儿童局灶性皮质发育不良Ⅱ型临床预测量表的构建及验证  

Development and validation of a clinical prediction scale for pediatric focal cortical dysplasia type Ⅱ

作　　者：周伯丞 孙宇[1] 刘庆祝[1] 于昊 刘畅[1] 王垚 王爽[1,2] 刘晓燕[1,2] 姜玉武[1,2] 蔡立新[1] Zhou Bocheng;Sun Yu;Liu Qingzhu;Yu Hao;Liu Chang;Wang Yao;Wang Shuang;Liu Xiaoyan;Jiang Yuwu;Cai Lixin(Pediatric Epilepsy Center,Peking University First Hospital,Beijing 102627,China;Department of Pediatrics,Peking University First Hospital,Beijing 102627,China)

机构地区：[1]北京大学第一医院儿童癫痫中心,北京102627 [2]北京大学第一医院儿科,北京102627

出　　处：《中华实用儿科临床杂志》2024年第8期579-583,共5页Chinese Journal of Applied Clinical Pediatrics

基　　金：国家自然科学基金(82071263)。

摘　　要：目的在儿童皮质发育畸形(MCD)疾病谱中构建预测局灶性皮质发育不良(FCD) Ⅱ型的临床预测量表。方法横断面研究。纳入2014年1月至2019年6月于北京大学第一医院儿童癫痫中心接受手术治疗,且术后病理诊断为MCD的患儿,采用随机数编号,将其分为训练集和验证集,在训练集中分析临床、电生理及影像数据,通过逻辑回归模型筛选出能够预测FCD Ⅱ型的变量,并构建评分量表,在验证集中对评分量表的诊断效能进行验证,确定最佳截断值,并进行一致性检验。结果共纳入381例患儿,其中训练集260例,验证集121例。在训练集中使用逻辑回归模型确定了5个与FCD Ⅱ型显著相关的独立影响因素:(1)癫痫发作起病年龄(<24个月);(2)额叶受累;(3)癫痫性痉挛发作;(4)癫痫家族史;(5)海马萎缩±信号改变。使用这5个变量构建FCD Ⅱ型预测量表,并在验证集中进行验证,曲线下面积为0.732。预测量表的最佳截断值为1分,约登指数为0.384,量表的阳性预测值为0.836,阴性预测值为0.500。病理诊断和FCD Ⅱ型预测量表对于FCD Ⅱ型的诊断一致性尚可(Kappa值为0.351),且二者差异无统计学意义(McNemar检验P=0.065)。结论 FCD Ⅱ型预测量表在临床上具有一定的实用性,应用此量表术前对MCD患者的病理类型进行预测,可以帮助医师选择合适的手术策略。ObjectiveTo construct a clinical prediction scale for focal cortical dysplasia(FCD)typeⅡin the malformation of cortical development(MCD)disease spectrum in children.MethodsA case-sectional study.From January 2014 to June 2019,patients who underwent surgery at the Pediatric Epilepsy Center of Peking University First Hospital and were pathologically diagnosed with MCD after surgery were enrolled and randomly divided into the training set and the validation set using random numbering.Clinical,electrophysiological,and imaging data of patients in the training set were analyzed.Variables that could predict FCD typeⅡwere screened out using a Logistic regression model,and a rating scale was constructed.The diagnostic efficiency of the scale was validated in the validation set to determine the optimum cut-off value,and a consistency test was performed.ResultsA total of 381 patients were enrolled in the study,with 260 in the training set and 121 in the validation set.Five clinical factors that exhibited a significant correlation with FCD typeⅡwere identified in the training set through the logistic regression model:(1)age of seizure onset(<24 months);(2)lesion involving the frontal lobe;(3)epileptic spasms;(4)family history of epilepsy;(5)hippocampal atrophy±signal change.Based on these 5 variables,the FCD typeⅡprediction scale was developed and validated in the validation set with an area under the curve of 0.732.The optimum cut-off value for the prediction scale was 1,at which point the Youden index was 0.384.The scale′s positive predictive value was 0.836,and the negative predictive value was 0.500.The diagnostic consistency between the pathological diagnosis and the FCD typeⅡprediction scale was acceptable(Kappa value=0.351),and there was no statistically significant difference between the two diagnostic methods(P value of the McNemar test=0.065).ConclusionsThe FCD typeⅡprediction scale has clinical practicability.The application of this scale to predict the pathological type of MCD before operation can

关 键 词：皮质发育畸形 局灶性皮质发育不良Ⅱ型 癫痫术前评估 儿童癫痫外科 

分 类 号：R742.1[医药卫生—神经病学与精神病学]