Abstract: Background　As a classic platelet aggregation inhibitor，clopidogrel has been widely used in secondary prevention and treatment of patients with ischemic stroke. However，some patients with clopidogrel treatment still have ischemic stroke，which might be attributed to CYP2C19 gene polymorphisms，but it remains controversial whether CYP2C19 gene polymorphisms are independently associated with increased risk of stroke in different age groups of patients. Objective　To explore the correlation between CYP2C19 polymorphisms and clinical outcomes in different age groups of patients with minor stroke caused by large-artery atherosclerosis（LAA）. Methods　A total of 479 inpatients with minor stroke due to LAA（169 cases≤60 years old，and 310 cases>60 years old） and one-year oral administration of clopidogrel were selected from Department of Neurology，the First Affiliated Hospital of Soochow University from April 2015 to April 2018. Baseline data were collected，including age，gender，NIHSS score，blood glucose，high-density lipoprotein（HDL），low-density lipoprotein（LDL），total cholesterol（TC），triacylglycerol（TG），white blood cell count，prevalence of diabetes，hypertension，CYP2C19 LOF alleles，smoking，and drinking. Recurrence of stroke or death within three months after stroke was obtained during follow-up. One-year clinical outcome was evaluated using the Modified Rankin Scale（mRS）（mRS score ≤2 indicating good outcome，and mRS score >2 indicating poor outcome）. The homogeneity test was used to analyze the heterogeneity of the results in different age groups. Multivariate Logistic regression was used to analyze the potential predictors of outcome in different age groups. Results　In patients aged ≤60 years old，significant differences in average NIHSS score，blood glucose，LDL，and white blood cell count，as well as prevalence of diabetes，hypertension，CYP2C19 LOF，smoking between good and poor outcome cases（P<0.05）. In patients aged>60 years，there were statistically significant differences in gender ratio，average blood glucose，HDL，LDL，and TG，as well as prevalence of diabetes，and hypertension between good and poor outcome cases（P<0.05）. Breslow-Day test found that there was an interaction effect between age and correlation of CYP2C19 LOF allele carriage with clinical outcome （χ2=6.352，P=0.012）. Tarone test also demonstrated that there was an interaction effect between age and correlation of CYP2C19 LOF allele carriage and clinical outcome（χ2=6.351，P=0.012）. The size of the interaction effect analyzed using binary Logistic regression showed that the difference in the size of the interaction effect was statistically significant〔P=0.013，OR=3.63，95%CI（1.31，10.07）〕. In addition，there was no heterogeneity in the correlation between CYP2C19 LOF alleles and stroke recurrence or death 3 months after stroke in patients of different age groups（P>0.05）. CYP2C19 LOF allele carriers had higher rate of stroke recurrence or death within 3 months after stroke than non-carriers（P=0.023）. Multivariate Logistic regression analysis showed that in patients ≤60 years old，CYP2C19 LOF alleles carriage，hypertension，diabetes，smoking，and elevated LDL may be independent predictors of clinical outcome（P<0.05）. Among those aged >60 years old，hypertension，diabetes，drinking，male，elevated blood glucose and HDL may be independent predictors of clinical outcome（P<0.05）. After controlling for confounding factors，such as CYP2C19 LOF alleles，hypertension，diabetes，smoking，drinking，gender，NIHSS score，blood glucose，TC，TG，LDL，HDL，and white blood cell count，it is still not certain whether CYP2C19 LOF was the independent predictor of stroke recurrence or death three months after stroke（P>0.05）. Conclusion　The correlation between CYP2C19 polymorphisms and long-term clinical outcome might be stronger in younger patients with minor stroke due to LAA treated with clopidogrel.

Key words: Stroke, Atherosclerosis, Age groups, CYP2C19 genetic, Genetic polymorphism, Cytochrome P450, Platelet count, Clopidogrel, Prognosis

摘要： 背景　氯吡格雷作为一种经典抗血小板聚集药物已广泛应用于缺血性脑卒中患者的二级预防及治疗中。然而部分患者接受氯吡格雷治疗后仍发生缺血性脑卒中事件。CYP2C19基因多态性被认为与此相关，然而不同年龄患者中该基因多态性是否为脑卒中的独立危险因素尚存争议。目的　探讨CYP2C19基因多态性与不同年龄层大动脉粥样硬化型小卒中患者口服氯吡格雷治疗后临床预后的相关性。方法　选取2015年4月—2018年4月在苏州大学附属第一医院神经内科住院治疗的大动脉粥样硬化型小卒中患者479例作为研究对象，年龄≤60岁169例，年龄>60岁310例。患者均接受氯吡格雷口服治疗为期1年。住院期间收集患者的基线资料，包括年龄、美国国立卫生研究院卒中量表（NIHSS）评分、血糖、高密度脂蛋白（HDL）、低密度脂蛋白（LDL）、总胆固醇（TC）、三酰甘油（TG）、白细胞计数、糖尿病、高血压、CYP2C19 LOF等位基因、吸烟、饮酒、性别。随访记录患者卒中后三个月卒中复发/死亡的情况，采用改良Rankin量表评分（mRS）评价1年期的临床预后，mRS≤2分为临床预后良好（预后良好组），mRS>2分为临床预后不良（预后不良组）。根据CYP2C19功能变异型（LOF）等位基因携带状态分为携带组和非携带组。使用同质性检验分析不同年龄层结果异质性，并使用多因素Logistic回归分析不同年龄层中与临床预后相关的潜在预测因子。结果　年龄≤60岁患者：预后良好组与预后不良组NIHSS评分、血糖、LDL、白细胞计数、糖尿病、高血压、CYP2C19 LOF等位基因、吸烟比例比较，差异有统计学意义（P<0.05）；年龄>60岁患者：预后良好组与预后不良组血糖、HDL、LDL、TG、糖尿病、高血压、性别比较，差异有统计学意义（P<0.05）。Breslow-Day检验：CYP2C19 LOF等位基因携带和临床预后的相关性与年龄间存在交互效应（χ2=6.352，P=0.012），Tarone's检验：CYP2C19 LOF等位基因携带和临床预后的相关性与年龄间存在交互效应（χ2=6.351，P=0.012）。使用二元Logistic回归法分析交互效应大小，结果显示，交互效应的大小比较差异有统计学意义〔P=0.013，OR=3.63，95%CI（1.31，10.07）〕。此外，不同年龄层患者CYP2C19 LOF等位基因与卒中后3个月卒中复发/死亡相关性差异并无异质性（P>0.05）。CYP2C19 LOF等位基因携带患者3个月卒中复发/死亡比例（9.7%）高于非携带患者（3.9%）（P=0.023）。多因素 Logistic回归分析结果显示，年龄≤60岁的患者中，CYP2C19 LOF等位基因、高血压、糖尿病、吸烟、LDL可能是临床预后的独立预测因子（P<0.05）；年龄>60岁的患者中，高血压、糖尿病、饮酒、男性、血糖、HDL可能是临床预后的独立预测因子（P<0.05）。在控制CYP2C19 LOF等位基因、高血压、糖尿病、吸烟、饮酒、性别、NIHSS评分、血糖、TC、TG、LDL、HDL、白细胞计数上述临床混杂因素后，尚不能肯定CYP2C19 LOF是该类研究患者卒中后3个月卒中复发/死亡的独立预测因子（P>0.05）。结论　在接受氯吡格雷治疗的大动脉粥样硬化型小卒中患者中，年轻患者的长期功能预后与CYP2C19 LOF基因的相关性可能较年老患者更强。

关键词: 卒中, 动脉粥样硬化, 年龄组, CYP2C19基因, 基因多态性, 细胞色素P450, 血小板计数, 氯吡格雷, 预后