T ing (CABG) compared with medical therapy for coronary

Transcription

1 Coronary Artery Bypass in Patients With Severely Depressed Ventricular Function Carmelo A. Milano, MD, William D. White, MPH, L. &chard Smith, PhD, Robert H. Jones, MD, James E. Lowe, MD, Peter K. Smith, MD, and Peter Van Trigt 111, MD Department of Surgery, Duke University Medical Center, Durham, North Carolina This study evaluates whether patients with coronary artery disease and severely depressed left ventricular ejection fraction benefit from coronary artery bypass grafting. From 1981 to 1991,118 consecutive patients with ejection fraction less than or equal to 0.25 underwent isolated coronary artery bypass grafting at Duke University Medical Center. Operative mortality was 11%. Ventricular arrhythmia requiring treatment was the most common postoperative complication (27%), followed by low cardiac output state (22%). Median length of postoperative hospitalization was 9 days. Kaplan-Meier estimate of survival at 1 year and 5 years was 77.2% and 57.5%, and was better than estimated survival with medical therapy alone. Survivors experienced significant improvement in angina class (p < O.OOOl), congestive failure class (p < O.OOOl), and follow-up ejection fraction (p < 0.005). Of 22 preoperative factors evaluated by univariate survival analysis, five were associated with significantly greater mortality: other vascular disease (p < 0.005), female sex (p < 0.005), hypertension (p < 0.005), elevated left ventricular end-diastolic pressure (p < 0.05), and depressed cardiac index (p < 0.05). Considering length of hospitalization, three factors showed significant adverse effect in a multivariate Cox model: time on cardiopulmonary bypass (p < 0.005), acute presentation (p < 0.005), and female sex (p < 0.01). These data and review of the literature suggest that patients with coronary artery disease and severely depressed ejection fraction benefit from coronary artery bypass grafting, and specific preoperative factors may help determine optimal treatment. (Ann Thorac Surg ) he long-term benefits of coronary artery bypass graft- T ing (CABG) compared with medical therapy for coronary artery disease (CAD) are more pronounced in patients with reduced left ventricular function [ 1-31, However, patients with severe generalized left ventricular dysfunction (ejection fraction [EF] ) have increased operative mortality, with earlier studies concluding that these risks are prohibitive [4]. Within the last two decades, however, improvements in anesthesia, myocardial protection, and postoperative pharmacologic and mechanical support have resulted in reduced operative mortality after CABG. These improvements may make CABG an effective treatment even for those patients with severe generalized left ventricular dysfunction. Patients with CAD and severely reduced left ventricular function may be candidates for cardiac transplantation. Indeed, ischemic cardiomyopathy is currently the most common indication for cardiac transplantation [5]. Unfortunately, the very limited number of organs has resulted in a long waiting list and exclusion of patients with comorbid disease or increased age [6]. The limitations of transplantation may make conventional bypass operation Accepted for publication April 21, Presented in part at the Fifty-eighth Annual Scientific Assembly of the American College of Chest Physicians, Chicago, IL, Oct 25-29, Address reprint requests to Dr Van Trigt, Department of Surgery, Duke University Medical Center, Box 3235, Durham, NC a more important therapeutic option in this patient population. In the present study, the results of CABG in patients with EF of 0.25 or less during the last 10 years at Duke University Medical Center (DUMC) were analyzed. Operative mortality, complications, length of postoperative hospitalization, angina and congestive heart failure (CHF) class, and late mortality were reviewed to help determine whether this population of patients benefits from CABG. In addition, survival after CABG was compared with estimated survival with medical therapy alone, which was obtained from a model generated in the analysis of a previous DUMC data set. Furthermore, preoperative and operative variables were evaluated as predictors of mortality and morbidity. Finally, previously published studies of patients with CAD and markedly reduced left ventricular EF treated medically and surgically were reviewed. Material and Methods A retrospective analysis was undertaken of all 118 patients whose preoperative EF was less than or equal to 0.25 and who underwent isolated CABG at DUMC from 1981 to Preoperative EF was calculated from either single or biplane ventricular angiography during left heart catheterization. Patients undergoing concomitant procedures such as valve replacement, aneurysm resection, or an arrhythmia operation were excluded. Follow-up was acquired on 116 of 118 patients (98.3%) in 1991, with a by The Society of Thoracic Surgeons /93/$6.00

2 488 MILAN0 ET AL Ann Thorac Surg median follow-up time of 26.5 months and cumulative follow-up person-years. Survival was estimated by the method of Kaplan and Meier [7]. In addition, survival estimates for medical therapy alone were calculated using a model based on a data set composed of patients who received cardiac catheterization and treatment for CAD at DUMC. This data set has been previously described and contains 5,809 patients treated from Jan 1, 1970, to Jan 1, Ages ranged from 30 to 85 years, and all patients had angina and at least single-vessel coronary disease by angiography; patients who had previously been treated with CABG, angioplasty, or thrombolytic therapy were excluded [8]. A survival model for this population was developed using the Cox proportional hazards method and the following information: age, sex, year of catheterization, left main stenosis, left anterior descending stenosis, number of diseased vessels, EF, angina class and pattern, history of myocardial infarction, cardiomegaly, ventricular arrhythmia, vascular disease, conduction defects, and degree of mitral insufficiency [8]. Applying these characteristics from the 118 patients in this review, the model was used to estimate survival with medical therapy alone at 1, 3, 5, 7, and 9 years from the time of hospitalization. In the surgical review group, preoperative and postoperative angina and CHF were rated using the classification of the Canadian Cardiovascular Society [9]. Data on postoperative angina and CHF class were acquired on 92 survivors either as documented in the medical record by DUMC cardiologists or cardiac surgeons or as determined through phone conversations in 1991 with patients who did not receive further follow-up at DUMC. Forty-eight survivors had follow-up measurements of left ventricular EF at DUMC, either at the time of follow-up catheterization or by radionucleotide ventriculography. Preoperative angina class, CHF class, and EF are compared with postoperative values using the paired t test. In all 118 operations, moderate systemic hypothermia (28" to 32 C) and either bubble or membrane oxygenators were employed. All patients underwent aortic crossclamping and received crystalloid cardioplegia (4 C). The number of grafts each patient received was determined by the surgeon at the time of operation; the majority of patients were completely revascularized. Data on 22 preoperative and operative variables selected as potential predictors of outcome were obtained from the medical record (Table 1). Each variable was evaluated as a predictor of survival by univariate analysis. A multivariate Cox model was employed to evaluate these variables as predictors of postoperative length of hospitalization, a measure of morbidity. These variables include hypertension defined as requiring medical therapy for at least 2 years, diabetes mellitus defined to include both insulin-dependent and non-insulin-dependent types, other vascular disease defined as either symptomatic cerebral or peripheral vascular disease, acute presentation defined as taken for operation directly from either an intensive care unit or a cardiac catheterization laboratory, pulmonary edema identified on physical examination and chest film, recent myocardial infarction docu- Table 1. Univariate Survival Analysis and Multivariate Analysis (Cox model) With Length of Postoperative Hospitalization as the Endpoint" Multivariate Length of Univariate Postop Survival Hospitalization Variable Percent (p Value) (p Value) Ageb >70 y y y <50 y Male Femaleb Hypertensionb Diabetes rnellitus Other vascular diseaseb Previous heart operation Previous PTCA~ Acute presentationb Class I11 or IV CHF Pulmonary edemab Recent MI Ventricular tachycardiab Preop IABP Class IV anginab Mitral regurgitation EF~ LVEDP~ 59 rnrn Hg mm Hg -29 rnrn Hg mm Hg 240 rnrn Hg Cardiac indexb 18.6% 44.9% 22.9% 14.4% 91.0% 9.0% 58.0% 33.0% 27.0% 7.6% 9.4% 51.7% 58.0% 36.0% 61.5% 14.7% 23.5% 80.0% 49.1% 33.9% 23.7%.3% 5.9% 16.1 % 5.5% 30.0% 38.2% 18.2% 7.3% Mean (L. min-'. m-*) Cardiology 29.0% Total 90.0% Mean aortic cross-clamp 56 f 22 time (rnin) Mean cardiopulmonary bypass timeb (rnin) Mean no. of graftsb 3.7? 1.2 IMA graft 84.6% <O.l <0.005 (0.01 <0.005 <0.005 (0.1 <o. 1 <0.05 <0.05 <0.1 <0.05 <0.005 <0.1 <0.005 a p values less than 0.1 are listed for univariate survival analysis. Variables included in multivariate length of stay analysis; final four predictors with p < 0.05 are listed. CHF = congestive heart failure; EF = ejection fraction; IABP = intraaortic balloon counterpulsation; IMA = internal mammary artery; LVEDP = left ventricular end-diastolic pressure; MI = myocardial infarction; PTCA = percutaneous transluminal coronary angioplasty.

3 Ann Thorac Surg MILANO ET AL 489 OPERATION WITH EF Number of Cua Number of Cases 1, 40 I I Ang.Claw CHF =c1w 1 E2 Claw I1 Claw Ill 0 c1.u IV Fig I. Preoperative angina (Ang.) and congestive heart failure (CHF) class (n = 117) Days to Discharge Fig 3. Days in the hospital after coronary artery bypass grafting (6 patients discharged). mented by creatine kinase MB criteria within 6 weeks of operation, sustained ventricular tachycardia documented by electrocardiographic monitor and requiring treatment with medication or cardioversion, and mitral regurgitation detected on left ventricular angiography. Age, EF, left ventricular end-diastolic pressure (LVEDP), cardiac index, aortic cross-clamp time, cardiopulmonary bypass time, and number of grafts are included in the analysis as continuous variables. Results Sixty-four percent of patients were 60 years or older; 105 were male and 13 female. Hypertension was present in 58%, diabetes mellitus was present in 33%, and other vascular disease occurred in 27%. Prior heart operation and prior percutaneous transluminal coronary angioplasty were present in 7.6% and 9.4%, respectively. Acute presentation occurred in 51.7%. Class I11 or IV CHF was present in 58.1%, and pulmonary edema occurred in 36%. Recent myocardial infarction was present in 61.5%, and sustained ventricular tachycardia occurred in 14.7%. Unstable or postinfarction angina was present in 70%, whereas class IV angina was present in 80%. The distribution of patients by angina class, CHF class, and angina Pmgnulve 17% mt-mi 42% Unstable 28% Fig 2. Preoperative angina pattern (n = 117). (MI = myocardial infarction.) pattern is shown in Figures 1 and 2. Preoperative intraaortic balloon counterpulsation was required in 23.5%. Coronary angiography revealed left main or triple-vessel disease in 78.8% (stenoses greater than 75% were considered clinically significant). Mean LVEDP was 23.6 mm Hg, and mean EF was 0.21; the distributions of both are shown in Table 1. Ninety percent of patients had measurement of cardiac index at the time of right heart catheterization (29% were obtained before the time of operation by cardiology, whereas an additional 71% were obtained in the operating room by a cardiac anesthesiologist before the skin incision); mean cardiac index was L * min-' * m-'. The average number of coronary bypass grafts was 3.7 * 1.2, and 84.6% received internal mammary artery grafts. Mean aortic cross-clamp time and cardiopulmonary bypass time were 56 * 22 and minutes. In-hospital mortality was 10.2% (12 patients). One additional patient died after discharge but within 30 days of operation, resulting in a total operative mortality of 11% (13 patients). The cause of death was ventricular arrhythmia in 6 of these 13 patients, cardiogenic shock in 4, multiorgan system failure in 2, and adult respiratory distress syndrome in 1. The median length of postoperative hospitalization for survivors was 9 days, and the distribution of length of postoperative hospitalization is shown in Figure 3. Ventricular arrhythmias, defined as premature ventricular complexes, ventricular tachycardia, or ventricular fibrillation requiring medication or defibrillation, were the most common postoperative complica- tion. Low cardiac output, defined as a cardiac index less than 2 L * min-' m-' for greater than 3 hours or need for intraaortic balloon counterpulsation to discontinue cardiopulmonary bypass, was the second most common complication; the eight most frequent complications are listed in Table 2. Postoperative angina class in 92 survivors improved an average of 2.1 f 1.2 classes (p < O.OOOl), whereas postoperative CHF improved an average of 0.73 * 1.5 classes (p < ). The distribution of preoperative and postoperative angina and CHF class is illustrated in Figures 4 and 5. Median follow-up for both CHF and angina class was 5.5 months. Forty-eight survivors also

4 490 MILANO ET AL Ann Thorac Surg Table 2. Postoperative Complications Complication Ventricular arrhythmia Low cardiac output Pneumonia Tracheostomy Leg infection Reoperation for bleeding Stroke Rise > 3.0 mg/dl in crea tinine Percent ~ 27.0% 22.0% 6.0% 4.3% 4.3% 3.4% 3.4% 3.4% had follow-up measurements of EF: mean follow-up was , and mean change in EF was (p < 0.005). The distribution of postoperative EFs compared with the preoperative distribution is shown in Figure 6. Figure 7 shows the Kaplan-Meier estimated survival and 95% confidence limits for the 118 CABG patients in this review. One-year and 5-year estimated survival is 77% and 58%, respectively. Also shown is model estimated survival for medical therapy alone at 1, 3, 5, 7, and 9 years from the time of hospitalization. At 1 year, surgical and medical survival are quite similar: 77% and 75%, respectively. At 5 years, the estimated medical survival falls well below surgical survival: 38% versus 58%, respectively. Of the 22 preoperative and operative variables, four were associated with significantly greater mortality: other vascular disease (p < 0.005), female sex (p < 0.005), hypertension (p < 0.005), and elevated LVEDP (p < 0.05). Cardiac index determined by cardiology before the day of operation was also significant ( p < 0.05) in the 34 patients for whom this measure was available. Four other variables were marginally significant (p < 0.1): advanced age, sustained ventricular tachycardia, preoperative intraaortic balloon counterpulsation, and prolonged time on cardiopulmonary bypass. Values of p less than 0.1 for the univariate survival analysis are listed in Table 1. Taking length of postoperative hospitalization as an endpoint, 14 No. of Cased Prcop Follow-up I I1 111 IV CHF Class Fig 5. Distribution of preoperative and follow-up congestive heart failure (CHF) class in 92 survivors. Median follow-up is 5.5 months. Preoperative versus follow-up: p = by paired t test. potential predictors with univariate p values less than 0.1 (see Table 1) were analyzed together in a multivariate Cox model using stepwise variable selection. The final model included three point predictors of increased length of stay: time on cardiopulmonary bypass (p < 0.005), acute presentation (p < 0.005), and female sex (p < 0.01) (history of preoperative angioplasty was marginally significant; p < 0.05). Comment Twenty studies of patients with markedly reduced EF (approximately <0.25) who underwent CABG were identified in the literature. Table 3 lists these in chronologic order. Each study examined a consecutive group of patients, and the EFs ranged from a mean of 0.18 to a mean of Thirty-day mortality ranged from 0% in two very small studies to 55%. Six studies were conducted within the last decade; in these, operative mortality ranged from 2.6% to 15%, with a mean of 9.1%. Compared with the very early studies, these more recent studies demonstrate improved operative mortality. The two most recent stud- Number Of CMm 35 No. Of CMea 80, I n 1 I1 I11 1v Angina Class Fig 4. Distribution of preoperative and follow-up angina class in 92 survivors. Median time to follow-up is 5.5 months. Preoperative versus follow-up: p = by paired t test. 5 n 4 0 IS Ejection Fraction (%) Fig 6. Distribution of preoperative and follow-up ejection fractions in 48 survivors. Preoperative versus follm-up: p < by paired t test.

5 Ann Thorac Surg MILANO ET AL CAB KY Survlvd 8 Yodeled Yedlul Survlrnl Estlmate.i % a t P ' 0.3 0' I Years Fig 7. Kaplan-Meier (KM) estimated survival for 118 patients undergoing coronary artery bypass grafting (CAB), shown with 95% confidence limits. Also shown is model estimated survival for medical therapy alone at 1, 3, 5, 7, and 9 years. ies contain large numbers of patients, but late mortality, predictors of late mortality, follow-up angina and CHF class, and comparison with medical therapy alone, which this study addresses, were not evaluated. In addition, many of these studies included patients who underwent concomitant valve replacement or left ventricular aneurysm operations and this complicates interpretation of their results; this study specifically excludes these patients requiring combined procedures. Seven studies were identified in which patients were managed entirely medically (Table 4). Each contains consecutive patients who received medical therapy alone either because they refused operation or because they i were not considered for operation. The most recent study was completed in 1981 and contains 59 patients. The largest contains 172. The EFs ranged from less than 0.25 to a mean of One-year mortality ranged from 10% to 50%. Late mortality ranged from 38% at 2 years in one study to 96% at 5 years in another. In six of the seven studies, the medically treated groups were compared directly with groups treated with CABG; in all six of these, the surgical groups had improved late survival. Recent series (within the last decade) of medically treated patients with severely depressed left ventricular function are not available in the literature. Prospective, randomized trials comparing medical therapy with CABG for patients with symptomatic CAD and EFs as low as 0.30 have shown a long-term survival benefit for those receiving CABG [2]. For patients with even greater impairment of ventricular function, management is unclear, and there is a reluctance of cardiologists to refer or surgeons to accept these patients for CABG. Furthermore, some of these patients (particularly those with more severe symptoms of CHF) are referred for cardiac transplantation, but donor organ numbers severely limit this form of treatment. In an effort to determine whether these patients with EF of 0.25 or less benefit from CABG for treatment of their symptomatic CAD, a review and analysis of 118 consecutive isolated CABG patients with preoperative EFs of 0.25 or less is presented. Operative mortality was 11%, and 5-year estimated survival was 58%. Although operative mortality is much higher than for routine CABG, it must be evaluated with regard to the unstable preoperative status of these patients (70% unstable or postinfarction angina, 61.5% re- Table 3. Review of the Literature for Coronary Artery Disease Patients With Severely Reduced Ejection Fraction Treated With Coronary Artery Bypass Grafting First Author Vlietstra [lo] Manley [ll] Yatteau [12] Oldham [4] Zubiate (131 Faulkner [14] Mitchel [15] Fox [16] Jones [17] Alderman [18] Mochtar [19] Zubiate [] Hochberg [21] Hochberg [21] Sanchez [22] Kron [23] Blakeman (241 Wong [25] Christakis [26] Hammermeister [27] No. of Patients Years EF ~1 196a <0.25 mean 0.22 < mean mean 0.25 C0. 0.2(M. 24 mean 0.28 mean 0.18 mean Day Mortality 16.0% 42.0% 55.0% 22.0% 4.0% 0.0% 0.0% 2.5% 8.0% 4.8% 5.0% 12.0% 37.0% 9.0% 2.6% 15.0% 9.0% 9.8% 9.2% Late Mortality 60% (2 y) 43% (5 y) 50% (2 y) 41% (6 y) 17% (2 y) 11% (1 y) 14% 10% (1 y) 37% (5 y) 30% (5 y) 50% (5 y) 42% (3 y) 85% (3 y) 24% (2 y) 17% (3 y) 30% (1 y) 23% (3 y)

6 492 MILANO ET AL Ann Thorac Surg Table 4. Review of the Literature for Coronary Artery Disease Patients With Severely Reduced Ejection Fraction Treated With Medical Therapy Alone One- No. of Year Late First Author Patients Years EF Mortality Mortality Vlietstra [lo] < % (2 y) Manley [ll] ~1 mean % 90% (6 y) Yatteau [12] < % 38% (2 y) Faulkner [ S75 mean 0. 10% 53% (2 y) Harris [28] < % 55% (5 y) Zubiate [] % 96% (5 y) Alderman [18] % 57% (5 y) cent myocardial infarction, 23.5% preoperative intraaortic balloon counterpulsation, 36% history of pulmonary edema). Probably a substantial portion of these patients would not have stabilized without surgical intervention. Late survival in these 118 patients after operation is better than their estimated survival with medical therapy alone; furthermore, their late survival is better than that of similar patients in previously published series who received medical therapy alone. In addition, after CABG survivors had significant improvement in angina symptoms, which probably reduced future hospitalizations and improved quality of life for these patients. Survivors also had significant improvement in CHF class. Some of this improvement undoubtedly resulted from medical treatment initiated during these patients hospitalizations; other patients probably preoperatively experienced reduced left ventricular function and symptoms of heart failure during episodes of ischemia, and with revascularization had improvement in their symptoms. This is further emphasized by improvement in follow-up EF measurements. In conclusion, although prospective, randomized trials for treatment for this population are not available, certain patients with symptomatic CAD and severely lowered EF benefit from CABG not only with regard to improvement of symptoms but probably also with improved long-term survival. Furthermore, this treatment does not appear to be associated with unacceptable morbidity. Patients with graftable vessels and severely reduced EF should not be denied CABG on the basis of their severely reduced left ventricular function; for many of these patients CABG is probably the most appropriate treatment. Prospective, randomized trials comparing CABG with medical therapy alone for this population should confirm these conclusions. The majority of preoperative factors analyzed in this study were poor predictors of total mortality; the most significant predictors were other symptomatic vascular disease, female sex, hypertension, elevated LVEDP, and depressed cardiac index. Patients with other vascular disease and hypertension probably experience some late mortality due to these conditions, and this may nullify the long-term benefit of CABG; medical therapy alone may be more appropriate for some of these patients. Other studies have noted depressed cardiac index and elevated LVEDP as important predictors of mortality after CABG in the low-ef population [29]. Patients with severely reduced EF and CAD may benefit from right heart catheterization with measurement of cardiac index before surgical consideration. Elevated LVEDP and depressed cardiac index in the setting of a very low EF signals a failure to compensate for depressed left ventricular function. This subset of patients may be best managed with cardiac transplantation provided they do not have other comorbid conditions. Cardiogenic shock in this study represented the second most common complication as well as the second most frequent cause of operative mortality. Proper employment of postoperative pharmacologic and mechanical support remains crucial in the management of these patients. Furthermore, methods to distinguish reversible myocardial dysfunction or stunned myocardium from irreversible dysfunction would be beneficial preoperatively in predicting not only which patients have sufficient reserve to tolerate surgical revascularization but also which are more likely to experience long-term benefits. Experimental efforts and clinical trials in this area could result in important diagnostic tools and should be supported. Finally, ventricular arrhythmias represented the most frequent cause of operative mortality as well as the most common complication. The presence of preoperative ventricular arrhythmias, however, did not strongly predict poor outcome. Therefore, regardless of preoperative history, careful assessment and treatment of ventricular arrhythmias is critical in the management of these patients. Through more careful assessment of preoperative risk factors and patient selection, as well as attention to common complications in this population, CABG may be offered to patients with very low EFs with reduced morbidity and mortality. References 1. Veterans Administrative Coronary Artery Bypass Surgery Cooperative Study Group. Eleven-year survival in the Veterans Administration Randomized Trial of Coronary Bypass Surgery for Stable Angina. N Engl J Med 1984;311:133>9. 2. Luchi RJ, Scott SM, Deupree RH, et al. Comparison of medical and surgical treatment for unstable angina pectoris: results of a Veterans Administration Cooperative Study. N Engl J Med 1987;316: Passamani E, Davis KB, Gillespie MJ, et al. A randomized trial of coronary artery bypass surgery: survival of patients with a low ejection fraction. N Engl J Med 1985;312:

7 Ann Thorac Surg MILAN0 ET AL Oldham HN, Kong Y, Bartel AG, et al. Risk factors in coronary artery bypass surgery. Arch Surg 1972;105: Gay WA, OConnell JB. Cardiac transplantation. In: Sabiston DC, Spencer FC, eds. Surgery of the chest. 5th ed. Philadelphia: Saunders, 1990: Evans RW, Maier AM. Outcome of patients referred for cardiac transplantation. J Am Coll Cardiol 1986;8: SAS Institute Inc. SASBTAT user s guide, version 6, 4th ed. vol2. Cary, NC: SAS Institute, Califf RM, Harrell FE, Lee KL, et al. The evolution of medical and surgical therapy for coronary artery disease: a 15 year perspective. JAMA 1989;261: Campeau L. Grading of angina pectoris. Circulation 1976;54: Vlietstra RE, Assad-Morel1 JL, Frye RL, et al. Survival predictors in coronary artery disease: medical and surgical comparisons. Mayo Clinic Proc 1977;52: Manley JC, King JF, Zeft HJ, Johnson WD. The bad left ventricle. Results of coronary surgery and effect on late survival. J Thorac Cardiovasc Surg 1976;72: Yatteau RF, Peter RH, Behar VS, Bartel AG, Rosati RA, Kong Y. Ischemic cardiomyopathy: the myopathy of coronary artery disease, natural history and results of medical versus surgical treatment. Am J Cardiol 1974;34: Zubiate P, Kay JH, Mendez M. Myocardial revascularization for the patient with drastic impairment of function of the left ventricle. J Thorac Cardiovasc Surg 1977;73: Faulkner S, Stoney WS, Alford WC, et al. Ischemic cardiomyopathy: medical versus surgical treatment. J Thorac Cardiovasc Surg 1977;74: Mitchel BF, Alivizatos PA, Adam M, Geisler GF, Thiele JP, Lambert CJ. Myocardial revascularization in patients with poor ventricular function. J Thorac Cardiovasc Surg 1975;69: Fox HE, May IA, Ecker RR. Long-term functional results of surgery for coronary artery disease in patients with poor ventricular function. J Thorac Cardiovasc Surg 1975;70: Jones EL, Craver JM, Kaplan JA, et al. Criteria for operability and reduction of surgical mortality in patients with severe left ventricular ischemia and dysfunction. Ann Thorac Surg 1978;25: Alderman EL, Fisher LD, Litwin P, et al. Results of coronary artery surgery in patients with poor left ventricular function (CASS). Circulation 1983;68: Mochtar B, Laird-Meeter K, Brower RW, Verbaan N, Haalebos MMP, Bos E. Aorto-coronary bypass surgery in 62 patients with severe left ventricular dysfunction-a follow-up study. Thorac Cardiovasc Surg 1985;33: Zubiate P, Kay JH, Dunne EF. Myocardial revascularization for patients with ejection fraction of 0.2 or less: 12 years results. West J Med 1984;140: Hochberg MS, Parsonnet V, Gielchinsky I, Hussain SM. Coronary artery bypass grafting in patients with ejection fractions below forty per cent. J Thorac Cardiovasc Surg 1983;86: Sanchez JA, Smith CR, Drusin RE, Reison DS, Malm JR, Rose EA. High-risk reparative surgery, a neglected alternative to heart transplantation. Circulation 1990;82(Suppl 4): Kron IL, Flanagan TL, Blackbourne LH, Schroeder RA, Nolan SP. Coronary revascularization rather than cardiac transplantation for chronic ischemic cardiomyopathy. Ann Surg 1989;210: Blakeman BM, Pifarre R, Sullivan H, Costanzo-Nordin MR, Zucker MJ. High risk heart surgery in the heart transplant candidate. J Heart Transplant 1990;9:46% Wong JW, Tong MC, Ong KK. Coronary artery bypass in patients with impaired left ventricular function. Ann Acad Med Singapore 1990;19: Christakis GT, Weisel RD, Fremes SE, et al. Coronary artery bypass grafting in patients with poor left ventricular function. J Thorac Cardiovasc Surg 1992;103: Hammermeister KE. Surgery for unstable angina (Department of the Veterans Administration Cardiac Surgery Risk Assessment Program). In: Morrison DA, Serruys PW, eds. Medically refractory rest angina. New York: Marcel Dekker, 1992: Harris PJ, Lee KL, Harrell FE, Behar VS, Rosati RA. Outcome in medically treated coronary artery disease. Ischemic events: nonfatal infarction and death. Circulation 1980;62:71% Hausmann H, Warnecke H, Schiessler A, et al. Predictors of survival in patients with left ventricular ejection fraction of 10-30% receiving coronary artery bypass grafting: analysis of preoperative variables in 177 patients. Circulation 1991; 84(Suppl 2):284.