Optimal Timing of Coronary Artery Bypass Graft Surgery After Acute Myocardial Infarction
Background To assess optimal timing for coronary artery bypass graft surgery (CABG) after an acute myocardial infarction (AMI), all patients undergoing CABG without associated procedures at our institution from January 1, 1991, to July 30, 1992, were reviewed. Patients were divided into three groups based on time from infarct to revascularization. The control group consisted of patients operated on for angina refractory to medical management. Relative risks (incident infarction group divided by incident control group) were established for need of vasopressors, new balloon to separate from bypass, perioperative myocardial infarction, and hospital mortality.
Methods and Results One hundred sixteen patients underwent CABG within 6 weeks of infarction. In the experimental group, 58 patients underwent CABG for non–Q-wave infarction, and 58 patients underwent CABG for Q-wave infarction. In the control group, 255 patients underwent surgery for angina without infarction. Patients were analyzed by group relative to the time between infarction and CABG. Patients were analyzed between infarction and CABG and assigned to one of three groups. Group 1 patients were revascularized within 48 hours; group 2, between 3 and 5 days; and group 3, after 5 days. Significance was determined by Fisher’s exact or Mantel-Haenszel χ2 test where appropriate. Multivariate analysis was performed on statistics that were significant. All patients within all groups after Q-wave or non–Q-wave myocardial infarction had a significantly higher risk of needing an intra-aortic balloon pump and vasopressors to be weaned from bypass and a greater incidence of perioperative MI compared with control patients. Surgical mortality is highest immediately after Q-wave infarctions.
Conclusions Patients with non–Q-wave infarction may undergo CABG relatively safely at any time. Acceptable timing for CABG after Q-wave infarction is after 48 hours.
The present study was initiated to provide data on optimal timing of early surgical revascularization for symptomatic patients after acute myocardial infarction (AMI). To provide a rational basis for comparison, the study group consisted of patients with acute Q-wave and non–Q-wave AMI who were compared with a control group consisting of patients without AMI who underwent elective or urgent coronary artery bypass graft surgery (CABG) at our institution during the same time period. Comparisons included establishing the relative risk, in experimental and control groups, for sustaining hospital mortality, requirement for intra-aortic balloon pump to wean from cardiopulmonary bypass or to sustain cardiac output in the postoperative period, perioperative myocardial infarction, postrevascularization requirement for vasopressors, and length of hospitalization. The same criteria for patient selection and technique for coronary revascularization, myocardial protection, and intraoperative and postoperative management were used for all patients during the study period. To provide a uniform database, all patients undergoing concomitant procedures such as valve replacement, aneurysmectomy, or electrophysiological surgery were excluded from the study. With these exceptions, all patients were operated on sequentially by the same group of surgeons. Three study groups were analyzed relative to the time between infarction and CABG to provide data for forming clinical and administrative decisions on the management of patients who remain symptomatic after AMI.
Clinical and administrative implications for early revascularization are increasingly important considering current health policies and trends emphasizing shorter hospitalizations with total quality management.
Eight hundred seventy-seven patients underwent CABG with or without other procedures on the University Service at our institution between January 1, 1991, and July 30, 1993. Of these, 368 patients underwent isolated CABG and serve as the basis for this report. One hundred sixteen consecutive patients underwent CABG within 6 weeks of AMI, and 252 control patients, without AMI, underwent elective or semielective CABG during the same time period. The mean age, sex, number of vessels involved, number of vessels grafted, and comorbid factors such as diabetes mellitus, cerebral vascular disease, peripheral vascular disease, and so on showed no statistically significant difference between experimental and control groups (Table 1⇓).
Patients with coronary artery disease treated for stable angina, progressive angina, unstable angina, postinfarction angina, angina at rest, or AMI and patients with three-vessel disease or left main stem disease without symptoms were included. Patients undergoing CABG for refractory cardiogenic shock were also included, although a low ejection fraction was not necessarily considered an indication for early operation. To assess the effect of surgery on naturally occurring disease, patients operated on for iatrogenically produced AMI after angioplasty failure were excluded. Also, CABG combined with other cardiac procedures were excluded. Most patients with AMI who had no contraindication and presented within 6 hours of pain received thrombolytic therapy as well as intravenous nitrates, β-blockers and antiarrhythmic agents if indicated. Ten of the 116 AMI patients underwent unsuccessful attempts at relieving the obstruction with balloon angioplasty. A trial with the intra-aortic balloon pump usually preceded CABG in patients who failed medical management. The decision as to timing of surgery for an AMI patient was based on the severity of clinical symptoms.
Q-wave infarction was defined as ST-segment changes that progressed to new Q waves in addition to a creatine phosphokinase (CPK)-MB isoenzyme elevation of more than 10 IU/L. Non–Q-wave infarction was defined as ST-segment and T-wave abnormalities that did not progress to pathological Q waves but showed abnormal elevation of CPK-MB isoenzyme of more than 10 IU/L. Perioperative infarction was defined as a CPK-MB fraction of more than 70 IU/L or new Q waves on ECG from the time the patient was brought to the operating room until the time of discharge or death. Inotropic support was defined as the need for the use of vasopressors to wean from cardiopulmonary bypass or for treating low cardiac output. A new intra-aortic balloon was counted as an intra-aortic balloon placed in the operating room to wean a patient from cardiopulmonary bypass.
Complete revascularization was the goal for all patients. Myocardial protection was achieved by moderated systemic hypothermia (28°C) and blood or crystaloid cardioplegia at 4°C with St Thomas’ solution given antegrade through the ascending aorta at 20-minute intervals. A new intra-aortic balloon pump was placed only for patients who could not be weaned from bypass. Within the 48-hour group, one patient did not receive an intra-aortic balloon pump before surgery because of occlusive peripheral vascular disease. Inotropic support was given for a cardiac index of less than 2.1 L · min−1.
Grouping of Patients and Statistical Analysis
Patients with AMI were subdivided into subgroups based on time intervals after myocardial infarction and included the first 48 hours, 3 to 5 days, and 6 to 42 days. Significance was determined by Fisher’s exact or Mantel-Haentszel χ2 test where appropriate, and multivariate analysis was performed on statistics that were significant. Data from multivariate analysis were then compared with those of control patients. Statistical analyses were performed using the Statistical Analysis System.
The hospital mortality in patients undergoing elective CABG was 2.7% (6 of 255). These patients served as the baseline for evaluating the AMI patients. The predicted mortality based on the Cleveland Clinic Severity Score was 4%.1
Patients operated on for AMI showed important differences and similarities both within the infarction subgroup and in comparison with control patients. For example, the incidence of women requiring emergency surgery within 48 hours was almost twice that of the remaining AMI group and control patients. As one might expect with critically ill patients requiring emergency coronary surgery, the internal mammary artery was used to graft the left anterior descending coronary artery only half as many times in the <48-hour group (Table 2⇓). On the other hand, there was only a 0.5 New York Heart Association and a 0.6 Canadian Cardiovascular Society class difference between the AMI and control group (Table 1⇑).
When the study group was analyzed by relative risk, ie, incident experimental divided by incident controls, the resulting data showed that a patient with a non–Q-wave infarction could be operated on safely at any time after AMI with a surgical mortality rate similar to that of control patients (3.4% versus 2.4%, respectively; P=NS). A patient with a Q-wave myocardial infarction, however, had a 50% perioperative mortality if operated on within the first 48 hours. After the first 48 hours (ie, excluding the <48-hour group), the surgical mortality fell to 7.7% (4 of 52) over the remaining 40-day period (Table 3⇓). χ2 analysis of hospital mortality for the Q-wave AMI group after 48 hours plus the non–Q-wave AMI group versus the control group was not significantly different. There also was no significant difference between the non–Q-wave group and control patients or between the non–Q-wave group and the Q-wave group after 48 hours. Although comparison of the Q-wave group after 48 hours with control patients was not significant, the probability was .07. Therefore, there does not appear to be a gain in safety by waiting longer than 48 hours to operate on a symptomatic patient with new Q waves. Nevertheless, the surgical mortality in the Q-wave group (including the <48-hour group) remained approximately fourfold that of the non–Q-wave group or control group throughout the 42-day period (P=.001). In most AMI patients, however, the incidence of requiring the intra-aortic balloon or vasopressors in the perioperative period was significantly higher than that in control patients. Although no significant difference was noted for the use of the intra-aortic balloon within the first 48 hours, the reason was probably due to our definition as most patients within this group had balloons placed before surgery and therefore the new balloon insertion incidence was lower than that in the group 3 to 5 days after AMI (Table 3⇓). The risk of extending the AMI was also significantly higher in the 48-hour AMI group for both Q-wave and non–Q-wave infarction than it was for developing a perioperative infarction in control patients. There was no significant difference in length of hospital stay (12.9±5.9 days versus 13.9±5.9 days [excluding 30-day outliers]). The generally long hospital stay reflects the bias of referring physicians to keep patients requiring surgery in the hospital until surgery. There also was no significant difference in days on the ventilator (2.3±8.5 days versus 3.3±8 days) for the control group or AMI group, respectively.
When the data were analyzed by multilogistic stepwise regression, age >65 years, clinical status, Q-wave infarction, and surgery within 48 hours showed the highest risk for hospital mortality.
We initiated this retrospective study to evaluate our experience and examine risk factors associated with early surgery in symptomatic patients after AMI. To provide a reliable and stable baseline, the AMI patients were evaluated against our own control patients.
Our analysis establishes relative risk for each time category for mortality, vasopressor use, intra-aortic balloon requirement, and perioperative myocardial infarction. The study also documents between type of infarction and supports the contention that non–Q-wave AMI has a better prognosis than Q-wave AMI.2 3 Non–Q-wave AMI has an early favorable prognosis with medical management and with early surgical management as well.4
The greatest decline in overall relative risk for hospital mortality, use of vasopressors, and perioperative myocardial infarction occurred after 48 hours. The relative risk of new balloon insertion was not significant due to an institutional bias to insert the intra-aortic balloon pump before surgery when operating on AMI patients with low ejection fractions. Within the Q-wave 48-hour group, one patient did not receive an intra-aortic balloon pump before surgery because of occlusive peripheral vascular disease: two balloons were placed for hemodynamic reasons and one for the relief of pain.
Three patients died within the Q-wave 48-hour group. Two were the result of cerebral events, and one was due to cardiogenic shock. The high stroke rate is curious and may be a function of the age group5 : one patient was 66 years old, and the other was 72 years old and had received thrombolytic therapy. Two patients were operated on in full cardiac arrest with external compression being performed during prepping. Both were in the Q-wave group. One was in the >48-hour group and survived, and the second was in the 6- to 42-day group and died.
In the early 1970s, surgical revascularization within the first 2 months of AMI was associated with a 14.5% to 20% mortality.6 The recommendation then was to wait 30 days after AMI before revascularization. In the late 1970s and early 1980s, improvements in myocardial protection and intra-aortic balloon counterpulsation reduced surgical mortality. Multivariate analysis of this data revealed then, as now, that advanced age, clinical status, and depressed myocardial function were associated with a higher mortality rate.7 In the late 1980s and early 1990s, advances in medical therapy, including thrombolysis and catheter-mediated angioplasty, produced a new group of patients who were referred for surgery only after failing to improve with the use of these techniques. Nevertheless, improvements in surgical and postoperative management permitted surgery in these critically ill patients with improving results.8 Some patients with transmural infarction and extraordinary circumstances may require revascularization within 48 hours, such as those with ongoing, severe unstable angina.9 In these patients, the surgical mortality will be significantly higher than after 48 hours but not necessarily higher than if the surgery were not performed. Even though the number of patients in the <48-hour group was small, the increased mortality was statistically significant compared with control patients. Our data indicate that revascularization after 48 hours is acceptable timing after Q-wave AMI infarction, and, under most circumstances, there is little to be gained by waiting a specified additional length of time. Non–Q-wave AMI in symptomatic patients may be revascularized at any time with no significant increase in mortality compared with elective patients.
- Copyright © 1995 by American Heart Association
Braunwald. Textbook of Cardiovascular Medicine, 4th ed. Philadelphia, Pa: WB Saunders; 1992.
Naunheim KS, Kesler KA, Kanter KR, Fiore AC, McBride LR, Pennington DG, Barner HB, Kaiser GC, Willman VL. Coronary artery bypass for recent infarction: predictors of mortality. Circulation. 1988;78(suppl I):I-122-I-128.