文献速递癫痫

特定亚群棘波可预测癫痫手术成功结局

A Subpopulation of Spikes Predicts Successful Epilepsy Surgery Outcome.
PubMed 文献 · Ann Neurol · 2022-12-10

研究分析2个癫痫中心83例手术患者的立体脑电图,在清醒、非快速眼动睡眠和快速眼动睡眠期自动检测棘波,评估135种棘波特征以判别致痫区并预测手术结局。清醒期伴前置γ活动的棘波率对手术结局分类的4折AUC为0.755±0.07,优于发作起始区(AUC=0.563±0.05,p=0.015)和涟波率(AUC=0.537±0.07,p=0.006);棘波-γ率超过1.9/分的通道位于致痫区的概率为80%。

要点

  • 清醒期伴前置γ活动的棘波率对手术结局分类的AUC为0.755±0.07。
  • 该特征优于发作起始区(AUC=0.563±0.05,p=0.015)和涟波率(AUC=0.537±0.07,p=0.006)。
  • 棘波-γ率超过1.9/分的通道有80%概率位于致痫区。

核心结论

  • 切除清醒期高棘波-γ率脑区与高概率实现无发作相关。
  • 该棘波-γ率可用于确定需切除的棘波通道最小数量。
  • 该特征易于视觉分析,可辅助临床术前评估。

研究设计与样本

  • 分析2个癫痫中心83例手术患者的立体脑电图片段,其中37例为Engel IA。
  • 在清醒期、非快速眼动睡眠期和快速眼动睡眠期自动检测棘波。
  • 评估基于频率、形态、传播和能量的135种棘波特征。
  • 采用4折交叉验证确定判别致痫区的最佳特征或特征组合。

主要发现

  • 清醒期伴前置γ活动的棘波率对手术结局分类的4折AUC为0.755±0.07。
  • 发作起始区的AUC为0.563±0.05,棘波-γ率优于发作起始区(p=0.015)。
  • 涟波率的AUC为0.537±0.07,棘波-γ率优于涟波率(p=0.006)。
  • 棘波-γ率超过1.9/分的通道位于致痫区的概率为80%。
  • 组合特征未改善结果。

关键数值与推荐

  • 清醒期伴前置γ活动的棘波率对手术结局分类的4折AUC为0.755±0.07。
  • 发作起始区的AUC为0.563±0.05。
  • 涟波率的AUC为0.537±0.07。
  • 棘波-γ率超过1.9/分的通道位于致痫区的概率为80%。

临床怎么用

  • 切除清醒期高棘波-γ率脑区与高概率实现无发作相关。
  • 该棘波-γ率可用于确定需切除的棘波通道最小数量。
  • 该特征易于视觉分析,可辅助临床术前评估。

局限性

  • 原文未报告局限性

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

OBJECTIVE: Epileptic spikes are the traditional interictal electroencephalographic (EEG) biomarker for epilepsy. Given their low specificity for identifying the epileptogenic zone (EZ), they are given only moderate attention in presurgical evaluation. This study aims to demonstrate that it is possible to identify specific spike features in intracranial EEG that optimally define the EZ and predict surgical outcome. METHODS: We analyzed spike features on stereo-EEG segments from 83 operated patients from 2 epilepsy centers (37 Engel IA) in wakefulness, non-rapid eye movement sleep, and rapid eye movement sleep. After automated spike detection, we investigated 135 spike features based on rate, morphology, propagation, and energy to determine the best feature or feature combination to discriminate the EZ in seizure-free and non-seizure-free patients by applying 4-fold cross-validation. RESULTS: The rate of spikes with preceding gamma activity in wakefulness performed better for surgical outcome classification (4-fold area under receiver operating characteristics curve [AUC] = 0.755 ± 0.07) than the seizure onset zone, the current gold standard (AUC = 0.563 ± 0.05, p = 0.015) and the ripple rate, an emerging seizure-independent biomarker (AUC = 0.537 ± 0.07, p = 0.006). Channels with a spike-gamma rate exceeding 1.9/min had an 80% probability of being in the EZ. Combining features did not improve the results. INTERPRETATION: Resection of brain regions with high spike-gamma rates in wakefulness is associated with a high probability of achieving seizure freedom. This rate could be applied to determine the minimal number of spiking channels requiring resection. In addition to quantitative analysis, this feature is easily accessible to visual analysis, which could aid clinicians during presurgical evaluation. ANN NEUROL 2023;93:522-535.

来源:Ann Neurol,查看原文。版权归原作者所有。

引用

  1. Thomas J, Kahane P, Abdallah C, et al. A Subpopulation of Spikes Predicts Successful Epilepsy Surgery Outcome. Ann Neurol. 2022. PMID: 36373178. doi:10.1002/ana.26548

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。