Evidence for a protective role of the human leukocyte antigen class II region in early rheumatoid arthritis.

OBJECTIVE To analyse the distribution of predisposing DQ3 (DQB1*03(*04)/DQA1*03) and DQ5 (DQB1*0501/DQA1*01), shared epitope encoding and protective DRB1 alleles in patients with early rheumatoid arthritis (RA) and undifferentiated arthritis (UA). METHODS Consecutive patients enrolled in an early arthritis clinic were DNA-typed for HLA-DQ and DR. RA patients (n=195), UA patients (n=160) and controls from the same region (n=306) were sorted according to their DQ-DR phenotypes: DQ3 vs DQ5 and the presence or absence of a protective DERAA-positive DRB1 molecule. The three groups were also sorted according to the shared epitope (SE) hypothesis. RESULTS We observed the association of both DQ3 and DQ5 with RA. DQ3/3 homozygous individuals had the highest risk of developing disease [odds ratio (OR)=20.00]. Conversely DQ5, but not DQ3, was associated with undifferentiated arthritis (OR=2.15 vs 1.25). Consistent with these differences, DQ3-positive individuals had significantly more active disease at the first visit at the outpatient clinic than DQ5-positive patients. DRB1 alleles encoding a DERAA motif in their third hypervariable region provided a strong dominant protection against RA among DQ5-positive individuals and decreased arthritis activity among DQ3-positive patients. Individuals with SE-positive DR1, DR4 and DR10 alleles were also predisposed to RA, DR4/4 homozygous individuals having the highest risk of developing RA (OR=11.00). CONCLUSION The DQ3-DR4/9 haplotypes are associated with RA. The DQ5-DR1/10 haplotypes are associated with less active disease, i.e. UA, and DERAA encoding DRB1 alleles modulate either predisposition to or the severity of RA. We propose that HLA polymorphism influences not only the initiation or perpetuation of RA but also protection against the disease.