Abstract 1340: The Effect of Rheumatic Disease Treatment on the Risk of Myocardial Infarction: Increased Risk from Rofecoxib, Valdecoxib and Celecoxib; Decreased Risk from Anti-TNF Therapy
Authors:
F. Wolfe, K. Michaud
Background
RA patients are at an increased risk for myocardial infarction (MI) and other clinical cardiovascular events. This risk may be secondary to the RA disease state (i.e. chronic inflammation, decreased physical activity), shared risk factors, RA therapeutics, or a combination of these. RA therapeutics have the potential to exert both positive and negative effects on cardiovascular risk, as presented here by Wolfe et al from data from the National Data Bank for Rheumatic Diseases.
Methods
Patients were participants in the National Data Bank for Rheumatic Diseases, a longitudinal community-based registry of RA and non-RA patients receiving a variety of immunomodulating therapies, including biologics. Data are derived from questionnaires periodically mailed to participants and treating rheumatologists. For this investigation, the association of rheumatic pharmacotherapies (TNF inhibitors, non-selective and COX-2 selective NSAIDs, methotrexate, and acetaminophen) with risk of incident MI, confirmed by medical records, was explored. Analyses were adjusted for baseline HAQ, prednisone use, hypertension, diabetes, and lifetime smoking exposure.
Results
Participants enrolled between 1998 and 2006 were included, encompassing 20,148 RA and 5,195 non-RA patients with non-inflammatory rheumatic conditions (i.e. osteoarthritis) with 95,670 patient-years of follow-up. Incident MI was associated with baseline HAQ, prednisone, use, hypertension, diabetes, prior MI, and smoking in all analyses. Adjusting for these factors, rofecoxib and valdecoxib use was associated with a significantly increased risk of MI in RA and non-RA patients, with the risk of incident MI increased 2.6-fold in rofecoxib users compared to non-users (95% CI 1.9 – 4.2) and 2.3-fold in valdecoxib users compared to non-users (95% CI 1.3 – 4.2). Use of celecoxib, non-selective NSAIDs, and acetaminophen did not significantly increase the risk of incident MI in users compared to non-users in adjusted analyses. However, compared to non-users of NSAIDs with no cardiovascular risk factors, the risk of incident MI was 7.6-fold higher for rofecoxib users (95% CI 4.5 – 12.9) and 3.1-fold higher for celecoxib users (95% CI 2.0 – 4.8).
TNF inhibitors were used in 56% of RA participants and their use was associated with a significant 30% reduction in risk of incident MI, regardless of agent (etanercept, infliximab, or adalimumab) after adjustment. Methotrexate use was not associated with a reduced risk of incident MI.
Conclusions
COX-2 inhibitors and prednisone increase the risk of incident MI, while TNF inhibitors decrease the risk, in RA patients.
Editorial Comment
These data support other findings that suggest that TNF inhibition may help to protect against MI risk. Whether this mechanism is achieved by a direct effect on stabilizing vulnerable plaques or indirectly, via decreasing corticosteroid use or improving physical function and cardiac fitness, is not known. As with all observational studies, selection bias may account for some of the apparent protective benefit, as TNF inhibitors are often not prescribed to patients with many non-RA comorbidities (and thus a higher risk of MI), making TNF inhibitor use itself appear protective when it may not be. Selection cannot be adjusted for in this type of analysis, as the reasons for using, or not using, therapies in patients are often complex, not explicit, and difficult to quantify.
The finding that COX-2 inhibitors, particularly rofecoxib, increase the risk of MI is not surprising given the events of the last few years. However, the finding that celecoxib use was associated with an increased risk compared to the lowest risk group (those not receiving any COX-2 inhibitor with no cardiovascular risk factors) suggests that all COX-2 inhibitors should be used with care, particularly in patients with other risk factors. This is consistent with the current FDA mandated “black box” warning contained in the prescribing information for celecoxib.
