Abstract
Measurement of serum rheumatoid factor (RF) by conventional methods in patients with rheumatoid arthritis (RA) has repeatedly identified a subpopulation of patients without detectable RF. Previous investigators have consistently confirmed the association of HLA–DR4 with seropositive RA, but studies of seronegative RA have been limited and contradictory. We studied 140 randomly selected patients from Alabama, all of whom had either classic or definite RA, and we were able to obtain complete HLA typing for 110 of these individuals. Eighty were consistently seropositive (on at least 2 separate occasions) and 30 were consistently seronegative (on at least 3 separate determinations). There was no statistically significant difference between the seronegative RA patients and 123 control subjects in the distribution of DR antigens. In seropositive RA, there was a significant increase in DR4 (P<0.001; relative risk = 8.02; attributable risk = 49.2%) and a significant decrease in DR3 (P<0.001; relative risk = 0.14) and DR7 (0.01>P>0.001; relative risk = 0.33). The clinical data also distinguished between seropositive RA and seronegative RA; subcutaneous nodules (37.5%) and vasculitis (6.3%) were present only in seropositive RA. DR4 positivity did not correlate with any of the clinical variables measured in the seropositive RA group. In contrast, DR4 in the seronegative RA group was associated with more destructive disease. The data suggest that seronegative RA represents a disease entity clinically and immunogenetically distinct from seropositive RA. Moreover, our results indicate that DR4 may be a previously undisclosed marker for disease severity in seronegative RA.