轻型急性缺血性卒中患者卒中前认知损害与MRI征象的相关性

Correlation between pre-stroke cognitive impairment and MRI markers in patients with minor acute ischemic stroke

目的探讨神经退行性疾病和血管性疾病的MRI征象与卒中前认知损害(pre-stroke cognitive impairment,PSCI)的相关性。方法回顾性纳入2019年3月至2021年12月扬州大学附属医院神经内科和临沂金锣医院神经内科收治的首次发病且年龄≥60岁的轻型急性缺血性卒中患者。利用MRI图像采用二分法视觉评分分析脑小血管病和神经退行性变影像学征象,前者包括脑白质高信号、血管源性腔隙病灶、脑微出血和血管周围间隙扩大,后者包括全脑皮质萎缩和内侧颞叶萎缩。根据老年人认知功能减退知情者问卷(Informant Questionnaire on Cognitive Decline in the Elderly,IQCODE)评分值将患者分为PSCI组(≥3.31分)和非PSCI组(<3.31分),对两组临床基线资料和MRI征象进行比较。采用多变量logistic回归模型分析MRI征象与PSCI的相关性,采用受试者工作特征(receiver operator characteristic,ROC)曲线分析MRI征象对PSCI的预测价值。结果共纳入221例患者,PSCI组77例(34.8%),非PSCI组144例(65.2%)。单变量分析显示,两组年龄、受教育年限、病理性脑白质高信号、内侧颞叶萎缩以及存在≥1种异常MRI征象的患者比例差异有统计学意义(P均<0.05)。多变量logistic回归分析显示,年龄较大[优势比(odds ratio,OR)1.089,95%置信区间(confidence interval,CI)1.034~1.146;P=0.001]、受教育年限<6年(OR 3.134,95%CI 1.534~6.401;P=0.002)、内侧颞叶萎缩(OR 2.911,95%CI 1.385~6.121;P=0.005)和存在≥1种异常MRI征象(OR 2.823,95%CI 1.305~5.938;P=0.007)为PSCI的独立危险因素。ROC曲线分析表明,内侧颞叶萎缩和存在≥1种异常MRI征象预测PSCI的曲线下面积均较小(分别为0.595和0.584),但二者和受教育年限联合预测时的曲线下面积最大(0.818,95%CI0.756~0.880;P<0.001),其预测PSCI的敏感性和特异性分别为79.9%和71.4%。结论PSCI发生率较高,内侧颞叶萎缩结合其他异常MRI征象对PSCI有一定的预测价值。

Objective To investigate the correlation between MRI markers of neurodegenerative diseases and vascular diseases and pre-stroke cognitive impairment(PSCI).Methods Patients with minor acute ischemic stroke at first onset and aged≥60 years admitted to the Department of Neurology,the Affiliated Hospital of Yangzhou University and the Department of Neurology,Linyi Jinluo Hospital from March 2019 to December 2021 were retrospectively enrolled.The imaging markers of cerebral small vessel disease and neurodegeneration were analyzed by dichotomy visual score.The former included cerebral white matter hyperintensities,vasogenic lacunar lesions,cerebral microbleeds,and enlarged perivascular space,and the latter included global cortical atrophy and medial temporal lobe atrophy.According to the score of Information Questionnaire on Cognitive Decline in the Elderly(IQCODE),the patients were divided into PSCI group(≥3.31 points)and non-PSCI group(<3.31 points).The clinical baseline data and MRI markers of both groups were compared.Multivariate logistic regression model was used to analyze the correlation between MRI markers and PSCI,and receiver operator characteristic(ROC)curve was used to analyze the predictive value of MRI markers to PSCI.Results A total of 221 patients were enrolled in the study,including 77 patients(34.8%)in the PSCI group and 144(65.2%)in the non-PSCI group.Univariate analysis showed that there were significant differences in age,years of education,pathological white matter hyperintensities,medial temporal lobe atrophy,and the proportion of patients with≥1 abnormal MRI markers between the two groups(all P<0.05).Multivariate logistic regression analysis showed that older age(odds ratio[OR]1.089,95%confidence interval[CI]1.034-1.146;P=0.001),years of education<6 years(OR 3.134,95%CI 1.534-6.401;P=0.002),medial temporal lobe atrophy(OR 2.911,95%CI 1.385-6.121;P=0.005),and presence of≥1 abnormal MRI markers(OR 2.823,95%CI 1.305-5.938;P=0.007)were the independent risk factors for PSCI.ROC curve analysis showed that the area under the curve of medial temporal lobe atrophy and the presence of≥1 abnormal MRI markers for predicting PSCI were both smaller(0.595 and 0.584 respectively),but the area under the curve was the largest when the two and years of education were combined(0.818,95%CI 0.756-0.880;P<0.001),and its sensitivity and specificity for predicting PSCI were 79.9%and 71.4%respectively.Conclusions The incidence of PSCI is high.Medial temporal lobe atrophy combined with other abnormal MRI markers has a certain predictive value for PSCI.