脑膜瘤术后癫疒间发作的危险因素分析及预警模型建立

Risk factors analysis and warning model establishment for postoperative seizures in patients with meningioma

目的探讨脑膜瘤术后癫疒间发生的危险因素,并建立风险预警模型。方法回顾性分析148例脑膜瘤病人的临床资料,根据术后有无癫疒间发作分组。通过单因素和多因素分析确定癫疒间发作的危险因素,并建立预警模型,同时应用拟合优度检验和ROC曲线评价模型效能。结果脑膜瘤术后癫疒间发作共32例,其中疒间性新发18例,疒间性再发14例。多因素统计分析示:饮酒、肿瘤MRI强化、脑膜尾征、肿瘤直径、水肿程度、肿瘤黏连是术后癫疒间发作的独立危险因素(P<0.05)。亚组分析示:饮酒、肿瘤直径、水肿程度、肿瘤黏连是术后疒间性新发的独立危险因素(P<0.05);而肿瘤直径、术前癫疒间发作类型是术后疒间性再发的独立危险因素(P<0.05)。拟合优度检验和ROC曲线分析示:术后癫疒间发作、疒间性新发、疒间性再发的Logit (P)模型预警效能良好(P<0.001)。结论脑膜瘤癫疒间发作与多种危险因素有关,且各种危险因素在不同癫疒间发作特点的病人间存在差异。临床上可依据上述风险因素构建预警模型,提高脑膜瘤管理。

Objective To investigate the risk factors related to postoperative seizures in meningioma, and establish risk warning models. Methods Clinical data of 148 patients with meningioma were analyzed retrospectively, and the patents were divided into two groups according to whether or not existing postoperative seizure. The risk factors of postoperative seizures were determined by univariate and multivariate analyses. The warning models were established and their effectiveness was assessed by goodness-of-fit test and receiver operating characteristic(ROC) curve. Results Postoperative seizures occurred in 32 patients, including 18 cases of postoperative new seizures and 14 recurrent seizures. Statistical analysis showed that drinking, tumor enhancement, dural tail sign,tumor size, edema degree, tumor adhesion were the independent risk factors of postoperative seizures(P<0.05). In further subgroup analysis, postoperative new seizures were associated with drinking, tumor size, edema degree, tumor adhesion, while recurrent seizures were associated with tumor size and preoperative seizure type(P<0.05). Goodness-of-fit test and ROC curve analysis revealed that 3 risk warning models had a good fitness of data and accuracy in predicting seizures(P<0.001). Conclusions There are so many factors affecting operative seizures in meningioma, and risk factors are different in patients with a variety of seizure characteristics. The risk warming models can be established based on these factors to improve meningioma management.