Antiplatelet Therapy Following Myocardial Infarction in Patients With Diabetes

Abstract

DESPITE ADVANCES IN THE CARE OF PATIENTS WITH diabetes with myocardial infarction (MI), recurrent ischemic events remain common. Although many explanations exist for this excess risk among patients with vs those without diabetes, one mechanism appears to be heightened platelet reactivity. It seems logical that antiplatelet therapy would be of benefit in diabetes or that more potent antiplatelet therapy would be required in diabetes. In this issueof JAMA,Anderssonet al present acareful analysis of patients who have survived 30 days after discharge for a first-time MI in the Danish nationwide registry. The authors examine the benefit of the oral antiplatelet agent clopidogrel in 7247 patients with vs 51 604 without medication-treated diabetes. Consistent with other data, patients with diabetes had a significantly higher rate of mortality and reinfarction (25% vs 15%) over the follow-up period of up to 1 year. Notably, the apparent beneficial effect of clopidogrel vs no clopidogrel was lower in patients with than in those without diabetes: adjusted hazard ratios of 0.89 vs 0.75 for all-cause mortality, 0.93 vs 0.77 for cardiovascular mortality, and 1.00 vs 0.91 for the composite of death or reinfarction. This analysis was not randomized, and the usual limitations of observational studies apply. However, it is a bit puzzling that the association of clopidogrel use with clinical outcomes seems stronger for mortality than for MI. The degree of mortality benefit associated with clopidogrel reported in this study also exceeds that reported in randomized clinical trials. Among patients undergoing percutaneous coronary intervention (PCI), which was expectedly the majority of these MI patients, there was significant and similar benefit associated with clopidogrel use in each of the end points for patients with or without diabetes, which, to an extent, undercuts the authors’ hypothesis of differential benefit. Of note, the authors did not find a differential degree of benefit associated with aspirin use in patients with diabetes vs those without, so their results seem specific to clopidogrel. However, why any patient who had an MI, especially those treated with PCI, would be receiving neither clopidogrel nor aspirin is unclear. It is possible that these were patients who experienced bleeding or were nonadherent due to any of a large number of reasons, and possible that these types of confounders may have contributed to the higher observed mortality. Limitations of observational data notwithstanding, are the results plausible? Is there something about patients with diabetes that makes them less likely to respond to standard antiplatelet therapy? A number of different data sets besides this one suggest that the answer is likely yes. Platelet reactivity is known to be higher in patients with coronary artery disease who have than in those who do not have diabetes, and several different platelet signaling pathways appear to be up-regulated. Prasugrel and ticagrelor are potential options for patients who require dual antiplatelet therapy after MI, such as those receiving stents. The pivotal trials that preceded the regulatory approval of these antiplatelet agents found significant benefit in the overall trial populations tested. Indeed, the European cardiology acute coronary syndrome guidelines give preference to these agents over clopidogrel, although the US cardiology guidelines leave the choice to the physician based on patient characteristics. The subgroups of patients with diabetes in TRITON-TIMI 38 and PLATO showed larger absolute benefits with the newer agents than did the subgroups without diabetes. Other than concerns about adverse effects, such as bleeding and higher cost, it seems that use of one 1 of these newer antiplatelet agents may be preferable to clopidogrel in patients with an MI and an approved indication for dual antiplatelet therapy, particularly those patients with diabetes. However, the CURE study also found larger absolute benefit of clopidogrel than placebo following acute coronary syndrome among patients with vs those without diabetes. The totality of data from the randomized trials of antiplatelet therapy does not show clear evidence of a significant statistical interaction between diabetic status and benefit but do support greater absolute risk reductions in the cohorts of patients with diabetes. What is uncertain is whether the greater absolute risk reductions observed in patients with diabetes are due to preferentially greater effect in patients with diabetes or because these patients have higher event rates and therefore derive more