Stopping Versus Continuing Aspirin Before Coronary Artery Surgery
Myles PS, Smith JA, Forbes A, et al. N Engl J Med. 2016;374(8):728–737.
Reviewers: Muhammad Raza1; Ahmed Awad, MD2
- Medical Student, Cooper Medical School of Rowan University, Camden NJ
- Assistant Professor of Anesthesiology, Cooper Medical School of Rowan University, Camden, NJ
A lack of clinical research and trials on the safety of preoperative aspirin in coronary artery surgery has led to different guidelines on the matter. The debate lies within the question of aspirin benefiting coronary-graft flow and minimizing graft thrombosis vs the risk of surgical bleeding. Historically, aspirin was discontinued 5–7 days before surgery and restarted 24 hours afterward. However, there is evidence suggesting preoperative or immediate postoperative administration of aspirin could lower mortality and/or the rate of severe complications.
Study Design: Double-blind and multicenter study of 2127 patients on the day of coronary artery surgery were randomly assigned in a 2x2 factorial design to preoperatively receive 100 mg aspirin or placebo and tranexamic acid or placebo, all with or without anxiolytic medication. Eligibility included adult patients at elevated risk for severe perioperative complications due to age or coexisting conditions who were having coronary artery surgery (either on-pump or off-pump) with or without other surgeries such as valvular placement. The attending anesthesiologists, surgical team, postoperative interviewers, and end-point adjudicators were blinded concerning the assignment of groups. There also was a preliminary safety analysis after enrollment of 824 patients to determine whether there was a significant risk of bleeding.
Exclusion Criteria: Patients taking aspirin regularly before the study or who took aspirin up to 4 days before coronary artery surgery were excluded, as were those who did not consent or had poor language comprehension.
Interventions and Outcomes: The key interventions of note were computer-generated randomization of patients, discontinuation of warfarin and clopidogrel at least 7 days prior to surgery, and 1:1 administration of aspirin or placebo 1–2 hours before surgery. The primary outcomes measured were death and thrombotic events such as nonfatal myocardial infarction, stroke, pulmonary embolism, renal failure, or bowel infarction in the 30-day postoperative window. The primary outcomes had set definitions that needed to be met to have a complication/event classified accordingly. The secondary outcomes measured were death, nonfatal myocardial infarction, major hemorrhage, cardiac tamponade, and transfusion requirement. The secondary outcomes had set definitions that needed to be met to have a complication/event classified accordingly.
Although 5784 patients were eligible (at 19 centers from March 2006 to January 2013 in five countries), only 2100 were enrolled. The aspirin group had 1047 patients randomly assigned to it, whereas 1053 were randomly assigned to the placebo group.
Primary Outcomes: Within 30 days following surgery, 202 patients (19.3%) died or suffered from a thrombotic event in the aspirin group while 215 patients (20.4%) experienced primary outcomes in the placebo group. Relative risk with aspirin was 0.94; 95% CI, 0.80–1.12; P=.55. Ignoring renal failure (RF), 172 patients (16.4%) died or suffered from a thrombotic event other than RF in the aspirin group while 192 patients (18.3%) experienced primary outcomes other than RF in the placebo group. Relative risk with aspirin was 0.90; 95% CI, 0.75–1.09; P=.30.
Secondary Outcomes: Within 30 days following surgery, 144 patients (13.8%) had a detectable myocardial infarction in the aspirin group while 166 patients (15.8%) had a detectable myocardial infarction in the placebo group. Relative risk: 0.87; 95% confidence interval (CI), 0.71–1.07; P=.20. Other notable secondary outcomes were cardiac tamponade (1.1% in aspirin and 0.4% in placebo [P=.08]) and major hemorrhage (1.8% in aspirin and 2.1% in placebo [P=.75]). Both groups were similar in the incidence of death, pulmonary embolism, renal failure, stroke, and bowel infarction.
Predetermined subgroups were age, sex, diabetic status, history of myocardial infarction, existence of unstable angina or lack thereof, EuroSCORE operative risk, left ventricular function, risk of intraoperative bleed, on-pump or off-pump surgery, and cross-clamp time. Twelve men died in the aspirin group as opposed to three men in the placebo group (P=.046). The other subgroups did not demonstrate significant findings.
The authors found no significant difference between aspirin vs placebo before surgery in increasing or mitigating the risk of death, thrombotic complications, reoperative necessity, surgical bleeding, or transfusion requirement. They did find an increased rate of myocardial infarction, which the authors attributed to a very lenient definition of myocardial infarction (accepting troponin elevation despite absence of Q wave detection).
Although the potential benefits of aspirin before surgery were briefly discussed in the paper, the research only focused on identifying the harms of aspirin before surgery. What is more relevant and unique to the study, however, is the fact that a low dose of aspirin was used. The authors cited a recent meta-analysis of 13 randomized trials with 2399 patients that saw a lower rate of myocardial infarction but a higher rate of adverse bleeding. Ultimately, if the definition of myocardial infarction adheres to a more standard guideline, and the study were repeated at the same dosage, perhaps we could begin to see evidence supporting the cardioprotective effects of perioperative aspirin.