Background The increased risk of cardiovascular disease in rheumatoid arthritis (RA) has been linked to markers of systemic inflammation, as well as to traditional risk factors.
Objectives To evaluate whether the level of response to a first tumor necrosis factor inhibitor (TNFi) is associated with the risk of acute coronary syndrome (ACS) in RA.
Methods From an underlying cohort of patients with RA and no previous ischemic heart disease, who started TNFi therapy 2001-2010 (n=7,704) identified from the National Patient Register and the Swedish Biologics Register, we identified 5,599 patients (mean age 57 years, 76% women) eligible for evaluation. From the Population Register 27,891 matched general population referents were identified. The level of EULAR response was evaluated at a visit 5±3 months after TNFi start. Covariates were extracted from the Patient register (comorbidity, joint surgery, disease duration) and Statistics Sweden (educational level and work disability). The outcome, incident ACS, was defined as a primary discharge diagnosis of myocardial infarction, or unstable angina, or myocardial infarction as the underlying cause of death, occurring during the first, or the first two years after the response visit. Incidence rates were calculated and Cox Proportional Regression models were utilized for risk estimations. In 661 patients (12%) data was incomplete, lacking one or more variables necessary for calculating EULAR response. For risk comparisons within the patient cohort, this was managed using multiple imputation. Sensitivity analyses were performed using the first and the best DAS28 measurement, respectively, within the evaluation time window.
Results A good EULAR response was observed in 37%, a moderate response in 37% and no response in 26% of the patients. The 33 ACS occurring among the patients during the 1st year resulted in the crude incidence rate (IR; 95%CI) 7.2 (5.1-10.1) per 1,000 person-years, and the 59 ACS during the first two years in the IR 6.9 (5.3-8.9). The hazard ratio (HR; 95%CI) adjusted for age, sex, inclusion year and county comparing patients with good compared with no response was 0.3 (0.1-0.8) for the risk of ACS during the 1st year and 0.4 (0.2-0.8) for the first two years. Only marginal changes in risk estimates were noted after further adjustments, and in the sensitivity analyses. No statistically significant difference in the risk was observed between patients with moderate and no response. The risk of ACS among patients with good response compared with the general population was for the 1st year HR 0.9 (0.3-2.4) and the first two years HR 1.2 (0.6-2.2). The corresponding HRs among moderate responders were 3.1 (1.8-5.4) and 2.5 (1.7-3.8), and for non-responders 3.5 (1.9-6.3) and 2.8 (1.8-4.4), compared with the general population.
Conclusions Good EULAR response on a first TNFi in patients with RA was associated with a significantly lower risk for ACS one and two years after the response visit, in comparison with no response. In patients with good response on therapy no significant increase in the risk for ACS was detectable compared with the risk in general population.
Disclosure of Interest L. Ljung Speakers bureau: fees for lectures from Bristol Myers Squibb and Abbvie, L. Jacobsson Speakers bureau: fees for lectures from Abbvie, Pfizer and UCB., S. Rantapää-Dahlqvist: None declared, J. Askling: None declared