No abstract

No abstract

We have defined HLA-DQA, DQB, DR and complement C4 variants in 92 subjects with SLE and 73 controls. Subjects with SLE showed an increased frequency of HLA-DQA*0501 (P < 0.01 corrected, odds ratio (OR) = 4.97; 95% C.I. = 2.52-9.81), DR3 (P < 0.001, OR = 3.18; 95% C.I. = 1.67-6.04) and C4A*Q0 (P < 0.05, OR = 1.91; 95% C.I. = 0.999-3.65) vs controls. These increases were particularly marked in those subjects positive for antibodies to both Ro and La. HLA-DQB*0501 (P < 0.01 corrected, OR = 0.03), DQA*0101 (P = 0.0012 uncorrected, OR = 0.23) and DR7 (P = 0.0018 uncorrected, OR = 0.28) were decreased in frequency in SLE. SLE patients with disease onset prior to age 30 yr were more likely to possess a DR3-bearing haplotype (P < 0.05 corrected) than those with onset after age 30 yr. No significant associations were found in patients with circulating antibodies to double-stranded DNA, Ro alone, U1 RNP, Sm or in those SLE patients with renal disease or vasculitis. The different associations found in different clinical and immunological subsets of SLE support the concept that SLE contains a variety of immunogenetic subgroups. Analysis of the associations between SLE and DR3, DQA*0501 and C4A*Q0 using the empirical logistic test suggests that the association of SLE with HLA-DQA*0501 is likely to be primary to the associations with both DR3 and C4A*Q0 (P < 0.001). Our results therefore raise the possibility that genes within the HLA-DQ region may have a direct effect upon susceptibility to SLE.