Interactions of Cigarette Smoking with NAT2 Polymorphisms Impact Rheumatoid Arthritis Risk in African Americans

Over the last thirty years, numerous investigations have confirmed the existence of an epidemiologic link between cigarette smoking and rheumatoid arthritis (RA) susceptibility [1–9]. The risk of RA that is attributable to smoking is most striking among those with heavier cumulative exposures, especially among individuals developing anti-citrullinated protein antibody (ACPA) positive disease [1, 10]. Among individuals of European ancestry, the attributable risk of RA due to ever smoking has been estimated to approach 1 in 6 [6]. In other words, 16% of all RA cases could be prevented in this population with the elimination of cigarette smoking as an environmental health hazard. The risk of ACPA positive RA appears to be even higher among cigarette smokers carrying HLA-DRB1 shared epitope (SE) containing alleles. In a recent Swedish case-control study, the attributable risk of ACPA positive RA due to heavy smoking among individuals homozygous for HLA-DRB1 SE approached 1 in 2 [11]. In this study, individuals homozygous for SE with a cumulative smoking exposure exceeding 20 pack-years were ~40-fold as likely to have ACPA positive RA compared to never smokers negative for SE. In addition to being associated with disease risk in populations of European ancestry, our group recently observed strong associations of heavy cigarette smoking with RA risk in African Americans [12]. Similar to findings from populations of European ancestry, we found that heavy smoking defined as more than 10 pack-years of exposure was associated with a more than two-fold risk of RA in African Americans. In this study, we observed an attributable RA risk due to heavy smoking of 16%, similar to the attributable risk due to ever smoking found in a prior prospective cohort study of older women from Iowa [6]. In addition to the effect of HLA-DRB1 SE, variation in genes encoding drug metabolizing enzymes (DMEs) could also impact the effect of smoking on RA risk. This hypothesis is consistent with the growing epidemiologic literature showing that the incidence of other smoking-related illnesses is heightened in individuals inheriting polymorphisms that detrimentally impact DME function and ultimately impair xenobiotic metabolism and elimination of smoking-related toxins. For instance, lung cancer risk is higher in cigarette smokers homozygous for a deletion polymorphism in glutathione S-transferase Mu-1 (GSTM1-null) compared to smokers expressing functional GSTM1, a risk that appears to be most striking in Asian populations [13, 14]. Likewise, both breast [15, 16] and bladder cancer [17] risk appear to be higher among cigarette smokers inheriting haplotypes associated with reduced N-acetyltransferase (NAT)-2 function. Variation in genes encoding alternative DME’s including NAT1 and epoxide hydrolase (EPXH1) have been suggested to mediate the risk of smoking related illnesses including select malignancies and chronic lung disease [18–21]. To date, potential interactions between these DME genes and smoking in RA risk have been subject to only limited study [22, 23] with prior investigations conducted exclusively in populations of European ancestry. The lack of such studies in African Americans represents a major knowledge gap, particularly since heavy smoking is strongly associated with disease risk in this population and smoking rates appear to be on the rise among African Americans [24]. In the present case-control study, we examined whether the risk of RA in African Americans attributable to cigarette smoking differs based on the presence of polymorphisms in a select number of DME genes. Specifically, we have examined whether there is evidence of significant DME gene-smoking interactions impacting RA risk in this understudied population.