Participants
The baseline screening of this study was performed between January 1, 2013 and December 30, 2014. In this prospective study, we recruited consecutive 33279 subjects aged 50 years or older from 17 regional general hospitals in China. The inclusion criteria were: (1) 50 years of age or older, (2) a detection of carotid and vertebral arteries by CTA, (3) long-term residents, (4) agreement to participate in the study. The exclusion criteria were: (1) withdrawn consent immediately; (2) intracranial hemorrhage and stroke; (3) Marfan syndrome, loeys Dietz syndrome and fibromuscular dysplasia; (4) participants who died or lost during the follow-up period; (5) patients who did not complete the follow-up due to withdrew consent.
Participants were randomly selected and indicated by CTA scanning. Participants were patients with transient ischemia attack (TIA) episodes, asymptomatic lacunar infarction, carotid plaque by ultrasound examination, accompanied by hypertension, diabetes, hyperlipidemia, and obesity, long-term smoking and alcohol abuse, so they needed CTA to determine the presence of stenosis, occlusion, and plaque formation in the carotid and vertebral arteries and to take appropriate management measures for stroke prevention, including stent implantation or carotid endarterectomy or drug treatment (such as statins, antiplatelet drugs). Because all participants have indications for CTA scanning, and this study is to clarify the relationship between arterial variations and cerebral infarction, it was approved by the ethics committee. In addition, all participants provided written informed consents.
Of the 33279 patients, 138 had withdrawn consents immediately, 306 had intracranial hemorrhage, 78 had Marfan syndrome, loeys Dietz syndrome and fibromuscular dysplasia. A total of 522 patients were also excluded, 32757 patients were followed up for 5 years. A flowchart of this process is presented in Fig. 1.
Procedures
Data were collected from all subjects using a standardized case report form. Clinical assessments were gathered from the patients' medical records. The collected medical data included age, gender, smoking status, alcohol consumption status and the presence of hypertension (systolic/diastolic blood pressure > 140/90 mmHg or being on antihypertensive treatment), diabetes mellitus (fasting blood glucose > 110 mg/dl or being on antidiabetic treatment), hyperlipidemia (total cholesterol > 200 mg/dl or LDL > 125 mg/dl), coronary artery disease or stroke. Blood pressure was measured, and electrocardiograms and chest X-rays were performed on-site. Fasting blood samples were collected to measure the glucose and dyslipidemia (total cholesterol levels, triglycerides levels, high-density lipoprotein cholesterol and lower-density lipoprotein cholesterol blood concentration). The diagnoses of diseases, including hypertension, diabetes mellitus, dyslipidemia, coronary artery disease, peripheral vascular disease and stroke, were based on the International Classification of Diseases, 10th Revision (ICD-10). We defined an ischemic stroke as a clinically definite stroke (symptomatic cerebral infarctions) in a patient whose brain imaging showed positive evidence of a relevant ischemic lesion. Each stroke patient was examination by transcranial Doppler (TCD) to document the cerebral hemodynamic condition within 1 week of symptom onset.
Image evaluation of arterial variations and stenosis
Arterial morphological variations were identified and luminal stenosis measurements were obtained by two independent, experienced radiologists who were blinded to the subjects' clinical information and worked at separate workstations. Variations of the carotid and vertebral arteries were classified into tortuosity, kinking and coiling using the criteria established by Weibel and Fields [3, 4]. Tortuosity is defined as a non-rectilinear stretch of an artery with an angulation >90-165°. Kinking is described as an acute angle between two arterial segments (≤ 90°). Coiling is defined as a 360° angulation of an artery on its transverse axis (“coil” configuration) [11]. Fig. 2 shows the tortuosity, kinking and coiling of carotid and vertebral arteries as determined by CTA. Extracranial carotid and vertebral arterial stenosis was categorized as either < 50% or > 50% as determined by CTA.
All the scans were performed using a dual-source, 64-slice, multidetector CT scanner (LightSpeed; GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA). The representative sample parameters were as follows: 0.625-mm section thickness, 0.33-sec gantry rotation time, 0.984 pitch, 120 kV tube voltage and 120 mA tube current. A single 80-mL injection of nonionic contrast medium (350 mg I/mL iomeprol; Bracco Omnipaque; Milan, Italy) was administered at an injection rate of 2.0 mL/sec through an 18-gauge intravenous catheter using a power injector (Medrad Stellant; Medrad, Pittsburgh, PA, USA).
Diagnosis of ischemic stroke
The diagnosis of ischemic stroke is based on neurologic deficit and brain CT findings. Patients with ischemic stroke is diagnosed by brain CT at baseline. Patients with ischemic stroke is diagnosed by brain CT or magnetic resonance imaging (MRI) during follow-up.
Follow-up
All patients were followed up annually for 5 years from January 2015 to December 2019. During the follow-up period, 2642 subjects died, 539 subjects were lost, 257 subjects withdrew consent, 29319 subjects completed follow-up. Subjects died from any tumor 584 (tracheal, bronchus and lung cancer 250, digestive system tumor 330, other tumors 178), died from stroke 608 (ischemic stroke 432, intracerebral hemorrhage 151, subarachnoid hemorrhage 25), died from heart diseases 654 (ischemic heart disease 514, heart failure 61, hypertensive heart disease 79), other causes of death 796 (COPD 242, lower respiratory infection 98, Alzheimer’s disease and other dementia 77, cirrhosis and other chronic liver disease 73, chronic kidney disease 59, self-harm 44, falls 25).
Statistical analysis
We performed t tests for independent normally distributed continuous data and the chi-square test for categorical data. Proportions were used for categorical variables, and the mean value ± standard deviation was used for continuous variables. Trend analyzes were performed using the linear-by-linear association test. Arterial variations were divided into carotid arterial variations and vertebral arterial variations according to the site of the variation. The carotid arterial variations observed in our study included variations in the common carotid arteries and the internal carotid arteries (ICAs). To assess the relationship of arterial variations with the risk of ischemic stroke, we compared the demographic and clinical characteristics of patients with and without arterial variations.
Cox proportional-hazards models were used to assess the associations [hazard ratio (HR)] of arterial variations with the risk of stroke. Cox proportional-hazards models were generated using the following candidate variables: age, sex, hypertension, dyslipidemia, diabetes mellitus, coronary artery disease, current smoking, current drinking, carotid arterial kinking and coiling and vertebral arterial kinking and coiling and arterial stenosis. The associations were first analyzed without adjusting for covariates. The associations were further analyzed following the adjustment for age and other covariates. All of the tests were two-sided, and a P value of 0.05 was considered statistically significant.