April 26, 2011 — A new study from the Perioperative Ischemic Evaluation (POISE) trial suggests that monitoring non–cardiac-surgery patients for asymptomatic myocardial infarctions (MIs) in the first few days after surgery could dramatically reduce their short-term mortality risk [1].

The study shows that "consistently, all over the world, people are at substantial risk of suffering a heart attack after surgery, and if they do, they're at substantial risk to die or suffer a major event in the coming days, and we need to do a better job to detect them and manage them," primary author Dr PJ Devereaux (McMaster University, Hamilton, ON) told heartwire ."We want to make the broader cardiology world aware that this is a huge emerging epidemic that's going to confront cardiology, because there are 200 million adults having surgery every year in the world. . . . We need to get a lot more aggressive about monitoring for these events and recognizing that the majority of people won't have symptoms when they have these events."

Devereaux and colleagues followed 8351 patients at 190 centers in 23 countries with four cardiac biomarker assays for three days postsurgery. All of the patients were part of the original POISE trial, reported by heartwire ,which showed that beta blockers reduce the risk of MI but increase the risk of severe stroke and overall death in patients undergoing non–cardiac surgery (including orthopedic, cancer, and noncardiac vascular surgeries). Results of the new study by Devereaux et al are published in the April 19, 2011 issue of the Annals of Internal Medicine.

"Surgery is sort of the ultimate stress test. It does everything that is relevant to causing acute coronary syndrome. It's very proinflammatory, and it activates the sympathetic system, coagulation, and platelets. That's why we have this problem of people having myocardial infarction after surgery, [and yet until now] there's not that much research on the outcomes of heart attacks after surgery," Devereaux told heartwire. Patients and physicians are often unaware of an MI during this early postoperative period because most of the patients are on high doses of narcotics that "blunt the discomfort of the surgery but may mask ischemic symptoms," Devereaux said.

Within 30 days of randomization, 415 patients in the study (5.0%) showed evidence of a perioperative MI, defined as either autopsy findings of acute MI or an elevated level of a cardiac biomarker or enzyme assay plus ischemic symptoms, development of pathologic Q waves, ischemic changes on electrocardiography, coronary artery intervention, or cardiac imaging evidence of MI. Nearly three-quarters of the MIs happened within 48 hours of the surgery, but almost two-thirds of the MI patients did not did not experience ischemic symptoms. In fact, patients with a periprocedural MI without ischemic symptoms had a higher mortality rate (12.5%) than those who had symptoms (9.7%).

The short-term prognosis for patients who suffer periprocedural MIs is very poor, with 11.6% mortality at 30 days postprocedure compared with 2.2% for patients who did not suffer a periprocedural MI (p<0.001). Furthermore, Devereaux noted that a recent meta-analysis by his group found that people who suffer a periprocedural MI continue to be at higher risk for death than those who do not for at least a year after the surgery.

Nobody thinks twice about being incredibly assertive about managing an MI in the emergency room . . . [and yet] those MIs have a much better prognosis than these MIs, and we're ignoring these MIs for the most part.

Regression analysis of the data showed that relatively simple therapies could have prevented many of these deaths. In the study, patients on aspirin had about half the 30-day mortality risk as those not on aspirin, while statins reduced the 30-day mortality rate by about three-quarters. Only 64.8% of patients who suffered an MI in the trial were on aspirin, only 17.8% were receiving clopidogrel or ticlopidine, 52.0% were receiving a statin, and 55.4% were receiving an ACE inhibitor or angiotensin-receptor blocker.

"Patients expect us to look for things that are modifiable and change their risks of very serious events quickly, and perioperative MI is definitely in that category," Devereaux said. "Nobody thinks twice about being incredibly assertive about managing an MI in the emergency room, which is completely appropriate, but those MIs have a much better prognosis than these MIs, and we're ignoring these MIs for the most part."

Commenting on the study by Devereaux et al, Dr Adrian Banning (John Radcliffe Hospital, Oxford, UK) told heartwire ,"We are not optimizing medical therapy before surgery. There are existing guidelines and risk scores that are probably underused. Preoperative testing for ischemia and revascularization is probably overused in a minority of patients, leaving an occult majority without simple medical measures that are likely to be beneficial--including aspirin, statins, and good perioperative blood-pressure control."

More Research Needed to Clarify Who Is at Risk and How to Treat Them

Devereux's group is currently enrolling patients into the 40 000-patient prospective cohort VISION study, which is intended to define the optimal approach for predicting major perioperative vascular events, the extent to which troponin measurement after surgery can identify asymptomatic MIs, and these patients' risk of vascular-related death within one year.

The first 20 000 patients in the study have been monitored with "fourth-generation" troponin assays, and the next 20 000 will be monitored with higher-sensitivity troponin assays. Commenting on the research, Dr Stephen Ellis (Cleveland Clinic, OH) pointed out that with the advent of highly sensitive troponin tests, more research will be needed to define what degree of troponin change is clinically important. "I'm sure there's some level of troponin where you see a bump that doesn't mean anything." For example, Banning and colleagues recently completed a study that suggests the current standard troponin cutoff used to detect an MI has been arbitrarily set too low and leads to an overestimate of the number of MIs.

Dr John French (University of New South Wales, Australia) added that future research should also try to risk-stratify these patients by collecting both pre- and postprocedural troponin levels. Elevated preprocedural troponin may also be a risk marker, he told heartwire .

Devereaux hopes there will also soon be a large national trial to evaluate the best way to manage non–cardiac-surgery patients in the vulnerable perioperative period. Ellis agreed that "we don't really understand the benefits of some of the medical treatments that we have in our armamentarium in this patient population. . . . There may be some other treatments that are less utilized at present that could cut down on the incidence of perioperative infarction."

Banning agreed that further research is needed to understand how to prevent these perioperative MIs, not merely detect them. "Troponin measurement postoperatively can help define a risk group with adverse outcome, [but] it is uncertain that we can influence that adverse outcome once the event has happened in those patients already on optimal medical therapy," he said. "There will be a group identified by routine troponin testing where this event is the first declaration of occult coronary disease, and perhaps this group potentially has the most to gain."

Although the best approach to managing these patients has yet to be clearly defined, Devereaux emphasized that "in the short term, there's a lot of intuitive things that we can do better that will likely improve the outcomes, and there's lots of reasons to be optimistic that, even if we just start monitoring them, we can improve the outcomes." Devereaux recommends that physicians caring for a surgery patient order a troponin test sometime between six and 12 hours after the surgery and then repeat tests for the first three days after surgery.

His institution has made perioperative MI prevention a priority for its cardiologists. "We've changed cardiology from regular cardiology to cardiology and perioperative vascular medicine," he said. All surgery patients' cardiac biomarkers are monitored, and the patients are triaged to the coronary care unit or less-intensive care based on their MI risk. He expects his group will be able to present data on the impact of this approach within the next year.

This study was supported by the Canadian Institutes of Health Research, the Commonwealth Government of Australia’s National Health and Medical Research Council, the Instituto de Salud Carlos III in Spain, the British Heart Foundation, and AstraZeneca.

References

1. Devereaux PG, Xavier D, Pogue J. Characteristics and short-term prognosis of perioperative myocardial infarction in patients undergoing noncardiac surgery. Ann Intern Med 2011; 154:523-528. Abstract

Clinical Context

Clinical Implications

• The perioperative MI rate after noncardiac surgery is 5%, and more than half of patients die within 48 hours after surgery.
• The risks for 30-day mortality and cardiovascular morbidity are increased with perioperative MI, regardless of whether patients are symptomatic.

CME Test