Myocardial infarction complicating aortic valve replacement

Myocardial infarction complicating aortic valve replacement One hundred consecutive aortic valve replacements were studied. Fifteen patients had a myocardial infarction as a result of the operation, and four of the five deaths in the series stemmed from this group. In the four deaths from infarction, autopsy revealed occlusion of a main coronary artery. This was attributable to coronary perfusion in three instances. All of the 11 survivors who sustained an infarct were free of angina and left ventricular failure 6 weeks after the operation. Patients with infarcts had longer bypass times and larger aortic systolic gradients than the patients who did not have an infarct. It is suggested that an infarct can occur as the result of occlusion of a main coronary artery; this is a fatal event commonly related to trauma from the coronary perfusion cannula. Alternatively, infarction may result from regional ischemia, perhaps without vessel occlusion, and is associated with long bypass times and with large aortic valve gradients. In such cases the prognosis is good. However, myocardial infarction was the major cause of death in this series.

G. P. Sharratt, M.D., M.R.C.P.,* P. Rees, M.B., B.S., M.R.C.P.,** and N. Conway, M.B., B.S., F.R.C.P.,*** Southampton, England

.L/uring aortic valve replacement the aorta is crossclamped and the coronary arteries are commonly cannulated and perfused. However, the operation is performed in an age group in which coronary artery disease is prevalent, and an eighteen per cent incidence of myocardial infarction has been reported1 following aortic valve replacement. This investigation was undertaken to assess the incidence of myocardial infarction as a complication of aortic valve replacement, to determine its effect on the prognosis, and to investigate the factors related to its occurrence. Clinical material and methods One hundred consecutive aortic valve replacements were reviewed. Four of the patients also had replacement of the ascending aorta, but no patient had any other valve surgery or saphenous vein coronary artery From the Wessex Regional Cardiothoracic Centre, Western Hospital, Southampton, England. Received for publication Sept. 29, 1975. Accepted for publication Jan. 14, 1976. Address for reprints: G. P. Sharratt, Western Hospital, Oakley Road, Southampton, S09 4WQ, England. *Lecturer in Medicine, Southampton University. "Registrar in Cardiology, Hammersmith Hospital, London, England. ***Senior Lecturer in Cardiology, Southampton University.

bypass grafts performed. There were 24 women and 78 men with a mean age of 53.4 years (range 16 to 73). Thirty-eight patients had dominant aortic stenosis, 23 patients had pure aortic incompetence, and 39 patients had mixed stenosis and incompetence. Thirty patients had a homograft valve inserted and 70 patients a prosthetic valve, of which there were 35 StarrEdwards, 11 Bjork-Shiley, 11 Braunwald-Cutter, and 8 Lillehei-Kaster valves. The duration of left and right coronary ischemia was calculated as the total time during which the aorta was cross-clamped and the respective coronary artery was not perfused. The standard twelve-lead electrocardiogram (ECG) was reviewed preoperatively, immediately postoperatively, during the recovery period, and 6 weeks after the operation. A myocardial infarction was considered to have occurred if the postoperative electrocardiogram showed the development of new pathological Q waves. Results Ten patients had evidence of a myocardial infarction on the ECG taken in the immediate postoperative period. Eleven patients developed left bundle branch block (LBBB) postoperatively. This problem resolved in 10, and, when normal conduction had returned, 3 patients were found to have the ECG changes of a myocardial infarction. Because LBBB persisted in one 869

8 70

The Journal of Thoracic and Cardiovascular Surgery

Sharratt, Rees, Conway

Table I. Comparison of patients with and patients without a myocardial infarction complicating aortic valve replacement Patients with preop. angina

Myocardial infarction No myocardial infarction Significance of difference

Patients with heavy calcification

%

No.

%

No.

Aortic valve gradient

60

9

80

12

92 ± 38

116 ± 3 5

45

39

50

42

66 ± 40

98 ± 27

p > 0.05

p < 0.05

p < 0.05

p > 0.05

Bypass time (min.)

patient, it is not known if he had sustained an infarct. Eleven patients had LBBB on the preoperative ECG. This abnormality resolved postoperatively in 5 patients, and in these there was no evidence of myocardial infarction. In the other 6 LBBB persisted; thus it is not known if a myocardial infarction had occurred preoperatively or postoperatively. Two patients died during surgery, and at autopsy a myocardial infarction was present. Although no ECG was available for these patients, they were included in the infarct group. We therefore considered that 15 patients had had a myocardial infarction as a result of their operation. Five of the 100 patients studied died before postoperative week 6. Four of them were among the 15 patients with a myocardial infarction, a mortality rate of 27 per cent. Only one of the patients who died did not have a myocardial infarction, a mortality rate of 1.2 per cent. This man had active rheumatoid aortitis, and the aortic suture line ruptured on the 15 postoperative day. Of the four deaths associated with myocardial infarction, two occurred during surgery. At autopsy there was occlusion of the main left coronary artery, by thrombus in one case and dissection in the other. One patient died 24 hours after valve replacement, and at autopsy there was occlusion of the right coronary artery by an intimal hemorrhage 2 cm. from its ostium. The fourth death occurred on the tenth postoperative day and was due to extensive anterior infarction which followed dissection of the aorta produced at operation by the aortic cannula. The aorta was medionecrotic. Of the 11 patients who had a myocardial infarction and survived, 8 had inferior infarcts and 3 had anterior infarcts. All of these 11 patients were well, with no

angina or clinical or radiological evidence of left ventricular failure 6 weeks after the operation. On comparing the 15 patients who had a myocardial infarction with those who had no infarction, we found no difference between the mean age, sex distribution, incidence of cigarette smoking, or preoperative diastolic hypertension. The percentage of homograft valve replacements was equal in the two groups, and the preoperative cardiac indices and left ventricular enddiastolic pressures were comparable. Preoperative angina was more frequent in the patients who sustained an infarct, as was heavy valve calcification, but the differences were not significant (Table I). Ninety-three per cent of the infarct group had some degree of aortic stenosis as compared to 76 per cent of the noninfarct group, and the peak-to-peak aortic systolic gradient was significantly greater in the infarct group (Table I). The bypass time in the patients who sustained an infarct was significantly longer than that in those who did not (Table I). There was no difference between the right and left coronary ischemic times of the two groups, except that the right coronary ischemic time in the patients who sustained an inferior infarct was 47 ± 36 minutes and that in the patients who had no myocardial infarction was 34 ± 30 minutes. This difference does not, however, achieve statistical significance. Preoperative coronary arteriography was performed in only 16 patients. Three patients who sustained a myocardial infarction had normal coronary arteriograms. Discussion Myocardial infarction during aortic valve replacement appears to be a not uncommon complication. It occurred in 15 per cent of this series and accounted for four of the five deaths. These four deaths were due to occlusion of the main left coronary artery in 3 and of the main right coronary artery in one. In 2 cases there was local dissection of the coronary artery, presumably related to trauma from the coronary artery cannula. This complication is reported to be a universally fatal event.2 In a third case there was thrombus at the origin of the left coronary artery, without evidence of dissection, and in the fourth case an aortic dissection had occluded the left coronary artery. The 11 patients who had a myocardial infarction and survived seem to be a different group. They all recovered uneventfully and are free of both angina and left ventricular failure. Inferior infarcts predominate, and this would be unusual if they were atherosclerotic in origin. When atherosclerosis is the underlying cause,

Volume 71 Number 6 June, 1976

approximately 55 per cent of the infarcts are anterior and 30 per cent inferior.3 Coronary risk factors are not relevant, because there was no difference in the age, sex, smoking habits, or the incidence of diastolic hypertension between patients who developed an infarct and those who did not. However, those developing infarcts did have significantly longer bypass times and more severe aortic stenosis. There was also a greater incidence of preoperative angina and heavy valve calcification in the infarct group, and the patients who sustained an inferior infarct had had long periods of right coronary artery ischemia. We postulate that myocardial infarction can occur during aortic valve replacement in one of two ways. Uncommonly, it can occur as a result of occlusion of a main coronary artery during the operation. This complication is fatal, although presumably prompt recognition and bypass grafting may allow survival. More usually, however, myocardial infarction can occur in a relatively benign form. The infarct may be due to local muscle necrosis, possibly without vessel occlusion, and this has been demonstrated in one of our cases by postoperative coronary arteriography. Long bypass periods, particularly with inadequate coronary perfusion, and large aortic valve gradients appear to be contributory factors. The association of large valve gradients may be explained by greater left ventricular hypertrophy, although we recognize that the aortic valve gradient alone is an inadequate measure of the severity of stenosis and the degree of left ventricular hypertrophy. LBBB is well recognized as a complication of aortic valve disease and aortic valve replacement.1' 4' 5 Preoperative LBBB resolved postoperatively in some

Aortic valve replacement

871

patients in this series. Although it is not unreasonable to expect LBBB to resolve, we do not believe that this has been widely documented.1 More important, we have found that LBBB is usually a temporary phenomenon when produced by the operation. However, the possibility that it may be masking the ECG evidence of a myocardial infarction should not be overlooked. Myocardial infarction may thus occur in a benign form during aortic valve replacement, but it may also be a fatal event and in this series accounted for four of the five deaths. We would like to thank Dr. A. Johnson, Mr. K. Ross, and Mr. J. L. Monro for permission to study their patients and Miss D. Lansley and Miss S. Trevett for secretarial help. REFERENCES 1 Follath, F., and Ginks, W. R.: Changes in the QRS Complex After Aortic Valve Replacement, Br. Heart. J. 34: 553, 1972. 2 Bulkley, B. H., and Roberts W. C: Isolated Coronary Arterial Dissection: A Complication of Cardiac Operations, J. THORAC. CARDIOVASC. SURG. 67: 148, 1974. 3 Tucker, H. H., Carson, P. H. M., Bass, N. M., Sharratt, G. P., and Stock, J. P. P.: Results of Early Mobilisation and Discharge After Myocardial Infarction, Br. Med. J. 1: 10, 1973. 4 Kloster, F. E., Bristow, J. D., and Griswold, H. E.; Medical Problems in Mitral and Multiple Valve Replacement, Progr. Cardiovasc. Dis. 7: 504, 1965. 5 Gannon, P. G., Sellers, R. D., Kanjuh, V. I., Edwards, J. E., and Lillehei, C. W.: Complete Heart Block Following Replacement of the Aortic Valve, Circulation 33, 34: 152, 1965 (Suppl. I).