Investigation of Combined Carotid Endarterectomy and Coronary Artery Bypass Graft Surgery Outcomes and Adverse Event Risk Factors in the Vascular Quality Initiative

Abstract

Objective:

The purpose of this study was to investigate outcomes of simultaneous CEA and CABG utilizing the Vascular Quality Initiative (VQI). Additionally, we seek to investigate risks for both perioperative and long-term mortality and adverse neurological outcomes.

Methods:

All carotid endarterectomies in the VQI between January 2003 and May 2022 were queried. We identified 171,816 CEA in the database. We extracted 2 cohorts from these CEA. The first group was patients who underwent simultaneous carotid endarterectomy (CEA) and coronary artery bypass (CABG) (N = 3137). The second group encompassed patients who underwent CABG or percutaneous coronary artery angioplasty/stent within 5 years of ultimately undergoing CEA (N = 27,387). We investigated the following outcomes in a multivariable fashion: 1. Risks for mortality in long term follow-up for both cohorts combined; 2. Risks for ischemic event in the cerebral hemisphere ipsilateral to the CEA site after index hospital admission in follow up for both cohorts combined. Tertiary outcomes are also investigated in the manuscript.

Results:

On multivariable analysis, patients undergoing simultaneous combined CEA and CABG had equivalent long-term survival to patients who underwent coronary revascularization within 5 years of ultimately undergoing CEA. Five-year survival is noted to be 84.5% vs 86% with a Cox regression non-significant P-value (.203). Significant multivariable risks for reduced long term survival (P < .03 for all) included: advancing age (HR 2.48/year); smoking history (HR 1.26); Diabetes (HR 1.33); history of CHF (HR 1.66); history of COPD (HR 1.54); baseline renal insufficiency at the time of surgery (HR 1.30); anemia (HR1.64); lack of preoperative aspirin (HR 1.12); and lack of preoperative statin (HR 1.32); lack of patch placement at CEA site (HR 1.16); perioperative MI (HR 2.04); perioperative CHF (1.66); perioperative dysrhythmia (HR 1.36); cerebral reperfusion injury (HR 2.23); perioperative ischemic neurological event (HR 2.48); and lack of statin at discharge (HR 2.04). Amongst patients with documented neurological status in follow up, combined CEA and CABG had over 99% freedom from ischemic cerebral event ipsilateral to the CEA site after discharge.

Conclusions:

Combined CEA and CABG provides excellent long-term mortality prevention in patients with co-existing severe coronary and carotid atherosclerosis. Simultaneous CEA and CABG provides equivalent stroke prevention and long-term survival to both a cohort of patients undergoing coronary revascularization within 5 years of CEA and patients undergoing isolated CEA or CABG in the literature. The two most impactful modifiable risk factors towards long-term stroke and mortality prevention for patients undergoing simultaneous CEA-CABG are patch placement at CEA site and adherence to statin medication therapy.

Introduction

Society for Vascular Surgery clinical practice guidelines note that there is weak evidence to guide the optimal management of concomitant significant carotid artery occlusive disease in the setting of severe coronary artery disease in need of bypass surgery (CABG).¹ Consideration towards simultaneous carotid endarterectomy (CEA) and CABG is recommended for patients with either symptomatic carotid stenosis greater than 50% or bilateral carotid stenosis over 70% whilst in need of CABG.¹

Other options for carotid occlusive disease meeting indication for intervention while requiring CABG include carotid artery stenting followed by CABG and CEA preceding CABG.^2,3 A recent Society of Thoracic Surgeons database analysis of perioperative outcomes of combined CABG and CEA for patients with over 80% carotid stenosis revealed equivalency for stroke and mortality for CABG with and without CEA. This analysis did not investigate long term outcomes.⁴ There is a notable lack of high-volume, randomized data on this clinical scenario. Management of the combination of these atherosclerotic pathologies thus remains controversial. The primary focus of prior investigations has been towards perioperative outcomes of stroke, myocardial infarction (MI) and mortality.

The purpose of this study was to investigate long term outcomes of simultaneous CEA and CABG utilizing the Vascular Quality Initiative (VQI). Using the database, we created a second cohort of patients who had undergone coronary revascularization within 5 years of CEA to be utilized in comparison. We hypothesized that combined CEA and CABG provides equivalent long term-stroke and mortality prevention relative to the analogous cohort we created. Similarly, we hypothesized that simultaneous CABG and CEA protects vs stroke in follow up and long term mortality with equivalency relative to patients with advanced carotid disease requiring CABG who were treated with different paradigms in the literature. Additionally, we seek to investigate risks for both perioperative and long-term mortality and adverse neurological outcomes.

Methods

Loyola University Chicago’s institutional review board approval was sought and waived for retrospective review of prospectively gathered de-identified data. VQI national research advisory board approval of this project was obtained. IBM SPSS version 28.0 software was used for statistical analysis.

All carotid endarterectomies in the VQI between January 2003 and May 2022 were queried. We identified 171,816 CEA in the database. We extracted 2 cohorts from these CEA. The first group was patients who underwent simultaneous CEA and CABG (N = 3137). The second batch encompassed patients who underwent CABG or percutaneous coronary artery angioplasty/stent (PCI) within 5 years of ultimately undergoing CEA (N = 27,387). VQI does not record PCI or CABG after an incident CEA, thus the second cohort we created is the most analogous group feasible with respect to simultaneous carotid stenosis and symptomatic coronary artery disease warranting intervention. Further there is no medical therapy cohort in VQI and thus comparison of patients undergoing medical management of significant carotid stenosis to those undergoing intervention is not feasible.

Pre-existing demographics and co-morbidities were included in the analysis along with key modifiable patient management variables. Variables in the VQI were merged into dichotomous (yes/no) status. Patients were placed into a category for any history of smoking, hypertension, diabetes mellitus, coronary artery disease (CAD), chronic obstructive pulmonary disease (COPD), and congestive heart failure (CHF). Chronic renal insufficiency was defined by a baseline creatine level of greater than 1.3 mg/dL. Anemia was defined by a hemoglobin level of less than 10 mg/dL. Symptomatic status at the time of CEA was defined by an ischemic stroke or TIA in the ipsilateral cerebral hemisphere within 183 days of surgery. Patients were categorized as no patch vs a single cohort of patch regardless of the material. Patients who underwent eversion endarterectomy were categorized as patch patients such that the “no patch” category applies exclusively to patients undergoing conventional endarterectomy without patch. Aspirin and P2Y12 inhibitor medications were merged into a single category for the presence of absence of any antiplatelet medication both at the time of discharge as well as at the time of long-term follow-up. Similarly, we investigated both discharge statin and statin usage at the time of follow up as separate variables. Surgeons are de-identified in the VQI database and thus whether the surgeon performing the CEA was trained as a vascular surgeon or a cardiothoracic surgeon is not available. All of the procedures were performed in centers which participate in the CEA module of the VQI.

We created a category for greater than 70% recurrent stenosis at the CEA site. Patients before 2012 in VQI were placed into re-stenosis categories by the individual institutional registrars. Duplex velocity criteria were recorded beginning in 2012. We categorized patients with either an ICA/CCA ratio of greater than 4 or those with an end diastolic velocity of over 100 cm/sec as having greater than 70% recurrent stenosis. The same velocity criteria were used to create a category of yes/no for contralateral stenosis over 70%. With respect to the ipsilateral carotid stenosis, we defined greater than 70% carotid stenosis as any patient with an ICA/CCA ratio of greater than 4, a CCA or ICA peak systolic velocity of greater than 100 cm/second, or a CTA, MRA or invasive angiogram interpretation as greater than 70% stenosis.

Perioperative cerebral ischemia in the VQI was not delineated based on laterality before 2012. Thereafter, hemispheric delineation was performed as was duration of symptoms. We merged these variables over the eras to create a single variable representing any cerebral ischemic event in the perioperative period. Cerebral ischemic events in this study included both transient ischemic attack (TIA) and ischemic stroke.

Comparison of baseline demographics, co-morbidities, and perioperative management variables was performed between the 2 cohorts (simultaneous CEA-CABG vs CABG/PCI within 5 years of CEA) in a univariable fashion. Mean age was investigated using t test and chi-squared testing was utilized for categorical variables of gender, presence/absence of baseline co-morbidities, and perioperative management variables.

We then investigated the following six outcomes:

1. Risks for mortality in long term follow-up for both cohorts combined

2. Risks for ischemic event in the cerebral hemisphere ipsilateral to the CEA site after index hospital admission in long-term follow up in both cohorts combined

3. Risks for mortality within 90 days of operation for both cohorts combined

4. Risks for ischemic neurological event during the index hospital admission for both cohorts combined

5. Risks amongst simultaneous CEA and CABG patients for the combined outcome of 90 day mortality and perioperative ischemic neurological event

6. Risks amongst simultaneous CEA and CABG patients for mortality in long term follow up

For outcomes #1 and 2 above, we first investigated univariable risks with chi-squared testing for categorical variables and t test for age. Those variables with significant univariable P-values of 0.05 or less were then utilized in time dependent multivariable Cox regression analysis for the given outcome. Again, CEA-CABG combined vs CABG/PCI preceding CEA as a dichotomous variable was investigated as 1 of the risk factors.

For outcomes #3 and 4 above, we first investigated univariable risks with chi-squared testing for categorical variables and t test for age. Those variables with significant univariable P-values of 0.05 or less were then utilized in multivariable binary logistic regression analysis for the given outcome. CEA-CABG combined vs CABG/PCI preceding CEA as a dichotomous variable was investigated as 1 of the risk factors.

For outcomes #5 and 6 above we investigated baseline demographics and co-morbidities utilizing binary logistic regression for the combined perioperative adverse event outcome (#5) and Cox regression for the long-term mortality outcome (#6).

The above multivariable analyses are an effective means to weigh discrepancies in the frequency of co-existing deleterious co-variates.

Kaplan Meier curves were created for visual aid in comparison of mortality and long-term ipsilateral cerebral ischemic event in the simultaneous CEA-CABG vs CAD revascularization proceeding CEA cohort. Curves were also created for freedom from mortality in patients discharged with and without statin medications after surgery and with and without patch angioplasty at the time of CEA. Finally, curves were created for age by decade category for patients undergoing simultaneous CEA and CABG.

Results

Comparison of baseline characteristics and perioperative management variables between the 2 cohorts is presented in Table 1. The simultaneous CEA-CABG cohort had statistically significant (P < 0.05) more: males (71.8% vs 67.7%), absence of preoperative aspirin (18.4% vs 11.9%), absence of preoperative statin (18.4% vs 11.9%), contralateral carotid stenosis over 70% (26.1% vs 19.6%), absence of patch at CEA site (9.2% vs 7.9%), and lack of statin at discharge (11.3% vs 10.1%). The coronary revascularization before CEA group had significantly (P < 0.05) more hypertension (94.4% vs 91.1%), lack of antiplatelet at discharge (10.5% vs 5.3%), lack of antiplatelet at long term follow up (LTFU) (44.9% vs 36.6%), and CEA being performed for symptomatic status (24.9% vs 13.5%). Amongst patients undergoing combined CEA-CABG 11.8% were treated with eversion endarterectomy. Amongst patients undergoing combined CEA-CABG 94.1% had a greater than 70% ipsilateral carotid stenosis confirmed preoperatively. There were 164 patients comprising the 5.9% of combined CEA and CABG patients without confirmed greater than 70% stenosis. Amongst these 164 patients, 40% (66/164) had no degree of stenosis listed in the database, and 84% (82/98) of the remaining patients were confirmed to have symptomatic carotid lesions.

Table 1.

Comparison of Characteristics and Treatment Variables Between the Combined CEA and CABG Cohort and the Coronary Revascularization within 5 Years of CEA Cohort.

Variable	Coronary Revascularization Prior to CEA (N = 27 387)	Combined CEA and CABG (N = 3137)	P-Value
Mean age	69.85 Years (STDEV 9.19)	68.79 Years (STDEV 8.81)	0.033
Male	67.7%	72%	<0.001
Hx of smoking	75.5%	76%	0.601
Hypertension	94.4%	91%	<0.001
Diabetes mellitus	45.5%	44%	0.268
Symptomatic carotid	24.9%	14%	<0.001
Hx of CHF	21.8%	23%	0.153
Hx of COPD	26.2%	25%	0.055
Hx of renal insufficiency	23.1%	22%	0.248
No pre-operative ASA	12.3%	17%	<0.001
No pre-operative statin	11.9%	18%	<0.001
Contralateral carotid stenosis >70%	19.6%	26%	<0.001
Yes patch used	92.1%	91%	0.01
Shunt used	50.9%	50%	0.227
Protamine used	72.5%	73%	0.657
Yes EEG used	28.2%	31%	<0.001
Stump pressure used	12.0%	8%	<0.001
No antiplatelet at D/C	10.5%	5%	<0.001
No statin at D/C	10.1%	11%	0.034
No antiplatelet at LTFU	44.9%	37%	<0.001
No statin at LTFU	32.5%	32%	0.892

Abbreviations: CEA = carotid endarterectomy; CABG = coronary artery bypass graft; Hx- = History of; CHF = Congestive Heart Failure; COPD = chronic obstructive pulmonary disease; D/C = discharge; EEG.

Risks for mortality in long term follow-up are summarized in Table 2. On multivariable analysis, patients undergoing simultaneous combined CEA and CABG had equivalent long-term survival to patients who underwent coronary revascularization within 5 years of ultimately undergoing CEA. Figure 1 displays the freedom from event curves for our 2 primary outcomes. Figure 1(A) represents 5 year survival curve comparison between the 2 cohorts. Five-year survival is noted to be 84.5% vs 86% with a Cox regression non-significant P-value (0.203). Figure 1(B) exhibits freedom from cerebral ischemia ipsilateral to the carotid endarterectomy site after initial hospitalization in long term follow up.

Table 2.

Risks for Mortality in Long Term Follow Up. Cox Regression Analysis Utilizing Significant Univariable Risks for Event for All Study Patients.

Variable		Long Term Mortality (n)	Univariable P-Value	Hazard Ratio (95% CI)	Multivariable P-Value
Mean age	Yes mortality in LTFU	71.73 YO (STDEV 8.95)	<0.001	2.48 (2.08 - 2.95)/year	<0.001
	No mortality in LTFU	69.50 YO (STEDV 9.09)
Gender	Male	10.2% (2114/20823)
	Female	11.2% (1095/9743)	0.004	1.06 (0.99 - 1.14)	0.120
Hx of smoking	Never smoking	8.5% (640/7535)
	Yes smoking	11.2% (2569/23031)	<0.001	1.26 (1.15 -1.37)	<0.001
Hypertension	No HTN	9.5% (172/1803)
	Yes HTN	10.6% (3037/28763)	.171
Diabetes mellitus	No DM	9.2% (1542/16700)
	Yes DM	12.0% (1667/13866)	<0.001	1.33 (1.24 - 1.43)	<0.001
Hx of CHF	No CHF	9.0% (2152/23876)
	Yes CHF	15.8% (1057/6690)	<0.001	1.66 (1.54 - 1.79)	<0.001
Hx of COPD	No COPD	9.1% (2062/22612)
	Yes COPD	14.4% (1147/7954)	<0.001	1.54 (1.43 - 1.66)	<0.001
Baseline renal insufficiency	No renal insufficiency	9.1% (2147/23525)
	Yes renal insufficiency	15.1% (1062/7041)	<0.001	1.30 (1.21 - 1.41)	<0.001
Anemia at surgery	No anemia	8.9% (2443/27354)
	Yes anemia	23.8% (766/3212)	<0.001	1.64 (1.49 - 1.79)	<0.001
Preoperative aspirin	No preop ASA	11.8% (458/3890)	0.005	1.12 (1.01 - 1.24)	0.029
	Yes preop ASA	10.3% (2751/26676)
Preoperative statin	No preop statin	13.3% (510/3842)	<0.001	1.32 (1.17 - 1.49)	<0.001
	Yes preop statin	10.1% (2699/26724)
Symptomatic carotid	No symptomatic carotid	10.5% (2448/23307)	0.962
	Yes symptomatic carotid	10.5% (761/7259)
Anesthesia group	Local/Regional	11.9% (257/2155)	0.025	1.06 (0.93 - 1.21)	0.388
	General	10.4% (2952/28411)
Contralateral carotid stenosis >70%	No contralateral stenosis	10.9% (2659/24376)	<0.001	1.04 (0.94 - 1.14)	0.441
	Yes contralateral stenosis	8.9% (550/6190)
Patch use	No patch use	27.9% (682/2444)	<0.001	1.16 (1.05 - 1.28)	0.004
	Yes patch use	9.0% (2527/28122)
Shunt use	No shunt use	10.7% (1612/15070)	0.265
	Yes shunt use	10.3% (1597/15496)
Protamine use	No protamine use	11.7% (984/8376)	<0.001	1.02 (0.94 - 1.10)	0.693
	Yes protamine use	10% (2218/22121)
Return to OR	No rexploration	10.4% (3134/30048)
	Yes rexploration	14.2% (67/471)	0.008	1.01 (0.79 - 1.29)	0.962
EEG	No EEG	10.0% (2179/21798)
	Yes EEG	11.7% (1020/8708)	<.001	1.1 (1.02 - 1.19)	.019
Stump pressure	No stump pressure	10.8% (2913/26967)	<0.001	1.07 (0.95-1.21)	0.279
	Yes stump pressure	8.1% (286/3535)
Postoperative MI	No postop MI	10.4% (3133/30252)
	Yes postop MI	24.2% (76/314)	<0.001	2.04 (1.61 - 2.58)	<0.001
Dysrhythmia	No dysrhythmia	10.2% (3034/29604)
	Yes dysrhythmia	18.1% (169/933)	<0.001	1.36 (1.15 - 1.61)	<0.001
Postoperative CHF	No CHF	10.3% (3132/30295)
	Yes CHF	28.8% (70/243)	<0.001	1.49 (1.17-1.91)	0.001
Cerebral reperfusion injury	No injury	10.4% (3179/30472)
	Yes injury	35.5% (22/62)	<0.001	2.23 (1.45 - 3.44)	<0.001
Combined CABG and CEA	No combined CABG and CEA	10.2% (2792/27387)
	Yes combined CABG and CEA	13.1% (410/3137)	<0.001	1.08 (0.96 - 1.21)	0.203
Perioperative ischemic neuro event	No event	10.2% (3064/29979)
	Yes event	24.7% (145/587)	<0.001	2.48 (2.08 - 2.95)	<0.001
Antiplatelet at D/C	No antiplatelet at D/C	11.8% (176/1488)	<.001	1.07 (0.997 - 1.16)	.059
	Yes antiplatelet at D/C	10.4% (3033/29018)
Statin at D/C	No statin at D/C	21.5% (672/3124)	<0.001	2.04 (1.82 - 2.29)	<0.001
	Yes statin on D/C	9.2% (2537/27442)
Smoking at LTFU	No smoking at LTFU	10.4% (2700/25845)
	Yes smoking at LTFU	10.8% (509/4721)	0.49
LTFU anticoagulation	No anticoagulation at LTFU	10.5% (2943/28033)	0.996
	Yes anticoagulation at LTFU	10.5% (266/2533)
LTFU antiplatelet	No antiplatelet at LTFU	10.4% (1424/13478)	0.73
	Yes antiplatelet at LTFU	10.6% (1785/17088)
LTFU statin	No statin at LTFU	10.5% (1039/9921)	0.919
	Yes statin at LTFU	10.5% (2170/20645)
Recurrent carotid stenosis >70%	No recurrent stenosis	10.5% (3124/29816)
	Yes recurrent stenosis	11.3% (85/750)	0.45
Ipsilateral ischemic event post DC	No ischemic event	10.3% (3111/30272)
	Yes ischemic event	33.3% (98/294)	<0.001	1.06 (0.85 - 1.31)	0.608

Abbreviations: CEA = carotid endarterectomy; CABG = coronary artery bypass graft; YO = years old; Hx− = History of; CHF = Congestive Heart Failure; COPD = chronic obstructive pulmonary disease; D/C = discharge; EEG = electroencephalogram; CAD = coronary artery disease; Revasc = myocardial infarction.

Figure 1.

(A) Freedom from mortality 5 year follow (B) Freedom from ischemic event ipsilateral to CEA site after index hospitalization in 5 year follow up.

Significant multivariable risks for reduced long term survival amongst baseline characteristics (P < 0.03 for all) included: advancing age; smoking history; Diabetes; history of CHF; history of COPD; baseline renal insufficiency at the time of surgery; anemia; lack of preoperative aspirin; and lack of preoperative statin. The following perioperative management variables and adverse events were associated with significantly increased long term mortality (P < 0.01 for all): lack of patch placement at CEA site; perioperative MI; perioperative CHF; perioperative dysrhythmia; cerebral reperfusion injury; perioperative ischemic neurological event; and lack of statin at discharge. Figure 2(A) depicts the survival curve impact of lack of patch at the CEA site. The separation of survival curves persisted significantly beyond the 5-year period and thus a 10 year curve is depicted. The impact of a lack of statin at the time of index hospitalization discharge is presented in a survival curve in figure 2(B) (HR 2.04, P < 0001).

Figure 2.

(A) 10 year survival in patients with respect to patch placement at carotid endarterectomy site (B) 5 year survival in patients with and without statin at time of discharge.

Table 3 displays the results of risk analysis for cerebral ischemic event ipsilateral to the CEA site in long term follow up after the index hospitalization. Patients undergoing simultaneous CEA and CABG had over 99% freedom from this event in follow up. However, there was significantly limited data on this particular variable with information available in the database on just 50% of patients at 1 year. Multivariable risks for ipsilateral cerebral ischemia in follow up (P < 0.04 for all) included: anemia at the time of surgery; symptomatic carotid lesion within 6 months of the time of surgery; lack of antiplatelet at discharge; lack of statin at discharge; lack of statin at LTFU; patients taking anticoagulation at LTFU; and smoking at time of LTFU.

Table 3.

Risks for Ischemic Event Ipsilateral to CEA in Follow Up. Cox Regression Analysis Utilizing Significant Univariable Risks for Event for All Study Patients.

Variable		Long Term Mortality (n)	Univariable P-Value	Hazard Ratio (95% CI)	Multivariable P-Value
Mean age	Age with ipsilateral neuro event LTFU	69.05 (STDEV 8.92)	0.192
	Age without ipsi neuro event LTFU	69.74 (STDEV 9.10)
Gender	Male	1.0% (198/20823)
	Female	1.0% (96/9743)	0.774
Hx of smoking	No smoking	1.0% (72/7535)
	Yes smoking	1.0% (222/23031)	0.948
Hx of HTN	No HTN	1.6% (29/1803)
	Yes HTN	0.9% (265/28763)	0.004	1.37 (0.93 - 2.01)	0.111
Hx of DM	No DM	1.0% (172/16700)
	Yes DM	0.9% (122/13866)	0.181
Hx of CHF	No CHF	1.0% (246/23876)	0.021	1.20 (0.88 - 1.64)	0.247
	Yes CHF	0.7% (48/6690)
Hx of COPD	No COPD	1.0% (224/22612)	0.385
	Yes COPD	0.9% (70/7954)
Baseline renal insufficiency	No renal insufficiency	0.9% (219/23525)
	Yes renal insufficiency	1.1% (75/7041)	0.311
Anemia at surgery	No anemia	0.5% (126/27354)
	Yes anemia	5.2% (168/3212)	<0.001	4.28 (3.28 - 5.59)	<0.001
Preoperative aspirin	No preop ASA	0.8% (33/3890)
	Yes preop ASA	1.0% (261/26676)	0.437
Preoperative statin	No preop statin	1.2% (47/3842)	0.076
	Yes preop statin	.9% (247/26724)
Symptomatic carotid stenosis	No symptoms	1.2% (283/23307)	<0.001	2.75 (1.47 - 5.16)	0.002
	Yes symptoms	0.2% (11/7259)
Anesthesia groups	Local/Regional	1.1% (23/2155)	0.603
	General	1.0% (271/28411)
Contralateral carotid stenosis >70%	No contralateral stenosis	1.1% (274/24376)	<0.001	1.04 (0.64 - 1.70)	0.876
	Yes contralateral stenosis	0.3% (20/6190)
Patch use	No patch use	8.2% (201/2444)	<0.0001	10.62 (7.84 - 14.38)	<0.001
	Yes patch use	0.3% (93/28122)
Shunt use	No shunt use	1.0% (154/15070)
	Yes shunt use	0.9% (140/15496)	0.289
Combined CABG and CEA	No combined CABG/CEA	1.0% (271/27387)
	Yes combine CABG/CEA	0.7% (23/3137)	0.164
Return to OR	No reexploration	0.9% (285/30048)
	Yes reexploration	1.9% (9/471)	0.034	1.19 (0.81 - 2.32)	0.615
EEG use	No EEG use	0.7% (161/21798)
	Yes EEG use	1.5% (133/8708)	<0.001	1.25 (0.98 - 1.58)	0.069
Stump pressure	No stump pressure	1.0% (283/26967)	<0.001	1.69 (0.92 - 3.12)	0.094
	Yes stump pressure	0.3%(11/3535)
Postoperative MI	No postop MI	1.0% (291/30252)
	Yes postop MI	1.0% (3/314)	0.991
Perioperative ischemic neuro event	Yes periop ischemic event	1.0% (293/29979)	0.047	5.83 (0.82 - 41.62)	0.079
	No periop ischemic event	0.2% (1/587)
Dysrhythmia	No dysrhythmia	1.0% (287/29 604)	0.5
	Yes dysrhythmia	0.8% (7/933)
Postoperative CHF	No postop CHF	1.0% (291/30295)
	Yes postop CHF	1.2% (3/243)	0.663
Cerebral reperfusion injury	No injury	1.0% (294/30472)
	Yes injury	0 (0/62)	0.437
Antiplatelet on DC	No antiplatelet on DC	0.7% (10/1488)	0.24
	Yes antiplatelet on DC	1.0% (284/29079)
Statin on DC	No statin on DC	3.2% (99/3124)	<0.001	1.42 (1.09 - 1.86)	0.01
	Yes statin on DC	0.7% (195/27442)
Statin at LTF	No statin at LTFU	0.7% (73/9921)
	Yes statin at LTFU	1.1% (221/20645)	0.005	1.66 (1.25 - 2.20)	<0.001
Smoking at LTF	No smoking at LTFU	0.9% (231/25845)
	Yes smoking at LTFU	1.3% (63/4721)	0.004	1.44 (1.08 - 1.91)	0.012
Anticoagulation at LTFU	No anticoagulation at LTFU	0.9% (248/28033)
	Yes anticoagulation at LTF	1.8% (46/2533)	<0.001	1.75 (1.27 - 2.41)	<0.001
Antiplatelet at LTFU	No antiplatelet at LTFU	1.0% (138/13478)	0.324	1.26 (.99 - 1.60)	0.063
	Yes antiplatelet at LTFU	0.9% (156/17088)

Abbreviations: CEA = carotid endarterectomy; CABG = coronary artery bypass graft; Hx- = History of; CHF = Congestive Heart Failure; COPD = chronic obstructive pulmonary disease; D/C = discharge; EEG = electroencephalogram; CAD = coronary artery disease; Revasc = revascularization; MI = myocardial infarction; LTFU = Long Term Follow Up.

Results for risks of mortality within 90 days for the entire study population are displayed in Supplemental Table 4. Combined CEA and CABG had an associated 4.4% 90-day mortality rate. Amongst pre-operative variables, lack of statin was the most potent risk factor for 90-day mortality (OR 3.35, P < 0.001). Other significant pre-existing risks on multivariable analysis for operative mortality included: advancing age, Diabetes, CHF, COPD, anemia, and symptomatic carotid lesion (all P < 0.01). Adverse events of perioperative events of: neurological ischemic event; myocardial infarction (MI); postoperative CHF; cerebral reperfusion injury; dysrhythmia; and lack of aspirin or statin at discharge (all P < 0.01) also had strong association with 90-day mortality. In the combined CEA and CABG cohort 70.9% of cases were deemed elective and had an associated 3.8% mortality rate. Urgent operative status was noted in 28.4% of the combined CEA-CABG patients with an associated 5.1% 90-day mortality risk. Finally, 0.7% of the CEA-CABG group underwent emergent operation with a 25% 90-day mortality.

Results for risks of neurological ischemic event during index hospital admission are listed in Table 5. Simultaneous CEA-CABG was associated with a 4.0% risk of ischemic neurological event perioperatively. Significant (P < 0.05) multivariable risks for the outcome included hypertension, Diabetes, absence of preoperative statin, symptomatic carotid lesion, and lack of protamine at the time of surgery. Adverse events of postoperative MI, dysrhythmia, CHF, and cerebral reperfusion injury similarly were associated with ischemic perioperative neurological event.

Table 5.

Risks for Neurologic Ischemic Event During Operative Admission. Binary Logistic Regression Utilizing Univariable Risks for All Patients Included in Study.

Variable		Perioperative Ischemic Neurological Event (n)	Univariable P-Value	Adjusted Odds Ratio (95% CI)	Multivariable P-Value
Mean age with neuro event	Mean age with neuro event	69.74 (STDEV 9.60)	0.981
	Mean age without neuro event	69.74 (STDEV 9.09)
Gender	Male	1.9% (390/20823)
	Female	2.0% (197/9743)	0.376
Hx of smoking	No hx of smoking	1.8% (133/7535)
	Yes hx of smoking	2.0% (454/23031)	.258
Hx of htn	No HTN	1.2% (21/1803)
	Yes HTN	2.0% (566/28763)	0.016	1.71 (1.10 - 2.68)	0.018
Hx of DM	No DM	1.7% (291/16700)
	Yes DM	2.1% (296/13866)	0.013	1.20 (1.01 - 1.42)	0.035
Hx of CHF	No CHF	1.9% (442/23876)
	Yes CHF	2.2% (145/6690)	0.096
Hx of COPD	No COPD	1.9% (423/22612)
	Yes COPD	2.1% (164/7954)	0.285
Hx of renal insufficiency	No renal insufficiency	1.9% (450/23525)
	Yes renal insufficiency	1.9% (137/7041)	0.86
Hx of anemia	No hx of anemia	2.2% (71/3212)
	Yes hx of anemia	1.9% (516/27354)	0.206
Preoperative aspirin	No preop ASA	2.3% (88/3890)	0.096
	Yes preop ASA	1.9% (499/26676)
Preoperative statin	No preop statin	2.4% (93/3842)	0.016	1.29 (1.02 - 1.62)	0.031
	Yes preop statin	1.8% (494/26724)
Symptomatic carotid lesion	No symptoms	1.6% (364/23307)
	Yes symptoms	3.1% (223/7259)	<0.001	2.11 (1.77 - 2.51)	<0.001
Anesthesia	Local/Regional	1.3% (27/2155)
	General anesthesia	2.0% (560/28411)	0.019	1.24 (0.83 - 1.86)	0.296
Contralateral carotid stenosis >70%	No contralateral stenosis	1.9% (471/24376)	0.766
	Yes contralateral stenosis	1.9% (116/6190)
Patch use	No patch use	2.0% (48/2444)	0.87
	Yes patch use	1.9% (539/28122)
Shunt use	No shunt use	2.1% (327/15496)	0.014	1.18 (0.97 - 1.43)	0.101
	Yes shunt use	1.7% (260/15070)
Protamine use	No protamine use	2.2% (182/8376)	0.049	1.22 (1.02 - 1.47)	0.029
	Yes protamine use	1.8% (404/22121)
EEG use	No EEG use	1.8% (396/21798)
	Yes EEG use	2.2% (191/8708)	0.031	1.16 (0.95 - 1.43)	0.148
Combined CEA and CABG	CEA only with prior CAD revasc	1.7% (461/27387)
	Yes combined CEA/CABG	4.0% (126/3137)	<0.001	2.37 (1.89 - 2.97)	<0.001
Stump pressure use	No stump pressure	2.0% (543/26967)	0.002	1.41 (1.02 - 1.94)	0.038
	Yes stump pressure use	1.2% (44/3535)
Postoperative MI	No postoperative MI	1.9% (567/30252)
	Yes postoperative MI	6.4% (20/314)	<0.001	2.33 (1.41 - 3.86)	0.001
Hx of dysrhythmia	No dysrhythmia	1.8% (535/29 604)
	Yes dysrhythmia	5.4% (50/933)	<0.001	1.52 (1.07 - 2.17)	0.019
Postoperative CHF	No postoperative CHF	1.9% (565/30295)
	Yes postoperative CHF	8.2% (20/243)	<0.001	2.51 (1.51 - 4.18)	<0.001
Cerebral reperfusion injury	No injury	1.8% (553/30472)
	Yes injury	50% (31/62)	<0.001	40.58 (23.80 - 69.22)	<0.001

Abbreviations: CEA = carotid endarterectomy; CABG = coronary artery bypass graft; Hx- = History of; CHF = Congestive Heart Failure; COPD = chronic obstructive pulmonary disease; D/C = discharge; EEG = electroencephalogram; CAD = coronary artery disease; Revasc = revascularization; MI = myocardial infarction.

Finally, analysis was performed to investigate just those patients undergoing simultaneous CEA and CABG with respect to baseline characteristics and the outcomes of combined perioperative cerebral ischemia and 90 day mortality (Table 6) and long term mortality (Table 7).

Table 6.

Risks Amongst Baseline Characteristics for Combined Outcome of Either Perioperative Ischemic Neurological Event or 90 Day Mortality for Patients Undergoing Simultaneous CEA and CABG. Binary Logistic Regression.

Variable		Percentage Experiencing Adverse event (n)	Adjusted odds ratio (95% CI)	Multivariable P-value
Mean age of those with adverse event	Yes neuro event or 90 Day mortality	74.32 YO (STDEV 7.43)	1.08/year (1.06-1.11)	<0.001
	No neuro event or 90 Day mortality	68.63 YO (STDEV 8.82)
History of smoking	Never smokers	4.1% (28/676)
	Yes smoking history	4.4% (92/2088)	1.295 (.814-2.06)	0.275
Hypertension	No hypertension	2.7% (7/257)
	Yes hyptertension	4.5% (113/2507)	1.34 (0.603-2.97)	0.474
Diabetes mellitus	No diabetes	4.4% (68/1557)	1.16 (0.756-1.64)	0.583
	Yes diabetes	4.3% (52/1207)
History of congestive heart failure	No CHF history	3.0% (65/2152)
	Yes CHF history	9.0% (55/612)	2.61 (1.77-3.85)	<0.001
Chronic obstructive pulmonary disease	No COPD	4.1% (85/2097)
	Yes COPD	5.2% (35/667)	1.15 (0.744-1.77)	0.534
Anemia (HGB <10 mg/dl)	No anemia	3.8% (90/2343)
	Yes anemia	7.7% (30/421)	1.37 (0.871-2.14)	0.175
Gender	Male gender	3.8% (76/1989)
	Female gender	5.7% (44/775)	1.58 (1.05-2.37)	0.028
Contralateral carotid stenosis >70%	No contralateral stenosis >70%	4.5% (91/2040)	1.06 (0.680-1.64)	0.811
	Yes >70% contralateral stenosis	4.0% (29/724)
Chronic renal insufficiency	No renal insufficiency	3.3% (72/2153)
	Yes renal insufficiency	7.9% (48/611)	1.76 (1.17-2.64)	0.006
Symptomatic carotid lesion within 6 Months	Asymptomatic carotid	4.1% (99/2289)
	Symptomatic carotid	5.6% (21/376)	1.68 (1.02-2.77)	0.042

Table 7.

Risks for Mortality in Follow Up Amongst Baseline Characteristics of Those Undergoing Simultaneous CEA and CABG. Cox Regression Analysis.

Variable		Percent Experiencing Mortality in Follow Up (n)	Hazard Ratio (95% CI)	Multivariable P-Value
Mean age	Yes mortality	72.49 YO (STDEV 8.07)	1.05/year (1.04-1.07)	<0.001
	No mortality	68.32 YO (STDEV 8.82)
Smoking history	Never smokers	11.1% (75/676)
	Yes smoking history	14% (292/2088)	1.27 (0.98-1.66)	0.075
Hypertension	No hypertension	12.8% (33/257)
	Yes hypertension	13.3% (334/2507)	1.12 (0.78-1.61)
Diabetes mellitus	No diabetes	12.3% (192/1557)
	Yes diabetes	14.5% (175/1207)	1.13 (0.92-1.40)	0.242
Congestive heart failure history	No CHF	11.1% (239/2152)
	Yes CHF	20.9% (128/612)	1.61 (1.29-2.01)	<0.001
Chronic obstructive pulmonary disease	No COPD	11.4% (239/2097)
	Yes COPD	19.2% (128/667)	1.69 (1.35-2.12)	<0.001
Anemia (HGB <10 mg/dl)	No anemia	11.2% (262/2343)
	Yes anemia	24.9% (105/421)	1.64 (1.30-2.07)	<0.001
Gender	Male gender	12.6% (250/1989)
	Female gender	15.1% (117/775)	1.14 (0.91-1.43)	0.27
Contralateral carotid stenosis >70%	No contralateral stenosis >70%	13.6% (277/2040)	1.08 (0.85-1.37)	0.527
	Yes >70% contralateral stenosis	12.4% (90/724)
Chronic renal insufficiency	No renal insufficiency	10.9% (234/2153)
	Yes renal insufficiency	21.8% (133/611)	1.48 (1.19-1.86)	<0.001
Symptomatic carotid lesion within 6 Months	No symptomatic carotid	13.3% (317/2388)
	Yes symptomatic carotid	13.3% (50/376)	1.44 (1.07-1.95)	0.018

Abbreviations: CEA = carotid endarterectomy; CABG = coronary artery bypass graft; CHF = congestive heart failure; COPD = chronic obstructive pulmonary disease; HGB = hemoglobin; YO = years old.

Significant baseline characteristics associated with either perioperative cerebral ischemia or 90 day mortality for those undergoing combined CEA and CABG (multivariable P < 0.05 for all) included: advancing age; history of CHF; female gender; baseline renal insufficiency; and symptomatic carotid lesion within 6 months of surgery.

Multivariable risks for mortality in long term follow up amongst those undergoing simultaneous CEA and CABG included: advancing age; history of CHF; anemia, baseline renal insufficiency; and symptomatic carotid lesion at the time of surgery (P < 0.02 for all). Figure 3 stratifies long term mortality by age decades with notation of decline in survival with aging that increases most significantly after 80 years old.

Figure 3.

Long term survival by age at surgery for combined CEA and CABG.

For the long-term survival analysis the mean follow up was 3096 days (8.48 years). For follow up neurological event data points the mean duration from surgery to the data point was 396 days. This discrepancy exists because survival data in VQI is captured both by institutional abstractors and social security death index matching while neurological outcomes are via abstractors only.

Discussion

Severe coronary artery disease requiring revascularization frequently coexists with high grade or symptomatic carotid artery stenosis. There is debate regarding whether a staged or synchronous surgical approach is appropriate to treat the combined pathologies. Herein, we utilized the VQI to assess the perioperative and long-term outcomes following combined CEA and CABG. We then compared these results to patients who underwent coronary revascularization (CABG or PCI) within 5 years of CEA. The VQI does not record the occurrence of CABG after CEA and thus the cohort we created is the best representation of patients who had both coronary and carotid atherosclerosis of such severity that they required intervention on both within 5 years. We proved equivalency between the cohorts for our 2 primary outcomes as hypothesized (Figure 1). We found no difference in long term survival between the 2 cohorts. Similarly, there was no difference in the rate of ischemic event ipsilateral to CEA site after index hospital admission between the 2 cohorts.

Perioperative Cerebral Ischemia and Mortality

With respect to perioperative outcomes, a true comparison of the cohorts would require combining the stroke and mortality rates of the CABG and the subsequent CEA in the second cohort to the perioperative simultaneous CEA-CABG outcomes. This isolated CABG outcome data is not available in VQI, but literature review does allow for meaningful comparison. In our study, simultaneous CABG and CEA was associated with 4.4% 90-day mortality and 4.0% perioperative ischemic cerebral ischemic event rate (Stroke and TIA combined). Prior studies have documented the risk of post-operative stroke after CABG in those with bilateral high-grade carotid stenosis to increase from 2% in those without significant carotid disease to 5%.⁵ A met-analysis by Chan et al⁶ included 67953 patients from 11 articles between 1997 – 2017. Their study showed the combined CEA and CABG to have a 30-day mortality rate of 4.4%, and post-operative stroke in 3.64%. They identified no difference in mortality at 1 year follow-up relative to staged approaches. In a meta-analysis of 16712 patients undergoing carotid stenting followed by CABG vs combined CEA-CABG equivalent perioperative stroke rates were identified with combined stroke and TIA rates of approximately 4% in both groups. The same study noted increased perioperative mortality with the combined approach (OR 1.8).⁷ Perioperative stroke has a clear immediate mortality association in our study (13% vs 1.4%) which is consistent with the adverse effect of stroke after CEA in the literature.⁸

Prior single institution studies have documented relative perioperative equivalence of these complex patients requiring either staged CEA and CABG or combined CEA-CABG.^9–12 In a robust analysis which included a cohort of patients with significant cerebrovascular disease undergoing CABG without simultaneous CEA, perioperative stroke and mortality rates were noted to be 2.7 and 3.3%, however these carotid lesions were not symptomatic.¹³ Fifteen percent of combined CEA-CABG patients in our study were symptomatic. Further, TIA combined with stroke is included in the perioperative cerebral ischemic event outcome in our study. Therefore, the perioperative stroke rate is less than 4.0%. Overall, we thus believe the combined CEA-CABG cohort perioperative outcomes herein indicate quality safety with respect to perioperative stroke and mortality compared to individuals treated in sequence in the literature. The perioperative neurological outcomes in this study are consistent with prior pooled analysis of combined CEA and CABG summarized in the European Society for Vascular Surgery Guidelines.¹⁴ Regardless of approach, patients requiring CABG with co-existing severe carotid atherosclerosis present a significantly higher risk group of patients than individuals who undergo CABG whilst not having severe carotid stenosis.

Long-Term Freedom From Mortality and Post Discharge Freedom From Cerebral Ischemia

At long term follow-up, our study indicates combined CEA-CABG shows no significant difference in mortality or neurologic ischemic event in comparison to patients undergoing CEA who underwent coronary revascularization within 5 years (Figure 1). We noted patients undergoing combined CEA-CABG to have a 5 year survival rate of over 84%. This is similar to isolated CABG in which long term survival approximates 86% at 5 years and freedom from late ipsilateral neurologic event approximates 98%.^15,16 Together, this suggests that those undergoing combined CEA/CABG do not have increased perioperative or long-term adverse cerebral events when in comparison to CABG alone. Equally or more supportive of combined CEA and CABG is the fact that the 5 year-survival rate of 84% herein is equivalent to the 5 year survival rate of patients in the VQI undergoing CEA in isolation for asymptomatic stenosis irrespective of CAD status (a cohort with a widely held lower co-morbidity index).¹⁷ We thus conclude combined CEA/CABG is an excellent option for these combined pathologies with high quality long mortality prevention and post hospital discharge ipsilateral cerebral ischemia prevention.

Impact of Perioperative Variables and Demographics on Outcomes

Variables related to adverse outcomes included no lack of statin at hospital discharge, lack of patch at carotid endarterectomy site, symptomatic carotid lesion, advanced age, female gender, history of CHF, history of chronic renal insufficiency and anemia (Hgb <10).

No statin at discharge was associated with increased 90-day mortality, long term mortality, perioperative ischemic event and post discharge ipsilateral ischemic event. A survival curve is depicted in Figure 2(B) demonstrating 5-year survival for no statin at discharge to approximate 79% while statin at discharge approximates 89%. This is of little surprise as the adoption of statin use has improved medical outcomes in patients with high-grade asymptomatic carotid artery stenosis and in patients undergoing CABG.^18,19 It is also interesting to note that in Minisandram et al¹⁹’s retrospective study reviewing the 30-year experience with combined CABG/CEA, they found perioperative stroke decreased in later years of CABG/CEA, at a time when statin use was also increasing. Therefore we recommend all patients without contraindications should be discharged with statin to reduce perioperative and long term adverse events.

Absence of patch placement at the CEA site was associated with increased long-term mortality and post discharge neurologic ischemic event. Figure 2(A) depicts 10-year survival for no patch placement to approximate 71% while patch placement approximates 80%. We recently identified lack of patch placement at CEA to be the lead predictor of long term ipsilateral ischemic event after index hospital admission amongst 171 816 CEA in the VQI.²⁰ This finding held true in this study for those with severe coronary disease. The European Society for Vascular Surgery Guidelines recommend routine patch use over primary as multiple systematic reviews have demonstrated significant reduction in restenosis and ipsilateral stroke.²¹ Other US studies have demonstrated similar findings of reduced long-term rates of ipsilateral stroke, reduced risk of stroke or death and reduced risk of restenosis.²² Patch placement at the CEA site in combined CABG-CEA is thus recommended based on our findings.

Symptomatic carotid lesion within 6 months was associated with increased 90-day mortality and perioperative ischemic event. Symptomatic carotid lesion was associated with a 5.6% combined perioperative neurologic event or 90 day mortality (Table 6) relative to 4.1% in those with asymptomatic carotid status for those undergoing combined CEA-CABG. This is of little surprise as symptomatic carotid lesions have been long identified with increased perioperative stroke risk following CEA relative to asymptomatic lesions.^23–25 Symptomatic status for the carotid lesion did not impact long term survival or ipsilateral cerebral ischemic events after discharge in both cohorts combined but it did hit multivariable significance for the CEA-CABG group in isolation (Table 6, Table 7). Interestingly, contralateral carotid artery stenosis was not associated with any adverse outcomes on the six outcomes in our study. Previous studies have demonstrated increased risk of stroke in patients with bilateral carotid artery stenosis.²⁴

Advanced age was associated with perioperative mortality, long term follow-up mortality, risk for 90-day mortality combined with 90-day ischemic event for combined CEA/CABG, and risk for long term mortality for combined CEA/CABG. Figure 3 nicely displays the increased long-term mortality associated with progressive age decade in those undergoing combined CEA-CABG in our study. In our previous investigation regarding risk for 5 year mortality in patients with asymptomatic carotid stenosis undergoing CEA, we found patients over 80 with significant co-morbidities to be at over 40% risk of mortality.²⁵ Patients over 80 undergoing isolated CABG are also at increased risk of mortality.²⁶ This should not be interpreted to conclude that octogenarians are not candidates for combined CEA-CABG as there is no medically managed cohort for comparison and these patients are presenting with clinically advanced coronary and carotid disease. Rather, the data reflect these pathologies are associated with increased mortality rate when combined with the aging process that induces frailty and other co-morbidity development.

Female gender was significantly associated with combined risk of 90-day mortality and neurologic ischemic event for our combined CEA/CABG cohort (Table 6). Regarding isolated CEA irrespective of CAD, our previous study did not show gender differences either.¹⁶ For isolated CABG, studies have shown female gender is an independent indicator for early and late mortality.^26,27 The putative reason being older average age, smaller coronary artery diameter and increased co-morbidities. Therefore, gender differences in our combined CEA/CABG cohort is likely attributable to the coronary disease and CABG component of their pathologies.

History of CHF or chronic renal insufficiency were both associated with 90-day mortality for all patients combined in our study. Similarly, CHF or renal insufficiency history in the combined CEA-CABG cohort in isolation were associated with the combined 90-day mortality-neurologic ischemic event and long-term morality. (Table 4, Table 6). Previous studies have indicated that in both isolated CEA and isolated CABG, patients with CHF and chronic renal insufficiency are at higher risks of adverse outcomes.^16,28

Table 4.

Risks for Mortality within 90 Days of Operation. Binary Logistic Regression Utilizing Univariable Risks for All Patients Included in Study.

Variable		Perioperative Mortality (n)	Univariable P-Value	Adjusted Odds Ratio (95% CI)	Multivariable P-Value
Mean age of those with mortality	Yes mortality within 90 Days	72.61 Years (STDEV 9.37)	<0.001	1.03 (1.02 - 1.05)/YR	<0.001
	No mortality within 90 Days	69.69 Years (STDEV 9.08)
History of smoking	Never smoker	1.5% (112/7535)	0.229
	Active or former smoker	1.7% (389/23031)
Gender	Male	1.5% (313/20823)
	Female	1.9% (188/9743)	0.006	1.21 (.99 - 1.48)	0.063
Diabetes mellitus	No DM	1.4% (227/16700)
	Yes DM	2.0% (274/13866)	<0.001	1.32 (1.09-1.60)	0.005
Hypertension	No HTN	1.6% (29/1803)	0.916
	Yes HTN	1.6% (472/28763)
Hx of CHF	No CHF	1.2% (284/23876)
	Yes CHF	3.2% (217/6690)	<0.001	2.17 (1.78-2.65)	<0.001
Hx of COPD	No COPD	1.4% (324/22612)
	Yes COPD	2.2% (177/7954)	<0.001	1.46 (1.19 - 1.78)	<0.001
Baseline renal insufficiency	No renal insufficiency	1.3% (317/23525)
	Yes renal insufficiency	2.6% (184/7041)	<0.001	1.46 (1.19 - 1.80)	<0.001
Anemia at surgery	No anemia	1.4% (382/27354)
	Yes anemia	3.7% (119/3212)	<0.001	1.45 (1.15 - 1.84)	0.002
Pre-operative aspirin	No ASA	1.9% (74/3890)	0.166
	Yes ASA	1.6% (427/26676)
Pre-operative statin	No statin	2.1% (81/3842)	0.014	3.35 (2.47 - 4.56)	<0.001
	Yes statin	1.6% (420/26724)
Symptomatic carotid stenosis	No symptoms	1.5% (338/23307)
	Yes symptoms	2.2% (163/7259)	<0.001	1.76 (1.43 -2.17)	<0.001
Anesthesia	Local regional	1.8% (38/2155)	0.637
	General anesthesia	1.6% (463/28411)
Contralateral carotid stenosis >70%	No contralateral stenosis	1.6% (398/24376)
	Yes contralateral stenosis	1.7% (103/6190)	0.863
Patch use	No patch use	2.0% (50/2444)	0.099
	Yes patch use	1.6% (451/28122)
Shunt use	No shunt use	1.6% (245/15070)	0.856
	Yes shunt use	1.7% (256/15496)
Protamine use	No protamine use	1.6% (138/8376)	0.923
	Yes protamine use	1.6% (361/22121)
Return to operating room	No return	1.6% (486/30048)
	Yes return	3.2% (15/471)	0.008	1.4 (.80 - 2.50)	.234
EEG use	No EEG use	1.7% (361/21798)	0.71
	Yes EEG use	1.6% (139/8708)
Stump pressure use	No stump pressure	1.7% (458/26967)	0.025	1.20 (0.86 - 1.66)	0.295
	Yes stump pressure	1.2% (42/3535)
Combined CEA and CABG	No (CEA only, prior CAD tx)	1.3% (364/27387)
	Yes (combined surgery)	4.4% (137/3137)	<0.001	2.48 (1.93 - 3.18)	<0.001
Perioperative cerebral ischemic event	No periop neuro event	1.4% (424/29979)
	Yes periop neuro event	13.1% (77/587)	<0.001	5.45 (3.98 - 7.47)	<0.001
Postoperative MI	No post-op MI	1.5% (461/30252)
	Yes post-op MI	12.7% (40/314)	<0.001	4.51 (2.93 - 6.94)	<0.001
Postoperative dysrhythmia	No postop dysrhythmia	1.4% (417/29604)
	Yes postop dysrhythmia	8.6% (80/933)	<0.001	2.22 (1.61 - 3.10)	<0.001
Postoperative CHF	No postop CHF	1.5% (463/30 295)
	Yes postop CHF	13.6% (33/243)	<0.001	2.15 (1.34 - 3.42)	0.001
Cerebral reperfusion injury	No injury	1.6% (479/30 472)
	Yes injury	25.8% (16/62)	<0.001	4.15 (2.02 - 8.52)	<0.001
Antiplatelet on D/C	No antiplatelet	1.7% (486/29078)	0.06
	Yes antiplatelet	1.0% (15/1488)
Statin on D/C	No statin	6.3% (197/3124)	<0.001	8.91 (6.93 - 11.46)	<0.001
	Yes statin	1.1% (304/27442)

Abbreviations: CEA = carotid endarterectomy; CABG = coronary artery bypass graft; Hx- = History of; CHF = Congestive Heart Failure; COPD = chronic obstructive pulmonary disease; D/C = discharge; EEG = electroencephalogram; CAD = coronary artery disease; Revasc = revascularization; MI = myocardial infarction; YR = year; Periop = perioperative.

We found anemia to be strongly associated with perioperative 90-day mortality, long term mortality, and neurologic ischemic event in follow up amongst all study patients (Table 2, Table 3, Table 5) and long-term morality in the isolated combined CABG/CEA cohort (Table 7). Anemia has been used as a surrogate for frailty and has been predictive of mortality for those undergoing isolated CEA.^16,29 Similar findings have been reported for those experiencing acute coronary syndrome.^30–32 Here we also confirm its predictive nature.

Study Strengths and Weaknesses

We identified some weaknesses in our study. There is no data in the carotid module on postoperative renal function and thus this information is not available in the study. Because VQI does not record the occurrence of coronary intervention after an index CEA in the carotid module, we created a control cohort of patients who underwent coronary revascularization (either CABG or PCI) within 5 years of ultimately undergoing CEA. While this does create a group of patients who have severe coronary and carotid atherosclerosis, they are not necessarily presenting simultaneously. This is thus an imperfect comparison. That stated, the long-term stroke prevention and mortality prevention documented in this study with combined CEA-CABG is very favorable towards stroke and mortality prevention in patients undergoing either CEA or CABG in isolation (without the other pathology being present) in the literature. Additionally, VQI is a collection of patients who underwent surgery. There is no medical therapy control or comparison. As medical therapy has developed enhanced stroke reduction for patients with asymptomatic carotid stenosis, the benefit may be reduced. With respect to asymptomatic patients, there is evidence in the literature with respect to ipsilateral stroke risk in those undergoing major cardiac surgery. Specifically, in a meta-analysis by Naylor and Bown, the risk of ipsilateral stroke in patients with a unilateral, asymptomatic 50% to 99% carotid stenosis undergoing CABG was found to be 2%.

As another weakness, VQI is unable to capture patients that may have presented to an outside hospital with stroke at any time after index hospital admission which may falsely deflate the long term ipsilateral cerebral ischemic event rate. While VQI provides high volume long-term mortality data points due to social security death index matching, there is a steep decline in the long term follow up neurological events data after 1 year. For the mortality analysis the mean follow up was 3096 days relative to a mean time to data point of just 396 days on the neurological outcomes data. These last 2 study weaknesses limit the statistical power in making long term neurological event conclusions.

It is worth noting that amongst those patients undergoing combined CEA and CABG only 26% had confirmed bilateral high grade stenosis and 14% were confirmed to have symptomatic carotid lesions. Thus 60% of the patients did not meet the most recent SVS guidelines with respect to indication for the combined procedure.¹ That stated, the SVS guidelines on the topic note poor quality evidence in support of the guideline. This supports the need for data such as that herein which document quality long term outcomes for the combined procedures in patients with advanced combined carotid and coronary atherosclerotic disease.

The relative strengths of this study include high volume analysis with statistical power regarding long term survival, this being the first long term analysis of combined CEA-CABG outcomes utilizing the VQI, and a meaningful data set in support of continued usage of the combined approach. Fortunately, long term ischemic events after CEA are rare, and thus a large study volume is necessary to identify meaningful risks with statistical significance. We created a novel comparison cohort utilizing patients who had undergone coronary revascularization before subsequent CEA and the combined CEA and CABG group is documented herein to have equivalent mortality protection in long term follow up.

Conclusions

Combined CEA and CABG provides excellent long-term mortality prevention in patients with co-existing severe coronary and carotid atherosclerosis. Simultaneous CEA and CABG provides equivalent stroke prevention and long-term survival to both a cohort of patients undergoing coronary revascularization within 5 years of CEA and patients undergoing isolated CEA or CABG in the literature. The 2 most impactful modifiable risk factors towards long-term stroke and mortality prevention for patients undergoing simultaneous CEA-CABG are patch placement at CEA site and adherence to statin medication therapy.