癫痫患者诊断与药物选择工具EpiPick的验证研究

Validation of the EpiPick tool for diagnosis and drug selection in epilepsy patients

目的利用已建的癫痫患者数据库,验证网页版癫痫诊断与抗癫痫发作药物(ASM)选择工具EpiPick用于国内癫痫患者的效能。方法回顾性收集2017年1月至2020年12月在昆明医科大学第一附属医院综合癫痫中心就诊的10岁以上确诊癫痫、定期随访和信息完整患者的临床资料。根据EpiPick工具推荐与患者实际使用的首个ASM是否一致,将患者分为EpiPick组和临床组。分析比较两组服用首个ASM后的药物保留率、Engel评分与30个月内连续无发作的累积概率,绘制Kaplan-Meier生存曲线。再检验EpiPick工具提供的诊断结果与临床实际诊断的癫痫发作类型的一致性。结果共纳入364例癫痫患者,其中EpiPick组237例,临床组127例。EpiPick组和临床组的药物保留率分别为67.9%(161/237)、56.7%(72/127),差异有统计学意义(χ²=4.534,P=0.039);EpiPick组的Engel评分Ⅰ、Ⅱ、Ⅲ、Ⅳ级分别占47.3%(112/237)、14.8%(35/237)、12.7%(30/237)、25.3%(60/237),显著优于临床组的32.3%(41/127)、11.8%(15/127)、11.0%(14/127)、44.9%(57/127),两组间比较差异有统计学意义(χ²=14.968,P=0.002)。在开始服首个ASM后的1个月、6个月、12个月、30个月及以上时,EpiPick组的累积无发作率分别为73.8%、61.2%、53.2%、50.6%,而临床组分别为52.0%、44.1%、40.2%、33.5%,logrank检验结果显示差异有统计学意义(HR=0.644,95%CI 0.476~0.871,P<0.001)。按癫痫发作类型——局灶性发作(196例)和全面性发作(168例)进行分组后,EpiPick组开始服用ASM后的1个月、6个月、12个月、30个月及以上累积无发作率仍显著高于临床组(局灶性发作患者两组间比较:HR=0.654,95%CI 0.443~0.964,P=0.004;全面性发作患者两组间比较:HR=0.586,95%CI 0.361~0.954,P=0.014)。全部364例患者中,293例临床诊断发作分类与EpiPick工具的分型结果一致,EpiPick组与临床组相比较,全面性发作分类一致性为83.9%(104/124),局灶性发作为78.8%(189/240;Kappa=0.591,P<0.001)。结论EpiPick网页工具更适用于推荐首个ASM,可为中国非癫痫专病医师在癫痫药物治疗决策时提供简便全面的参考工具。

Objective Using the established epilepsy patient database to validate the efficacy of the web-based epilepsy diagnosis and anti-seizure medications(ASM)selection tool,EpiPick,for domestic epilepsy patients.Methods The retrospective collection of clinical data was conducted on patients aged 10 and above who were diagnosed with epilepsy at the Comprehensive Epilepsy Center of the First Affiliated Hospital of Kunming Medical University from January 2017 to December 2020,with regular follow-up and complete information.According to the first ASM recommended by the EpiPick tool and whether they are consistent with the actual ASM used by patients,patients were divided into EpiPick group and clinical group to verify the effectiveness of the EpiPick tool in selecting ASM.The drug retention rate,Engel score,and cumulative probability of no consecutive episodes within 30 months after using the first ASM were compared between the 2 groups,and Kaplan-Meier survival curves were drawn.Finally,the diagnostic results provided by the EpiPick tool were compared with the actual types of epileptic seizures diagnosed clinically,and consistency tests were performed.Results A total of 364 epilepsy patients were included,including 237 in the EpiPick group and 127 in the clinical group.The ASM retention rates of patients in the EpiPick group and clinical group were 67.9%(161/237)and 56.7%(72/127),respectively,with statistically significant differences(χ²=4.534,P=0.039).GradesⅠ,Ⅱ,ⅢandⅣaccording to the Engel scores in the EpiPick group patients who took the first ASM after diagnosis accounted for 47.3%(112/237),14.8%(35/237),12.7%(30/237),and 25.3%(60/237),respectively,compared to the clinical group of 32.3%(41/127),11.8%(15/127),11.0%(14/127),and 44.9%(57/127),respectively.There was a statistically significant difference in Engel scores between the 2 groups(χ²=14.968,P=0.002).The cumulative seizure-free rates in the EpiPick group at the 1st,6th,12th,30th month and above after starting the first ASM were 73.8%,61.2%,53.2%,and 50.6%,respectively,which in the clinical group were 52.0%,44.1%,40.2%,and 33.5%,respectively.The logrank test showed a statistically significant difference in the cumulative probability of consecutive seizure freedom between the 2 groups(HR=0.644,95%CI 0.476-0.871,P<0.001).After grouping by seizure type[focal seizures(196 cases)and generalized seizures(168 cases)],the cumulative seizure-free rates at the 1st,6th,12th,30th month and above after starting ASM were significantly higher in the EpiPick group than in the clinical group(comparison between the 2 groups in patients with focal seizures:HR=0.654,95%CI 0.443-0.964,P=0.004;comparison between the 2 groups in patients with generalised seizures:HR=0.586,95%CI 0.361-0.954,P=0.014).Among 364 patients,293 cases were clinically diagnosed with seizure classification consistent with the classification results of EpiPick tool.Agreement between the algorithm and the experts in classifying generalized seizures was 83.9%(104/124),which in classifying focal seizures was 78.8%(189/240;Kappa=0.591,P<0.001).Conclusion Web-based EpiPick tool is suitable to be used to select the first ASM,and is portable for Chinese non-epilepsy specialists to choose ASM for epilepsy patients.