Background:The HLA shared epitope (SE) and smoking are the best known genetic and environmental risk factors for rheumatoid arthritis (RA) development; however, at which pre-RA stage they exert their effect is unknown. The following stages are discerned: an asymptomatic stage in which autoimmune responses can develop, a symptomatic stage (clinically suspect arthralgia (CSA)), and development of clinically apparent inflammatory arthritis (IA). Studies in the general asymptomatic population revealed contrasting results on the associations between SE-alleles and smoking and the presence of anti-citrullinated protein antibodies (ACPA). Furthermore, studies on these risk factors in the symptomatic pre-RA phase are scarce and these data might teach us whether SE-alleles and smoking are involved in symptom development and/or progression to clinical arthritis.Objectives:We aimed to determine at which pre-RA stage SE and smoking exert their effect. In this respect, the analyses were focused on the presence of ACPA, but associations for other anti-modified protein antibodies (anti-carbamylated and anti-acetylated protein antibodies (anti-CarP and AAPA, respectively)) were also studied.Methods:Results from the literature on the association of SE and smoking with ACPA in the asymptomatic population were summarized in inverse-variance weighted meta-analyses. In addition 577 CSA-patients were studied. Associations of SE and smoking with IgG ACPA were studied at baseline (CSA-onset), to assess an effect on symptom development. Additionally, patients were monitored for the development of clinically apparent inflammatory arthritis (IA) for median 2 years and associations of SE, smoking and auto-antibodies with progression to IA were determined. Analyses were stratified for ACPA-status and associations in ACPA-positive patients were validated in meta-analyses with other arthralgia-cohorts. Finally analyses were repeated for anti-CarP and AAPA.Results:Meta-analyses showed that SE is not associated with ACPA-positivity in the asymptomatic population (OR 1.06 (95%CI 0.69-1.64)), whereas smoking was associated (OR 1.37 (1.15-1.63)). At CSA-onset, both SE and smoking associated with ACPA-positivity (OR 2.08 (1.24-3.49) and OR 2.41 (1.31-4.43), respectively). During follow-up of CSA-patients SE associated with IA-development (HR 1.86 (1.23-2.82)), in contrast to smoking. SE conferred risk for IA-development in ACPA-negative CSA-patients (HR 1.71 (0.99-2.96)) and in ACPA-positive patients (CSA-cohort HR 1.29 (0.67-2.47); meta-analysis three arthralgia-cohorts HR 1.52 (1.08-2.15)). Investigating the other autoantibodies revealed that SE and smoking were not associated with anti-CarP or AAPA-positivity at CSA-onset; longitudinally AAPA associated with progression to IA independent from ACPA and RF (HR 1.79 (1.02-3.16)), whilst anti-Carp did not.Conclusion:SE and smoking act in partly different pre-RA stages. Although SE does not associate with ACPA in the general population, it does mediate symptom-development and further progression to clinical arthritis. Smoking confers risk to development of ACPA and/or joint symptoms, but is not further involved in IA-development. The time-specific biologic pathways that are underlying need further exploration. These data enhance the understanding of the timing of key genetic and environmental risk factors in the trajectory of RA development.Disclosure of Interests:None declared
Environmental risk factors of pregnancy outcomes: a summary of recent meta-analyses of epidemiological studies
