed by the European Regional Development Fund of the European Union.
Disclosure of Interest None Declared