Harmful Effects of NSAIDs among Patients with Hypertension and Coronary Artery Disease: Bavry AA, Khaliqu A, Gong Y, et al. Am J Med 2011;124:614–20.

Abstract

Although selective cyclooxygenase-2 inhibitors have been shown to increase the risk of myocardial infarction (MI) and aspirin has been shown to be cardioprotective, the effect of chronic use of non-steroidal anti-inflammatory drugs (NSAIDs) on adverse cardiovascular outcomes and blood pressure has not been well established. This observational study was a post hoc analysis from the International Verapamil Trandolapril Study (INVEST), an international randomized trial comparing the effects of a calcium antagonist (verapamil SR) vs. a beta-blocker (atenolol) on hypertension in stable coronary artery disease, that investigated the association between chronic NSAID use, blood pressure, and adverse cardiovascular outcomes, defined as all-cause mortality, non-fatal MI, or non-fatal stroke, among patients with hypertension and coronary artery disease. Patients had scheduled follow-up visits every 6 weeks for the first 6 months, then biannually for a mean of 2.7 years and were asked every visit about current NSAID use. Patients were divided into two groups: chronic NSAID users if they reported use at every visit and non-chronic NSAID users (occasional or never used). The study found 882 self-reported chronic NSAID users and 21,694 non-chronic users (including 14,408 never NSAID users and 7286 intermittent users). The primary outcomes (all-cause mortality, stroke, or MI) occurred at a rate of 4.4 events per 100 patient-years for chronic NSAID users and 3.7 events per 100 patient-years for non-chronic users, with adjusted hazard ratio (HR) of 1.47 (95% confidence interval [CI] 1.19–1.82, p=0.003). Propensity-matched chronic NSAID users, when compared to non-chronic NSAID users, still demonstrated higher incidences of the primary outcomes, with an adjusted HR of 1.60 (95% CI 1.18–2.17, p<0.0023). The study found that this 47% increase in the first occurrence of the primary outcomes in the chronic NSAID group was due to an increase in cardiovascular mortality in the chronic users when compared to the non-chronic users, adjusted HR 2.26 (95% CI 1.70–3.01, p<0.0001). The study also found an overall 90% increase in all-cause mortality (adjusted HR 1.89, 95% CI 1.53–2.35, p≤0.0001), specifically a 66% increase in total MI (adjusted HR 1.66, 95% CI 1.21–2.28, p=0.0017), but no significant difference in total stroke when comparing chronic users to non-chronic users. The study did find slightly but statistically significant lower systolic and diastolic blood pressures in the chronic NSAID users group. The researchers recognized that without information on specific NSAID types and dosages, the findings represent a class effect, and that placebo-controlled trials of specific agents are needed to determine the exact safety profile of individual agents. They concluded that despite lower mean blood pressures, the net effect from the multivariate modeling and propensity matching was an overall increase in adverse cardiovascular events, and they support the American Geriatric Society Panel's recommendation of acetaminophen as the first-line agent in chronic pain in the elderly until randomized trial data for specific agents are available.