The role of carotid artery stenosis in predicting stroke after coronary artery bypass grafting in a Chinese cohort study

Current guidelines give priority to surgical treatment of carotid artery stenosis (CAS) before coronary artery bypass grafting (CABG), especially in symptomatic patients. Carotid artery stenting is an alternative treatment for narrowing of the carotid arteries. This study sought to demonstrate the role of severe CAS in predicting stroke after CABG and assess the efficacy of carotid artery stenting in preventing postoperative stroke in a Chinese cohort. From 2015 to 2021, 1799 consecutive patients undergoing isolated CABG surgery were retrospectively recruited in a Chinese cohort. The predictive value of severe CAS in postoperative stroke and carotid stenting in preventing postoperative stroke was statistically analyzed. The incidence of postoperative stroke was 1.67%. The incidence of CAS with stenosis ≥ 50% and ≥ 70% was 19.2% and 6.9%. After propensity matching, the incidence of stroke was 8.0% in the severe CAS group and 0% in the non-severe CAS group. We successfully established an optimal predictive nomogram for predicting severe CAS in patients undergoing CABG. Carotid artery stenting was found ineffective in preventing postoperative stroke. The present study provides the incidence of CAS and postoperative stroke in a Chinese cohort, identifies severe CAS as an independent risk factor for postoperative stroke after CABG, constructs a nomogram predicting the incidence of severe CAS, and evaluates the effectiveness of carotid artery stenting in preventing postoperative stroke after CABG.


CABG and support techniques
The CABG surgical approach (off-pump CABG: median sternotomy or anterior or lateral minimally invasive direct coronary artery bypass [MIDCAB] approach) is determined by the patient cardiac symptoms and the severity of the CAD or CAS.In most cases in this series, the median sternotomy approach is the standard strategy.The MIDCAB approach is preferred for grafting isolated proximal disease of the left anterior descending branch or the first diagonal artery.The operation of carotid artery stenting is determined by the neurologist's assessment and the severity of the CAS (≥ 70%).Percutaneous carotid artery stenting (transfemoral approaches) is performed in most cases one week before CABG.All patients were treated with dual antiplatelet (DAPT) therapy with aspirin and clopidogrel before receiving carotid artery stenting for two days and after the procedure.All carotid artery stent implantation procedures in this study were performed by the same neurosurgeon and their team, using conventional interventional techniques.The patient selection criteria for carotid artery stent implantation in this study were confirmed vascular stenosis of 70% or more by cerebral angiography or clinical symptoms with vascular stenosis of 50% or more.

Nomogram
Nomograms are increasingly used for prognostic analysis as a simpler, more intuitive, and more advanced method 21 .Univariate and multivariate logistic regression models were used to explore potential factors and estimate their weights by severe CAS.Variables with p < 0.05 and those possible predictor variables in the univariate model were entered into the multivariate logistic regression model and Forward: LR was performed with probabilities of entry and removal of 0.05 and 0.10, respectively.Based on these important risk factors, the screening was candidate nomogram models with appropriate predictive power.The predictive performance of the nomogram and other models to predict severe CAS rates was quantified using the area under the curve (AUC).The capability of the nomogram was also tested by fitting a well-calibrated curve.The clinical utility of nomograms was also carefully investigated using decision curve analysis (DCA) to compensate for the limitation that ROC curves do not achieve optimal sensitivity and specificity simultaneously.

Propensity matching
To assess the association between severe CAS and stroke, we used propensity matching to group patients with and without severe CAS.A parsimonious model was first developed using multivariate logistic regression, as described in the previous section.Clinically correlated variables that were not found to be substantially relevant to postoperative stroke were then appended to the fitted model to generate a propensity-matched model.The propensity matching model was quantified using Kdensity plots and propensity matching test plots.) in occluded unilateral groups, and 0 of 3 (0%) in occluded bilateral groups (Fig. 2).There was a statistically significant difference in the incidence of postoperative stroke in the severe CAS group (P < 0.001), severe unilateral CAS group (P = 0.007), severe bilateral CAS group (P = 0.001), and occluded unilateral CAS group (P = 0.033) compared with all patients.It indicates that the greater the degree of stenosis, the higher the incidence of perioperative stroke, especially in the severe unilateral, severe unilateral, and occluded unilateral groups.

Risk factors for severe CAS
Severe carotid stenosis (70%-99%) was found in 117 of 1702 patients in our patient population (6.87%).CAS.And multivariate logistic regression analysis found the following independent risk factors for severe CAS: age ≥ 65 years, SBP, DBP, CKD, and preoperative hemoglobin.

Nomogram model for severe CAS
To develop an optimal nomogram model, we evaluated the individual and combined performance of these five factors using ROC analysis (Fig. 3).The individual AUCs for age ≥ 65 years, SBP, DBP, CKD, preoperative hemoglobin, and prediction model were 0.598, 0.558, 0.554, 0.518, 0.606, and 0.681, respectively.In addition, as seen in Fig. 4, each of the chosen biomarkers was assigned a proportional score based on its value on the nomogram.
To confirm the generality of the nomogram in predicting the incidence of severe CAS, we screened the nomogram for comprehensive validation.The well-fitted calibration curves showed high agreement in predicting the incidence of severe CAS, as shown in Fig. 5 (Hosmer-Lemeshow P value = 0.982).Also, DCA curves were created in Fig. 6.Regardless of the threshold, the nomogram model performed well across the various predictors, which ensured that maximum clinical benefit was achieved.Overall, the DCA curves indicate that the nomogram model is feasible to make a valuable and favorable assessment.

Propensity matching
To assess the association between severe CAS and postoperative stroke, we used a propensity-matched approach to group patients with and without severe CAS.A total of 8 factors (including age ≥ 65 years, male, hypertension, peripheral vascular disease, cerebrovascular accident, chronic kidney disease, left main artery disease, and carotid stenting) were included to form a propensity-matched model.The propensity matching model was quantified using Kdensity plots (Fig. 7 shows the Kdensity before matching and Fig. 8 shows the Kdensity after matching) and propensity matching test plots (Fig. 9).These tests showed that the two groups after propensity matching achieved a good match.The clinical and demographic characteristics of the patients after propensitymatched analysis by severe CAS were shown in Table 4.In univariate logistic regression analysis, a statistically significant difference in ICU length of stay (p = 0.041) was found.After matching, stroke occurred in 6 of 75 patients in the severe CAS group (8.0%[95% CI, 1.86%-14.14%])and in 0 of 75 patients in the non-severe CAS group (0%) (p = 0.028).

Evaluating of carotid artery stenting
The symptomatic status of 13 patients with severe CAS was shown in

Discussion
Stroke has continued to be one of the potentially destructive complications of CABG, with significant clinical and economic implications for patients and healthcare systems 6,22,23 .In previous studies, incidence of stroke was 3.0% 24 , 1.7% 25 , 1.6% 6 among patients undergoing CABG mainly in American and European area.Over the past 40 years, the incidence of stroke has declined despite increasing patient risk profiles, which benefited from improvements in preoperative assessment, surgical techniques, and postoperative care.The incidence and risk of CAS in Chinese patients undergoing CABG is still unclear.Thus, we compiled 1799 consecutive patients who underwent isolated CABG surgery at Shandong Provincial Hospital.We observed a similar incidence of postoperative stroke (1.67%) in our Chinese CABG cohort.Patients with stroke after CABG had a longer hospital stay and longer time in the ICU than patients without stroke.However, there was no statistical difference in the in-hospital survival between the two groups.
It is believed that carotid stenosis is one of the risk factors for stroke after CABG 5,25,26 .In addition, a previous meta-analysis of US and European CAS studies showed that the perioperative risk of stroke after cardiac surgery varied from 3.8% to 7.4% for patients with ≥ 50% CAS and increased to 2% to 9.1% for patients with ≥ 70% CAS 27 .In our study, the incidence of postoperative stroke after CABG was found to be similar, with 3.7% of patients with ≥ 50% CAS and 7.7% of patients with ≥ 70% CAS.A higher incidence of postoperative stroke was found in the severe unilateral group (P = 0.007), the severe bilateral group (P = 0.001), and the occluded unilateral group (P = 0.033) compared to the overall patients.It was shown that a more severe stenosis was associated with a higher  incidence of postoperative stroke, particularly in the severe unilateral, severe unilateral, and occluded unilateral groups.After propensity matching, the incidence of stroke was 8.0% in the severe CAS group and 0% in the non-severe CAS group (p = 0.028).The present study is consistent with the prevailing view that severe CAS is an independent risk factor for postoperative stroke.The incidences of significant CAS stenosis (≥ 50%) were 12.8% to 22.1% and severe CAS stenosis (≥ 70%) were 4.6% to 5.0% in patients with CABG in the United States and European countries 20,28 .In our Chinese CABG cohort, the incidence of CAS with stenosis ≥ 50% and ≥ 70% was 19.2% and 6.9%, which were consistent with those of previous studies.
An optimal predictive nomogram incorporating age ≥ 65 years, SBP, DBP, CKD, and preoperative hemoglobin to predict severe CAS in patients with CABG was successfully established and carefully evaluated.The available evidence suggests that the nomogram could effectively predict patient prognosis, and its simplicity and intuitive nature facilitate the interpretation by clinical staff 29,30 .To our knowledge, the present study is the first report on the development of a nomogram for predicting severe CAS in patients with CABG using a number of baseline tests.In a multivariate logistic regression model, we observed that age ≥ 65 years, SBP, DBP, CKD, and preoperative hemoglobin were all independently associated with the incidence of severe CAS.Satisfactory accuracy was observed when the above five variables were included in the nomogram model (AUC = 0.681).In recent years, many studies have reported that the incidence of CAS increases with age 31,32 .In Durand DJ's study, age over 65 years was shown to be a significant predictor of CAS 33 .In previously conducted studies, SBP or DBP was considered as a predictor of carotid stenosis 32,34,35 .In addition, Puz P et al. found that CKD was an independent risk factor for symptomatic internal CAS 36 .It could be interpreted that CKD is independently associated with carotid atherosclerosis 37 .Furthermore, Dijk JM and other colleagues demonstrated that increased hemoglobin levels were associated with reduced severity of atherosclerosis, assessed as the presence of ≥ 50% CAS 38 .Fortunately,   www.nature.com/scientificreports/our findings showed that age ≥ 65 years, SBP, DBP, CKD, and hemoglobin were also significantly associated with the incidence of severe CAS in patients undergoing CABG, in agreement with these published findings.To avoid the limitations of a single predictor and to obtain higher prediction accuracy, this study combined five tested predictors to form a nomogram model.Our data confirmed that the nomogram was more effective at predicting severe CAS than any single predictor (AUC = 0.681).Moreover, DCA curves have been commonly used in many studies to assess the efficacy of specific clinical approaches 39,40 .In this study, we also used DCA curves to examine the underlying clinical effects of the nomogram, and our findings suggest that the nomogram was more valuable than other indicators in predicting the incidence of severe CAS.Carotid endarterectomy is considered to be an effective treatment for both symptomatic patients and asymptomatic patients with CAS 41,42 .Carotid artery stenting is another treatment technique.However, evidence for the effect of carotid stenting in patients with severe CAS undergoing CABG is lacking, and its effectiveness preventing postoperative stroke remains controversial 43 .Although there were 117 patients with severe carotid artery stenosis confirmed by carotid ultrasound in the study, the number of patients who ultimately underwent carotid artery stenting did not exceed 30.It was because the degree of stenosis confirmed by cerebral angiography did not meet the above criteria in some of the remaining patients.And other patients could not tolerate carotid artery surgery or combined surgery due to severe conditions (such as frequent angina attacks or severe heart failure).Our study suggested that carotid stenting is not effective in preventing postoperative stroke.Possible explanations are as follows.First, carotid stenosis may be a marker of high atherosclerotic burden and stroke risk, rather than a direct stroke mechanism in most patients.Second, there was relatively few data on carotid artery stenting in our study.Third, combined CABG and carotid stenting are at higher risk than CABG alone.
The study has several limitations.First, the study design was retrospective and non-randomized.In addition, data from a single medical center were analyzed and only selective patients were included.Due to less than thirty patients who underwent carotid artery stent implantation in this study, sample size was indeed limited.Therefore, the results of the study may not be generalizable to other Asian populations and studies with larger sample sizes and higher-quality are needed to confirm our findings.Second, we were unable to determine whether the etiology of each stroke was embolic, thrombotic, or hypoperfused.In this study, all stroke patients were diagnosed with ischemic stroke.However, due to the clinical situations in China and the difficulty in differential diagnosis of stroke etiology, it is challenging to accurately determine whether the stroke is caused by thrombosis, embolism, plaque rupture, or perioperative hypoperfusion and etc.In theory, we could differentiate the etiology of stroke to some extent based on the course of the disease and cranial magnetic resonance imaging (MRI).However, in practice, it is difficult to perform these assessments.Especially since most patients have steel wires fixating the

Figure 1 .
Figure 1.Incidence of carotid artery stenosis.Note The incidence of carotid stenosis was classified by the degree of stenosis and unilateral or bilateral.

Figure 2 .
Figure 2. Incidence of postoperative stroke.Note The incidence of postoperative stroke was divided according to the degree of carotid stenosis and unilateral or bilateral.

Figure 6 .
Figure 6.DCA curves of the predictive nomogram model constructed from age ≥ 65 years, SBP, DBP, CKD, and hemoglobin.NoteThe horizontal axis represents the threshold value, which is the reference probability of whether a patient receives treatment, and the vertical axis represents the net benefit rate after subtracting the disadvantage from the advantage.A larger net benefit for the same threshold probability means that the patient receives the greatest benefit using the model's diagnosis.the closer the curve in the DCA plot is to the top, the higher the value of the model's diagnosis.DCA, decision curve analysis; SBP, systolic blood pressure; DBP, diastolic blood pressure; CKD, chronic kidney disease.

Figure 7 .
Figure 7. Kdensity plots of the propensity-matched model for severe carotid stenosis.Note Figure 7 shows the Kdensity before matching.

Figure 8 .
Figure 8. Kdensity plots of the propensity-matched model for severe carotid stenosis.Note Figure 8 shows the Kdensity after matching.

Figure 9 .
Figure 9. Propensity-matched test plots for the severe carotid stenosis model.Note Propensity-matched test plots for the severe carotid stenosis model.

Table 1 .
Carotid artery stenosis definition and degree of severity.PSVICA, peak systolic velocity of the internal carotid artery; PSVCCA , peak systolic velocity of the common carotid arteries.PresentationStatistical analyses were performed using SPSS statistical software version 25.0 and Stata statistical software version 16.0.Continuous variables are expressed as mean ± SD or median and interquartile variance (IQR).Categorical data were expressed as frequencies or percentages.The t test or Mann-Whitney U test and χ2 test or Fisher's exact test were employed to test for the presence of significant differences between groups.Values less than 0.05 were considered statistically significant for all tests as two-sided tests.

Table 3 .
Clinical characteristics of patients after coronary artery bypass grafting classified by severe carotid artery stenosis.Intensive Care Unit; Af, Atrial Fibrillation.The categorical variables in the table are presented by the number of cases (with percentage) and the continuous variables are expressed by the median (with interquartile range) or mean (with standard deviation).P Value: Compare the patients with and without severe carotid artery stenosis.P values were the results of unpaired t-test or Mann-Whitney U test for continuous variables, and χ2 test or Fisher's exact test for categorical variables.

Table 4 .
Clinical characteristics of patients after propensity matching analysis.BMI: Body Mass Index; SBP: Systolic Blood Pressure; DBP: Diastolic Blood Pressure; MAP: Mean Arterial Pressure; COPD: Chronic Obstructive Pulmonary Disease; CRP: C-Reactive Protein; BNP: B-Type Natriuretic Peptide; PT: Prothrombin Time; INR: International Normalized Ratio; SCr: Serum Creatinine; eGFR: Estimated Glomerular Filtration Rate; LVEF: Left Ventricular Ejection Fraction; MIDCAB: Minimally Invasive Direct Coronary Artery Bypass; OPCABG: off-pump Coronary Artery Bypass grafting; ICU: Intensive Care Unit; Af, Atrial Fibrillation The categorical variables in the table are presented by the number of cases (with percentage) and the continuous variables are expressed by the median (with interquartile range) or mean (with standard deviation).P Value: Compare the patients with and without severe carotid artery stenosis.P values were the results of unpaired t-test or Mann-Whitney U test for continuous variables, and χ2 test or Fisher's exact test for categorical variables.

Table 5 .
The Symptomatic status of patients with severe carotid artery stenosis.