Tumour necrosis factor alpha (TNF alpha) single nucleotide polymorphisms (SNPs) in the idiopathic inflammatory myopathies (IIMs)

Aim: To investigate links between the TNF gene, and the haplotype associations of TNF, HLA class I and II, and the IIMs Methods: DNA samples were obtained from 286 Caucasian IIM patients, consisting 111 polymyositis (PM) (median age of onset 50.5 [±14.5] years), 98 dermatomyositis (DM) (49 [±13.3] years), and 77 polymyositis/overlap disease (PM/OL) (46 [±12.1] years), recruited via the Adult Onset Myositis Immunogenetic Collaboration (AOMIC). All had definite myositis, as per the Tanimoto et al criteria. Four polymorphic TNF positions (-1031, -863, -308, +489) were typed using a fluorescence-based primer method (SNaPshotTM). The cohort had previously been typed for HLA-B and –DRB1. Data was compared to that of 185 ethnically matched controls. Genotype frequencies for each TNF SNP were tested for Hardy-Weinberg equilibrium in the controls. A haplotype trend regression analysis was also performed. This uses the Expectation/Maximization (EM) algorithm to estimate haplotype frequencies, and then relates haplotype probability distribution to observed outcome, using a regression-based approach. Results: TNF-308A was a significant risk factor in all IIM patients combined (30.4% of patients vs 15.7% controls, odds ratio [OR] 2.3, 95% confidence interval [CI] 1.6-3.3). The association was still significant when stratified by IIM subgroup, the strongest being in the PM/OL group (OR 2.85, 95% CI 1.8-4.5), but no subgroup differences were apparent. In most cases, a single copy of the TNF-308A allele was sufficient to increase the frequency, suggesting a dominant mode of inheritance. Furthermore, TNF-1031T was a weak but significant risk factor in DM (85.7% patients vs 75.7% controls, OR 1.93 (95% CI 1.2-3.1), the association being lost after correction for multiple comparisons. The combination of B*08/TNF-308A/ DRB1*03 was present at a higher frequency in the IIMs, compared to controls (controls=15.7%, PM=29.4%, DM=20.3%, PM/OL=20.7%). Haplotype trend regression analyses revealed significant associations with the 4 SNP haplotypes of TNF for all IIM subgroups (p