TNF Inhibitor Use Associated with a Reduced Risk of Myocardial Infarction in Rheumatoid Arthritis Patients?

Background:

RA is associated with an increase in cardiovascular (CV) events, such as myocardial infarction (MI) and sudden cardiac death.  Inflammatory cytokines, such as TNF-α, are implicated in atherothrombosis and the RA disease state, potentially explaining the increased risk of CV events in RA.  The use of TNF inhibitors was associated with a reduced risk of cardiovascular events in RA patients in one study, although it is not clear whether these effects are due to unique features of TNF inhibitors or are a consequence of non-specific suppression of inflammation.  Here, Dixon et al (Arthritis Rheum 2007; 56(9): 2905) explore these issues in analyses of participants enrolled in the British Society for Rheumatology Biologics Register (BSRBR).

Methods:

The BSRBR is a registry of patients initiated on biologic therapy in the UK with a parallel comparison cohort of non-TNF inhibitor treated patients enrolled.  Participants undergo a comprehensive baseline assessment and follow-up assessments by questionnaire (to participants and treating rheumatologists) every 6 months thereafter.  For these analyses, only RA patients were included.  Information on incident MIs was gathered from patient report and verified using hospital records and national mortality registries.
Rates of incident MI were compared between TNF-inhibitor treated and untreated groups, and within groups according to response to treatment.  Only first MIs and those occurring while undergoing active treatment were analyzed.

Results:

There were 10,755 RA subjects included in the analyses, of whom 8,659 were treated with TNF inhibitors.  Compared to TNF inhibitor treated subjects, DMARD treated subjects were older (mean age 60 vs 56 years at enrollment; p = 0.002) with a lower proportion of females (72% vs. 76%; p=0.01), lower DAS28 scores (mean 5.0 vs. 6.6; p <0.001), and lower HAQ scores (mean 1.5 vs. 2.1; p<0.001).  Among CV factors in comparison to TNF inhibitor treated subjects, DMARD treated subjects reported prior MI and angina nearly twice as frequently TNF inhibitor treated subjects, were significantly more likely to be receiving lipid lowering and anti-platelet medications, had a significantly lower proportion of corticosteroid treatment (19.3% vs. 43.7%; p <0.001), and were slightly less frequently treated with NSAIDs (62% vs. 66%; p = 0.001).
Median follow-up time was relatively short; 1.66 years in the TNF inhibitor cohort and 1.34 years in the DMARD cohort.  During follow-up, 63 verified MIs were reported in the TNF inhibitor cohort (4.8 per 1000 person-years) compared to 17 in the DMARD cohort (5.9 per 1000 person-years).  After adjusting for demographic and cardiovascular risk factors, there was no difference in the adjusted rate of MI between the TNF inhibitor and DMARD groups.

Among the TNF inhibitor treated subjects (n=7,515), 5,877 (78%) were classified as responders to therapy at 6 months.  The incidence rate of MI was 3.5 per 1000 person-years in 6-month responders (95% CI 2.5 – 4.9) vs. 9.4 per 1000 person-years in non-responders (95% CI 5.5-15.0).  After adjusting for demographic and cardiovascular risk factors, there was a 64% reduction in MI risk in 6-month TNF inhibitor responders compared to non-responders (95% CI 0.19 – 0.69).  In gender stratified analyses, reduced MI risk was seen in male and female responders vs. non-responders; however, the risk reduction was significantly lower in men only.

Conclusions:

TNF inhibitor use was not uniquely associated with a reduction in incident MI over the risk reduction observed with response to treatment.

Editorial Comment:

These data suggest that there may be no unique benefit to TNF inhibitor therapy for MI reduction in RA patients over and above their ability to normalize disease activity.  This is not discouraging, as it suggests that any DMARD regimen effective at managing disease activity in an individual patient will confer some degree of CV protective benefit.  The observation that response to TNF inhibitor therapy is associated with a reduction in MI risk is consistent with other reports demonstrating CV mortality reduction with methotrexate use.  Inflammation is general, and inflammatory cytokines like TNF-α in particular, are associated with destabilization of vulnerable atherosclerotic plaques; a potential link between inflammation in RA and MI.  However, the mechanisms of inflammation associated atherogenesis may be different from those of atherothrombosis, and thus, more study is needed to clarify the role of RA therapeutics on CV disease risk.

Jon Giles, MD

About Jon Giles, MD