Outcome of coronary endarterectomy: a case-control study

Background . Despite early reports showing a high incidence of postoperative morbidity and mortality, coronary endarterectomy continues to be used as an adjunct to coronary artery bypass grafting, particularly in diffusely diseased coronary arteries. The changing nature of patients and improvements in modern cardiac surgery call for a reevaluation of the role of coronary endarterectomy.

 Methods . Data from the 56 patients, who underwent coronary endarterectomy in our institution between January 1993 and August 1996, were reviewed retrospectively and compared with a control group of 56 patients matched for age, sex, LV function, and angina class. In the endarterectomy group, there were 47 men and nine women, with a mean age of 59.6 years. The mean follow-up time was 21 months. Indications for operation were angina in 45 (80.3%), angina with signs or symptoms of cardiac failure in 3 (5.4%), and prognosis in 8 (14.3%) asymptomatic patients.

 Results . Fifty (89%) patients had one, four (7.2%) had two, and 2 (3.6%) patients had three coronary arteries endarterectomized. Of these 64 endarterectomies, 33 (51.5%) involved the right coronary artery, 20 (31.3%) the left anterior descending artery, and 11 (17.2%) branches of the circumflex artery. There were three (5.4%) nonfatal myocardial infarctions in the endarterectomy group, but none in the control group ( p > 0.05). Two patients (3.6%) in the endarterectomy group, but none in the control group, died within the first 30 days ( p > 0.05). Actuarial survival and incidence of recurrent angina were similar in the two groups.

 Conclusions . In current cardiac surgical practice, coronary endarterectomy displays satisfactory rates of postoperative morbidity and medium term results in selected groups of patients.

 

 

 

 

 

 

 

Forty years ago, coronary endarterectomy was introduced as a method of treating coronary artery disease and was performed initially without cardiopulmonary bypass or associated coronary artery bypass grafting (CABG) [1] . The use of coronary endarterectomy, frequently as an adjunct to CABG, was adopted by several institutions in the early years after this initial report. Although the procedure was shown to produce relief from angina, early studies reflecting the experience in the 1960s and 1970s recorded an increased postoperative morbidity and early mortality. It followed, therefore, that the role of coronary endarterectomy became controversial [2] [3] [4] [5] . More recent reports suggest that, since the 1980s, the procedure is being performed in a much smaller proportion of patients undergoing CABG [6] [7] [8] in comparison to earlier decades [9] . Although these studies also reported encouraging results in terms of survival and morbidity, the main indication for coronary endarterectomy remains the presence of diffusely diseased coronary arteries that are unsuitable for distal grafting. Improvements in the management of cardiopulmonary bypass and the ischemic myocardium, together with the increasing referral of diffusely diseased coronary arteries for CABG do, however, call for a reevaluation of the role of coronary endarterectomy.

 

The aim of the present study is to evaluate the outcome of coronary endarterectomy, in our institution, performed in association with primary CABG in the mid 1990s.

 

Patients and methods

 

The study group

 

Between January 1993 and August 1996, 1366 patients underwent isolated primary CABG in our institution. Of these, 56 (4.1%) patients required at least one coronary endarterectomy. In view of the previous literature reports, we are conservative about the indications for coronary endarterectomy in our institution. It is only performed on occluded or nearly occluded vessels that have long, multiple stenoses that extend distally. In the absence of widely accepted preoperative criteria, the decision to endarterectomize a coronary artery is made intraoperatively, and is based on technical considerations in combination with information given by the coronary angiogram. Although the pattern of atherosclerosis can be assessed based on the coronary angiogram, we find that preoperative prediction of the patients who will need coronary endarterectomy is rarely possible.

 

Data concerning the hospital stay of the endarterectomized patients were extracted from our prospective computerized clinical database. In addition, the clinical notes of these patients were reviewed retrospectively. When the information provided by the clinical notes was not adequate, the patient's general practitioner was contacted by telephone or mail.

 

[Table 1] summarizes the preoperative clinical features of the study patients. The mean follow-up time was 21 months (range 2−48 months), amounting to a total of 100.5 patient years. Indications for operation were angina in 45 (80.3%) patients, angina with signs or symptoms of cardiac failure in 3 (5.4%), and prognosis in 8 (14.3%) asymptomatic patients.

 

Control group

 

For every patient who had isolated primary CABG and underwent coronary endarterectomy, one non-endarterectomized patient who underwent isolated primary CABG, matched for age, sex, and, when possible, for LV function and angina class, was selected. Data of the patients comprising the control group were obtained as described for the study group. The preoperative clinical data of the control group are also shown in . There were no significant differences between the two groups with regard to the preoperative features of the patients.

 

For comparison, preoperative clinical information concerning the total group of 1366 patients who underwent isolated primary CABG during the study period is also shown in .

 

Surgical technique

 

Cardiopulmonary bypass was sustained for all operations using a hollow fiber membrane oxygenator (Bard William Harvey, Haverhill, MA). In the endarterectomy group, myocardial management was achieved with intermittent ischemia and ventricular fibrillation in 29 (51.8%), antegrade warm (32°C) blood cardioplegia in 14 (25%), antegrade crystalloid cardioplegia (St Thomas') in 8 (14.3%), and antegrade cold (6°C) blood cardioplegia in 5 (8.9%) patients. The frequency of the different myocardial management techniques was comparable in the control group ( [Table 2] ). Core temperature during cardiopulmonary bypass was maintained at moderate hypothermia between 30° and 32°C. None of the patients received aprotinin or amicar during the operation.

 

All endarterectomies were performed manually and the conduit was anastomosed to the endarterectomized artery, end to side, in all cases. The artery was opened at a chosen site and endarterectomy was only performed when the artery was almost completely or completely occluded or the vessel wall contained heavily calcified plaques. The arteriotomy was usually approximately 15 mm in length, but was extended up to 40 mm in a few cases. A Watson-Cheyne dissector was used to develop a plane between the media and the core of the atheroma. Gentle traction was then used to extract the atheroma proximally. Attention was then turned to the distal artery and, with graduated traction, once again the atheromatous core was teased out of the vessel. In this case, however, care was taken to ensure that the distal extracted tip of the atheroma had a tapered end. If this was not the case, the arteriotomy was extended distally until a satisfactory result was achieved. No attempt was made to guarantee the extraction of atheromatous cores from all the individual branches of the artery.

 

Peri- or postoperative myocardial infarction was defined by persistent electrocardiographic changes such as new Q waves, loss of R wave progression, new intraventricular conduction defect or by new echocardiographic evidence of wall kinetic abnormality. All patients received a routine evaluation with electrocardiogram immediately postoperatively and on days two and five. In the presence of suspected or proved myocardial infarction, patients received several electrocardiograms and echocardiography. Serial postoperative creatine kinase levels were not routinely obtained.

 

Statistical analysis

 

Nominal data were analyzed using <Рисунок: chi> 2 test and interval data using t test. Actuarial survival rates were calculated by Kaplan-Meier survival analysis and actuarial survival curves were compared by log-rank test. Statistical significance was determined by p < 0.05.

 

Results

 

The number of grafts in the two groups is shown in . Fifty (89%) patients had one, 4 (7.2%) patients had two, and 2 (3.6%) patients had three coronary arteries endarterectomized. Of these 64 endarterectomies, 33 (51.5%) involved the right coronary artery (RCA), 20 (31.3%) the left anterior descending (LAD) artery, and 11 (17.2%) the major obtuse marginal branch of the circumflex artery.

 

Early morbidity

 

The hospital complications observed are presented in . [Table 3] There were three (5.3%) nonfatal myocardial infarctions in the endarterectomy group, but none in the control group ( p > 0.05).

 

The endarterectomy group had a higher incidence of pleural effusions requiring draining ( p = 0.04). The rate of postoperative dysrhythmias, nonfatal strokes, and sternal wound infections requiring treatment with antibiotics was similar between the two groups ( p > 0.05).

 

Early mortality

 

There were no early deaths in the control group, but two patients (3.6%) died within the first 30 postoperative days in the endarterectomy group ( p > 0.05). One of these patients was a 64−year-old woman with poor LV function (EF < 0.3). The insertion of an intra aortic balloon pump was required preoperatively due to low cardiac output. After completion of the operation, she developed ventricular fibrillation that responded to internal cardiac massage and cardioversion. The patient was transferred to the intensive care unit (ICU) on a high dose of inotropic support, but never regained consciousness and died on the 11th postoperative day.

 

The other early death occurred in a 77−year-old man, also with poor LV function (EF < 0.3), who underwent urgent simultaneous CABG with repair of an abdominal aortic aneurysm. He required abdominal reexploration for bleeding from the proximal anastomosis of the aortic graft, subsequently developed low cardiac output syndrome and renal failure, and died 48 hours postoperatively.

 

The endarterectomy group displayed similar early mortality (3.6%) to the total CABG population in our hospital (3.2%).

 

Actuarial survival and follow-up

 

There was no significant difference in the actuarial survival between the two groups ( [Figure 1] ). Three of the patients who underwent coronary endarterectomy died after discharge from hospital. One of these patients was a 45−year-old woman from overseas, suffering from renal failure requiring hemodialysis, who died twelve months postoperatively, after she returned to her home country. Sepsis was reported as the cause of death. A 79−year-old woman died of large bowel perforation 14 months after cardiac surgery. The only late death that was due to cardiac reasons occurred in a 65−year-old male patient with known preoperative poor left ventricular function, who died of cardiac failure three months postoperatively.

 

Of the 51 surviving patients, 38 (74.5%) remained free from angina throughout the follow-up time. In the control group, there were two deaths after hospital discharge and 38 (70.3%) of the 54 surviving patients reported freedom from angina at follow-up ( p > 0.05).

 

In the endarterectomy group, three patients underwent recatheterization postoperatively. A 40−year-old female developed recurrent angina 18 months after revascularization of the obtuse marginal and right coronary arteries with endarterectomy of the right coronary artery. Coronary catheterization revealed complete occlusion of the right coronary artery venous graft. A 50−year-old male was recatheterized, for recurrent angina, one year after CABG. However, all grafts and native coronary arteries were patent. A 46−year-old male underwent recatheterization for recurrent angina, 18 months after CABG with endarterectomy of the right coronary artery. This patent but progressive disease was demonstrated in the left anterior descending artery. There were no reoperations during the study period.

 

Comment

 

The rationale for performing coronary endarterectomy is to undertake as complete a revascularization as possible in coronary arteries with diffuse flow-limiting atherosclerotic stenoses [6] . Early reports of a high incidence of postoperative morbidity and mortality [2] [3] [4] [5] and the introduction of alternative coronary interventional procedures such as angioplasty, stenting, and atherectomy [10] are potential explanations for the fact that endarterectomy is currently performed more infrequently [7] [9] in comparison to earlier series [6] [8] [11] in many centers. Although refinements in technique, such as long arteriotomy and vein patch, have been described [12] [13] , coronary endarterectomy remains a rather crude surgical technique and has not found wide acceptance. Reduced postoperative morbidity and mortality after cardiac surgery, as a consequence of recent advances in the management of cardiopulmonary bypass and the myocardium, and the successful management of one and two vessel coronary artery disease through interventional cardiology, has resulted in an increasing number of patients with diffuse coronary artery disease being referred for CABG. There is, therefore, a need to reassess the early and medium term postoperative outcomes in patients undergoing primary CABG with coronary endarterectomy in modern cardiac surgery.

 

Preoperative characteristics

 

The patients who underwent coronary endarterectomy in this study were not different in preoperative terms to the control group or the total CABG group. The higher incidence of preoperative myocardial infarctions among the endarterectomized patients might reflect the fact that coronary artery disease is more diffuse in this group, even though this did not result in poorer LV function.

 

Early mortality and morbidity

 

The incidence of early (most commonly 30 day) mortality reported in previous studies has been consistently between 2%−8%, throughout the last fifteen years [6] [7] [9] [11] [13] [14] [15] [16] [17] [18] [19] . These figures were higher than in patients undergoing CABG without coronary endarterectomy in the same institutions but, nevertheless, demonstrate a significant reduction in mortality in comparison to the early years of endarterectomy. Recently, Djalilian and Shumway [7] reported two (3.1%) early deaths in 64 patients who underwent coronary endarterectomy, between 1988 and 1992. Our operative mortality of 3.6% in the endarterectomized group is higher when compared to the control group, but the difference is not statistically significant. This could reflect improved operative outcomes with contemporary management strategies. Early mortality was also reported to be higher after endarterectomy of the LAD artery [9] and in patients undergoing endarterectomy in more than one coronary artery [6] . Although only 31.3% of our patients had LAD artery endarterectomies, and although the number of multiple endarterectomies was low (11%), both our early deaths followed endarterectomy in more than one vessel and both included endarterectomy of the LAD artery. In line with previous experience, our strategy is to avoid endarterectomy of the LAD artery except in those situations where it is essential for technical reasons. Although it was not a significant predictor of mortality in our series, the fact that both our deaths were in LAD artery endarterectomized patients means that we shall continue to attempt to avoid this particular procedure.

 

Myocardial infarction is the main perioperative complication in patients undergoing coronary endarterectomy with a reported incidence from 1.5%−19% [6] [7] [9] [13] [14] [15] [18] , probably depending on the diagnostic criteria. In our study, three (5.6%) of the 56 endarterectomized patients, but none in the control group, suffered a non-fatal perioperative myocardial infarction. These low mortality and morbidity rates and small patient numbers preclude any reliable univariate analysis of the risk factors associated with postoperative early mortality and myocardial infarction. Patients suffering other causes of early morbidity, such as sternal wound infection and stroke, had a prolonged hospital stay, but eventually enjoyed full recovery. Overall, the two groups displayed similar rates of early morbidity, with the exception of pleural effusions requiring drainage, which were more common among endarterectomized patients ( p = 0.027). This difference is not related to the duration of CPB ( p > 0.05) or any other clinical parameters and is probably due to statistical error Type II, because endarterectomy is unlikely to be the cause of increased intrapleural hemorrhage. The incidence of sternal wound infection requiring treatment remained low and did not reach a statistically significant difference between the two groups.

 

Actuarial survival

 

Previous reports displayed actuarial survival rates of 96.7% at one year [15] and 71%−92% at five years after coronary endarterectomy [6] [7] [9] [11] [15] [18] . Multiple endarterectomies have been shown to be associated with shorter long-term survival [6] . The 90% 4−year survival rate in this study is in agreement with these earlier findings. Our control group did slightly, but not significantly, better.

 

Follow-up

 

The 74.5% postoperative freedom from angina at 21 months mean follow-up time is comparable to previous reports [6] [9] [18] , although Sommerhaug and colleagues . [13] showed 99% of their patients remaining in angina class I at a mean time of 20 months, after long coronary arteriotomy, endarterectomy, and reconstruction of the vessel. According to previous reports employing postoperative recatheterization, early (within 12 months following operation) patency rates of endarterectomized coronary arteries vary from 66% to 90% [7] [8] [11] [14] . Late (after 12 months following operation) patency rates are reported between 56% and 75%, depending on the time span between CABG and recatheterization [7] [18] [11] [14] [15] [16] . Recatheterization following endarterectomy is not routine in our institution and the small number of patients who underwent repeat catheterization after endarterectomy does not allow us to comment on our patency rates. It is encouraging, however, that only three patients had recurrent angina severe enough to justify repeat catheterization, although freedom from angina does not necessarily imply patency of grafts and native vessels.

 

The anticoagulation protocol after coronary endarterectomy has evolved in our institution from exclusive use of warfarin (to achieve an international normal ratio of 2.5 — 3.5), to exclusive early (300 mg daily starting within eight hours following operation) use of aspirin, to currently employing both warfarin (INR at 2.0 mg) and early low dose aspirin (75−150 mg daily) for a period of three months. The strategy of inhibiting both arms of thrombus formation may produce further improvement in the outcome following endarterectomy and result in lower medium-term mortality in these high-risk patients. After the first three postoperative months, the administration of warfarin is discontinued, whereas aspirin, in doses of 150−300 mg daily is given indefinitely.

 

In summary, the present study demonstrates that coronary endarterectomy remains a procedure performed in a small percentage of patients undergoing primary CABG with the right coronary artery as the most commonly endarterectomized vessel. The procedure has a low rate of postoperative mortality and morbidity, and relief from angina is obtained in the majority of patients. As the long-term outcome of coronary endarterectomy is known to be moderate [6] [9] , these satisfactory medium- term results do not support the universal use of coronary endarterectomy, but suggest that there is potential benefit associated with this procedure in an increasing number of patients with advanced complex coronary artery disease.

 

References

 

 

 

 

 

 

1.Bailey C.P., May A., Lemmon W.M.. Survival after coronary endarterectomy in man. JAMA 1957;164:641−646.

 

2.Sawyer P.N., Kaplitt M., Sobel S.. Experimental and clinical experience with coronary gas endarterectomy. Arch Surg 1967;95:736−742.

 

3.Kaplitt M.J., Robinson G.. Coronary gas endarterectomy. Am Heart J 1971;81:136−140.

 

4.Sabiston D.C., Ebert P.A., Friesinger G.C.. Proximal endarterectomy, arterial reconstruction for coronary occlusion at aortic origin. Arch Surg 1965;91:758- .

 

5.Effler D.B., Groves L.K., Sone F.M.. Endarterectomy in the treatment of coronary artery disease. J Thorac Cardiovasc Surg 1964;47:98−108.

 

6.Brenowitz J.B., Kayser K.L., Johnson W.D.. Results of coronary artery endarterectomy and reconstruction. J Thorac Cardiovasc Surg 1988;95:1−10.

 

7.Djalilian A.R., Shumway S.J.. Adjunctive coronary endarterectomy. Ann Thorac Surg 1995;60:1749−1754.

 

8.Hochberg M.S., Merrill W.H., Michaelis L.L.. Results of combined coronary endarterectomy and coronary bypass for diffuse coronary artery disease. J Thorac Cardiovasc Surg 1978;75:38−46.

 

9.Livesay J.J., Cooley D.A., Hallman G.L.. Early and late results of coronary endarterectomy. J Thorac Cardiovasc Surg 1986;92:649−660.

 

10.Lau K.W., Sigwart U.. Novel coronary interventional devices. American Heart J 1992;123:497−506.

 

11.Qureshi S.A., Halim M.A., Pillai R.. Endarterectomy of the left coronary system. Analysis of a 10 year experience. J Thorac Cardiovasc Surg 1985;89:852−859.

 

12.Ladowski J.S., Schatzlein M.H., Underhill D.J.. Endarterectomy, vein patch, and mammary bypass of the anterior descending artery. Ann Thorac Surg 1991;52:1187−1189.

 

13.Sommerhaug R.G., Wolfe S.F., Reid D.A.. Early clinical results of long coronary arteriotomy, endarterectomy and reconstruction combined with multiple bypass grafting for severe coronary artery disease. Am J Cardiol 1990;66:651−659.

 

14.Goldstein J., Cooper E., Saltups A.. Angiographic assessment of graft patency after coronary endarterectomy. J Thorac Cardiovasc Surg 1991;102:539−545.

 

15.Kay P.H., Brooks N., Magee P.. Bypass grafting to the right coronary artery with and without endarterectomy. Br Heart J 1985;54:489−494.

 

16.Minale C., Nikol S., Zander M.. Controversial aspects of coronary endarterectomy. Ann Thorac Surg 1989;48:235−241.

 

17.Shapira N., Lumia F.J., Gottdiener J.S.. Adjunct endarterectomy of the left anterior descending coronary artery. Ann Thorac Surg 1988;46:289−296.

 

18.Ivert T., Welti R., Forssell G.. Coronary endarterectomy- angiographic and clinical results. Scand J Thorac Cardiovasc Surg 1989;23:95−102.

 

19.Wallsh E., Franzone A.J., Clauss R.H.. Manual coronary endarterectomy with saphenous bypass. Ann Thorac Surg 1981;32:451−457.