Background B-cell activating factor of the TNF family (BAFF) plays a role in (auto)antibody production. Patients with idiopathic inflammatory myopathies (IIMs or myositis) have elevated levels of BAFF in the serum (sBAFF) which associate with disease activity. Previously an association was shown between rs9514828 (-871 C/T) SNP (single nucleotide polymorphism) in the promoter region of the BAFF gene and idiopathic thrombocytopenic purpura and with phenotypic diversity in autoimmune rheumatic diseases but with inconsistent results. An association between the haplotype from the four SNPs located upstream of the BAFF gene and presence of myositis was found in one cohort of Czech patients (CZE).
Objectives To replicate our findings in an independent cohort of myositis patients from Sweden (SWE) and to study correlations between particular SNPs and serum levels of BAFF.
Methods Two cohorts of patients with myositis (CZE: n=311, age (years) 54.6, females 74% and SWE: n=307, age 54.1, females 64%) and healthy controls (CZE: n=103, age 42, females 69% and SWE: n=325, age 52.0, females 74%) were included in the study. SNPs (rs9514827, rs3759467, rs1041569 and rs9514828) were analysed by direct DNA sequencing and levels of sBAFF were measured by ELISA. The χ2 or Fisher tests for analysis of allelic, haplotype and genotype associations and nonparametric tests for group comparisons with Bonferroni's correction were used.
Results The four SNPs were in strong linkage disequilibrium and formed four common haplotypes (TTAC, CTAT, TCAC, TTTT) in both patients and controls, compatible to results reported from other populations. A significantly higher frequency of the TTTT haplotype was present in the CZE myositis patients compared to healthy controls (16% vs. 10%; OR=1.73, 95%CI=1.05-2.85, p<0.02), but not in the SWE cohort (patients 13% vs. controls 17%, p=0.07). CZE and SWE myositis patients did not differ significantly in allelic, genotype or haplotype frequencies, but a difference between the control groups was observed. SWE controls had higher frequencies of the rs1041569 T-allele and rs3759467 TC-genotype compared to CZE HC (19% vs 12% and 33% vs 19% respectively; p<0.05 for both). No associations were found between alleles or genotypes and serum levels of BAFF for any of the studied SNPs.
Conclusions We found different patterns of associations between BAFF genetic polymorphism and myositis in two European populations. Our initial finding in Czech population could not be replicated in an independent cohort from Sweden. The difference between the control groups requires further analyses.
Acknowledgements IGA -MZ Czech Republic NT/12438-4, ESF in the framework of the EuMyoNet, Krystufkova O. and Svitalkova T. contributed equally
Disclosure of Interest O. Krystufkova Grant/research support: IGA -MZ Czech Republic NT/12438-4, ESF in the framework of the EuMyoNet, T. Svitalkova: None declared, H. Hulejova: None declared, M. Svetla: None declared, L. Plestilova: None declared, O. Pecha: None declared, H. Mann: None declared, A. Notarnicola: None declared, P. Venalis: None declared, L. Padyukov: None declared, P. Novota: None declared, I. Lundberg: None declared, J. Vencovsky: None declared