Elevated triglyceride-glucose index predicts poor outcome in patients with intracranial atherosclerotic stenosis after extracranial and intracranial bypass

Abstract

Background and purpose

The triglyceride-glucose (TyG) index, a novel reliable biomarker for IR that incorporates blood glucose and triglyceride, is linked to intracranial atherosclerotic stenosis (ICAS). In this study, we aimed to further investigate the association between the TyG index and the outcomes of ICAS patients following extracranial-to-intracranial (EC-IC) bypass grafting.

Methods

489 ICAS patients who underwent EC-IC bypass between Jan 2009 and Jan 2022 at our hospital were retrospectively collected. The major adverse cardiac and cerebrovascular events (MACCEs), and anastomotic restenosis, both of which are critical factors leading to poor prognosis of ICAS patients after EC-IC bypass, were mainly recorded and analyzed. Kaplan–Meier survival curve and Log-rank tests were sequentially conducted. Cox regression model was used to investigate the association between the TyG index and MACCEs & anastomotic stenosis. C-statistics, continuous net reclassification improvement (NRI), and integrated discrimination improvement (IDI) evaluated the incremental predictive value of the TyG index.

Results

A higher incidence of MACCEs and anastomotic stenosis was found in higher-tertile TyG index group. The TyG index was significantly associated with an increased risk of MACCEs and anastomotic stenosis, independent of confounding factors, with a value of HR (1.30, 95%CI 1.10-1.51, p < 0.001) and (1.27, 95%CI 1.16-1.40, p < 0.001) respectively. The area under the curve (AUC) in the model with the TyG index for predicting the occurrence of MACCEs and anastomotic stenosis were 0.708 (95%CI 0.665-0.748) and 0.731 (95%CI 0.689-0.770) respectively. The addition of the TyG index significantly improved the global performance of the baseline model according to the C-statistics, NRI, and IDI (All p < 0.05).

Conclusions

Higher TyG levels were associated with poorer outcomes in ICAS patients after EC-IC bypass. TyG could be a key factor in managing ICAS risk and standardizing the indications for EC-IC bypass.

Introduction

Stroke is one of the main causes of disability and mortality, in China alone, with an average of 1 new stroke every 12 s and 1 stroke death every 21 s [1]. Ischemic stroke, with about 2 million new incidents per year, accounts for 75%-90%, which is caused mainly by intracranial atherosclerotic stenosis (ICAS) [2,3].

As a plausible treatment strategy for ICAS patients, extracranial-intracranial (EC-IC) bypass is aimed at restoring blood flow perfusion to prevent stroke risk by the anastomosis of the superficial temporal artery (STA) to the middle cerebral artery (MCA) [4,5]. Our previous studies, along with othes, demonstrated the good efficacy of EC-IC bypass in reducing the recurrence of ischemic strokes among ICAS patients [6,7]. However, there are still some patients who do not improve significantly or even experience a poor prognosis [8]. Anastomotic stenosis of STA & MCA and dysfunction of intracranial collateral circulation endpoints could be major factors for unimproved even worsen prognosis. Healthy vascular is crucial for the patency of anastomoses and collateral circulation formation, beyond the surgical technique. Thus, it’s of great significance to investigate the potential factors affecting vascular health and reducing the efficacy of EC-IC bypass for further benefiting clinical management and prognosis for ICAS patients.

Insulin resistance (IR) has been recognized as a crucial factor in metabolic disorders, diabetes, and atherosclerotic cardiocerebrovascular disease (CCVD) [9,10]. Previous studies suggested that IR-induced lipid metabolism disorders could also accelerate intracranial or carotid atherosclerosis formation [11]. IR has been identified as a significant risk factor for the progress of ICAS and increasing ischemic stroke [12,13]. The hyperinsulinemic-euglycaemic clamp test and homeostasis model assessment of IR (HOMA-IR) are regarded as the gold standard method for diagnosing IR. Still, the technique is difficult to implement in clinical practice and large epidemiological investigations because it is a time-consuming, complex process, and costly [14]. The metabolic score for insulin resistance (METS-IR), a simple, convenient, and reliable marker for resistance insulin (IR), has been regard as a predictor of cardiovascular diseases (CVD), even the prognosis of patients after coronary artery bypass graft [15–18]. Both HOMA-IR and METS-IR are calculated from blood glucose and insulin levels. As we all know, it’s not only hyperglycemia, but also high hyperlipidemia are major risk factors for vascular health. Meanwhile, ICAS patients are often accompanied by hyperlipidemia. Thus, further investigating the index including blood glucose and blood lipid holds promise as a potential breakthrough in this field.

Recently, the triglyceride-glucose (TyG) index, derived from triglyceride (TG) and fasting plasma glucose (FPG) level, has been proposed as a reliable and simple surrogate marker of IR and demonstrated a high concordance with the HOMA-IR [19–22]. Accumulating studies demonstrated the validity and feasibility of the TyG index in predicting the severity of atherosclerosis, CVD, and other metabolic diseases [23,24]. Recently, a higher TyG index was also reported to be intimately associated with the development of ICAS [12,13,25–27]. Thus, we speculated that TyG could be a reliable indictors for the prognosis of ICAS patients after EC-IC bypass, and contribute to improve stroke risk stratification in ICAS patients, as well as to refine clinical indications and perioperative management for EC-IC bypass grafting.

Methods and materials

Patients selection

As a retrospective observational cohort study, this study adhered to the declaration of Helsinki and was approved by the Medical Ethics Committee of Third Affiliated Hospital of Sun Yat-sen University (20200218101). The requirement for informed consent was waived by the Medical Ethics Committee of third affiliated hospital of Sun Yat-Sen University due to the retrospective design of this study and no identifiable patient information. This study adhered to the STROBE guidelines, which have been included and presented as supplementary material. ICAS patients who underwent EC-IC bypass grafting at our hospital between January 2009 and January 2022 were retrospectively and consecutively enrolled with a follow-up period of 2 years, as presented in Figure 1.

Figure 1.

Flowchart of patient selection. ICAS, intracranial atherosclerotic stenosis; EC-IC, extracranial-intracranial; NIHSS, National Institutes of Health Stroke Scale; mRS, modified Ranks Score; TyG, triglyceride-glucose.

The inclusion criteria were as follows: (I)18 ≤ age ≤ 75 years old; (II) ICAS at the internal carotid artery middle cerebral artery and basilar artery based on head magnetic resonance angiography (MRA) or CT angiography (CTA), and further confirmed by digital subtraction angiography (DSA) according to the standard of Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) [28]; (III) modified Rankin Scale (mRS) score of 0 to 2; (IV) No massive cerebral infaction (>50% of the MCA territory) demonstrated by CT or MRI; (V) Reduced cerebral perfusion: MTT > 4s, rCBF < 0.95 (symptomatic side/asymptomatic side); (VI) transient ischemic attack (TIA, acute onset of neurologic deficit persisting <24h) or ischemic stroke persisting more than 24h in the terriory of ICA, MCA or basilar artery within 12 months prior to EC-IC bypass, which indicators are as follows: (i) stenosis of internal carotid artery (ICA) and middle cerebral artery (MCA) >50%, demonstrated by digital subtraction angiography (DSA); (ii) reduced cerebral perfusion related to responsible vessels; (iii) ineffective medical treatment.

The exclusion criteria were as follows: other neurovascular disease (such as cerebral aneurysm or arteriovenous malformation) conditions likely to cause focal cerebral ischemia patients with severe impairment of consciousness, NIHHS ≥ 9 or mRS > 3, any diseases likely to lead to death within 2 years, thrombolysis or mechanical thrombectomy, coagulopathy, surgery, trauma, bleeding, malignant tumor, serious injury of liver or kidney, uncontrolled hypertension (systolic BP > 180mmHg, diastolic BP > 110mmHg) and diabetes mellitus (>16.7 mmol/l), or incomplete clinical data and cerebrovascular angiography.

Definition of triglyceride-glucose index

The triglyceride-glucose (TyG) index was defined using the formula as previously reported: Ln [triglyceride (mg/dL) × fasting blood glucose (mg/dL) ÷2] [22,29,30]. Patients were grouped into 3 groups according to the tertiles of the TyG index (T1: TyG index < 8.662; T2: 8.662 ≤ TyG index < 9.401; T3: TyG index ≥ 9.401).

Data collection

The study collected clinical baseline information and the imaging data were collected at initial admission and following follow-up time. The evaluation of neurologic function before and after the operation was conducted using the modified Rankin Scale (mRS) and the National Institute of Health Stroke Scale (NIHSS). Additionally, cerebral hemodynamics were comprehensively assessed through the application of CT perfusion (CTP). All patients underwent laboratory tests, CTA/CTP, and DSA per follow-up period.

The study collected clinical baseline information such as age, gender, diabetes mellitus (DM), history of hypertension, cardiovascular diseases (CVD), drinking status, smoking, and medicating history. In addition, examination and laboratory tests were recorded, consisting of initial systolic blood pressure (SBP) and diastolic blood pressure (DBP), National Institutes of Health Stroke Scale (NIHHS), modified Ranks Score (mRS), body mass index (BMI), fasting blood glucose (FBG), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol (TC), C-reactive protein (CRP). On the morning of the second day after admission, blood samples were harvested from all patients on an empty stomach by the same professional nurse. The imaging data including CTA, MRA, and DSA were collected at initial admission and following follow-up time.

The cerebral blood flow perfusion from CTP parameters (TTP, MTT, CBF, and CBV) was graded into 4 stages as previous standard: stage 1 with prolonged TTP and normal MTT, CBF, and CBV; stage 2 with prolonged TTP and MTT, normal CBF, and normal or mildly elevated CBV; stage 3 with prolonged TTP and MTT as well as decreased CBF, normal or mildly decreased CBV; stage 4 with prolonged TTP and MTT as well as decreased CBF and CBV. Hypertension was defined as repeated systolic blood pressure (SBP) ≥140mmHg and/or diastolic blood pressure (DBP) ≥90mmHg, or a history of hypertension with anti-hypertensive drugs. Diabetes milieus were diagnosed as one of the following criteria: (1) previously diagnosed by other physicians or the use of lowering glucose at initial admission; (2) with any characteristic symptoms of diabetes, such as thirst, polyphagia, polyuria, and weight loss with any blood glucose >11.1 mmol/L; (3) a fasting blood glucose level > 7.0 mmol/L; 4) a 2-h blood glucose after a 75 g glucose load >11.1 mmol/L in an oral glucose tolerance test after fasting 8 h overnight.

Surgical procedures

The superficial temporal artery-to-middle cerebral hemispheres (STA-MCA) bypass was chosen as the main surgical procedure of EC-IC bypass as described in our previous studies [6,31]. The neurosurgeons operated with a minimum of 10 years of experience in cerebrovascular surgery. In the STA-MCA bypass operation, we preferred parietal branches and the frontal and occipital branches as substitutes to anastomose with long-straight and few-branches MCA branches on the surface of the brain. Cerebellar perfusion is used as a reference to evaluate bilateral cerebral perfusion. All patients underwent CTA and CTP examination on the first day after surgery to check for postoperative hemorrhage or infarction and perfusion.

Statistical analysis

All statistical analyses were performed using SPSS software (Windows version 25.0, IBM). All tests were 2-tailed and a p-value < 0.05 was defined as statistical significance. The continuous variables were described as mean ± standard deviation when consistent with a normal distribution, otherwise as median (interquartile range, IQR). The categorical variables were described as numbers and percentiles (%). The ANOVA was applied to compare the difference in normal-distribution continuous variables between groups, and the Kruskal-Wallis H test was used to analyze the difference in nonnormal-distribution continuous variables. The Chi-squared test or Fisher’s exact test was applied to analyze the difference in categorical variables between groups.

The Kaplan–Meier survival curve and Log-rank tests were used to investigate the differences in incident rates of endpoints between TyG tertiles and to plot time-to-endpoint curves. Benjamin-Hochberg algorithm was used to multiply corrected P values for pairwise comparisons. Univariate Cox regression analysis was performed to identify prognostic predictors in ICAS patients who underwent EC-IC bypass. Multivariate Cox proportional hazards regression analysis was performed to investigate the association of the TyG index in developing each endpoint by estimating the Hazard ratio (HR) and the 95% confidence interval (CI). Covariates were included in three-stage models: Model 1 adjusted for age and sex; Model 2 adjusted for variables with p < 0.05 in the univariate analysis including age, BMI, hypertension, diabetes, smoking, dyslipidemia, previous stenting therapy, multi-vessel stenosis, TG, TC, HDL-C, LDL-C, FPG, CRP, LP(a); Model 3 adjusted for all variables including age, BMI, hypertension, diabetes, smoking, drinking, SBP, DBP, dyslipidemia, previous MI, previous stenting therapy, multi-vessel stenosis, insulin use, aspirin, lipid-lowering therapy, oral hypoglycemic drugs, antihypertensive agents, TG, TC, HDL-C, LDL-C, FPG, Uric acid, CRP, LP(a).

The receiver operating characteristic (ROC) curves and the area under the curve (AUC) were used to assess whether the addition of the TyG index optimizes the predictive value of the traditional risk model. Concordance statistics (C-statistics) was used to examine model discrimination and the incremental predictive information yielded by the addition of the TyG index. Furthermore, two measures, net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were used to further evaluate the risk reclassification.

Results

The primary endpoint was a composite of major adverse cardiac and cerebrovascular events (MACCEs), including non-disabling MACCEs and disabling MACCEs (mRS ≥ 3 or NIHSS ≥ 7; an increase of at least 1 point in the mRS score and 4 points in the NIHSS score from prestroke baseline) during the follow-up periods after EC-IC bypass. Moreover, we meticulously documented the occurrence of anastomotic stenosis during postoperative follow-up after a successful EC-IC bypass procedure according to the standard of Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) [28].

Baseline characteristics of patients

At first, a total of 723 ICAS patients who underwent EC-IC bypass were collected. Exclusions included 45 patients aged over 75, 18 with an NIHSS score ≥9 at admission, and 13 with an mRS score >3 at admission, as they were unlikely to benefit from EC-IC bypass. Additionally, 43 patients with surgical or severe complications were excluded to minimize the influence of confounding factors, such as variations in surgical technique. A further 28 patients with incomplete data and 89 patients lost to follow-up were also excluded, leaving a final cohort of 489 patients for analysis (Figure 1).

489 ICAS patients treated with EC-IC bypass (average age 67.47 ± 15.16 years; 62.58% male), a correlation between TyG index tertiles and various metabolic parameters (BMI, lipid profiles, FPG, CRP) as well as multi-vessel stenosis incidence (all p < 0.001) (Table S1) was observed. The highest TyG index tertiles were associated with increased prevalence of hypertension, diabetes, smoking, and insulin or oral hypoglycemic drug use, as well as lower improvement of cerebral perfusion (Table S1). Anastomotic stenosis was more frequent in the highest TyG tertile (6.34%, p < 0.05) (Table S1). Short- and long-term mRS and NIHSS scores were significantly improved by EC-IC bypass, with more pronounced benefits observed in patients within the lowest TyG index tertiles (p < 0.001) (Table 1).

Table 1.

The comparison between clinical presentation and TyG index.

Variable	T1 (158)	T2 (162)	T3 (169)	P-value
mRS
Admission	1.75 ± 0.57	1.84 ± 0.58	1.91 ± 0.63	0.654
3 months	1.59 ± 0.61	1.72 ± 0.63	1.84 ± 0.74	<0.001
Admission-3 months	0.16 ± 0.59	0.12 ± 0.61	0.07 ± 0.71	0.446
P value	0.018	0.075	0.345
2 years	1.08 ± 0.61	1.31 ± 0.66	1.49 ± 0.71	<0.001
Admission-2 years	0.67 ± 0.54	0.53 ± 0.55	0.42 ± 0.61	<0.001
P value	<0.001	<0.001	<0.001
NIHSS
Admission	4.91 ± 0.72	4.97 ± 0.70	5.03 ± 0.76	0.332
3 months	3.85 ± 0.65	4.01 ± 0.63	4.36 ± 0.70	<0.001
3 months-admission	1.06 ± 0.60	0.96 ± 0.57	0.67 ± 0.64	<0.001
P value	<0.001	<0.001	<0.001
2 years	2.62 ± 0.62	2.73 ± 0.65	2.98 ± 0.69	<0.001
2 years-admission	2.29 ± 0.58	2.24 ± 0.61	2.05 ± 0.59	<0.001
P value	<0.001	<0.001	<0.001

Data are shown as mean ± standard deviation (SD) for continuous variables. mRS, modified Rank’s scale; NIHHS, National Institutes of Health Stroke Scale; P values in bold are < 0.05.

The baseline characteristics of patients in the present study stratified by the MACCEs were presented in Table S2. 33.74% of patients experienced at least one MACCE, with incidence rates increasing across TyG index tertiles: 19.39% in T1, 33.33% in T2, and 46.06% in T3 (p < 0.001). Additional factors associated with MACCEs included elevated levels of BMI, SBP, LDL-C, CRP, TG, TC, FPG, LP(a), and higher rates of hypertension, diabetes, smoking, previous stenting therapy and elderly age. These patients also had lower HDL-C levels and multi-vessel stenosis rate than no MACCEs group (all p < 0.05).

In sum, these indicate the potential causal association between TyG and the prognosisof ICAS with EC-IC bypass.

High TyG index increased the risk of MACCEs

During the follow-up period, 165 patients (33.74%) experienced MACCEs, including 137 (28.01%) non-disabling events (T1: 34, 6.95%; T2: 44, 9.00%; T3: 59, 12.07%) and 28 (5.73%) disabling events (T1: 4, 0.82%; T2: 6, 1.23%; T3: 18, 3.68%). A significant increase in MACCE incidence with higher TyG index tertiles (log-rank test, p < 0.0001) (Figure 2a), consistent across non-disabling (p = 0.0009) (Figure 2b) and disabling (p = 0.03) (Figure 2c) categories were revealed by Kaplan-Meier analysis. The results suggested that a higher TyG index was associated with an increased risk of Major Adverse Cardiovascular and Cerebrovascular Events (MACCE). Specifically, Kaplan-Meier analysis showed that higher TyG index tertiles were linked to a higher incidence of both non-disabling and disabling MACCEs, indicating that patients with elevated TyG levels were at greater risk for these adverse events.

Figure 2.

(a) Kaplan-Meier survival curves for the primary MACCEs, for non-disabling MACVEs (b), and for disabling MACVEs (c) across the TyG index tertiles. (d) The receiver operating characteristic (ROC) curves and AUC for the prediction of MACCEs.

TyG: triglyceride-glucose; MACCEs: major adverse cardiac and cerebrovascular events. ROC, curve receiver operating characteristic curve; AUC, the area under the receiver operating characteristic curves. P values in bold are < 0.05.

TyG index as significantly associated with MACCEs (HR per SD increase: 1.30, 95%CI: 1.15-1.47, p < 0.001) (Table S3) was identified by univariate cox regression. After adjusting for age and sex (model 1), the TyG index was a significant predictor for MACCEs. After adjusting for both variables with p < 0.05 in univariate analysis (in model 2) and other all potential confounders (in model 3), regardless of whether the TyG index was a categorical or continuous variable, maintained the TyG index’s predictive value for MACCEs (Table 2). In addition, high TyG index tertiles correlated with increased risk of both non-disabling and disabling MACCEs (Table 3). Improved MACCE prediction with the TyG index [AUC (95%CI): 0.708 (0.665-0.748)] was demonstrated by ROC analysis (Figure 2d), enhancing model 3 performance [C-statistic (95%CI): 0.698 (0.657-0.740), p = 0.004; continuous NRI (95%CI): 0.286 (0.159-0.413), p = 0.003; IDI (95%CI): 0.013 (0.004-0.022), p = 0.005] (all p < 0.05) (Table S4). The model sensitivity without compromising specificity [event NRI (95%CI): 0.181 (0.113-0.249), p = 0.009, but non-event NRI (95%CI): 0.105 (−0.178-0.388), p = 0.345] (Table S4) was improved by the addition of the TyG index, highlighting its potential as a valuable risk predictor for MACCEs.

Table 2.

Multivariate Cox regression analyses for MACCEs.

TyG index	Model 1		Model 2		Model 3
	HR (95%CI)	P-value	HR (95%CI)	P-value	HR (95%CI)	P-value
Per Unit increase	1.85 (1.24-2.52)	<0.001	1.73 (1.18-2.47)	<0.001	1.76 (1.21-2.49)	<0.001
Per SD increase	1.36 (1.16-1.63)	<0.001	1.28 (1.07-1.49)	<0.001	1.30 (1.10-1.51)	<0.001
Tertile 1	1.0 (Reference)		1.0 (Reference)		1.0 (Reference)
Tertile 2	1.25 (1.14-1.53)	0.021	1.23 (1.13-1.51)	0.043	1.18 (1.10-1.46)	0.089
Tertile 3	2.45 (1.58-3.72)	<0.001	2.21 (1.37-3.34)	<0.001	2.18 (1.31-3.27)	<0.001
P-value for trend	<0.001		<0.001		<0.001

Model 1: adjusted for age and sex.

Model 2: adjusted for variables with p < 0.05 in the univariate analysis including age, BMI, hypertension, diabetes, smoking, dyslipidemia, previous stenting therapy, Multi-vessel stenosis, TG, TC, HDL-C, LDL-C, FPG, CRP, LP(a).

Model 3: adjusted for all variables including age, BMI, hypertension, diabetes, smoking, drinking, SBP, DBP, dyslipidemia, previous MI, previous stenting therapy, multi-vessel stenosis, insulin use, aspirin, lipid-lowering therapy, oral hypoglycemic drugs, antihypertensive agents, TG, TC, HDL-C, LDL-C, FPG, Uric acid, CRP, LP(a).

MACCEs, major adverse cardiac and cerebrovascular events; P values in bold are < 0.05.

Table 3.

Multivariate Cox regression analyses for non-disabling and disabling MACCEs.

TyG index	Non-disabling MACCEs		Disabling MACCEs
	HR (95%CI)	P-value	HR (95%CI)	P-value
Per Unit increase	1.42 (1.21-2.33)	0.001	1.52 (1.12-2.01)	0.007
Per SD increase	1.14 (1.03-1.27)	0.009	1.17 (1.01-1.39)	0.037
Tertile 1	1.0 (Reference)		1.0 (Reference)
Tertile 2	1.45 (0.97-2.27)	0.080	1.28 (1.05-1.55)	0.012
Tertile 3	2.27 (1.46-3.50)	<0.001	2.21 (1.37-3.34)	<0.001
P-value for trend	<0.001		<0.001

MACCEs, major adverse cardiac and cerebrovascular events; P-values in bold are < 0.05.

Collectively, these data suggested that the TyG index was a significant predictor of MACCEs, with its predictive value robust across various models and adjustments; it enhanced risk prediction and improved model performance, demonstrating its potential as a valuable tool for assessing MACCE risk.

High TyG index predicted anastomotic restenosis of EC-IC bypass

Baseline characteristics of patients with and without anastomotic stenosis were stratified in Table S5. During follow-up, 31 (6.34%) patients (T1: 5, 16.13%; T2: 9, 29.03%; T3: 17, 54.84%) experienced anastomotic stenosis, with higher TyG index tertiles correlating with increased incidence (p = 0.034) (Table S5), and an elevated cumulative incidence of anastomotic stenosis (log-rank test, p = 0.0108) (Figure 3a).

Figure 3.

(a) Kaplan-Meier survival curves for the anastomotic stenosis across the TyG index tertiles. (b) The receiver operating characteristic (ROC) curves and AUC for the prediction of anastomotic stenosis. TyG: triglyceride-glucose; P values in bold are < 0.05. ROC, curve receiver operating characteristic curve; AUC, the area under the receiver operating characteristic curves; TyG, triglyceride-glucose.

The significant associations between anastomotic stenosis and age, diabetes, smoking, multi-vessel stenosis, TG, TC, LDL-C, FPG, and the TyG index (HR per SD increase: 1.31, 95%CI: 1.18-1.45, p < 0.001), was identified by univariate Cox regression analysis (Table S6). Further, the TyG index’s predictive value after adjusting for other potential confounders (Table 4) was confirmed by multivariable analysis. In model 3, the AUC (95%) [0.731 (0.689-0.770) vs 0.653 (0.609-0.695), p = 0.041)] (Figure 3b), C-statistic [0.698, 95%CI (0.671-0.725), p = 0.001], Continuous NRI [0.271, 95%CI (0.146-0.396), p = 0.030] and IDI [0.012, 95%CI (0.005-0.019), p = 0.009], were significantly improved by the addition of the TyG index, enhancing the model specificity and sensitivity (Table S7).

Table 4.

Multivariate Cox regression analyses for anastomotic stenosis.

TyG index	Model 1		Model 2		Model 3
	HR (95%CI)	P-value	HR (95%CI)	P-value	HR (95%CI)	P-value
Per Unit increase	1.73 (1.38-2.08)	<0.001	1.56 (1.13-1.99)	<0.001	1.57 (1.15-2.01)	<0.001
Per SD increase	1.36 (1.23-1.50)	<0.001	1.27 (1.15-1.39)	<0.001	1.27 (1.16-1.40)	<0.001
Tertile 1	1.0 (Reference)		1.0 (Reference)		1.0 (Reference)
Tertile 2	1.16 (1.01-1.32)	0.029	1.13 (0.99-1.27)	0.088	1.15 (0.95-1.34)	0.178
Tertile 3	2.28 (1.25-3.37)	0.004	2.07 (1.19-2.95)	0.006	2.09 (1.16-3.16)	0.011
P-value for trend	<0.001		<0.001		<0.001

Model 1: adjusted for age and sex.

Model 2: adjusted for variables with p < 0.05 in the univariate analysis including Age, diabetes, smoking, Multi-vessel stenosis, TG, TC, LDL-C, and FPG.

Model 3: adjusted for all variables including age, BMI, hypertension, diabetes, smoking, drinking, dyslipidemia, previous MI, previous stenting therapy, multi-vessel stenosis, insulin use, aspirin, lipid-lowering therapy, oral hypoglycemic drugs, antihypertensive agents, TG, TC, HDL-C, LDL-C, FPG, Uric acid, CRP, LP(a).

P-values in bold are < 0.05.

In sum, these showed that the TyG index was significantly associated with anastomotic stenosis, with higher TyG tertiles correlating with increased incidence; it enhanced the predictive accuracy and model performance, improving both specificity and sensitivity for identifying anastomotic stenosis.

Discussion

The prognostic significance of the TyG index in ICAS patients undergoing EC-IC bypass was examined in this study. Patients in the highest TyG index tertile had a 2.18-fold increased risk of MACCEs and a 2.09-fold increased risk of anastomotic stenosis compared to those in the lowest tertile. Thus, the TyG index can serve as a valuable tool for risk stratification in ICAS patients undergoing EC-IC bypass. By identifying patients with elevated TyG levels, clinicians can better predict the risk of adverse outcomes, including MACCEs and anastomotic stenosis. This can contribute to more personalized treatment plans and monitoring strategies, potentially improving patient outcomes and guiding decision-making for surgical interventions.

Previous studies have highlighted the limited efficacy of EC-IC bypass in ICAS patients, potentially due to broad inclusion criteria such as non-severe stenosis and unsuitable vascular lesions [8,32,33]. The Carotid and Middle Cerebral Artery Occlusion Surgery Study (CMOSS) findings also questioned its effectiveness for symptomatic atherosclerotic occlusion of the ICA or MCA, suggesting a need for more precise patient selection [34]. However, carefully selected ICAS patients with moderate to severe stenosis and non-disabling stroke symptoms showed significant improvements in cerebral perfusion and clinical outcomes after EC-IC bypass [6,7,31,34]. This highlighted the controversy surrounding EC-IC bypass as a treatment and underscores the importance of identifying factors that affect its clinical success. Theoretically, EC-IC bypass under strict adherence to surgical indicators can increase intracranial cerebral blood flow and significantly improve ischaemic symptoms in patients with ICAS. However, anastomotic restenosis, severe perioperative complications and dysfunction collateral & microcirculation could be the major factors leading to a bad prognosis of ICAS patients after EC-IC bypass.

In this study, we focused on endogenous metabolisms that are strongly associated with the occurrence and prognosis of ICAS by affecting vascular health. This includes blood glucose, blood lipids, and insulin resistance. These factors impair vascular endothelial cell function, exacerbate inflammation and oxidative stress, and accelerate the atherosclerotic process, thereby increasing the risk of cardiocerebrovascular events [35–37]. IR, known for impairing insulin-mediated glucose uptake, can also impair cerebral perfusion and increase the risk of adverse events, and is linked to ischemic stroke’s pathogenesis by promoting proinflammatory and prothrombotic changes in ICAS [38]. Studies, including those by Hoscheidt SM et al. have shown that higher IR levels correlate with reduced artery blood flow and cerebral perfusion, notably in the internal carotid arteries [39]. Furthermore, IR’s association with worse postoperative outcomes in cardiovascular disease (CVD) patients suggests it may hinder cerebral perfusion and neurological recovery in ICAS patients after EC-IC bypass [40,41]. The TyG index, which reflects IR, dyslipidemia, and blood glucose, has been connected to cardiocerebrovascular diseases and identified as a significant predictor for ICAS and adverse outcomes in cardiac surgery patients [13,23–25,42,43]. Here, we confirmed that TyG could be a critical predictor of MACCEs and anastomotic stenosis in ICAS patients following EC-IC bypass, suggesting that metabolic health significantly influences the success of EC-IC bypass. For clinicians and policymakers, this underscores the importance of incorporating metabolic assessments into patient evaluations and decision-making processes to enhance treatment outcomes, thereby enhancing risk stratification and the appropriateness of surgical indications for these patients.

Our study’s strengths include the robust predictive value of the TyG index for major adverse cardiovascular and cerebrovascular events (MACCEs) and anastomotic stenosis, as demonstrated by comprehensive statistical analyses and model enhancements. Unlike some studies that question the efficacy of EC-IC bypass, here we focused on optimizing preoperative surgical indication for EC-IC bypass by incorporating TyG index of glucose and lipids to assess vascular health. This approach aims to exclude ICAS populations that may not benefit from EC-IC bypass surgery, offering valuable insights into how metabolic parameters influence surgical outcomes and improving risk stratification and surgical management of ICAS.

However, the study’s limitations include its retrospective, single-center design and the lack of a control group. The focus on specific patient populations and a single baseline measurement may limit generalizability. Additionally, measuring the TyG index only at baseline and excluding patients without comprehensive imaging data could introduce bias. Future research should address these limitations by conducting larger, multicenter trials to validate our findings. Investigations should explore the relationship between the TyG index and long-term outcomes, including the mechanisms underlying its predictive value. Moreover, evaluating the integration of the TyG index into clinical guidelines and its potential impact on patient care and outcomes for ICAS patients undergoing EC-IC bypass is essential. Further exploration of additional biomarkers and refinement of risk stratification models could enhance the predictive accuracy and clinical utility of metabolic indices.

Conclusion

The TyG index could be a critical predictor of MCAVEs and anastomotic stenosis in ICAS patients undergoing EC-IC bypass grafting, enhancing risk stratification and refining surgical indications for these patients.

Funding Statement

This study is supported by the “Five and five” Project of the Third Affiliated Hospital of Sun Yat-Sen University (No.2023WW504); the Clinical Research 5010 Program of Sun Yat-Sen University (No.2020001); the Clinical Research Program of the Third Affiliated Hospital of Sun Yat-Sen University (No. YHJH201902).

Ethics approval and consent to participate

This study was approved by the Medical Ethics Committee of our hospital (20200218101). The requirement for informed consent was waived by the Medical Ethics Committee of third affiliated hospital of Sun Yat-Sen University due to the retrospective design of this study and no identifiable patient information.

Authors contributions

Jun Sun, Chuan Chen and Hui Wang conceived and designed the research. Jun Sun, Qiuhua Zeng, and Zhimin Wu analyzed the data and drafted the manuscript. Jun Sun, Zhimin Wu, Qiuhua Zeng, Lixin Huang, and Tao Sun collected the data and performed the research. Cong Ling, Chuan Chen, Baoyu Zhang and Hui Wang performed EC-IC bypass surgery. Hui Wang received the funding support. All authors reviewed and revised the manuscript and finally approved the version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.