Background/Purpose  African American (AA) patients with rheumatoid arthritis (RA) have been shown to have worse disability scores, increased disease activity scores, and reduced use of disease modifying anti-rheumatic drugs.  The frequency of antinuclear antibodies (ANA) in Caucasian (CAU) RA has been reported, but ANA occurrence in AA RA patients is unknown.  ANA positive CAU RA patients have an increased autoimmune genetic risk score and this phenotype is associated with coexistent Sjögren syndrome (SS). There is a high frequency of anti-SSA antibodies in SS. We hypothesized that AA RA patients have an increased frequency (compared to CAU RA patients) of ANA, anti-SSA, and other connective tissue disease (CTD) associated autoantibodies.    

Methods  We assayed plasma samples from two published RA cohorts: (1) AA patients in the Consortium for the Longitudinal Evaluation of African-Americans with Early Rheumatoid Arthritis (CLEAR II) and (2) CAU patients from the Treatment of Early Aggressive Rheumatoid Arthritis (TEAR) Trial.  All patients met the revised 1987 American College of Rheumatology classification criteria for RA.  A total of 587 CLEAR and 390 TEAR samples were tested for ANA by indirect immunofluorescence (IIF) and the multiplex bead array (MBA) Bioplex 2200 ANA Screen.  Differences in prevalence between AA and CAU RA patients was assessed using tests of proportions with chi-square and Fisher’s exact tests.   Multivariable logistic regression was used to adjust results for age, sex, body mass index (BMI), modified HAQ, RF and anti-CCP positivity. 

Results  ANA by IIF was positive (≥ 1:80) significantly more frequently in AA RA patients (187/587 or 32%) compared with CAU RA patients (65/390 or 17%, p<0.001).  The pattern of nuclear staining among patients with a positive ANA significantly differed by racial/ethnic group (p<0.001), with AA RA patients having higher frequencies of speckled staining (64% vs. 46%; p=0.01) where as CAU RA patients had higher frequencies of nucleolar staining (10% vs. 3%; p=0.03).  Anti-SSA was detected in twice as many AA RA patients (80/587 or 14% vs. 22/390 or 6%, p<0.001). Systemic lupus associated antibodies (anti-Sm, anti-RNP, anti-Sm/RNP, and anti--chromatin) were also higher in this population (see Table 1). Higher prevalence of ANA and anti-SSA remained significant (p<0.001) after adjustment. In the AA RA patients, logistic regression showed only RF was associated with ANA and anti-SSA without previous TNF use or disease duration significant.

Conclusion We found that the prevalence of ANA, anti-SSA, and other CTD associated autoantibodies is higher in AA RA patients than in a CAU RA cohort. The increase in ANA prevalence is greater than the published difference in ANA by race (15% vs. 3%) and suggests the existence of a meaningful serological subset of AA RA patients beyond the currently recognized serologic subsets identified by RF and anti-CCP antibodies.