Background/Purpose: Subjects with rheumatoid arthritis (RA) are at increased risk of myocardial infarction (MI) compared to subjects without RA, with the increased risk potentially driven by inflammation. Anti tumour necrosis factor (TNFi) drugs may modulate the risk and severity of the MI. The aim of this analysis was to compare the risk and severity of MI in subjects treated with TNFi to non-biologic drugs (nbDMARDs) using linked data from BSRBR-RA (an ongoing prospective study of the safety of biologic therapy in RA) and the Myocardial Ischaemia National Audit Project (MINAP), a national database of hospitalisations with MI in England and Wales.

Methods: This analysis included subjects with RA starting TNFi (etanercept, infliximab, adalimumab) and a biologic-naïve comparator cohort receiving nbDMARDs recruited between 2001-2008; all followed via physician and patient questionnaires and flagged with the national death register. All subjects were also linked to MINAP using deterministic matching which provided both additional events and further event data regarding MI severity. Events from both datasets were verified using the American Heart Association/European Society of Cardiology criteria for MI with additional criteria of thrombolysis/angioplasty and MI listed as the underlying cause of death on death certificates, coded using International Classification of Diseases 10. When estimating incidence of MI, subjects were censored at first MI, death, 90 days following discontinuation of TNFi, last physician follow-up or 04/20/2010, whichever came first. The risk of MI was compared between subjects receiving nbDMARDs and TNFi using a Cox regression model, adjusted on deciles of propensity scores (PD) (Table). In a subset of MIs with additional event data from MINAP; MI phenotype and severity (using surrogates: cardiac arrest, peak creatine kinase (CK) levels and length of hospital stay) were compared between treatment groups using descriptive statistics.

Results: Using the linked data, a total of 252 verified first MIs were analysed: 58 in 3058 nbDMARD subjects and 194 in 11200 TNFi-treated subjects. The PD-adjusted hazard ratio (HR) of MI in TNFi referent to nbDMARD was 0.61 (95%CI 0.41, 0.89). The estimate was unchanged when all follow-up time following first dose of TNFi was included. There were 143 MIs (nbDMARD: 35, TNFi: 108) with additional MINAP data. No statistical differences were observed in MI severity when compared between nbDMARD and TNFi-treated subjects, although there was a trend towards higher peak CK levels recorded in TNFi-treated subjects (Table).

Conclusion: Subjects receiving TNFi drugs had a decreased risk of MI compared to subjects receiving nbDMARDs over the medium term which could relate to attributes of the drug itself or better overall disease control. The severity of MIs at presentation appears to be similar between both treatment groups.

 Table