脑小血管病影像学标志预测自发性脑出血患者的血肿增大

Cerebral small vessel disease imaging markers predict hematoma expansion in patients with spontaneous intracerebral hemorrhage

目的探讨脑小血管病(cerebral small vessel disease,CSVD)影像学标志与自发性脑出血(spontaneous intracerebral hemorrhages,sICH)患者早期血肿增大(hematoma expansion,HE)的相关性。方法回顾性纳入2015年1月1日至2019年12月31日期间在青岛大学附属医院神经内科住院的sICH患者。所有患者均在发病6 h内接受颅脑非增强CT扫描(noncontrast CT,NCCT),在初次NCCT检查后24 h内进行复查以明确是否发生HE,在发病48 h内完成颅脑MRI检查。HE定义为NCCT复查显示血肿体积较基线增大>33%或>6 ml。应用NCCT评估形态和密度异常征象,包括混合征、漩涡征、黑洞征、岛征和卫星征。应用MRI评估CSVD影像学标志,包括腔隙性梗死(lacunar infarcts,LIs)、血管周围间隙扩大(enlarged perivascular space,EPVS)、脑白质高信号(white matter hyperintensities,WMHs)、脑微出血(cerebral microbleeds,CMBs)和脑皮质表面铁沉积(cortical superficial siderosis,CSS)。应用多变量logistic回归分析确定HE的独立危险因素。应用受试者工作特征(receiver operator characteristic,ROC)曲线评估影像学标志对sICH患者HE的预测能力。结果共纳入216例sICH患者,年龄(57±15)岁,男性113例(61.6%),88例(40.7%)发生HE,123例(56.9%)存在NCCT征象,122例(56.5%)存在CMBs,143例(66.2%)存在WMHs,44例(20.4%)存在CSS,25例(11.6%)存在LIs,31例(14.4%)存在EPVS。HE组基线血肿体积、血钙、入院时改良Rankin量表和美国国立卫生研究院卒中量表评分以及NCCT征象、CMBs、WMHs和CSS检出率均显著高于非HE组(P均<0.05)。多变量logistic回归分析显示,血钙[优势比(odds ratio,OR)0.040,95%置信区间(confidence interval,CI)0.004~0.238;P=0.001]、任何NCCT征象(OR 3.275,95%CI 1.492~7.188;P=0.003)、CMBs 4级(OR 3.591,95%CI 1.146~11.250;P=0.028)、CSS(OR 3.008,95%CI 1.214~7.452;P=0.017)、NCCT征象+CMBs 3级(OR 3.390,95%CI 1.035~11.102;P=0.044)、NCCT征象+CMBs 4级(OR 5.473,95%CI 1.352~22.161;P=0.017)、NCCT征象+CSS(OR 3.544,95%CI 1.215~10.336;P=0.021)是sICH患者HE的独立危险因素。ROC曲线分析显示,NCCT征象、CMBs和CSS预测HE的敏感性分别为81.8%、64.8%和34.1%,特异性分别为60.2%、60.9%和89.1%。NCCT征象+CMBs分级以及NCCT征象+CSS的预测敏感性较单一影像学标志降低(分别为59.1%和30.7%),而特异性则较单一影像学标志增高(分别为78.1%和93.7%)。结论CSVD影像学标志与sICH患者发生HE的风险密切相关。严重CMBs和CSS是sICH患者发生HE的独立危险因素。NCCT征象联合CSVD影像学标志预测HE的特异性增高,但敏感性降低。

Objective To investigate the correlation between the imaging markers of cerebral small vessel disease(CSVD)and early hematoma expansion(HE)in patients with spontaneous intracerebral hemorrhage(sICH).Methods Patients with sICH admitted to the Department of Neurology,the Affiliated Hospital of Qingdao University between January 1,2015 and December 31,2019 were enrolled retrospectively.All patients received noncontrast CT(NCCT)within 6 h after onset.Within 24 h after the initial NCCT examination,they were reexamed to determine whether HE occurred,and brain MRI examination was completed within 48 h after onset.HE was defined as the increase of hematoma volume on NCCT reexamination by>33%or>6 ml compared with the baseline.NCCT was used to evaluate the abnormal morphology and density signs,including blend sign,swirl sign,black hole sign,island sign,and satellite sign.MRI was used to evaluate CSVD imaging markers,including lacunar infarcts(LIs),enlarged perivascular space(EPVS),white matter hyperintensities(WMHs),cerebral microbleeds(CMBs),and cortical superficial siderosis(CSS).Multivariate logistic regression analysis was used to determine independent risk factors for HE.The receiver operator characteristic(ROC)curve was used to evaluate the predictive ability of imaging markers for HE in patients with sICH.Results A total of 216 patients with sICH were included.Their age was 57±15 years,113(61.6%)were male,88(40.7%)had HE,123(56.9%)had NCCT signs,122(56.5%)had CMBs,143(66.2%)had WMHs,44(20.4%)had CSS,25(11.6%)had LIs,and 31(14.4%)had EPVS.The baseline hematoma volume,blood calcium,the modified Rankin Scale score and the National Institutes of Health Stroke Scale score at admission,and detection rates of NCCT signs,CMBs,WMHs and CSS in the HE group were significantly higher than those in the non-HE group(all P<0.05).Multivariate logistic regression analysis showed that the blood calcium(odds ratio[OR]0.040,95%confidence interval[CI]0.004-0.238;P=0.001),any NCCT signs(OR 3.275,95%CI 1.492-7.188;P=0.003),CMBs grade 4(OR 3.591,95%CI 1.146-11.250;P=0.028),CSS(OR 3.008,95%CI 1.214-7.452;P=0.017),NCCT signs+CMBs grade 3(OR 3.390,95%CI 1.035-11.102;P=0.044),NCCT signs+CMBs grade 4(OR 5.473,95%CI 1.352-22.161;P=0.017),and NCCT signs+CSS(OR 3.544,95%CI 1.215-10.336;P=0.021)were the independent risk factors for HE in patients with sICH.ROC curve analysis showed that the sensitivity of NCCT signs,CMBs and CSS for predicting HE were 81.8%,64.8%and 34.1%,respectively,and the specificity were 60.2%,60.9%and 89.1%,respectively.The predictive sensitivity of NCCT signs+CMBs and NCCT signs+CSS(59.1%and 30.7%,respectively)was lower than that of single imaging marker,while the specificity(78.1%and 93.7%,respectively)was higher than that of single imaging marker.Conclusions The imaging markers of CSVD are closely associated with the risk of HE in patients with sICH.Severe CMBs and CSS are the independent risk factors for HE in patients with sICH.The specificity of NCCT signs combined with CSVD imaging markers for predicting HE is increased but the sensitivity decreased.