Coronary Bypass for Relief of Persistent Pain Following Acute Myocardial Infarction

Transcription

1 Coronary Bypass for Relief of Persistent Pain Following Acute Myocardial Infarction Ellis L. Jones, M.D., Thad F. Waites, M.D., Joe M. Craver, M.D., James M. Bradford, Ph.D., John S. Douglas, M.D., Spencer 8. King, M.D., David K. Bone, M.D., Edward R. Dorney, M.D., Stephen D. Clements, M.D., Tom Thompkins, P.A., and Charles R. Hatcher, Jr., M.D. ABSTRACT Between January, 1976, and April, 1980, 116 patients had urgent myocardial revascularization for clinical instability within 30 days of acute myocardial infarction (MI). Group 1 (8 patients) had coronary bypass grafting within 24 hours of acute MI; Group 2 (20 patients) had coronary bypass grafting 2 to 7 days after acute MI; and Group 3 (88 patients) had coronary bypass grafting 8 to 30 days after infarction. Indications for operation were persistent or recurrent pain (8l%), pain plus ventricular arrhythmias (12%), and pain plus compelling anatomy. The incidence of single-vessel, triple-vessel, and left main coronary artery disease was 28%, 31%, and 12%, respectively. There were no hospital deaths in the series. The incidence of inotropic requirements, postoperative intraaortic balloon pumping, ventricular arrhythmias, and perioperative infarction was higher in patients operated on within 7 days of acute MI than for patients having coronary bypass grafting after this time. There have been 5 late deaths during a mean follow-up of 14 months. Actuarial survival was 97% at 18 months. Seventy-one percent of patients are presently pain free. Graft patency was 84% in 17 patients recatheterized after coronary bypass grafting and in 14 patients, grafts placed into the area of infarction were patent. This study suggests that the frequency of perioperative complications will be increased in patients operated on within one week of acute MI, but after this period, coronary bypass grafting can be ac- From the Departments of Surgery and Medicine, Emory University School of Medicine, Woodruff Medical Center, Atlanta, GA. Our greatest appreciation to Mrs. Judy Evans of the Emory University Hospital Data Bank who helped in the compilation of much of these data. Presented at the Twenty-seventh Annual Meeting of the Southern Thoracic Surgical Association, Nov 13-15, 1980, White Sulphur Springs, WV. Address reprint requests to Dr. Jones, Emory University Clinic, 1365 Clifton Rd, NE, Atlanta, GA complished with the same morbidity as that of elective operation. The unstable clinical course of some patients following acute myocardial infarction (MI) has been documented in recent reports [l-81. At the Massachusetts General Hospital, early mortality of patients with transmural MI was 29% and of those with subendocardial infarction, 14% [71. Although immediate prognosis after subendocardial infarction appears fairly good, Madigan and co-workers [9] found the incidence of unstable angina or later transmural infarction to be 46% and 21%, respectively. If left ventricular power failure and cardiogenic shock occur as a complication of acute infarction, mortality has been exceptionally high with medical therapy, approaching 80 to 90% in most series [lo-141. Recently there has been renewed interest in applying the coronary bypass operation to selected patients early after acute MI in order to prevent extension of infarction and preserve myocardial muscle mass and left ventricular function [7, 8, 15-18]. In 1978, we [15] reported the early results in 35 patients undergoing urgent myocardial revascularization for persistent pain within 30 days of acute MI; there was good initial success. We have continued to implement this mode of therapy in selected patients and have extended this early series to a total of 116 patients. The clinical profile of the patient population, criteria for operation, and methods of reducing early surgical mortality form the basis of this report. Material and Methods Between January, 1976, and April, 1980, 116 patients had urgent myocardial revascularization for clinical instability within 30 days of acute MI. There were 88 men and 27 women with a mean age of 52.9 years. Infarction was by The Society of Thoracic Surgeons

2 34 The Annals of Thoracic Surgery Vol 32 No 1 July 1981 defined as prolonged ischemic pain accompanied by electrocardiographic (ECG) evidence of new 0.04 second Q waves (84 patients), or acute S-T depression or T wave inversion or both of these (32 patients) associated with elevation of creatine phosphokinase (CPK) to twice normal levels for our laboratory (290 p per milliliter). In 7 of the 32 patients diagnosed as having subendocardial MI, the CPK-MB fraction was greatly elevated in association with marked elevation of total CPK and ECG S-T segment and T wave changes. No patient in this series underwent emergency operation secondary to a complication of cardiac catheterization. Patients were arbitrarily divided into three groups defined by the interval between infarction and revascularization. Group 1 (8 patients) underwent operation within 24 hours of acute MI; Group 2 (20 patients) had revascularization 2 to 7 days (mean, 4.6 days) following infarction; and Group 3 (88 patients) had revascularization 8 to 30 days (mean, 18.5 days) after infarction. Clinical instability was defined as persistent pain despite maximal therapy with vasodilator drugs, narcotics (Group l), and propranolol; recurrent pain following infarction (Groups 2 and 3); or persistent or recurrent pain associated with ventricular arrhythmias (premature ventricular contractions, ventricular tachycardia, or ventricular fibrillation). Of the entire series, 89% ( ) of patients received propranolol therapy prior to revascularization. The average daily dose was 122 mg with little difference among the groups. Intravenously administered nitroglycerin was used in 33% (381116) of the patient population and was given more frequently to patients in Groups 1 and 2 (50% and 45%, respectively) than to those in Group 3 (28% or 25 of 88 patients). Six percent of the patients received lidocaine for control of troublesome ventricular arrhythmias (7/116), but only patients in Groups 2 and 3 received this drug. Preoperative intraaortic balloon pumping (IABP) was used in 3.4% of the series (41116) and was restricted to patients in Groups 2 and 3. Variations in the numbers of patients for any given analysis were due to missing data for that particular variable. Missing data usually were attributable to clinical evaluation and cardiac catheterization performed outside our medical facility. Statistical analysis for differences among the three groups of patients was determined by chi-square testing. Special Techniques for Cardiac Catheterization All patients underwent selective coronary arteriography and left ventriculography by the single catheter femoral percutaneous technique [19, 201. Medical therapy was not interrupted prior to catheterization. Left ventricular pressure was recorded, and biplane cine left ventriculography was performed in the 30-degree right anterior oblique and 60-degree left anterior oblique projections with injections of 32 to 40 ml of Renografin-76 (diatrizoate meglumine and diatrizoate sodium) over four seconds. Nitroglycerin was administered sublingually before and after left ventricular injection when ischemic pain occurred or when left ventricular end-diastolic pressure (LVEDP) was greater than 20 mm Hg. In selected patients with markedly impaired wall motion, a separate ventriculogram was performed after nitroglycerin had been given. Ejection fraction was calculated from the right anterior oblique view by means of the area-length method [21]. After intravenous administration of 0.6 mg of atropine, each coronary artery was selectively catheterized and 35 mm cinecoronary angiographic studies were performed in multiple oblique and longitudinally angulated views with 4 to 8 ml of Renografin-76 injected by hand. Sublingually administered nitroglycerin was used to enhance distal opacification of totally occluded vessels and to exclude coronary spasm. Treatment of ischemic pain during the study was directed at cautious reduction of the heart rateblood pressure product and left ventricular wall tension by administering nitroglycerin sublingually and meperidine intravenously, elevating the upper part of the trunk, and giving 1 mg increments of propranolol intravenously. Sufficient time was allowed between coronary injection for relief of pain and return of S-T segments, heart rate, and LVEDP to values measured before stress. Transient hypotension after coronary injections was common but was effectively treated by having the patient cough to increase the aortic pressure momentarily and

3 35 Jones et al: Coronary Bypass for Pain Relief after MI " INF. ANT TRANSMURAL SUBENDOCARDIAL Fig 1. Infarction pattern in 111 patients undergoing myocardial revascularization after acute infarction. (INF. = inferior; ANT. = anterior.) enhance clearing of the contrast agent from the coronary circulation. The number of coronary injections was minimized without compromising the objective of obtaining clear definition of coronary anatomy. Although it is our policy to use the intraaortic balloon whenever necessary to support patients who are hemodynamically unstable, none in the present series required IABP prior to coronary arteriography. No complications of coronary arteriography were recognized. Coronary arteriograms were graded by observing the diseased segments in all views and comparing the diameter of the diseased segment with normal-appearing adjacent segments. Coronary artery lesions were considered significant if the luminal diameter at the obstruction was equal to or less than 50% of the diameter of normal adjacent segments. Clinical Profile Indications for operation were medically uncontrollable pain following infarction or pain with persistent ventricular arrhythmias in 108 patients. Arrhythmias consisted of tachycardia, ventricular fibrillation, or premature ventricular contractions. Ventricular tachycardia or fibril- lation was present in 5% of the total group (61116). Pain plus a ventricular arrhythmia as defined was present in 12% (14/116). There were 8 patients in whom pain following infarction was the initial complaint but who underwent myocardial revascularization primarily because of their young age, compelling anatomy-for example, high-grade left main obstruction-or both. There was no important difference in the indications for operation among the three groups. There was a higher incidence of transmural than subendocardial infarction in the entire series (Fig 1). If a transmural infarction occurred, it was more likely to be inferior or posterior (41%) than anterior (29%), except in Group 1 patients who had a very high incidence (71%) of anterior wall transmural infarction. This finding was statistically significant ( p < 0.01). Cardiac Catheterization LEFT VENTRICULAR FUNCTION. Mean ejection fraction for the entire series was 0.55 with no significant difference among the three groups. Mean ejection fraction was equal to or less than 0.40 in 17.5% (17/97) of patients and equal to or less than 0.35 in 10% of patients. No patient operated on within 24 hours of infarction had an ejection fraction equal to or less than 0.35 (Table 1). Left ventricular wall motion was

4 36 The Annals of Thoracic Surgery Vol 32 No 1 July 1981 Table 1. Preoperative Left Ventricular Function in Patients with Recent Acute Myocardial lnfarction Ejection Fraction LVEDP Patient S0.40 S0.35 S15 mm Hg 220 mm Hg Group Mean ( OIo ) (Yo) Mean (O%) (Yo) Series (N = 97) Group 1 (N = 7) Group 2 (N = 16) Group 3 (N = 74) " "p < LVEDP = left ventricular end-diastolic pressure. v) I- z U E s g 20 t p=ns I1 Ill Series NORMAL HYPOKINETIC AKlNETlC Fig 2. Preoperative left ventricular wall motion in patients with recent infarction. (NS = not significant.) normal in 44% of patients (461104) and hypokinetic or akinetic in 21% and 35% of patients, respectively. Fewer patients in Group l had akinesis of at least one wall segment than patients in Group 2 or Group 3 but this difference was not statistically significant (Fig 2). Resting LVEDP was normal in 73% (69/94) and abnormal in 27% of patients. It was equal to or greater than 20 mm Hg in 12% (11/94) of patients (see Table l). It was interesting that most patients with markedly abnormal LVEDP (320 mm Hg) were in Gro-up 1. CORONARY ARTERY DISEASE. At least 75% of the luminal area of the left main coronary artery was obstructed in 12% of the patients in the series (Table 2). A higher percentage of patients in Groups 1 and 2 had left main coronary artery disease than in Group 3 (25% and 23.5% versus 9%, respectively), and this was statistically significant (p < 0.05). The incidence of triplevessel disease was 31% (331105). The incidence of single-vessel disease was 28% for the series (291105) and highest for patients operated

5 37 Jones et al: Coronary Bypass for Pain Relief after MI Table 2. Distribution of Coronary Artery Disease in Patients with Coronary Bypass and Recent lniarction» Patient Group Series Group 1 Group 2 Group 3 Single Vessel Disease" (%) 28 (291105) 37.5 (3/8) 23.5 (4/17) 27.5 (22/80) Triple Vessel Disease" (%) 31 (331105) 37.5 (3/8) 23.5 (4/17) 32.5 (26/80) Left Main Coronary Artery Disease" (%) 12 (13/105) 25 (2/8) 23.5 (4/17) 9 (7/80) "Numbers in parentheses indicate numbers of patients. "No statistically significant difference between groups. "p < on within the first 24 hours of infarction (37.5%). There was at least one totally occluded artery in 46% of the entire series (53/116). The left anterior descending coronary artery was greatly obstructed in 81 % of patients (91/112) and was involved in all patients in Group 1. This artery was the only artery involved with a hemodynamically significant lesion in 21% of the series (24/112). Operation Precise monitoring of rate-pressure product, inferior and lead V 5 precordial EeG changes, pulmonary capillary wedge pressure, and cardiac output was performed in most patients before bypass. Swan-Ganz catheters were not routinely employed in those patients who had ventricular arrhythmias in the preoperative period. Anesthesia was maintained primarily by frequent doses of morphine sulfate, with diazepam and nitrous oxide, as indicated, to control blood pressure or heart rate. For most patients (except those with evidence of reduced cardiac output), propranolol therapy was continued until 12 hours prior to anesthetic induction. However, if either tachycardia or hypertension occurred before bypass, it was managed promptly with small doses of intravenously administered propranolol or trinitroglycerin drip, respectively. Adverse changes in the variables monitored were treated quickly to avoid acute deterioration in ventricular function prior to cardiopulmonary bypass. Otherwise, the operative tech- SERIES = 2.6 GRAFTS/PT Fig 3. Operative procedures in patients with recent acute myocardial infarction. (DCABG = double coronary artery bypass graft; TCABG = triple coronary artery bypass graft; SCABG = single coronary artery bypass graft.) nique in these patients was no different from that used in others having coronary artery bypass [22]. Moderate systemic hypothermia was employed, and left ventricular depressive factors such as ventricular distention and fibrillation were minimized. However, left ventricular venting was not used in the latter part of the series. No patient in the series had endarterectomy, infarctectomy, or aneurysmectomy combined with coronary bypass grafting. Myocardial preservation was accomplished with either a cold (4 C) hyperkalemic or hyperkalemic hyperosmolar solution injected into the aorta at the time of aortic crossclamping. The former was a potassium chloride solution (20 meq/l) buffered to ph 7.4, and the latter was a potassium chloride-mannitol solution with the following concentrations: potassium, 28 meq/l; bicarbonate radical, 9.3 meq/l; sodium, 95.7 meq/l; dextrose 50%,3.2 ml; mannitol 15%, 200 ml with sterile water to 1,000 ml; ph 8.1; osmolarity, 415 mosm at room temperature. A single aortic cross-clamp period was used for all distal anastomoses, regardless of the degree of ventricular dysfunction or the number of bypass grafts to be performed. Single coronary bypass grafting was performed in 18% of patients (20/112), and three or more coronary bypass grafts were performed in 54% of patients (61/112). The extent of grafting could not be determined in 4 patients. The incidence of multiple grafting was approximately the same for all groups. The average number of grafts per patient for the entire series was 2.62 (Fig 3).

6 38 The Annals of Thoracic Surgery Vol 32 No 1 July 1981 INOTROPES VENT. ARRHYTHMIAS NEW Q WAVE SERIES =5% SERIES = 9% 1% SERIES =3% SERIES =5% Fig 4. Complications following coronary artery bypass grafting in patients with recent infarction. There were no hospital deaths in the series. (VENT. == ventricular; IABP == intraaorticballoon pumpingj Results Hospital Mortality and Morbidity There were no hospital deaths in this series of patients. Inotropic drugs were required postoperatively in 5% of patients. The requirement for these drugs was higher for patients operated on within a week after infarction (Groups 1 and 2) than for patients in Group 3 (Fig 4). Inotropic requirements for patients having elective coronary bypass over a comparable period of time were approximately the same as those for patients having revascularization more than one week after infarction (see Fig 4). Of those patients requiring inotropic drugs postoperatively, 2 had a preoperative ejection fraction greater than 0.40 and 1 had an ejection fraction less than Contractility was normal in 2 of the patients requiring inotropic drugs postoperatively and abnormal in another patient. In the other 3 patients requiring inotropic support, preoperative ejection fraction was unknown. Intraaortic balloon pumping was used in 4 of the 116 patients in the preoperative period and in 3 additional patients to wean the patient from cardiopulmonary bypass at the conclusion of the operative procedure (see Fig 4). Postoperative use of IABP for elective operation over a comparable period of time was 1.2%. Postoperative use of IABP was much more frequent for patients in Groups 1 and 2 than Group 3. Of the 7 patients requiring postoperative use of IABP, 2 had triple-vessel disease and 1 had single-vessel disease. Left ventricular function (as determined by ejection fraction) was greater than 0.40 in 3 of the 7 patients requiring IABP and less than 0.40 in 2. Ejection fraction was unknown in the other 2 patients. Left ventricular contractility in these 7 patients was normal in 2, abnormal in 3, and could not be evaluated in the other 2. In 5 of the 7 patients requiring IABP who could be evaluated angiographically, 3 had at least one akinetic left ventricular segment. Postoperative pharmacological treatment of ventricular arrhythmias consisting of multifocal premature ventricular contractions, ventricular tachycardia, or fibrillation was required in 9% of the patients (10/116). Although not statistically significant, the incidence of arrhythmias was higher in patients operated on within 7 days of infarction (see Fig 4). Perioperative infarction (new Q wave) occurred in 4 of 80 patients (5%) in whom it could be evaluated by serial ECGs. This compared with 4.6% of patients having elective operation during the same period. Again, the incidence of periop-

7 39 Jones et al: Coronary Bypass for Pain Relief after MI Table 3, Analysis of 5 Late Deaths Left Ventricular Function Patient Vessel No, Disease EF Contraction Wall Motion 1 Single 0.59 Abnormal Akinetic 2 Double 0.35 Abnormal Akinetic 2 Triple 0,41 Abnormal Akinetic 4 Double 0,76 Abnormal Hypokinetic 5 Single 0,63 Normal Normal EF = ejection fraction; Contraction = overall angiographic left ventricular contraction pattern,...i c( > s a: :;:) III W > 5:;:) ~ :;:) U N= MONTHS AFTER SURGERY Fig5, Actuarial survival among 116 patients having myocardial revascularization after infarction, Table 4, Functional Status of Patients Having Coronary Bypass Soon after Acute Infarction Mean Late Follow- Cardiac Anginal Status (%) Patient up Deaths Im- Group (mos) (%) No Pain proved Series Group Group Group erative infarction was higher for patients in Groups 1 and 2 (see Fig 4), Late Follow-up There have been 5 late deaths during a short follow-up period (mean, 14 months). There were no deaths in Group 1, 2 deaths in Group 2, and 3 deaths in Group 3. Of the 5 patients who died, only 1 had an ejection fraction less than All patients but 1 dying late had abnormal left ventricular contractility on angiography (Table 3). Actuarial survival was 97% at 18 months (Fig 5). Seventy-one percent of patients (66/93) are presently pain free, and 28%, although improved, do complain of some degree of residual chest discomfort thought to represent angina pectoris. Only 1% of patients are worse or unchanged following revascularization (Table 4). Repeat cardiac catheterization was performed 1 to 19 months following operation in 17 patients. Graft patency was 84% (36/43) for the series, and in 76% of patients (13/17) all grafts were patent. In 14 of the 17 patients, grafts placed into the area of myocardial infarction were patent at the time of restudy. Mean ejection fraction following revascularization increased slightly (0.44 to 0.50) in those restudied. Normal wall motion by angiography was present in 47% of patients (8/17) after operation, compared with 23% of the same patients (8/35) at the time of original catheterization. Most patients demonstrated either an improvement in ejection fraction when this value had been depressed preoperatively or very little deterioration when ejection fraction had been high preoperatively. Only 1 patient exhibited severe

8 40 The Annals of Thoracic Surgery Vol 32 No 1 July 1981 deterioration in ejection fraction, and this patient had an MI after catheterization and just prior to operation. Improvement in segmental wall motion following revascularization occurred in 7 patients; there was no change in 6 patients, and deterioration of movement occurred in 2 patients. Comments The unstable clinical course of some patients following acute MI has been well documented [1,2, 7]. Overall hospital mortality has been reported as 10 to 15% during the first year and 20 to 30% by the third year after infarction. Over one-half of these deaths are sudden [8]. Stenson and associates [23] identified a group of patients with persistent or recurrent pain and S-T segment elevation after infarction in whom mortality was very high on medical therapy and observation alone. Infarct extension soon after the initial infarction accounts for a large number of these deaths. The true incidence of this phenomenon is unknown but has been reported to be between 33 and 86% [4, 5]. In 1979, Fraker and co-workers [3] retrospectively analyzed 458 patients admitted to their coronary care unit and found the incidence of infarct extension to be less than reported previously, but still around 13%. Extension of the infarction occurred an average of 3.4 days after admission to the hospital. Hospital mortality in these patients was 36% versus 9% for patients without infarct extension. One-year survival for patients with extension was 76% and 91% for matched patients without extension. Persistent or recurrent pain following acute infarction in our patients was interpreted as incomplete infarction distal to an obstructed artery (borderzone ischemia) or loss of collateral blood supply to a myocardial region away from the original infarction but supplied by another greatly obstructed artery. Schuster and Bulkley [6] corroborated this clinical supposition by describing two morphological classifications of infarction thought to be important in the production of postinfarction angina. They identified 20 patients in whom angina occurred in association with transient ischemic ECG changes after infarction. In 12 of these patients, the angina was associated with ECG changes in the ECG area of the original infarct. In the remaining patients, there was ECG evidence of ischemia developing at a site distant from the area of initial infarction which they interpreted as infarction-induced interruption of collaterals to a distant site supplied by a second chronically narrowed coronary vessel. In the majority of these patients, rest pain developed within the first 36 hours after infarction. Death was due not to cardiogenic shock but ventricular arrhythmias. This led the authors to conclude that patients in whom ischemia distant from the site of infarction develops following acute MI may be especially suited for therapeutic interventions directed at limiting ischemia, ischemia-related arrhythmias, and necrosis. In contrast, patients with ischemia in the infarct zone at postmortem examination appeared more likely to have acute MI with a large mass of irreversible necrotic myocardium, a setting in which any intervention would be of limited value. In recent years there has been much interest in the surgical treatment of patients during or shortly after acute MI [6, 7, 9, 15, 17, 18, 24-27]. Since the early reports of Keon [28], Hill [29], Dawson [30], Johnson [31], Loop [32], and their associates, interest in the clinical management of these patients has continued [7, 8, 17, 18]. Careful selection is important in assuring a successful outcome in patients undergoing coronary bypass shortly after MI. Continued or recurrent pain after infarction implies the event is not yet complete and suggests improved patient salvage may be attained by surgical intervention. Our data indicate this is particularly true if residual left ventricular function is adequate, suitable distal vessels are visualized on arteriograms, and medical therapy after catheterization is not pursued to such a point that infarction occurs again and irreparable left ventricular damage is done. Rogers and colleagues [8] found no statistical difference in three-year survival between patients treated surgically versus those treated nonsurgically following acute infarction. However, there was a trend toward better long-term survival in those patients with better left ventricular function and a remarkable absence of sudden death in the group having operation.

9 41 Jones et al: Coronary Bypass for Pain Relief after MI There were differences and similarities in the patient population of that study and the one reported here. Seventy-five percent of the patients in the series of Rogers and associates [8] had no angina at the time operation was performed, whereas in the present series almost all patients had pain as a major indication for surgical intervention. A major characteristic of our patient population was the relatively good left ventricular function found in a large proportion of individuals. Revascularization in these patients will have the lowest hospital mortality and the greatest expectation of a satisfactory long-term result. Retrospectively, several factors appeared to be important in our selection of patients for operation after infarction. First, no patient underwent coronary bypass within 24 hours of infarction if ejection fraction was equal to or less than Ejection fraction for the entire series was well within the normal range for most patients. Although not statistically significant, fewer patients operated on within the first 24 hours of infarction had angiographically demonstrated left ventricular akinesis compared with those operated on after this time. Another factor influencing our decision to operate very early in patients with pain after infarction was the presence of left main coronary obstruction. There was a significantly higher incidence of left main obstruction in those patients operated on within the first week of infarction. The low morbidity and mortality seen in our patients are probably a reflection of the aggressiveness of our medical and surgical staff in treating patients with persistent or recurrent pain after infarction in addition to selection of good candidates for surgical therapy. Our incidence of single-vessel disease was higher than that reported by Levine [7]/ Brundage [18], and their co-workers in which hospital mortality after revascularization was 8.8% and 9%/ respectively. In the study of Levine and colleagues [7], preoperative IABP was used in 72% of patients. Brundage and associates [18] reported a series of 22 patients having revascularization an average of 15.8 days after acute MI in which triple-vessel disease was present in 82%/ the mean age of the patient population was 61.4 years, and there was a high incidence of left main coronary artery stenosis. Whereas IABP was used by Levine and colleagues [7] for relief of angina and hemodynamic instability and for hemodynamic instability in the report of Brundage and associates [18], our use of this device before operation has been very limited, confined primarily to a few patients with intractable pain until the patient could undergo myocardial revascularization. Intraaortic balloon pumping has been used in some patients prior to cardiopulmonary bypass to reduce or terminate refractory ventricular arrhythmias, minimize ischemic trauma associated with anesthetic induction, and reduce left ventricular work associated with return of cardiac function at the conclusion of the operative procedure. Although intraaortic balloon support may be advantageous in stabilizing the patient's condition before cardiac catheterization, it was not used in any of our patients. We agree with others [27/ 33/ 34] that cardiac catheterization can be performed safely in patients soon after MI. The three groups of patients in this study were arbitrarily defined according to the infarction-revascularization interval to discern if hospital morbidity or mortality could be correlated with the temporal relationship between infarction and operation. In our initial series of 35 patients reported in 1978 [15], it appeared that difficulty in the perioperative period was unrelated to this interval. As the series was extended/ however, it was found that the incidence of postoperative inotropic requirements, IABP, ventricular arrhythmias, and perioperative infarction was found to be more frequent in those patients having myocardial revascularization in the first week after infarction. Although difficulty in the postoperative period appears to be increased in patients operated on within a week of infarction, hospital mortality and long-term survival do not appear to be affected adversely. Use of inotropic drugs or IABP in the postoperative period could not be predicted prior to operation by preoperative descriptors such as extent of vessel disease, ejection fraction, or angiographically documented wall motion abnormalities. Although the follow-up period is short and many of the patients had operation recently, we

10 42 The Annals of Thoracic Surgery Vol 32 No 1 July 1981 believe our major objectives of therapy have been realized: there have been no hospital deaths and only 5 late deaths with an 18-month actuarial survival of 97%. Although all but 1 patient dying late had preoperative angiegraphic left ventricular wall motion abnormality, preoperative descriptors such as multi-vessel disease and depressed ejection fraction were not predictive of poor late survival. Only a small percentage of patients had an unfavorable result, and the patency rate for most patients was gratifying. Our findings of a high patency rate in grafts performed to the area of infarction agree with the clinical studies of others [16, 17, 34]. Left ventricular function was preserved and frequently improved as demonstrated by postoperative changes in ejection fraction and left ventricular segmental wall motion [15]. However, this study is not proof that myocardial revascularization following infarction improves left ventricular function and regional wall motion. It is well known that during an acute ischemic event, with or without infarction, myocardial wall motion and ejection fraction can be markedly depressed but can be improved considerably once the initial insult has passed. This finding is true regardless of whether patients are treated medically or surgically. Although conclusive data regarding the best mode of therapy in this particular patient population would best be obtained by a medical and surgical randomized study, we have attempted to show that adequate revascularization can be accomplished safely and effectively in a group of patients previously deemed at very high operative risk. Improvement in anginal status in this series was practically identical to that of patients having elective operation. It might be argued that more persistent medical therapy would be effective in temporarily controlling ventricular arrhythmias, pain, or even reduced cardiac output. However, we believe that revascularization of the ischemic and injured myocardium can be accomplished effectively with low operative risk and offers the best chance of minimizing left ventricular muscle damage and patient mortality. This study suggests that the frequency of perioperative complications will be increased in patients operated on within a week of infarction but after this period, revascularization can be accomplished with the same morbidity as that of elective operation. Patients with infarction should not be treated medically for an arbitrary period of time before operation is considered because this will only enhance the chances of irreparable ventricular damage from infarct extension. Each patient must be evaluated individually with all clinical factors known to the responsible physicians. If these recommendations are followed, we believe that myocardial revascularization can be accomplished with very low mortality, morbidity, and satisfactory long-term survival. References 1. Madigan NP, Rutherford BD, Frye RL: The clinical course, early prognosis and coronary anatomy of subendocardial infarction. Am J Med 60:634, Levy WK, Cannorn DS, Cohen LS: Prognosis of subendocardial myocardial infarction. Circulation 51,52:Suppl 2:107, Fraker RD Jr, Wagner GS, Rosati RA: Extension of myocardial infarction: incidence and prognosis. Circulation 60:1126, Reid RR, Taylor DR, Kelly DT, et al: Myocardial infarct extension detection by precordial ST-T segment mapping. N Engl J Med 290:123, Kronenberg MK, Hodges M, Akiyama T, et al: S-T segment variations after acute myocardial infarction: relationship to clinical status. Circulation 54:756, Schuster EH, Bulkley BH: Ischemia at a distance after acute myocardial infarction: a cause of early postinfarction angina. Circulation 62:509, Levine FH, Gold HK, Leinbach RC, et al: Safe early revascularization for continuing ischemia after acute myocardial infarction. Circulation 60:1-5, Rogers WI, Smith LR, Oberman A, et al: Surgical vs nonsurgical management of patients after myocardial infarction. Circulation 62:1-67, Madigan NP, Rutherford BD, Barnhorst DA, Danielson GK: Early saphenous vein grafting after subendocardial infarction: immediate surgical results and late prognosis. Circulation 56:Suppl 2:1, Cronin RFP, Moore S, Marpole DG: Shock following myocardial infarction: a clinical survey of 140 cases. Can Med Assoc J 93:57, Killip I, Kimball JJ: Treatment of myocardial in-

11 43 Jones et al: Coronary Bypass for Pain Relief after MI farction in a coronary care unit: a two-year experience with 250 patients. Am J Cardiol 290: 457, 196? 12. Nielsen BL, Marner IL: Shock in acute myocardial infarction. Acta Med Scand 175:65, Scheidt S, Ascheim R, Killip T: Cardiogenic shock after acute myocardial infarction. Am J CardioI26:556, Lown B, Vassaux G, Hoop SB, et al: Unresolved problems in coronary care. Am J Cardiol 20:494, Jones EL, Douglas JS [r, Craver JM, et al: Results of coronary revascularization in patients with recent myocardial infarction. J Thorac Cardiovasc Surg 76:545, Berg R [r, Kendall RW, Duvoisin GE, et al: Acute myocardial infarction: a surgical emergency. J Thorac Cardiovasc Surg 70:432, Phillips SJ, Kongtahworn C, Zef RH, et al: Emergency coronary artery revascularization: a possible therapy for acute myocardial infarction. Circulation 60:241, Brundage BH, Ullyot DJ, Winokur S, et al: The role of aortic balloon pumping in postinfarction angina: a different perspective. Circulation 62:1 119, Schoonmaker FW, King SB III: Coronary arteriography by the single catheter percutaneous femoral technique: experience with 6,800 cases. Circulation 50:735, King SB III, Douglas JS Jr: Coronary arteriography and left ventriculography. In Hurst JW, Logue RB, Schlant RC, Wenger NK (eds): The Heart, Arteries, and Veins. Fourth Edition. New York, McGraw-Hill, Sandler H, Dodge HT: The use of single plane angiograms for the calculation of left ventricular volume in man. Am Heart J 75:325, Jones EL: Coronary artery bypass grafting: simplification and refinement of surgical technique. Ann Thorac Surg 30:84, Stenson RE, Flamm MD Jr, Zaret BL, McGowan RL: Transient ST-segment elevation with postmyocardial infarction angina: prognostic significance. Am Heart J 89:449, Bardet J, Rigaud M, Kahn JC, et al: Treatment of post-myocardial infarction angina by intraaortic balloon pumping and emergency revascularization. J Thorac Cardiovasc Surg 74:299, Kongtahworn C, Zeff RH, Iannone L, et al: Emergency myocardial revascularization during acute evolving myocardial infarction. Chest 72:403, Mundth ED, Buckley MJ, Leinbach RC, et al: Myocardial revascularization for the treatment of cardiogenic shock complicating acute myocardial infarction. Surgery 70:78, Cheanvechai C, Effler DB, Loop FD, et al: Emergency myocardial revascularization. Am J Cardiol 32:901, Keon WJ, Bedard P, Shankar KR, et al: Experience with emergency aortocoronary bypass grafts in the presence of acute myocardial infarction. Circulation 47,48:Suppl 3:151, Hill JD, Kerth WJ, Kelly JJ, et al: Emergency aortocoronary bypass for impending or extending myocardial infarction. Circulation 43,44:Suppl 1:105, Dawson JT, Hall RJ, Hallman GL, Cooley DA: Mortality in patients undergoing coronary artery bypass surgery after myocardial infarction. Am J Cardiol 33:483, Johnson WD, Flemma RJ, Lepley D Jr: Direct coronary surgery utilizing multiple-vein bypass grafts. Ann Thorac Surg 9:436, Loop FD, Cheanvechai C, Sheldon WC, et al: Early myocardial revascularization during acute myocardial infarction. Chest 66:478, Cohn LH, Godin R, Herman MV, Collins JJ [r: Aortocoronary bypass for acute coronary occlusion. J Thorac Cardiovasc Surg 64:503, Begg RF, Kooros MA, Magovern GJ, et al: The hemodynamics and coronary arteriography patterns during acute myocardial infarction. J Thorac Cardiovasc Surg 58:647, 1969