SAT0615 THE DIFFERENT EFFECTS OF CIGARETTE SMOKING ON ANTI-CITRULLINATED CYCLIC PROTEIN ANTIBODY AND RHEUMATOID FACTOR FORMATION IN RELATION TO SHARED EPITOPE ALLELES IN JAPANESE RA PATIENTS

Abstract

Background In rheumatoid arthritis (RA), cigarette smoking affects both rheumatoid factor (RF) and anti-citrullinated cyclic peptide/protein antibody (ACPA) formation, but its association in relation to HLA-DRB1 alleles, especially shared epitope (SE) alleles, have been controversial among different races1-5. Furthermore, the impact of cigarette smoking and its cessation on levels of RF and ACPA have not been well documented. Objectives To investigate the impact of cigarette smoking and cessation on RF and ACPA levels in relation to HLA-DRB1 alleles in the largest RA cohorts in Asians. Methods A total of 6,239 subjects from two independent Japanese cohorts were enrolled. Their precise smoking histories both before and after the onset of RA were collected in questionnaires. The latest RF and ACPA levels were used (mean disease duration 15.6 years). We defined top quadrant of levels of RF or ACPA as high levels. Associations between smoking status and positivities or high levels of ACPA or RF as well as effects of HLA-DRB1 alleles on the associations were investigated by multiple logistic regression models. Results Smoking at onset was an independent risk of not only RF and ACPA positivities (RF, odds ratio (OR) 1.52, 95% confidence interval (CI) 1.26-1.85, p=1.8x10-5; ACPA, OR 1.39, 95%CI 1.09-1.76, p=6.8x10-3), but also high levels of these autoantibodies, especially RF (OR 2.06, 95% CI 1.70-2.48, p=7.4x10-14; ACPA OR 1.29, 95%CI 1.06-1.57, p=1.2x10-2). The larger ORs of RF than ACPA suggests that RF is more sensitive to cigarette smoking than ACPA. The effects of cigarette smoking were significantly larger in males than in females. The patients who quited smoking before onset had no longer significant risks of high autoantibody levels compared to subjects who had never smoked (RF, OR 1.33, p=0.099; ACPA, OR 1.19, p=0.093), and the risk was gradually attenuated depending on cessation years (RF, 0-10 years OR 1.34, 10-20 years OR 1.31, > 20 years OR 0.97; ACPA, 0-10 years OR 1.38, 10-20 years OR 1.01, > 20 years OR 1.12). The effect of smoking on ACPA positivity and its high level was apparent only in the presence of SE alleles, while the effect on RF positivity and its high level was apparent despite the presence of SE alleles (Table below). Conclusion Cigarette smoking especially at RA onset is a significant risk of future high levels of ACPA and RF preferentially in males, and RF is more sensitive to smoking status than ACPA. The effect on ACPA is apparent only in the presence of SE alleles, indicating that an interaction between cigarette smoking and SE alleles affects ACPA formation. On the other hand, the effect of cigarette smoking on RF formation may be independent of SE alleles. Our study imply a novel potential mechanism of RA pathogenesis.