The clinical data and ApoE gene tests of inpatients, who underwent the head and neck digital subtraction angiography (DSA) or computed tomography angiography (CTA) in our department from June 2020 to January 2024, were retrospectively analyzed. The clinical data comprised demographic information and the cerebrovascular disease’s common risk factors. The tortuosity index (TI) of ECA for each patient were measured and calculated. Referring to the methods in previous literatures (11, 12), the included patients were divided into three groups based on the tertile distribution of ECA TI and two groups based on the median distribution. We compared the differences in clinical data and ApoE genotypes between these groups.

The inclusion criteria for this study were hospitalized patients who had completed both head and neck CTA/DSA and ApoE genotyping, without restriction based on disease entities. The exclusion criteria were defined as follows. (1) patients with poor imaging quality of CTA or DSA which could not complete the measure and calculate for the tortuosity index. (2) patients with the bilateral ECA occlusion. (3) patient is under 18 years old. This study was designed for retrospective research, and the formal consent from patients was not required. All research procedures involving human participants adhered to the ethical standards of the Declaration of Helsinki of 1964 and its subsequent amendments or similar ethical standards.

The DSA was performed using an Allura xper FD20 X-ray system (Philips, Netherlands), while the CTA was performed using a Brilliance 64 row 128 slice spiral CT (Philips, Netherlands). The three-dimensional volume rendering of the head and neck arteries were achieved using the Philips Extended Brilliance^™ workstation (Philips, Netherlands). Both DSA and CTA procedures were all using non-ionic contrast agents (iohexol, iopamidol or iodixanol) for imaging. The informed consent form for DSA or CTA procedures were signed by the enrolled patients or their next of kin.

The DICOM document of CTA and DSA images were downloaded and collected from picture archiving and communication system (PACS). The tortuosity index of the ECA was measured and calculated by two independent experts in cerebrovascular disease imaging using Sante DICOM Viewer 3D software (Sante soft LTD, Nicosia, Cyprus). After loading DICOM files or selected JPEG format images into the Sante DICOM Viewer software, the anteroposterior view of 3D volume-rendered images for bilateral ECA were choose to measure. The proximal end for measure was the bifurcation point of the aortic arch in the left ECA or the brachiocephalic trunk in the right ECA. The distal end of measurement was the ECA located at the entrance of the carotid canal. A line measurement tool was used to measure the actual length from the proximal end to the distal end along the curvature of the ECA, and a linear distance tool was used to measure the straight length between the two points. The calculation formula for the TI of ECA is as follows. TI = [actual length/straight length−1] × 100. Since relative numbers cannot be averaged, the higher value obtained from the bilateral ECA calculations was taken to represent the TI of the patient. The specific measurement and calculation methods refer to the previously published literature (13, 14) (Figure 1).

After extracting DNA from venous blood of patients, the ApoE genotypes was detected and analyzed by Fascan-48E Multichannel Fluorescence Quantitative Analyzer and SNP-U4 Human ApoE Gene PCR Detection Kit (TIANLONG Technology Ltd., Xi’an city, CHN). The genotype of each subject was one of six possible results: three homozygotes (ε2/ε2, ε3/ε3, ε4/ε4) or three heterozygotes (ε2/ε3, ε2/ε4, ε3/ε4).

All the data were statistically processed by using SPSS software (version 20.0, IBM Corporation, Armonk, NY). The Kolmogorov Smirnov test and Shapiro Wilk test was used to examine the normal distribution of continuous variables, and the Levene test is used to test the homogeneity of variance. The continuous variables that follow a normal distribution were described as Mean (SD), and inter group comparisons were performed by using the analysis of variance or the student’s t test. The continuous variables with skewed distribution were described as Median (IQR), and intergroup comparisons were performed by using Kruskal Wallis rank-sum test or Mann Whitney U test. The categorical variables were described as n (%), and inter group comparisons were conducted by using chi-square test, Fisher’s exact test, or Kruskal Wallis rank-sum test. The multivariate analysis was performed by using the ordinal logistic regression and binary logistic regression. The odds ratio (OR) and 95% confidence interval (CI) were calculated to analyze the independent risk factors for ECA tortuosity. The spearman correlation coefficient (r_s) was used to analyze the correlation between ApoE gene polymorphism and the degree of ECA tortuosity. p < 0.05 indicates the statistically significant.

A total of 238 patients with ischemic cerebrovascular disease or those with its risk factors were included in this study. According to the tertile distribution of ECA TI, there were 91 cases (38.2%) in the TI < 13 group, 65 cases (27.3%) in the TI 13–19 group, and 82 cases (34.5%) in the TI > 19 group. According to the median distribution of ECA TI, there were 127 cases (53.4%) in the TI ≤ 15 group and 111 cases (46.6%) in the TI > 15 group. The age, sex, hypertension, diabetes, total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), atrial fibrillation, smoking history, drinking history, and previous stroke history were analyzed and compared.

Among the 238 patients included, the distribution of ApoE genotypes is as follows: 4 cases of ε2/ε2 (1.68%), 40 cases of ε2/ε3 (16.8%), 1 case of ε2/ε4 (0.42%), 143 cases of ε3/ε3 (60.1%), 49 cases of ε3/ε4 (20.58%), and 1 case of ε4/ε4 (0.42%). Due to the rarity of genotypes ε2/ε2, ε2/ε4, and ε4/ε4, the six cases were excluded for further statistical analysis as refer to a previous research (15).

There were significant differences (p < 0.05) in age, female gender, and smoking history between the tertile groups and the median groups. In addition, there were significant differences in alcohol consumption between the median groups and in atrial fibrillation between the tertile groups. As for the genotypes of ApoE, ε3/ε4 (p = 0.029) showed statistically difference in the tertile groups for ECA TI. In contrast, no statistically significant differences were observed in either grouping methods for ε2/ε3 and ε3/ε3 (Table 1).

The variables demonstrated with statistical differences in univariate analysis, including age, gender, smoking history, alcohol consumption history, and the two target independent variables (ApoE genotype ε2/ε3 and ε3/ε4), were utilized into ordinal multiclassification logistic regression and binary logistic regression analyses to investigate the risk factors associated with ECA tortuosity.

After adjusting for age, female gender, and smoking, the ordinal multiclassification logistic regression for the tertile groups revealed that ε2/ε3 (OR = 1.116, 95% CI 0.550–2.264; p = 0.76) did not emerge as an independent risk factor, while ε3/ε4 (OR = 0.469, 95% CI 0.242–0.909; p = 0.025) demonstrated as a protective factor for ECA tortuosity. Neither ε2/ε3 (OR = 1.374, 95% CI 0.609–3.102; p = 0.444) nor ε3/ε4 (OR = 0.6, 95% CI 0.282–1.273; p = 0.183) showed the significant independent relationship to ECA tortuosity after adjusting for age, female gender, smoking, and alcohol consumption in binary logistic regression (Table 2).

The distribution of the ECA tortuosity index in ApoE genotypes (ε2/ε3, ε3/ε3, and ε3/ε4) is displayed in Figure 2. The Spearman’s rank correlation analysis revealed a significant negative correlation between the TI of ECA and ApoE genotypes (in sequential order: ε2/ε3, ε3/ε3, and ε3/ε4) (r_S = −0.149, p = 0.023).