开颅脑胶质瘤切除术后早期癫痫发作的危险因素及预测模型构建  

Risk factors of epileptic seizure after open brain glioma resection and its construction of prediction model

作　　者：葛瑾 杜伟[1] GE Jin;DU Wei(Department of Neurosurgery,the First Affiliated Hospital of Zhengzhou University,Zhengzhou 450000,Henan,China)

机构地区：[1]郑州大学第一附属医院神经外科,河南郑州450000

出　　处：《护士进修杂志》2024年第18期1947-1951,1965,共6页Journal of Nurses Training

摘　　要：目的构建开颅脑胶质瘤切除术后早期癫痫发作的预测模型。方法回顾2020年4月-2023年12月间我院收治的240例开颅脑胶质瘤切除术后患者的临床资料,根据术后早期有无癫痫发作分为癫痫组和非癫痫组,采用多因素logistic回归分析法分析引起早期癫痫发作的危险因素,并依据logistic回归分析的结果构建早期癫痫发作的风险预测模型,采用C-index、校准曲线、受试者工作特征(ROC)曲线进行模型验证。结果开颅脑胶质瘤切除术后早期癫痫发作率为19.17%。logistic回归分析显示,术前癫痫史、病理分级Ⅱ级、无AEDs预防用药、肿瘤最大径≥5 cm、累及皮层均是导致开颅脑胶质瘤切除术后早期癫痫发作的独立危险因素(P<0.05);构建预测模型logit(P)=-1.154+1.235×术前癫痫史+0.453×病理分级Ⅱ级+0.822×无AEDs预防用药+0.316×肿瘤最大径≥5 cm+1.421×累及皮层。Bootstrap法对模型内部验证,结果显示C-index=0.807(95%CI:0.725~0.892);绘制Calibration曲线分析显示,Hosmer-Lemeshowχ^(2)=0.612,P=0.322;绘制ROC曲线对模型内部预测效能验证,结果显示ROC曲线下面积(AUC)为0.893(95%CI:0.847~0.929,P<0.05),灵敏度93.48%、特异度80.93%、约登指数0.744。结论脑胶质瘤患者有术前癫痫史、病理分级低、肿瘤大、无预防性抗癫痫用药、累及皮层均是术后早期癫痫发作的危险因素,据此构建的预测模型区分度、拟合度好,对临床识别脑胶质瘤术后早期癫痫发作具有较高的预警价值。Objective To construct an prediction model for early epileptic seizures after craniotomy for open brain glioma resection.Methods The clinical data of 240 patients admitted to our hospital after craniotomy for glioma between April 2020 and December 2023 were reviewed,and the patients were divided into the epileptic and the non-epileptic groups according to the presence or absence of seizures in the early postoperative period.A multivariate logistic regression analysis was used to analyze the risk factors for early epileptic seizure,and a risk prediction model for early epileptic seizure was constructed based on the results of the logistic regression analysis.The model was evaluated using C-index,calibration curve,and receiver operating characteristic(ROC)curve.Results The early epileptic seizure rate after craniotomy for glioma resection was 19.17%.The results of logistic regression analysis showed that preoperative epilepsy history,pathological gradeⅡ,no prophylaxis with AEDs,maximum tumour diameter≥5 cm,and involvement of the cortex were all independent risk factors for early epileptic seizure after craniotomy for glioma(P<0.05).A prediction model was constructed logit(P)=-1.154+1.235×preoperative epilepsy history+0.453×pathological gradeⅡ+0.822×no prophylaxis with AEDs+0.316×maximum tumour diameter≥5 cm+1.421×involvement of the cortex.Bootstrap method was used to internal validation of the model,and the results showed that C-index was 0.807(95%CI was 0.725-0.892).The plotted calibration curve analysis showed Hosmer-Lemeshowχ^(2)=0.612,P=0.322.The ROC curve was plotted to validate the model's internal predictive efficacy,and its results showed that an area under the curve(AUC)was 0.893(95%CI was 0.847-0.929,P<0.05),a sensitivity of 93.48%,a specificity of 80.93%,and Youden index of 0.744.Conclusion The preoperative history of epilepsy,low pathological grade,large tumor,no preventive anti-epileptic drugs,and cortical involvement are all risk factors for early postoperative epilepsy in patients with

关 键 词：脑胶质瘤切除术 早期癫痫发作 危险因素 预测模型 

分 类 号：R739.41[医药卫生—肿瘤]