Abstract
Objectives
To compare the short- and long-term outcomes of coronary endarterectomy (CE) as an adjunct to coronary artery bypass grafting (CABG) with CABG alone in patients with diffuse coronary artery disease (CAD).
Methods
Postoperative complications, in-hospital and long-term mortality, and major adverse cardiovascular and cerebrovascular events (MACCE) were compared between 702 patients undergoing CE + CABG and 2808 propensity score–matched patients undergoing CABG alone at Tehran Heart Center between 2007 and 2016 and during a median follow-up of 98.6 months (96% confidence interval [CI], 97-99.9 months).
Results
Preoperative risk factors were perfectly balanced between the 2 groups. Compared to the CABG-only group, patients in the CE + CABG group had longer median aortic cross-clamp time (53.5 [interquartile range (IQR), 43-66.3] minutes versus 45 [IQR, 35-55] minutes; P < .001), longer cardiopulmonary bypass time (90 [IQR, 71-110] minutes vs 75 [IQR, 60-93] minutes; P < .001), longer intensive care unit stays (44.3 [IQR, 23.5-71] hours vs 25.3 [IQR, 22-48] hours; P < .001), longer duration of mechanical ventilation (10 [IQR, 8-13] hours vs 9.3 [IQR, 7-12] hours; P = .001), and higher rates of transfusion (46.9% vs 39.4%; P < .001) and in-hospital mortality (2.1% vs 0.8%; P = .003). Among short-term complications, pericardial involvement was more common in the CE + CABG group, but there were no significant differences in rates of cerebrovascular events, cardiac arrest, and infective complications. Despite comparable long-term mortality (hazard ratio [HR], 1.10; 95% CI, 0.93-1.31), patients undergoing CE + CABG had slightly higher rates of MACCE in their long-term follow-up (HR, 1.16; 95% CI, 1.01-1.33).
Conclusions
Despite higher rates of short-term morbidity and mortality, CE + CABG, if performed after meticulous patient selection and by experienced surgeons, could be an acceptable approach with favorable long-term outcomes in patients with diffuse CAD.
Key Words: coronary artery disease, coronary artery bypass, endarterectomy, hospital mortality, stroke, percutaneous coronary intervention
Graphical abstract
Short- and long-term outcomes of CE + CABG compared to CABG alone.
Central Message.
Despite higher rates of short-term morbidity and mortality, performing coronary endarterectomy as an adjunct to coronary artery bypass grafting surgery could be an acceptable approach, with favorable long-term outcomes in patients with diffuse coronary artery disease.
Perspective.
Coronary endarterectomy as an adjunct to coronary artery bypass grafting (CE + CABG) is performed in patients with diffuse CAD in whom complete revascularization cannot be obtained otherwise; however, concerns exist about its safety and long-term outcomes. Currently there are no guidelines on proper patient selection, and most previous studies are heterogeneous and of limited sample size. Here we aimed to compare the outcomes of CE + CABG with CABG alone.
Owing to increase in basic knowledge of the pathophysiology of coronary artery disease (CAD), recent advances in procedural techniques and devices, and increasing expertise of cardiologists, percutaneous coronary intervention (PCI) has become the treatment of choice in patients with focal or localized CAD.1 Coronary artery bypass grafting (CABG) is reserved for patients with more extensive and diffuse CAD with multivessel involvement and those with impaired left ventricular function.2,3 However, CABG cannot be performed safely and successfully in up to one-fourth of patients with complex CAD. Elderly patients with multiple risk factors, such as diabetes mellitus and dyslipidemia, who present with more severe and diffuse atherosclerotic disease and may have undergone previous PCIs have an elevated risk for surgery, and CABG may be associated with worse outcomes in these patients.4,5
Coronary endarterectomy (CE), first performed by Bailey in 1957, is a surgical adjunct to CABG in a carefully selected group of patients with diffuse CAD whose complete revascularization cannot be obtained with CABG alone.6,7 In this procedure, atheromatous plaque occluding the vessel is excised through an arteriotomy, providing sufficient blood supply to the myocardium. However, recent meta-analyses have demonstrated that despite similar long-term survival, CE + CABG is associated with increased 30-day mortality; greater postoperative complications, including myocardial infarction (MI), ventricular arrhythmias, renal dysfunction, cerebrovascular accident (CVA), and need for blood transfusion; increased length of stay; and reduced patency of endarterectomized vessels on follow-up angiography compared to CABG alone.8, 9, 10
Despite the ability of CE to improve the safety and success rate of CABG in patients with diffuse CAD, most surgeons are reluctant to perform CE because of the controversies surrounding the safety of the procedure, especially for the left anterior descending artery (LAD), which carries a high operative risk, in addition to the increased morbidity and mortality and reduced graft patency after CE.11,12 Currently there are no guidelines on the proper selection of patients who may benefit from CE + CABG, and most previous studies are heterogeneous and have a limited sample size. In the current study, we aimed to compare the outcomes of CE + CABG with a matched group of patients undergoing CABG alone at Tehran Heart Center, a leading tertiary care center for cardiovascular diseases in Tehran, Iran.
Materials and Methods
Study Population
The single-center Adult Cardiac Surgery Database and CABG Follow-up Registry of Tehran Heart Center were used for this study. Baseline demographics, risk factors, comorbidities, clinical and paraclinical data, and operative and postoperative details and follow-up data of 24,389 patients undergoing CABG at Tehran Heart Center between 2007 and 2016 were collected from these 2 databases. Patients with unavailable follow-up data, patients with missing data on variables used for propensity score matching (PSM), and those undergoing concurrent procedures, including valvular repair or replacement, ventricular aneurysm resection, and carotid endarterectomy, were excluded, leaving 23,933 patients for the current analysis.
The included patients were divided into 2 groups: those undergoing CE concomitantly with CABG (CE + CABG group; n = 702) and those undergoing CABG alone (CABG-alone group; n = 23,231). Because of the higher risk profile of patients in the CE + CABG group, 4:1 PSM was performed to match the 702 patients in this group with 2808 patients in the CABG-alone group. The follow-up protocol included in-person visits or telephone interviews at 4-6 months after operation, 1 year after operation, and annually thereafter.
Verbal informed consent was obtained from each patients to record their data in the patient registry and use these data anonymously in future studies. This study complies with the Declaration of Helsinki and was approved by the Ethics Committee of Tehran Heart Center (IR-THC-13799; approved April 23, 2018).
Procedure
According to the 2011 American College of Cardiology Foundation/American Heart Association guidelines, CABG is indicated in patients with high-grade blockage in any of the major coronary arteries, such as left main disease >50%, 2-vessel disease involving the LAD, 3-vessel disease with >70% blockage, 1-vessel disease with >70% blockage in a survivor of sudden cardiac death with ischemia-related ventricular tachycardia, and significant stenosis >70% in 1 or more vessels in a symptomatic patient despite maximal medical therapy, as well as patients in whom previous PCI failed to resolve the blockage.13 CABG was performed according to the standard techniques described in the latest guidelines.14,15
The main indication for CE is the presence of a long-segment, diffusely diseased coronary artery complicated by narrow segments or severe calcifications occluding the entire artery or causing multiple obstructions in a coronary artery and its side branches that make distal grafting infeasible.16,17 CE was performed in our institute with the close technique, in which a small incision is made on the adventitia of the target coronary artery, and after a small segmental dissection of the intima from the vessel wall, the atheromatous plaque is dissected from the native artery proximally and distally with gentle, steady traction.18,19 Patients with diffuse CAD and those undergoing CE + CABG received dual antiplatelet therapy for 3 months after the surgery, followed by lifelong aspirin therapy. In contrast, patients in the CABG-alone group received lifelong aspirin monotherapy, in accordance with the prevailing guidelines at the time.13
Study Endpoints and Definitions
Our study endpoints were in-hospital and long-term mortality, postoperative morbidities during hospitalization, and major adverse cardiovascular and cerebrovascular events (MACCE) during the median follow-up of 98.6 months (95% CI, 97-99.9 months). Postoperative morbidities were defined as renal failure (serum creatinine >1.4 mg/dL), inotrope use, blood transfusion, CVA, pericardial effusion or cardiac tamponade, prolonged mechanical ventilation (>24 hours), cardiac arrest, intensive care unit (ICU) stay, and infective complications, including deep sternal wound or leg infection, pneumonia, urinary tract infection, and septicemia. MACCE was defined as a composite of all-cause mortality, acute coronary syndrome, stroke or transient ischemic attack (TIA), and the need for revascularization.
Statistical Analysis
For the descriptive statistics, mean ± standard deviation or median and interquartile range (IQR) were calculated for variables with and without normal distribution, respectively. Frequency and percentage were also computed for categorical data. Comparisons of covariates for the CE + CABG and CABG-alone groups were done using either an independent t test or the Mann-Whitney U test for continuous variables and the Pearson χ2 test or Fisher exact test for categorical variables, as appropriate.
A PSM technique was used to reduce selection bias and control the confounding effects by matching participants with similar propensity scores based on observed covariates. The propensity score for the CE technique was determined using multiple logistic regression with 4:1 matching (4 control:1 CE) on variables including age, sex, body mass index, diabetes mellitus, hypertension, dyslipidemia, family history of CAD, current or former opium addiction or cigarette smoking, previous atrial fibrillation, previous MI, left ventricular ejection fraction (LVEF), peripheral vascular disease (PVD), previous CVA, renal failure, chronic obstructive pulmonary disease, previous PCI or CABG, number of grafts, urgent or emergent surgery, need for preoperative intra-aortic balloon pump (IABP) insertion, and off-pump surgery. Subsequently, the scores were stabilized using the crude probability of being in the CE-CABG group. The multivariate imputation by chained equations method was used for sensitivity analysis of the missing values.
The density distribution plot of the propensity scores, the standardized mean differences of the covariates, and the mean ± SD of weights were determined to assess appropriate balancing between the 2 study groups (Figures 1 and 2). After balancing the 2 groups, the standardized mean differences were all <0.1, and the high coverage in the density plot indicated that the groups were excellently balanced. The effect of CE on long-term mortality and MACCEs was assessed by applying a Cox proportional hazards model, with results are expressed as hazard ratio (HR) with corresponding 95% confidence interval (CI). All statistical tests were 2-sided, and P < .05 was considered to indicate statistical significance. Analyses were conducted using R version 4.2.2 (R Core Team) on Windows 10 × 64 (build 19,044), using the survival (version 3.5.5), cobalt (version 4.5.1), MatchIt (version 4.5.3), and ggsurvfit (version 0.3.0) packages.
Figure 1.
Coverage plot showing distributions of estimated propensity scores (PS) before and after propensity score matching. CE, Coronary endarterectomy; CABG, coronary artery bypass grafting.
Figure 2.
Covariate balance assessment by absolute standardized mean difference before (original) and after propensity score matching (PSM). After PSM, the differences between groups were <10% (dashed line), indicating good covariate balance. AF, Atrial fibrillation; BMI, body mass index; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; CVA, cerebrovascular accident; LVEF, left ventricular ejection fraction; MI, myocardial infarction; PCI, percutaneous coronary intervention; PVD, peripheral vascular disease.
Results
Baseline Characteristics
After excluding patients with unavailable follow-up data (n = 97) and those missing data for at least 1 matched variable (n = 359), 23,933 patients who underwent CABG at Tehran Heart Center between 2007 and 2016 were included in the current study. Among them, 702 patients (2.93%) underwent adjunct CE. The total study population had a mean age of 65.19 ± 9.93 years, and was 73.7% male. A previous history of MI was reported in 15.8% of patients, prior PCI in 7.1%, and previous CABG in 0.5%.
Before PSM, patients in the CE + CABG group were significantly younger (mean age, 60.73 ± 9.64 years vs 65.32 ± 9.9 years; P < .001) and had higher rates of opiate consumption (18.4% vs 15.3%; P = .026), diabetes mellitus (47.7% vs 39.8%; P < .001), PVD (3.8% vs 2.1%; P = .002), and renal failure (3.8% vs 2.5%; P = .026), in addition to lower median LVEF (45% [IQR, 40%-55%] vs 50% [40%-55%]; P < .001). After PSM, there were no statistically significant differences in baseline characteristics between the 2 groups except in age, which remained slightly lower in the CE + CABG group (mean, 60.73 ± 9.64 years vs 61.83 ± 9.87 years; P = .005). Baseline demographic and clinical characteristics of the study population before and after PSM are presented in Table 1.
Table 1.
Baseline characteristics of the study population before and after PSM
| Characteristic | Before PSM |
After PSM |
||||||
|---|---|---|---|---|---|---|---|---|
| Total cohort (N = 23,933) | CABG-only group (N = 23,231) | CE + CABG group (N = 702) | P value∗ | Total cohort (N = 3510) | CABG-only group (N = 2808) | CE + CABG group (N = 702) | P value† | |
| Age, y, mean ± SD | 65.19 ± 9.93 | 65.32 ± 9.9 | 60.73 ± 9.64 | <.001 | 61.61 ± 9.83 | 61.83 ± 9.87 | 60.73 ± 9.64 | .005 |
| Male sex, n (%) | 17,641 (73.7) | 17,113 (73.7) | 528 (75.2) | .358 | 2635 (75.1) | 2107 (75.0) | 528 (75.2) | .922 |
| BMI, kg/m2, mean ± SD | 27.38 ± 4.28 | 27.37 ± 4.27 | 27.64 ± 4.61 | .211 | 27.59 ± 4.4 | 27.58 ± 4.35 | 27.64 ± 4.61 | .778 |
| Cigarette smoking, n (%) | 4203 (17.6) | 4061 (17.5) | 142 (20.2) | .059 | 712 (20.3) | 570 (20.3) | 142 (20.2) | .967 |
| Opiate use, n (%) | 3685 (15.4) | 3556 (15.3) | 129 (18.4) | .026 | 634 (18.1) | 505 (18) | 129 (18.4) | .803 |
| Diabetes mellitus, n (%) | 9578 (40) | 9243 (39.8) | 335 (47.7) | <.001 | 1617 (46.1) | 1282 (45.7) | 335 (47.7) | .329 |
| Hypertension, n (%) | 13,477 (56.3) | 13,087 (56.3) | 390 (55.6) | .682 | 1967 (56) | 1577 (56.2) | 390 (55.6) | .756 |
| Dyslipidemia, n (%) | 16,786 (70.1) | 16,289 (70.1) | 497 (70.8) | .698 | 2479 (70.6) | 1982 (70.6) | 497 (70.8) | .913 |
| Family history of CAD, n (%) | 9539 (39.9) | 9264 (39.9) | 275 (39.2) | .707 | 1394 (39.7) | 1119 (39.9) | 275 (39.2) | .746 |
| Previous MI, n (%) | 3783 (15.8) | 3688 (15.9) | 95 (13.5) | .094 | 464 (13.2) | 369 (13.1) | 95 (13.5) | .782 |
| Previous AF, n (%) | 211 (0.9) | 209 (0.9) | 2 (0.3) | .086 | 11 (0.3) | 9 (0.3) | 2 (0.3) | .88 |
| Previous CVA/TIA, n (%) | 1431 (6) | 1390 (6) | 41 (5.8) | .875 | 205 (5.8) | 164 (5.8) | 41 (5.8) | >.999 |
| Peripheral vascular disease, n (%) | 516 (2.2) | 489 (2.1) | 27 (3.8) | .002 | 116 (3.3) | 89 (3.2) | 27 (3.8) | .372 |
| Previous renal failure, n (%) | 609 (2.5) | 582 (2.5) | 27 (3.8) | .026 | 123 (3.5) | 96 (3.4) | 27 (3.8) | .591 |
| Previous COPD, n (%) | 892 (3.7) | 863 (3.7) | 29 (4.1) | .566 | 136 (3.9) | 107 (3.8) | 29 (4.1) | .694 |
| Previous procedure, n (%) | .242 | .93 | ||||||
| No | 22,133 (92.5) | 21,477 (92.4) | 656 (93.4) | 3274 (93.3) | 2618 (93.2) | 656 (93.4) | ||
| PCI | 1689 (7.1) | 1648 (7.1) | 41 (5.8) | 202 (5.8) | 161 (5.7) | 41 (5.8) | ||
| CABG | 111 (0.5) | 106 (0.5) | 5 (0.7) | 34 (1) | 29 (1) | 5 (0.7) | ||
| LVEF, %, median (IQR) | 50 (40-55) | 50 (40-55) | 45 (40-55) | <.001 | 47.5 (40-55) | 47.5 (40-55) | 45 (40-55) | .458 |
Bold type indicates statistical significance.
PSM, Propensity score matching; CABG, coronary artery bypass grafting; CE, coronary endarterectomy; BMI, body mass index; CAD, coronary artery disease; MI, myocardial infarction; AF, atrial fibrillation; CVA, cerebrovascular accident; TIA, transient ischemic attack; COPD, chronic obstructive pulmonary disease; PCI, percutaneous coronary intervention; LVEF, left ventricular ejection fraction.
Wilcoxon rank-sum test, Pearson χ2 test, Fisher exact test, independent t test.
Univariate logistic regression considering the cluster of matched individuals.
Procedural Details
Elective surgery was performed in 94.8% of cases, and urgent or emergent procedures accounted for the remaining 5.2%. Three or 4 bypass grafts were commonly used in both groups. The left internal mammary artery was the most frequently used arterial conduit, used in 97% of patients undergoing CE + CABG and 98% of those undergoing CABG alone. A radial artery conduit was used in only 1.5% of the CE + CABG patients and 3.3% of the CABG-alone patients. Endarterectomy was most frequently performed on the right coronary artery in 326 patients (46%), followed by the LAD and diagonal branches in 177 patients (25.7%). The procedure was performed on the posterior descending artery, posterolateral branch, or right ventricular branch in 94 patients (13%) and on the left circumflex or obtuse marginal branches in 52 patients (7.4%). In 56 patients (7.4%), endarterectomy was performed on 2 or more vessels.
Even after PSM, patients in the CE + CABG group had significantly longer median aortic cross-clamp time (median 53.5 [IQR, 43-66.3] minutes vs 45 [IQR, 35-55] minutes; P < .001) and cardiopulmonary bypass (CPB) time (median, 90 [IQR, 71-110] minutes vs 75 [IQR, 60-93] minutes; P < .001). IABP was inserted during the surgery in 1.8% of the total cohort and more frequently in the CE + CABG group (3.9% vs 1.7%; P < .001). Procedural details before and after PSM are demonstrated in Table 2.
Table 2.
Procedural details before and after PSM
| Parameter | Before PSM |
After PSM |
||||||
|---|---|---|---|---|---|---|---|---|
| Total cohort (N = 23,933) | CABG-only group (N = 23,231) | CE + CABG group (N = 702) | P value∗ | Total cohort (N = 3510) | CABG-only group (N = 2808) | CE + CABG group (N = 702) | P value† | |
| Status of procedure, n (%) | <.001 | .178 | ||||||
| Elective | 22,694 (94.8) | 22,061 (95) | 633 (90.2) | 3116 (88.8) | 2483 (88.4) | 633 (90.2) | ||
| Urgent/emergent | 1239 (5.2) | 1170 (5) | 69 (9.8) | 394 (11.2) | 325 (11.6) | 69 (9.8) | ||
| Grafts, n (%) | <.001 | .805 | ||||||
| 1 | 400 (1.7) | 397 (1.7) | 3 (0.4) | 19 (0.5) | 16 (0.6) | 3 (0.4) | ||
| 2 | 2678 (11.2) | 2646 (11.4) | 32 (4.6) | 178 (5.1) | 146 (5.2) | 32 (4.6) | ||
| 3 | 9089 (38) | 8915 (38.4) | 174 (24.8) | 1156 (32.9) | 982 (35) | 174 (24.8) | ||
| 4 | 9460 (39.5) | 9139 (39.3) | 321 (45.7) | 1435 (40.9) | 1114 (39.7) | 321 (45.7) | ||
| 5+ | 2306 (9.6) | 2134 (9.2) | 172 (24.5) | 722 (20.6) | 550 (19.6) | 172 (24.5) | ||
| Off-pump CABG, n (%) | 2329 (9.7) | 1989 (8.6) | 340 (48.4) | <.001 | 1493 (42.5) | 1153 (41.1) | 340 (48.4) | <.001 |
| Aortic cross-clamp time (min), median (IQR) | 39 (30-50) | 39 (30-49) | 53.5 (43-66.3) | <.001 | 46 (36-57) | 45 (35-55) | 53.5 (43-66.3) | <.001 |
| Full perfusion time (min), median (IQR) | 68 (55-84) | 67 (54-83) | 90 (71-110) | <.001 | 80 (63-98) | 75 (60-93) | 90 (71-110) | <.001 |
| Intraprocedural IABP insertion, n (%) | 439 (1.8) | 418 (1.8) | 21 (3.9) | <.001 | 68 (2) | 47 (1.7) | 21 (3.9) | <.001 |
Bold type indicates statistical significance.
PSM, Propensity score matching; CABG, coronary artery bypass grafting; CE, coronary endarterectomy; IQR, interquartile range; IABP, intra-aortic balloon pump.
Wilcoxon rank-sum test, Pearson χ2 test, Fisher exact test.
Univariate logistic regression considering the cluster of matched individuals.
Short-Term Outcomes
Patients undergoing CE + CABG had longer ICU stays (median 44.3 [IQR, 23.5-71] hours vs 25.3 [IQR, 22-48] hours; P < .001) and required a longer duration of mechanical ventilation (median, 10 [IQR, 8-13] hours vs 9.3 [IQR, 7-12] hours; P = .001). Transfusions also were more common in this group (46.9% vs 39.4%; P < .001). Renal failure was significantly more frequent in the CABG-alone group (3.2% vs 1.5%; P = .021), while pericardial complications such as pericardial effusion (1.5% vs 0.9%) and cardiac tamponade (2.6% vs 1%; P = .001) occurred more often in the CE + CABG group. Despite similar rates of CVA/TIA, infections, and cardiac arrest in the 2 groups, the CE + CABG group had a significantly higher rate of in-hospital mortality (2.1% vs 0.8%; P = .003). The short-term outcomes of the patients before and after PSM are described in detail in Table 3.
Table 3.
Short-term outcomes of CABG versus CE + CABG before and after PSM
| Outcome | Before PSM |
After PSM |
||||||
|---|---|---|---|---|---|---|---|---|
| Total cohort (N = 23,933) | CABG-only group (N = 23,231) | CE + CABG group (N = 702) | P value∗ | Total cohort (N = 3510) | CABG-only group (N = 2808) | CE + CABG group (N = 702) | P value∗ | |
| ICU stay, h, median (IQR) | 28 (23-66) | 28 (23-65.2) | 44.3 (23.5-71) | <.001 | 26 (22-51) | 25.3 (22-48) | 44.3 (23.5-71) | <.001 |
| Mechanical ventilation duration, h, median (IQR) | 9.5 (7.3-12.5) | 9.5 (7.3-12.5) | 10 (8-13) | .033 | 9.5 (7.3-12) | 9.3 (7-12) | 10 (8-13) | .001 |
| Prolonged ventilation, n (%) | 585 (2.4) | 561 (2.4) | 24 (3.4) | .087 | 95 (2.7) | 71 (2.5) | 24 (3.4) | .19 |
| Intraoperative/postoperative inotrope use, n (%) | 3788 (15.9) | 3693 (15.9) | 95 (14.5) | .334 | 452 (13.1) | 357 (12.8) | 95 (14.5) | .23 |
| Intraoperative/postoperative transfusion, n (%) | 11,229 (47.1) | 10,922 (47.1) | 307 (46.9) | .934 | 1408 (40.8) | 1101 (39.4) | 307 (46.9) | <.001 |
| Renal failure, n (%) | 539 (2.3) | 529 (2.3) | 10 (1.5) | .202 | 100 (2.9) | 90 (3.2) | 10 (1.5) | .021 |
| CVA/TIA, n (%) | 150 (0.6) | 146 (0.6) | 4 (0.6) | >.999 | 14 (0.4) | 10 (0.4) | 4 (0.6) | .317 |
| Pericardial involvement, n (%) | .014 | .001 | ||||||
| Pericardial effusion | 158 (0.7) | 148 (0.6) | 10 (1.5) | 34 (1) | 24 (0.9) | 10 (1.5) | ||
| Cardiac tamponade | 459 (1.9) | 442 (1.9) | 17 (2.6) | 45 (1.3) | 28 (1.0) | 17 (2.6) | ||
| Infective complications, n (%) | 811 (3.4) | 789 (3.4) | 22 (3.4) | .986 | 97 (2.8) | 75 (2.7) | 22 (3.4) | .322 |
| Intraoperative/postoperative cardiac arrest, n (%) | 497 (2.1) | 479 (2.1) | 18 (2.7) | .227 | 90 (2.6) | 72 (2.6) | 18 (2.7) | .794 |
| In-hospital mortality, n (%) | 259 (1.1) | 245 (1.1) | 14 (2.1) | .008 | 37 (1.1) | 23 (0.8) | 14 (2.1) | .003 |
PSM, Propensity score matching; CABG, coronary artery bypass grafting; CE, coronary endarterectomy; ICU, intensive care unit; IQR, interquartile range; CVA, cerebrovascular accident; TIA, transient ischemic attack.
Wilcoxon rank-sum test, Pearson χ2 test, Fisher exact test.
Long-Term Outcomes
Cox proportional hazards regression analysis revealed no significant association between CE + CABG and long-term mortality before PSM (HR, 1.03; 95% CI, 0.89-1.19; P = .734) and after PSM (HR, 1.10; 95% CI, 0.93-1.31; P = .267). However, for MACCE, the HR increased from 1.11 (95% CI, 0.98-1.24; P = .089) before PSM to 1.16 (95% CI, 1.01-1.33; P = .042) after PSM, indicating a borderline significant association with increased MACCE risk. The results of Cox proportional hazards regression analysis assessing the effect of CE + CABG on long-term mortality and MACCE are reported in Table 4.
Table 4.
Cox proportional hazards regression for the effect of CE + CABG on long-term mortality and MACCE before and after PSM
| Parameter | Before PSM (n = 23,933) |
After PSM (n = 3510) |
||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Long-term mortality | 1.03 | 0.89-1.19 | .734 | 1.1 | 0.93-1.31 | .267 |
| MACCE | 1.11 | 0.98-1.24 | .089 | 1.16 | 1.01-1.33 | .042 |
Bold type indicates statistical significance.
PSM, Propensity score matching; HR, hazard ratio; CI, confidence interval; MACCE, major adverse cardiac and cerebrovascular events.
Kaplan-Meier survival curves for MACCE-free survival and mortality (Figure 3) demonstrated comparable declining trends for the CABG-alone and CE + CABG groups over time, with close alignment after PSM. These findings suggest a minimal impact of CE + CABG on long-term mortality but a potential borderline increase in MACCE risk.
Figure 3.
Kaplan-Meier survival curves before and after propensity score matching (PSM) for major adverse cardiac and cerebrovascular event (MACCE)-free survival (A and B) and mortality (C and D). The lines represent the survival probability over time for the coronary artery bypass grafting (CABG)-only group and the coronary endarterectomy (CE) + CABG group, and the shaded areas indicate the 95% confidence intervals around each survival estimate.
Discussion
This single-center retrospective cohort study evaluated the short- and long-term outcomes of CE + CABG compared to CABG alone, with a median follow-up of 8 years. Our findings indicate that CE + CABG is associated with longer aortic cross-clamp and CPB times, an increased need for IABP support, and higher rates of short-term morbidity and in-hospital mortality. Despite these short-term drawbacks, the long-term mortality rates were comparable in the 2 groups, although the CE + CABG group exhibited a slightly elevated risk of MACCE.
Introduced by Bailey and colleagues in the 1950s, CE is indicated for patients with diffusely diseased coronary arteries, particularly those with hard calcified plaques occluding the entire artery or causing multiple obstructions in one coronary artery and its side branches, making distal grafting infeasible.6,20 The reported incidence of CE varies widely, ranging from 3.7% to 42%, reflecting the lack of consistent, guideline-based indications for its use.21 While earlier studies reported poor postoperative outcomes for CE,11,22 more recent evidence suggests improved short- and long-term outcomes, especially in high-risk patients with complex and diffuse CAD in whom complete revascularization cannot be achieved with CABG alone.23, 24, 25
In our study, patients undergoing CE demonstrated a higher prevalence of comorbidities, such as diabetes mellitus, PVD, renal impairment, and lower baseline LVEF. These factors indicate a greater atherosclerotic burden, affecting not only the coronary arteries, but also the renal and peripheral vasculature. We used PSM to mitigate confounding factors, resulting in an excellent balance between the 2 groups except for a slight age difference.
Patients undergoing CE had significantly longer aortic cross-clamp time (median, 53.5 [IQR, 43-66.3] minutes vs 45 [IQR, 35-55] minutes) and CPB time (median, 90 [IQR, 71-110] minute vs 75 [IQR, 60-93] minutes), consistent with previous studies. These findings can be attributed to the added complexity of the endarterectomy procedure, more diffuse CAD, and higher overall plaque burden in this group.8,9,21,26 As in previous studies, we found higher rates of IABP insertion in the CE + CABG group (3.9% vs 1.7%; P < .001), which could be justified by the higher risk of cardiogenic shock in this group attributed to the more complex surgery and longer aortic cross-clamp and CPB times.27 IABP is one of the most frequently used circulatory assist devices in cardiac surgery to help reduce ventricular afterload and myocardial oxygen demand while improving diastolic perfusion pressure and coronary blood flow.28
Mechanical trauma from CE leads to the loss of endothelial integrity and function.8,29,30 This damage disrupts the ability of intact endothelium to produce vasoactive factors and counteract leukocyte adhesion and platelet aggregation, as well as the subsequent inflammation and thrombosis. Consequently, CE increases the tendency toward de novo thrombogenesis and embolization from dislodged atheromatous material, leading to an increased risk of MI and mortality.11,26 Longer operative and CPB times can cause ischemia-reperfusion injury, leading to accelerated neutrophil–platelet and platelet–endothelial interactions, subsequently accelerating inflammation and increasing the risk of short-term morbidity and mortality.22,31 The foregoing mechanisms, in addition to a higher perioperative atherosclerotic burden and longer aortic cross-clamp, cardioplegia, and CPB times, represent the underlying causes of the higher rates of in-hospital mortality (2.1% vs 0.8%) and short-term complications, and subsequent increased ICU length of stay (median, 44.3 [IQR, 23.5-71] hours vs 25.3 [IQR, 22-48] hours) observed in patients undergoing CE + CABG at our center, concordant with the results of previously published systematic reviews.8,9
Pericardial effusion (1.5% vs 0.9%) and cardiac tamponade (2.6% vs 1%) were more frequent in the CE + CABG group, potentially due to greater microvascular injury and inflammation associated with prolonged CPB and manipulation of diffusely diseased arteries.32 The higher rate of transfusions in the CE + CABG group (46.9% vs 39.4%) can be explained by the increased bleeding risk associated with prolonged and more complex surgery, platelet dysfunction induced by CPB, and dual antiplatelet therapy, in addition to a higher burden of baseline comorbidities, such as diabetes mellitus and renal failure, that can impair vascular integrity and hemostatic mechanisms.33 Systemic inflammatory response syndrome caused by more extensive surgical dissection and longer CPB times, in addition to the increased risk of transfusion-related acute lung injury caused by higher rates of blood transfusion, are potential explanations for the prolonged ventilatory support in the CE + CABG group (median, 10 [IQR, 8-13] hours vs 9.3 [IQR, 7-12] hours).26,34,35
Mechanical injury to the vascular endothelium during endarterectomy leads to increased thrombogenesis and atheroembolization, predisposing patients to postoperative CVA/TIA. A meta-analysis by Wang and colleagues10 reported a 48% increase in the incidence of CVA in patients undergoing CE + CABG compared to those with isolated CABG. The reported rates of CVA varied between 0.5% and 6% in previous studies,36,37 while we found a CVA/TIA rate of 0.6% in patients undergoing CE + CABG, with no significant difference from those undergoing CABG alone. The absence of an increased risk of CVA in our center could result from meticulous patient selection for endarterectomy, enhanced expertise of surgeons, use of advanced surgical techniques, postoperative care protocols, and use of dual antiplatelet therapy. Although the latest guidelines do not provide any recommendations regarding the optimal thromboprophylaxis strategy in patients undergoing CE,38,39 we used dual antiplatelet therapy for 3 months after surgery in our center, which appears to be effective in reducing the rate of postoperative thrombotic events in patients undergoing CE.40
Long-term mortality rates were similar in our 2 study groups, consistent with prior studies,9,20 while a slight increase in the risk of MACCE was observed in patients undergoing CE + CABG. Although increased short-term mortality and morbidity may raise concerns regarding the safety of CE, the comparable long-term survival is indicative of its beneficial effects, especially in elderly patients with multiple comorbidities and more severe and diffuse CAD who may have undergone multiple previous PCI and in whom complete revascularization could not be obtained otherwise. Given that incomplete revascularization is one of the most critical factors affecting long-term morbidity and mortality, necessary measures should be taken to improve the safety and efficacy of the procedure. Careful patient selection should be conducted by a multidisciplinary Heart Team that includes interventional and noninvasive cardiologists and cardiac surgeons while considering the patients’ symptoms, frailty, physiologic parameters, coronary anatomy, and myocardial viability.2,8,13 The risk of postoperative complications should be reduced by using optimized surgical techniques and aggressive thromboprophylaxis strategies.8 To better understand whether complete revascularization with endarterectomy is worth its additional short-term complications, choosing patients undergoing CABG with incomplete revascularization without CE or those with diffuse CAD undergoing medical management alone as a control group would make our comparison more appropriate.
This study is strengthened by its large sample size of more than 23,000 patients and rigorous PSM, effectively reducing selection bias and ensuring comparability between the CE + CABG and CABG-alone groups. The comprehensive analysis of both short- and long-term outcomes provides valuable insights into the procedural risks and benefits, offering a holistic view of CE's clinical implications. Furthermore, the study's focus on real-world data from a high-volume center enhances its relevance and applicability to similar clinical settings. Nonetheless, the study's retrospective design limits the ability to establish causality, and the single-center nature may restrict the generalizability of our findings to other populations or institutions with different surgical techniques and protocols. In addition, the lack of postoperative graft patency data is a critical limitation, owing to the invasive nature of coronary angiography, the high cost and potential risks (eg, contrast nephropathy, radiation exposure) associated with coronary computed tomography angiography, and the fact that such assessments are not routinely performed in asymptomatic patients. The lack of data on baseline CAD complexity between the 2 groups is another limitation; however, this was partially addressed by recognizing that the decision to perform CE was at the surgeon's discretion, guided by perioperative assessments indicating that patients undergoing CE + CABG had more severe and complex CAD that was not amenable to CABG alone. Furthermore, key clinical complexity factors from the SYNTAX Score II—including age, LVEF, presence of PVD, renal failure, and chronic obstructive pulmonary disease—were incorporated into the PSM process to balance baseline characteristics between the 2 groups. Finally, the reliance on registry data for long-term follow-up introduced potential reporting biases, underscoring the need for prospective, multicenter studies to validate these results and refine the role of CE in modern cardiac surgery.
In conclusion, CE + CABG can be an acceptable strategy in patients with diffuse CAD when performed with careful patient selection and by experienced surgical teams. Despite the procedure's higher short-term morbidity and mortality rates, the potential for favorable long-term outcomes makes it an acceptable option, particularly for high-risk patients who might otherwise face incomplete revascularization. This underscores the importance of a multidisciplinary approach to patient evaluation and the continued refinement of surgical techniques and antithrombotic strategies to optimize safety and efficacy.
Conflict of Interest Statement
The authors reported no conflicts of interest.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
Footnotes
Drs Bagheri and Firoozbakhsh contributed equally to this work.
This study complies with the Declaration of Helsinki and was approved by the Ethics Committee of Tehran Heart Center (IR-THC-13799, approved April 23, 2018). Verbal informed consent was obtained from the patients to record their data in the patient registry and use the data anonymously in future studies.
References
- 1.Fihn S.D., Blankenship J.C., Alexander K.P., et al. 2014 ACC/AHA/AATS/PCNA/SCAI/STS focused update of the guideline for the diagnosis and management of patients with stable ischemic heart disease: a report of the American college of cardiology/American Heart association task force on practice guidelines, and the American association for thoracic surgery, preventive cardiovascular nurses association, society for cardiovascular angiography and interventions, and society of thoracic surgeons. J Am Coll Cardiol. 2014;64(18):1929–1949. doi: 10.1016/j.jacc.2014.07.017. [DOI] [PubMed] [Google Scholar]
- 2.Windecker S., Kolh P., Alfonso F., et al. 2014 ESC/EACTS guidelines on myocardial revascularization: the task force on myocardial revascularization of the European society of cardiology (ESC) and the European association for cardio-thoracic surgery (EACTS) developed with the special contribution of the European association of percutaneous cardiovascular interventions (EAPCI) Eur Heart J. 2014;35(37):2541–2619. doi: 10.1093/eurheartj/ehu278. [DOI] [PubMed] [Google Scholar]
- 3.Elbardissi A.W., Balaguer J.M., Byrne J.G., Aranki S.A. Surgical therapy for complex coronary artery disease. Semin Thorac Cardiovasc Surg. 2009;21(3):199–206. doi: 10.1053/j.semtcvs.2009.08.006. [DOI] [PubMed] [Google Scholar]
- 4.Santini F., Casali G., Lusini M., et al. Mid-term results after extensive vein patch reconstruction and internal mammary grafting of the diffusely diseased left anterior descending coronary artery. Eur J Cardiothorac Surg. 2002;21(6):1020–1025. doi: 10.1016/s1010-7940(02)00074-x. [DOI] [PubMed] [Google Scholar]
- 5.Moshkovitz Y., Paz Y., Shabtai E., et al. Predictors of early and overall outcome in coronary artery bypass without cardiopulmonary bypass. Eur J Cardiothorac Surg. 1997;12(1):31–39. doi: 10.1016/s1010-7940(97)00129-2. [DOI] [PubMed] [Google Scholar]
- 6.Bailey C.P., May A., Lemmon W.M. Survival after coronary endarterectomy in man. J Am Med Assoc. 1957;164(6):641–646. doi: 10.1001/jama.1957.02980060017005. [DOI] [PubMed] [Google Scholar]
- 7.Parsonnet V., Gilbert L., Gielchinsky I., Bhaktan E.K. Endarterectomy of the left anterior descending and mainstem coronary arteries: a technique for reconstruction of inoperable arteries. Surgery. 1976;80(6):662–673. [PubMed] [Google Scholar]
- 8.Soylu E., Harling L., Ashrafian H., Casula R., Kokotsakis J., Athanasiou T. Adjunct coronary endarterectomy increases myocardial infarction and early mortality after coronary artery bypass grafting: a meta-analysis. Interact Cardiovasc Thorac Surg. 2014;19(3):462–473. doi: 10.1093/icvts/ivu157. [DOI] [PubMed] [Google Scholar]
- 9.Wang J., Gu C., Yu W., Gao M., Yu Y. Short-and long-term patient outcomes from combined coronary endarterectomy and coronary artery bypass grafting: a meta-analysis of 63,730 patients (PRISMA) Medicine (Baltimore) 2015;94(41) doi: 10.1097/MD.0000000000001781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang J., Gu C., Gao M., Yu W., Yu Y. Adjunct coronary endarterectomy increases cerebrovascular accident after coronary artery bypass grafting: a meta-analysis of 17,600 patients. Int J Cardiol. 2015;182:79–81. doi: 10.1016/j.ijcard.2014.12.061. [DOI] [PubMed] [Google Scholar]
- 11.Livesay J.J., Cooley D.A., Hallman G.L., et al. Early and late results of coronary endarterectomy: analysis of 3,369 patients. J Thorac Cardiovasc Surg. 1986;92(4):649–660. [PubMed] [Google Scholar]
- 12.Song Y., Xu F., Du J., Zhang J., Feng W. Coronary endarterectomy with coronary artery bypass graft decreases graft patency compared with isolated coronary artery bypass graft: a meta-analysis. Interact Cardiovasc Thorac Surg. 2017;25(1):30–36. doi: 10.1093/icvts/ivx045. [DOI] [PubMed] [Google Scholar]
- 13.Hillis L.D., Smith P.K., Anderson J.L., et al. 2011 ACCF/AHA guideline for coronary artery bypass graft surgery: a report of the American college of cardiology foundation/American heart association task force on practice guidelines. Circulation. 2011;124(23):e652–e735. doi: 10.1161/CIR.0b013e31823c074e. [DOI] [PubMed] [Google Scholar]
- 14.Eagle K.A., Guyton R.A., Davidoff R., et al. ACC/AHA guidelines for coronary artery bypass graft surgery: executive summary and recommendations: a report of the American college of cardiology/American heart association task force on practice guidelines (committee to revise the 1991 guidelines for coronary artery bypass graft surgery) Circulation. 1999;100(13):1464–1480. doi: 10.1161/01.cir.100.13.1464. [DOI] [PubMed] [Google Scholar]
- 15.Cohn L.H., Adams D.H. 5th ed. McGraw-Hill Education; 2017. Cardiac Surgery in the Adult. [Google Scholar]
- 16.Nishigawa K., Fukui T., Takanashi S. Coronary endarterectomy for the diffusely diseased coronary artery. Gen Thorac Cardiovasc Surg. 2014;62(8):461–467. doi: 10.1007/s11748-014-0414-x. [DOI] [PubMed] [Google Scholar]
- 17.Asimakopoulos G., Taylor K.M., Ratnatunga C.P. Outcome of coronary endarterectomy: a case-control study. Ann Thorac Surg. 1999;67(4):989–993. doi: 10.1016/s0003-4975(99)00094-6. [DOI] [PubMed] [Google Scholar]
- 18.Keogh B.E., Bidstrup B.P., Taylor K.M., Sapsford R.N. Angioscopic evaluation of intravascular morphology after coronary endarterectomy. Ann Thorac Surg. 1991;52(4):766–772. doi: 10.1016/0003-4975(91)91208-d. [DOI] [PubMed] [Google Scholar]
- 19.Tiemuerniyazi X., Yan H., Song Y., Nan Y., Xu F., Feng W. Mid-term outcomes of coronary endarterectomy combined with coronary artery bypass grafting. Interact Cardiovasc Thorac Surg. 2021;32(2):188–195. doi: 10.1093/icvts/ivaa252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wang C., Chen J., Gu C., Li J. Analysis of survival after coronary endarterectomy combined with coronary artery bypass grafting compared with isolated coronary artery bypass grafting: a meta-analysis. Interact Cardiovasc Thorac Surg. 2019;29(3):393–401. doi: 10.1093/icvts/ivz125. [DOI] [PubMed] [Google Scholar]
- 21.Tiruvoipati R., Loubani M., Lencioni M., Ghosh S., Jones P.W., Patel R.L. Coronary endarterectomy: impact on morbidity and mortality when combined with coronary artery bypass surgery. Ann Thorac Surg. 2005;79(6):1999–2003. doi: 10.1016/j.athoracsur.2004.12.041. [DOI] [PubMed] [Google Scholar]
- 22.Silberman S., Dzigivker I., Merin O., Shapira N., Deeb M., Bitran D. Does coronary endarterectomy increase the risk of coronary bypass? J Card Surg. 2002;17(4):267–271. doi: 10.1111/j.1540-8191.2001.tb01138.x. [DOI] [PubMed] [Google Scholar]
- 23.Janiec M., Ragnarsson S., Nozohoor S. Long-term outcome after coronary endarterectomy adjunct to coronary artery bypass grafting. Interact Cardiovasc Thorac Surg. 2019;29(1):22–27. doi: 10.1093/icvts/ivy363. [DOI] [PubMed] [Google Scholar]
- 24.Nishigawa K., Fukui T., Yamazaki M., Takanashi S. Ten-year experience of coronary endarterectomy for the diffusely diseased left anterior descending artery. Ann Thorac Surg. 2017;103(3):710–716. doi: 10.1016/j.athoracsur.2016.11.028. [DOI] [PubMed] [Google Scholar]
- 25.Qiu Z., Auchoybur Merveesh L., Xu Y., et al. The midterm results of coronary endarterectomy in patients with diffuse coronary artery disease. J Cardiothorac Surg. 2018;13(1):90. doi: 10.1186/s13019-018-0776-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.LaPar D.J., Anvari F., Irvine J.N., Jr., et al. The impact of coronary artery endarterectomy on outcomes during coronary artery bypass grafting. J Card Surg. 2011;26(3):247–253. doi: 10.1111/j.1540-8191.2011.01247.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ahmad I., Islam M.U., Rehman M.U., Khan B. Frequency of intra-aortic balloon pump insertion and associated factors in coronary artery bypass grafting in a tertiary care hospital. Pak J Med Sci. 2021;37(2):393–397. doi: 10.12669/pjms.37.2.3614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Khan T.M., Siddiqui A.H. StatPearls Publishing; 2025. Intra-Aortic Balloon Pump. [Internet] [PubMed] [Google Scholar]
- 29.Akgun S., Selim Isbir C., Yildirim T., Civelek A., Arsan S. Results of adjunctive coronary endarterectomy in 548 patients. Heart Surg Forum. 2008;11(2):E66–E70. doi: 10.1532/HSF98.20071058. [DOI] [PubMed] [Google Scholar]
- 30.Zamir A.G., Iqbal A., Ahmad S.A., Rashid A., Khan A.A. Coronary endarterectomy for severe diffuse coronary artery disease. J Coll Physicians Surg Pak. 2009;19(6):338–341. [PubMed] [Google Scholar]
- 31.Kuebler W.M. Selectins revisited: the emerging role of platelets in inflammatory lung disease. J Clin Invest. 2006;116(12):3106–3108. doi: 10.1172/JCI30664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ashikhmina E.A., Schaff H.V., Sinak L.J., et al. Pericardial effusion after cardiac surgery: risk factors, patient profiles, and contemporary management. Ann Thorac Surg. 2010;89(1):112–118. doi: 10.1016/j.athoracsur.2009.09.026. [DOI] [PubMed] [Google Scholar]
- 33.Edmunds L.H., Jr. Blood-surface interactions during cardiopulmonary bypass. J Card Surg. 1993;8(3):404–410. doi: 10.1111/j.1540-8191.1993.tb00384.x. [DOI] [PubMed] [Google Scholar]
- 34.Dijkhuizen A., De Bruin R., Arbous S. Incidence, risk factors and outcome of TRALI after cardiac surgery. Crit Care. 2013;17(Suppl 2) [Google Scholar]
- 35.Vlaar A.P., Hofstra J.J., Determann R.M., et al. The incidence, risk factors, and outcome of transfusion-related acute lung injury in a cohort of cardiac surgery patients: a prospective nested case-control study. Blood. 2011;117(16):4218–4225. doi: 10.1182/blood-2010-10-313973. [DOI] [PubMed] [Google Scholar]
- 36.Ariyaratnam P., Javangula K., Papaspyros S., McCrum-Gardner E., Nair R.U. Long-term survival from 801 adjunctive coronary endarterectomies in diffuse coronary artery disease. Eur J Cardiothorac Surg. 2012;42(6):e140–e145. doi: 10.1093/ejcts/ezs510. [DOI] [PubMed] [Google Scholar]
- 37.Byrne J.G., Karavas A.N., Gudbjartson T., et al. Left anterior descending coronary endarterectomy: early and late results in 196 consecutive patients. Ann Thorac Surg. 2004;78(3):867–874. doi: 10.1016/j.athoracsur.2004.03.046. [DOI] [PubMed] [Google Scholar]
- 38.Piepoli M.F., Corrà U., Adamopoulos S., et al. Secondary prevention in the clinical management of patients with cardiovascular diseases. core components, standards and outcome measures for referral and delivery: a policy statement from the cardiac rehabilitation section of the European association for cardiovascular prevention & rehabilitation. Endorsed by the committee for practice guidelines of the European society of cardiology. Eur J Prev Cardiol. 2014;21(6):664–681. doi: 10.1177/2047487312449597. [DOI] [PubMed] [Google Scholar]
- 39.Patel M.R., Dehmer G.J., Hirshfeld J.W., Smith P.K., Spertus J.A. ACCF/SCAI/STS/AATS/AHA/ASNC 2009 appropriateness criteria for coronary revascularization: a report by the American college of cardiology foundation appropriateness criteria task force, society for cardiovascular angiography and interventions, society of thoracic surgeons, American association for thoracic surgery, American heart association, and the American society of nuclear cardiology endorsed by the American society of echocardiography, the heart failure society of America, and the society of cardiovascular computed tomography. J Am Coll Cardiol. 2009;53(6):530–553. doi: 10.1016/j.jacc.2008.10.005. [DOI] [PubMed] [Google Scholar]
- 40.Yan H., Tiemuerniyazi X., Song Y., Xu F., Feng W. Comparison of dual antiplatelet therapies after coronary endarterectomy combined with coronary artery bypass grafting: a cohort study. J Cardiothorac Surg. 2020;15(1):155. doi: 10.1186/s13019-020-01205-z. [DOI] [PMC free article] [PubMed] [Google Scholar]