Unilateral Versus Bilateral Internal Mammary Revascularization
Survival and Event-Free Performance
Background The influence of unilateral (UL) and bilateral (BL) mammary artery revascularization, within age groups ≤60 years and >60 years, on patient survival, ischemic-related events, and interventional management was studied in 1142 patients who had coronary artery bypass graft surgery between 1984 and 1992.
Methods and Results UL revascularization was performed in 765 (67%) and BL in 377 (33%) patients with supplemental vein grafts. The overall early and hospital mortality rate was 2.7%. For UL in the age group ≤60 years, it was 1.1%; for BL ≤60 years, 1.3% (P=NS); for UL >60 years, 4.3%; and for BL >60 years, 2.8% (P=NS). Twenty-five preoperative patient characteristics representing demographics, extent of disease, concomitant disease, ventricular dysfunction, previous surgery, and status did not differentiate the patient groups (P=NS). Patient survival at 5 years was not different: 94% for UL ≤60 years, 95% for BL ≤60 years, 91% for UL >60 years, and 86% for BL >60 years (P=NS). The freedom from ischemic-related events was not different at 5 years (P=NS). The freedom from recurrent angina was 78% for UL ≤60 years, 88% for BL ≤60 years, 82% for UL >60 years, and 83% for BL >60 years (P= NS). The myocardial infarction freedom was 98% for UL ≤60 years, 96% for BL ≤60 years, 99% for UL >60 years, and 97% for BL >60 years (P=NS). The freedom from sudden unexpected death and cardiac death did not differentiate the groups (P=NS). The freedom from angioplasty and reoperation did not differentiate the groups (P=NS). The freedom from all ischemic-related and interventional events was 76% for UL ≤60 years, 84% for BL ≤60 years, 81% for UL >60 years, and 79% for BL >60 years (P=NS). A trend exists for less angina pectoris in the bilateral population ≤60 years, which reflects in the trend in the freedom from overall events.
Conclusions UL and BL mammary artery revascularizations have the same early mortality regardless of age but do not reveal any advantage for BL revascularization at 5 to 7 years.
The success of myocardial revascularization is dependent on long-term graft patency and the rate of progression of native vessel disease. The conduit of choice for revascularization of the anterior descending coronary artery has been recognized for the past decade to be the in situ left internal mammary coronary artery (IMA).1 2 3 4 5 6 7 The IMA maintains patency of 80% to 90% for 10 to 15 years.1 2 3 4 8 9 The saphenous vein (SV) graft is prone to atherosclerosis, and patency can be reduced to 50% to 60% after a decade.1 2 3 There remains conflicting opinion as to the indications for single (simple) versus double (unilateral [UL] versus bilateral [BL]) IMA revascularization.3 4 8 9 10 11 12 13 14 15 The influence on long-term survival and freedom from recurrent ischemic events and reoperation remains without clear delineation. There is the opinion that selectivity is important in the use of double IMA revascularization, with obesity, diabetes mellitus, and advanced age being relative contraindications.11 The concerns of sternal wound complications and reoperative difficulties are also expressed as reasons to avoid double arterial revascularization with IMAs.
We did not begin using BL IMA revascularization until 1984. There has been no concerted opinion or pattern of practice for the use of single or double IMA grafting. In the present study, we document our experience and provide the results of 5 and 7 years, including age evaluation, of the use of single or double IMA revascularization.
From 1984 through 1992, 1142 patients underwent myocardial revascularization with UL and BL IMA and supplemental vein grafts. There were 957 men (83.8%) and 185 women (16.2%) (mean age, 59.8 years; age range, 31.8 to 82.9 years). The patient population comprised 765 (67%) who had UL IMA and 377 (33%) who had BL IMA. Of the total, 440 (38.5%) were ≤60 years of age, and 702 (61.5%) were >60 years of age (mean age of the ≤60 group, 52.0 years; range, 31.8 to 60.0 years; mean age of the >60 group, 67.5 years; range, 60.0 to 82.9 years).
The distribution of use of IMA grafts by vessel of insertion is demonstrated for both the UL and BL populations in Tables 1⇓ and 2⇓. There were 853 IMA grafts (1.12 per patient) in the UL population of 765 patients and 823 (2.18 per patient) in the BL population of 377 patients. The total number of grafts per patient for the UL group were 3.57 per patient (≤60, 3.50; >60, 3.52) and for the BL group were 3.9 per patient (≤60, 3.79; >60, 4.0). There was a minimal number of free IMA grafts and gastroepiploic artery grafts (Table 3⇓).
Twenty-five preoperative characteristics of the patients were used to distinguish the groups. The characteristics are part of the Society of Thoracic Surgeons National Cardiac Surgery Database collection form that was used for collection of data in this retrospective evaluation. The definitions of the terms are defined and published by the Society of Thoracic Surgeons. The most important parameters of the total patient population are illustrated in Tables 4⇓ and 5⇓. Of the total, 18.8% had diabetes mellitus, 6.5% had previous coronary artery bypass graft surgery, 27.9% had left main coronary artery stenosis (>50%), and 60.8% had moderate-to-severe wall motion abnormalities with ejection fraction below 45%. Of the total, 87.0% had elective or urgent operation, whereas of the remainder, 12.6% were classified as low emergent and 0.4% as high emergent. The majority of the patients, 93.2%, were having a first operative procedure, with 6.5% having a first reoperation and 0.4% having a second reoperation.
The UL and BL patient parameters did not distinguish the groups (P=NS). The predominant 14 characteristics and percentages for both groups are shown in Table 6⇓. There were no distinguishing features between the two groups.
The patient population was divided into two subgroups: 440 (38.5%) were ≤60 years of age, and 702 (61.5%) were >60 years of age. The mean age of the ≤60 group was 52.0 years (range, 31.8 to 60.0 years) (mean: UL, 52.2 years; BL, 51.7 years) and of the >60 group was 67.5 years (range, 60.0 to 82.9 years) (mean: UL, 67.6 years; BL, 67.1 years).
Data Management and Statistical Analysis
All data were compiled and stored on the Society of Thoracic Surgeons (Summit Medical Systems Inc) computerized database. Univariate analysis (χ2 and grouped t tests) was used to compare the characteristics of the groups of patients. Multivariate logistic regression analysis was not applied because the univariate analysis had shown no statistical significance. Actuarial life tables or the Cutler-Ederer method was used to evaluate overall survival and freedom from ischemic events such as angina pectoris, myocardial infarction, and death and interventions, percutaneous transluminal coronary angioplasty (PTCA), and reoperation for repeat revascularization. A set level of significance (α=.05) was chosen for decision making.
The early mortality (including hospital mortality beyond 30 days) was 2.7% (31 of 1142 patients). There were 26 deaths (2.2%) within 30 days and 5 in-hospital deaths beyond 30 days. The overall early mortality for the group ≤60 years was 1.1% (5 patients) and for the group >60 years it was 3.7% (26 patients) (P=NS). For the group ≤60 years, the early mortality for the UL patients was 1.1% (3 patients) and for BL was 1.3% (2 patients) (P=NS). For the group >60 years, the early mortality for UL was 4.3% (20 patients) and for BL was 2.8% (6 patients) (P=NS). The predominant causes of early mortality were perioperative myocardial infarction, cerebrovascular accident, low output syndrome, cardiac arrest, and sternal infection and mediastinitis.
Patient survival by age groups for UL and BL mammary revascularization is illustrated in Fig 1⇓. The 5- and 7-year survival is indicated, showing no difference within age groups for UL and BL mammary artery use or between groups (P=NS). The 5-year survival for ≤60 years UL is 94.4±1.5% and for BL is 94.8±2.2%; the rates for >60 years are 90.8±1.7% and 85.8±3.6%, respectively, for UL and BL (P=NS).
The incidence of exploration for postoperative bleeding and tamponade was 6% (70 patients); for UL, 5.1% (39 patients); and for BL, 8.2% (31 patients) (P<.05). There were 2 deaths for which bleeding and tamponade were contributing factors.
There were 11 cases of deep sternal wound infection and mediastinitis (1.0%; UL, 8; BL, 3). In the ≤60 year group, the rate was 1.1% (3 patients) for UL and 0.6% (1 patient) for BL (P=NS). In the >60 year group, the rate was 1.0% (5 patients) for UL and 0.9% (2 patients) for BL (P=NS). Of the 11 patients, there were 2 deaths (18%).
The freedom from ischemic-related events is demonstrated in Figs 2 through 5⇓⇓⇓⇓. The freedom from recurrent angina pectoris was not different (P=NS); 77.8±3.0% for UL ≤60 years, 88.1±2.6% for BL ≤60 years, 81.8±2.2% for UL >60 years, and 83.0±3.5% for BL >60 years (Fig 2⇓). The freedom in the ≤60 year group indicated a trend toward less angina with BL rather than UL mammary revascularization.
The freedom from myocardial infarction did not differentiate the groups at 5 and 7 years (P=NS). The freedom for UL ≤60 years was 98.2±0.9%; for BL ≤60 years, 96.2±2.0%; for UL >60 years, 98.8±0.6%; and for BL >60 years, 96.8±2.0% (P=NS) (Fig 3⇑). The freedom from sudden unexpected death did not differentiate the groups (P=NS), with freedom ranging between 97.8±1.8% and 100% (Fig 4⇑).
The freedom from cardiac-related death did not reveal any difference between the groups at 5 and 7 years (P=NS) (Fig 5⇑). The freedom was 96.1±1.3% for UL ≤60 years, 98.0±1.2% for BL ≤60 years, 96.5±0.8% for UL >60 years, and 92.0±2.8% for BL >60 years at 5 years (P=NS).
The freedom for interventional management in the patient subgroups is shown in Figs 6⇓ and 7⇓. The freedom from PTCA was not different at 5 years: 91.0±2.2% for UL ≤60 years, 95.5±1.9% for BL ≤60 years, 96.3±1.1% for UL >60 years, and 98.4±0.9% for BL >60 years (P=NS) (Fig 6⇓). Reoperation was performed infrequently during the observation period; the freedom was 98.2±0.9% for UL ≤60 years, 99.4±0.6% for BL ≤60 years, 98.0±1.1% for UL >60 years, and 99.4±0.6% for BL >60 years (P=NS) (Fig 7⇓).
The demonstration of freedom from all ischemic-related and management interventional events is presented in Fig 8⇓. The freedom from overall events was 75.7±3.0% for UL ≤60 years, 83.9±3.2% for BL ≤60 years. 80.9±2.2% for UL >60 years, and 79.0±3.9% for BL >60 years at 5 years (P=NS). The trend toward reduced freedom for the bilateral revascularization group ≤60 years is related to the noted trend in greater freedom from recurrent angina pectoris.
The results for the total UL and BL mammary revascularization groups, regardless of age groupings, for patient survival, ischemic-related events, and interventional management are presented at the 5-year interval (Table 7⇓). There also was no difference considering the two populations as a whole without consideration of age groupings (P=NS).
The role of BL internal mammary revascularization has not been fully clarified. The major contributions to the surgical management of coronary artery disease with internal mammary revascularization have been provided by the Cleveland Clinic Foundation,5 10 13 15 16 17 18 19 Columbia University,4 20 St Louis University,3 9 11 21 and the Good Samaritan Hospital and Medical Center, Portland, Ore.2 We formulated the present study to determine the early to intermediate-term performance of UL and BL mammary revascularization on patient survival and prevention of ischemic-related events within age categorization of patients.
The use of arterial revascularization with the IMA became more common practice after the 1984 reports by Grondin et al1 and Okies et al.2 The report from the Montreal Heart Institute by Grondin et al1 compared survival and patency of IMA grafts with those of SV grafts at 10 years. The patient survival at 10 years for IMA grafts was 84% compared with 70% for SV grafts. The patency at 1 year was 89% and 76% and at 10 years was 84% and 53%, respectively, for IMA and SV grafts. In the same year, Okies et al2 identified the left IMA as the graft of choice to the anterior descending coronary artery in a nonrandomized study. The study revealed, at 10 years, survival rates of 82% and 69%, event-free survival rates of 61% and 48%, and patency rates of 83% and 48%, respectively, for IMA and SV grafts.2 The 5-year survival rate of Okies et al2 was 92% for IMA compared with our overall IMA survival rate of the same value. Their event-free survival was 61%2 compared with our event-free survival rate including a reoperation rate of 80%. This fact could be explained by advanced myocardial preservation techniques with minimal ventricular muscle compromise at surgery became congestive heart failure was a factor, together with recurrent angina, myocardial infarction, and death in their event-free survival rate.
Additional reports3 4 5 7 documented comparisons between the performance of IMA and SV grafts. Barner et al,3 reporting in 1985, revealed a 5-year survival rate of 93% (comparable to our survival rate of 91%) and 10-year survival rate of 84%. The patency rate in this study was 88% at 5 and 10 years for IMA grafts and 74% at 5 years and 41% at 10 years for SV grafts. Cameron et al4 (including Green) revealed a 5-year survival rate for IMA and SV grafts of 91% and 86%, respectively, and, at 10 years, of 82% and 72%. Loop et al5 from the Cleveland Clinic Foundation reported on an extensive series of patients and found survival rates at 10 years for management of three-vessel disease with IMA grafts of 83% and with SV grafts of 71%. Survival was higher for one- and two-vessel disease.
In 1994, Edwards et al7 reported on the Society of Thoracic Surgeons database, which showed improved early survival with IMA grafts: 2.0% for IMA grafts and 4.5% for SV grafts. The early mortality from several reports4 7 8 9 10 11 ranges from 1.9% to 2.8% for UL IMA grafts and from 1.3% to 3.7% for BL IMA grafts. The early mortality in the authors’ series, expanding over 8 years from 1984, was not different for UL and BL IMA grafts: 3.0% and 2.1%, respectively. The early mortality for SV grafting in comparison reports has ranged from 4.1% to 4.5%.1 2 3 7 This may be due to more compromised and elderly patients receiving SV grafts. There appears to be no difference in early mortality with use of UL and BL IMA grafts. There remains concern that BL IMA grafting may still increase morbidity, ie, sternal infections and mediastinitis, bleeding requiring reexploration, and the need for prolonged ventilation. In the authors’ series, BL IMA grafting did not increase the incidence of deep sternal infections and mediastinitis (P=NS) but did increase the incidence of reexploration for bleeding and tamponade (P<.05).
The debate remains over the use of UL or BL IMA grafts.3 4 8 9 10 11 12 13 14 15 In 1990, Fiore et al,9 reporting on the St Louis experience, revealed that survival was extended and ischemic events were reduced with BL IMA grafts at 13 years. Survival for BL IMA grafts was 74%, and for UL grafts, 59%. The 10-year survival rate was 82% and 84%, respectively, favoring BL IMA grafting. Cameron et al4 reported survival in favor of BL IMA at 10 years: 90% versus 83%. In 1989, Johnson et al13 reported no improvement in survival between BL and UL IMA grafting. In the same year, Kirklin et al14 found improved survival with UL IMA graft. Galbut et al,8 in 1990, reporting on a 17-year experience with BL IMA revascularization in 1087 patients, revealed a 10-year survival rate of 80% and a 15-year rate of 60%. In the authors’ experience, the 5-year survival rate was similar; 92% for UL IMA and 90% for BL IMA grafting.
Freedom from ischemic-related events and reoperation has also been varied. In the authors’ series, there has been no difference between the UL and BL populations. The freedom from recurrent angina pectoris, at 5 years, was 80% for UL IMA and 85% for BL IMA (P=NS). The freedom from myocardial infarction rate was 99% and 97%, respectively. The overall event-free comparison, at 5 years, was 79% for UL IMA and 81% for BL IMA. Okies et al,2 reporting in 1984 on UL IMA graft to the anterior descending coronary artery, had an event-free survival of 92% at 5 years and of 61% at 10 years. This perspective is not comparable to our freedom from all ischemic-related events and cardiac death rate. Fiore et al9 reported, in 1990, differences in long-term performance between BL and UL IMA grafting. At 13 years, the freedom from myocardial infarction was 75% for BL IMA and 59% for UL IMA, the freedom from recurrent angina was 36% and 27%, and the freedom from overall ischemic events was 32% and 18%, respectively.
In other terms, ischemic events have been reported as percentage per year. In 1986, Cameron et al4 found the rate of myocardial infarction for UL IMA to be 1.8%/year and for BL IMA to be 1.1%/year compared with SV grafting (3.1%/year). Barner et al,3 reporting in 1985, had a rate of 1.5%/year for UL and BL grafting. The authors’ rate is 0.7%/year over a shorter time interval. Barner et al,3 in 1985, indicated a recurrent angina rate of 6.2%/year compared with our rate of 3% to 4%/year for undifferentiated IMA grafting. The recurrence of angina is more likely to be unrelated to IMA graft status, but to progression of native vessel disease or vein graft atherosclerosis.
The reintervention rate, including PTCA and reoperation, was reported by Barner et al3 to be 0.85%/year, comparable to our rate of 1.0%/year. Recent literature deals extensively with reoperative surgery and primary operative techniques to facilitate reoperation.16 17 18 19 20 21 22 In 1994, Joyce et al16 reported that reoperations with previous BL IMA grafting can be facilitated by retrograde cardioplegia or hypothermic circulatory arrest. Lytle et al,18 in 1994, documented that IMA grafting at the primary operation does not increase the risk of reoperation, nor does use of IMA grafting at reoperation increase mortality. Navia et al17 reported a high incidence of hypoperfusion syndrome with IMA conduits between an interrupted stenotic vein graft to the anterior descending coronary artery. In a subsequent editorial, Barner21 reviewed the problem of hypoperfusion syndrome and competitive flow and the development of IMA “string sign” of an IMA graft supplemented by a vein graft. Reoperative surgery is facilitated by the right IMA traversing the transverse sinus rather than the substernal approach. In 1994, Tector et al22 reported on the formation of T grafts (proximal right IMA free to left IMA) to facilitate reoperation.
Our experience provides early to intermediate-term results with comparable populations with UL and BL IMA grafts in two age groups. At the reported time frame, there is no apparent benefit to BL IMA grafting over the age of 60 years with a trend to less recurrent angina with BL grafting in the ≤60 year group. There has been limited documentation on age consideration; in 1989, Gardner et al6 found IMA grafting in the elderly to be an independent predictor of survival. We will continue to be selective in UL and BL IMA use. This series requires continuing evaluation to determine the long-term results of UL and BL IMA revascularization to 10 and 15 years.
Presented American Heart Association, 1994
- Copyright © 1995 by American Heart Association
Grondin CM, Campeau L, Lespérance J, Enjalbert M, Bourassa MG. Comparison of late changes in internal mammary artery and saphenous vein grafts in two consecutive series of patients 10 years after operation. Circulation. 1984;70(suppl I):I-208-I-212.
Okies JE, Page US, Bigelow JC, Krause AH, Salomon NW. The left internal mammary artery: the graft of choice. Circulation. 1984;70(suppl I):I-213-I-221.
Cameron A, Kemp HG, Green GE. Bypass surgery with the internal mammary artery graft: 15-year follow-up. Circulation. 1986;74(suppl III):III-30-III-36.
Lytle BW, Cosgrove DM, Loop FD, Borsh J, Goormastic M, Taylor PC. Perioperative risk of bilateral internal mammary artery grafting: analysis of 500 cases from 1971 to 1984. Circulation. 1986;74(suppl III):III-37-III-41.
Hall RJ, Elayda MA, Gray A, Mathur VS, Garcia E de Castro CM, Massumi A, Cooley DA. Coronary artery bypass: long-term follow-up of 22,284 consecutive patients. Circulation. 1983;68(suppl II):II-20-II-26.
Kirklin JW, Naftel DC, Blackstone EH, Pohost GM. Summary of a consensus concerning death and ischemic events after coronary artery bypass grafting. Circulation. 1989;79(suppl VI):VI-81-VI-91.
Loop FD, Lytle BW, Cosgrove DM, Woods EL, Stewart RW, Golding LAR, Goormastic M, Taylor PC. Reoperation for coronary atherosclerosis. Ann Thorac Surg. 1990;212:378-386.