A Comprehensive Update on Aspirin Management During Noncardiac Surgery

Neal S. Gerstein, MD, FASE; Cory L. Albrechtsen, BS; Nestor Mercado, MD, PhD; Joaquin E. Cigarroa, MD; Peter M. Schulman, MD


Anesth Analg. 2020;131(4):1111-1123. 

In This Article

Aspirin During the Perioperative Period

How aspirin should be optimally managed in patients undergoing noncardiac surgery is a vexing problem with significant clinical implications. Given the nonperioperative data demonstrating the beneficial role of aspirin in secondary prevention along with the adverse impact of a potential aspirin withdrawal syndrome and the proinflammatory hypercoagulable perioperative period—it is reasonable to assume that stopping aspirin might increase perioperative CV risk. Conversely, since perioperative adverse CV events are predominantly the result of supply-demand mismatch rather than acute thrombosis, it is plausible that continuing aspirin may confer little CV protection and may alternatively increase CV risk if excessive bleeding ensues.

In 2010, Oscarsson et al[18] published the first of several trials designed to determine the impact of aspirin cessation versus continuation in those at high risk for perioperative MACE. Two-hundred twenty patients were randomized to receive 75 mg aspirin or placebo starting 7 days preoperatively and continued 3 days postoperatively. Patients in the placebo group previously taking aspirin had it restarted on postoperative day 3. Included patients had to undergo elective high- or intermediate-risk noncardiac surgery and have at least one of the following cardiac risk factors: CAD, congestive heart failure, renal impairment, stroke, or insulin-dependent diabetes mellitus. Exclusion criteria were unstable CAD, decompensated heart failure, shock, allergy to aspirin, age <18 years, prior GI or intracranial bleeding, use of warfarin/clopidogrel/methotrexate, or presence of a coronary stent. The study was terminated early due to the contemporaneous release of ACC/AHA guidelines during the study recruitment period recommending perioperative aspirin continuation in high-risk patients along with a study amendment to exclude patients with coronary stents, both of which negatively affected recruitment ability.[40] Though the investigators reported that the primary end point (postoperative myocardial injury) was not significantly different between groups (3.7% aspirin versus 9.0% placebo; P = .10), aspirin continuation resulted in a 7.2% absolute risk reduction (95% CI, 1.3–13) and 80% relative risk reduction (95% CI, 9.2–95) for the trial's secondary end point of postoperative MACE. No significant differences in bleeding complications were observed between groups, although the trial was not powered for this outcome.

A second trial, Impact of Preoperative Maintenance or Interruption of Aspirin on Thrombotic and Bleeding Events After Elective Noncardiac Surgery (STRATAGEM), which was multicenter, randomized, and placebo controlled, enrolled 291 patients already taking an antiplatelet agent (predominantly aspirin) for secondary prevention and scheduled to undergo intermediate- or high-risk noncardiac surgery. The placebo group (n = 146) had their aspirin or alternative antiplatelet agent held 10 days preoperatively, and the treatment group (n = 145) had their antiplatelet agent continued or their regimen switched to 75 mg of aspirin starting 10 days preoperatively. All patients had their initial antiplatelet therapy restarted as soon as it was deemed safe postoperatively. Notable exclusion criteria included patients undergoing carotid endarterectomy (CEA), recent MACE, or presence of a drug-eluting stent (DES; but not bare metal stent [BMS]). Although the study was terminated early due to recruitment difficulties and thus underpowered, no difference in the primary outcome was found, which consisted of a composite score of major thrombotic and bleeding events within 30 days of surgery (0.67 aspirin versus 0.65 placebo group; P = .94).[41]

Due to the relatively small size of the 2 aforementioned studies and conflicting results, a larger trial was subsequently conducted in an attempt to answer the aspirin cessation versus continuation question more definitively. The 2014 PeriOperative ISchemia Evaluation-2 (POISE-2) trial was a randomized double-blind, multinational, multicenter controlled trial that evaluated the outcome of continuing or stopping low-dose (200 mg initially followed by 100 mg/d for 30 days) aspirin in 10,010 "at-risk" patients undergoing noncardiac surgery.[16] At-risk patients were those over 45 years old with 1 of 5 inclusion criteria: CAD, PAD, stroke, undergoing major vascular surgery (except CEA), or presence of ≥3 of 9 risk factors (≥70 years, undergoing major surgery, heart failure, TIA, diabetes, hypertension, serum creatinine >2.0 mg/dL, smoking within 2 years of surgery, emergent/urgent surgery). Patients were stratified into 2 initial groups based on whether they were taking aspirin before the study (continuation strata) or aspirin-naive (initiation strata). Patients in these 2 stratum were then randomized to receive either aspirin or placebo, resulting in 4 final groups: (1) no prior aspirin use followed by perioperative aspirin, (2) no prior aspirin use followed by placebo only, (3) prior aspirin use followed by perioperative aspirin, and (4) prior aspirin use followed by 7 days of placebo and then returning to aspirin for the final 23 days of follow-up. Important exclusion criteria included BMS within 6 weeks of randomization or DES within 1 year of randomization. No difference was found among groups or strata in the primary outcome of a composite of death or nonfatal MI within 30 days of surgery. Major bleeding, mostly at the surgical site, was greater in the aspirin group (4.6% vs 3.8%; P = .04); however, no differences were reported in either "clinically important hypotension" or "life-threatening" bleeding between groups. Because of these findings, the authors concluded that perioperative aspirin continuation does not confer a benefit in terms of MACE and may contribute to excess blood loss.

However, POISE-2 has been criticized for a number of shortcomings that temper its interpretations. Approximately one third of all study patients assigned to the aspirin group had a definitive primary or secondary indication for aspirin, and it is also unclear if patients in the continuation stratum were using aspirin for an appropriate ACC/AHA guideline-based indication. In addition, vascular surgery, which has the highest associated risk of MACE among all types of noncardiac surgery, accounted for only 4.9% of high-risk procedures. Moreover, patients undergoing CEA were excluded, and only those with a coronary stent were <10% of recruited subjects, both further affecting the generalizability of the results. Hence, it is likely that a majority of the patients in the study were at low rather than high baseline risk for perioperative MACE.[42] The results were also confounded by the postoperative administration of antithrombotic agents. By postoperative day 3, 65% of patients received prophylactic anticoagulation, 4.0%–4.5% of patients in both groups received therapeutic anticoagulation, and 1.2% of patients in both groups received a P2Y12 inhibitor.[42] Seven safety outcomes were separately evaluated in POISE-2, of which only major bleeding was significant, though no differences in life-threatening bleeding or significant hypotension were reported.

In 2014, based on POISE-2 and the other evidence available at the time, the ACC/AHA published their latest "Guideline on Perioperative Cardiovascular Evaluation and Management of Patients Undergoing Noncardiac Surgery," which concluded that patients without a coronary stent should not have aspirin initiated perioperatively.[43] However, it suggested that continuing aspirin is reasonable for patients already receiving it for secondary prevention of CV disease when major noncardiac surgery is planned and the perceived risk of a MACE is greater than the risk of bleeding. In part due to remaining uncertainty about the effect of perioperative aspirin withdrawal on higher-risk patients, a 2018 subgroup analysis of POISE-2 was conducted to specifically determine how aspirin should be managed in patients undergoing vascular surgery.[42,44] Of the 10,010 patients in POISE-2, 603 underwent vascular surgery, including 319 in the continuation stratum and 284 in the initiation stratum. No interaction with the primary outcome (ie, composite of death or MI at 30 days) was found for type of surgery or aspirin stratum and no difference between patients undergoing surgery for aneurysm or occlusive disease. There was also no subgroup interaction for major or life-threatening bleeding. Thus, the authors concluded that perioperative aspirin withdrawal did not increase CV or vascular occlusive complications and that it is therefore not necessary to start aspirin preoperatively in patients undergoing vascular surgery.[44] They also concluded that continuing aspirin in patients already receiving it might be reasonable, since its continuation did not increase the risk of major bleeding. However, they cautioned that since even clinically unimportant bleeding may be associated with MI, an argument could be made for stopping rather than continuing aspirin in patients undergoing vascular surgery but acknowledged that larger studies would be necessary to more definitively answer this question. Finally, the authors noted that patients undergoing CEA were entirely excluded from POISE-2. Since perioperative aspirin proved beneficial for CEA surgery in both a prior RCT (although there was no placebo group) and large observational trial, they recommended continuing aspirin in patients undergoing CEA.[45,46]

Patients With Coronary Stents Undergoing Noncardiac Surgery. Retrospective and observational data demonstrate that perioperative ischemic events and mortality occur more frequently when noncardiac surgery is performed soon after coronary stenting.[47] Patients undergoing percutaneous coronary intervention (PCI) revascularization following an ACS and who undergo noncardiac surgery shortly thereafter are at highest risk. For example, in a retrospective cohort study by Cruden et al,[48] the risk of perioperative death and ischemic cardiac events occurred more frequently when noncardiac surgery was performed within 42 days of stent implantation (42.4% vs 12.8% beyond 42 days; P < .001), especially in patients revascularized after an ACS (65% vs 32%; P = .037). In 2013, another retrospective cohort study was performed by Hawn et al[49] and found that time between PCI and surgery was associated with MACE (<6 weeks, 11.6%; 6 weeks to <6 months, 6.4%; 6–12 months, 4.2%; >12–24 months, 3.5%; P < .001). After adjustment, the 3 factors most strongly associated with MACE were nonelective surgical admission (adjusted odds ratio [AOR], 4.77; 95% CI, 4.07–5.59), history of MI in the 6 months preceding surgery (AOR, 2.63; 95% CI, 2.32–2.98), and a revised cardiac risk index >2 (AOR, 2.13; 95% CI, 1.85–2.44).[49] In 2018, a substudy of POISE-2 subjects was published evaluating 470 POISE-2 subjects with prior PCI to specifically evaluate the risk of stopping or continuing aspirin in these patients.[50] This analysis found that perioperative aspirin as compared with placebo reduced the risk of the primary outcome (composite of death and nonfatal MI). This beneficial effect, which differed from the effect found in patients without prior PCI, was driven by a reduction in MI (adjusted risk reduction = 5.9%, 95% CI, 1.0–10.8).[50] Finally, POISE-2 excluded patients within 6 weeks of BMS implantation or within 1 year of DES implantation. Because the POISE-2 investigators did not anticipate enrolling patients with a history of PCI who were not otherwise excluded, a PCI subgroup analysis was not initially planned or conducted. Thus, ACC/AHA recommendations for the perioperative management of patients with coronary stents did not include data from POISE-2 and instead were largely based on the preceding observational studies. In 2020, Sessler et al[51] reexamined POISE-2 data for outcomes at 1 year after enrollment. The authors reported similar findings with regard to aspirin's lack of effect on the primary outcome (composite of death or nonfatal MI) in the original entire cohort. However, in the subgroup of 470 patients with prior PCI, similar to the 30-day analysis,[50] perioperative aspirin continuation had a net benefit on the primary outcome that persisted at 1 year (prior PCI: HR = 0.58, 95% CI, 0.35–0.95 versus no history of PCI: HR = 1.03, 95% CI, 0.91–1.16; P for interaction = .033).[51]

The ACC/AHA concluded that the risk of stent thrombosis is highest during the perioperative period in the first 4–6 weeks after stent implantation, and the risk of DES thrombosis is likely elevated for patients undergoing noncardiac surgery for about 6 months after DES implantation. Consequently, the following recommendations were made: (1) patients undergoing noncardiac surgery during the first 4–6 weeks after BMS or DES implant should have dual antiplatelet therapy (DAPT) continued unless the risk of bleeding outweighs the risk of stent thrombosis, and (2) patients with coronary stents who must have a P2Y12 inhibitor stopped for noncardiac surgery should have aspirin continued whenever possible. The current generation of DES stents has a higher safety profile than BMS and first-generation DES. The risk of stent thrombosis perioperatively is likely lower with the current-generation stents. Certain patient and stent characteristics are associated with an increased risk of in-stent thrombosis (Table 5). These variables need to part of the decision-making process when considering stopping or continuing perioperative antiplatelet agents.

Finally, a 2018 Cochrane meta-analysis examined continuation versus cessation of antiplatelet agents (aspirin, clopidogrel, or both) in elective noncardiac surgery (primarily abdominal, orthopedic, urologic, and gynecologic) in patients with at least 1 CV disease risk factor.[52] The review included 5 RCTs (3 of which were ongoing studies) with a total of 666 adults and compared those who continued their single or dual antiplatelet regimen throughout the perioperative interval with those who discontinued their regimen at least 5 days preoperatively. Statistically significant distinctions were not found in any analyzed outcome, but the reported 95% CIs were also too wide to permit conclusions of equivalence. The authors did not find mortality differences between groups at 30 days (risk ratio = 1.21, 95% CI, 0.34–4.27) or at 6 months (risk ratio = 1.21, 95% CI, 0.34–4.27). There were also no significant differences in the incidence of a variety of ischemic events (peripheral ischemia, CVA, or MI) by 30 days postoperatively (risk ratio = 0.67, 95% CI, 0.25–1.77). Regarding blood loss, antiplatelet continuation versus cessation did not significantly impact blood loss requiring transfusion (risk ratio = 1.37, 95% CI, 0.83–2.26) or blood loss requiring additional surgery (risk ratio = 1.54, 95% CI, 0.31–7.58). With regard to mortality, bleeding necessitating surgery, and ischemic events, the authors described their findings as "low certainty" (true effect may be substantially different from the estimate of the effect) and for bleeding requiring transfusion, the findings are "moderate certainty" (true effect is likely close to the estimate of the effect but with the possibility that it is substantially different). Overall, these findings should be interpreted cautiously due to significantly limited evidence from a very small number of studies with few events leading to wide effect-estimate CIs.[52]