INTRACTABLE EPILEPSY--PRESURGICAL STRATEGY AND OUTCOME

外科手术越来越多地被用作治疗难治性疾病 癫痫，而前颞叶切除术(ATL)是最常见的外科手术 程序。然而，很少有对照或前瞻性研究。 用于定位可切除的致痫区域的测试或 ATL的综合结果。这项研究的长期目标是 开发一种策略，通过非侵入性测试识别最大的 可能患有顽固性癫痫的患者数量 ATL成功治疗并决定ATL的综合疗效 使用多种措施。预期的结果将减少 需要长期进行脑电检查的患者 在手术前，并将提供更完整的关键信息 和适当的术前知情同意程序。的具体目标 这项研究是为了；确定一种准确、高效和具有成本效益的 非侵入性检测的算法，它选择了大多数患者 ATL能成功治疗顽固性癫痫，2)确定 ATL使用多种功能、生活质量、 3)决定ATL的发病率和病死率。这个 这项研究是一项前瞻性的非随机对照试验，分为两部分 阶段；第一阶段测试非侵入性算法，第二阶段确定 综合结果。算法中的预测器变量基于 发作间期脑电、脑电和临床发作的分析 视频/脑电监测、神经检查、发烧病史 抽搐，磁共振成像(MRI)海马区异常， MR的肌酸/N-乙酰天冬氨酸比值增加 光谱学。加权测试标准的组合将决定 不需要颅内监测构象的ATL候选资格 致痫区域。第一阶段的成果衡量标准已经完成 停止癫痫发作。第二阶段的预测变量为ATL。这个 第二阶段的结果衡量标准是癫痫停止、生活质量、 神经心理功能、精神状态、职业和教育 成就和患者满意度。生活质量、神经心理学 功能和精神状态将由以前的 标准化和验证性测试。将重复所有结果衡量标准 每门课程每年进行一次，直到研究结束；招生期间 将是四年，这样所有患者将至少有一年的时间 后续行动。75%的受试者将被跟踪两次以上 好几年了。对照组将颞叶癫痫患者确诊为 推迟手术或最终不是手术候选人的算法 切除手术(即，双颞叶癫痫发作)，并随后进行 至少1年。第一阶段的统计分析将包括物流 回归分析，首次发作时间的比例风险模型， 以及一个循序渐进的过程。第二阶段的统计分析将包括 适当的分类格式(卡方检验或费舍尔精确检验)， 重复测量分析、交互作用评估和方差分析 协方差调整。

Surgery is being used increasingly as a treatment for medically intractable epilepsy, and anterior temporal lobectomy (ATL) is the most common surgical procedure. However, there are very few controlled or prospective studies of the tests employed to localize a resectable epileptogenic region or of comprehensive outcome from ATL. The long-term objectives of this study are to develop a strategy that will identify by noninvasive tests the largest possible number of patients with intractable epilepsy who can be successfully treated with ATL and determine comprehensive outcome from ATL using multiple measures. The expected results will decrease the number of patients who require long-term intracranial electroencephalography (EEG) prior to surgery and will provide critical information for a more complete and appropriate presurgical informed consent process. The specific aims of the study are to; 21) determine an accurate, efficient, and cost-effective algorithm of noninvasive tests that selects a majority of patients with intractable epilepsy who can be treated successfully by ATL, 2) determine the outcome from ATL using multiple measures of function, quality of life, and seizures, and 3) determine the morbidity and mortality from ATL. The study is a prospective, nonrandomized controlled trail divided into two phases; phase I tests the noninvasive algorithm and phase II determines comprehensive outcome. The predictor variables in the algorithm are based on interictal EEG, electrographic and clinical seizures analyzed by video/EEG monitoring, neurological examination, history of febrile convulsions, hippocampal abnormality on magnetic resonance imaging (MRI), and an increased temporal creatine/N-acetyl aspartate ration on MR spectroscopy. Combinations of weighted test criteria will determine candidacy for ATL without requiring intracranial monitoring conformation of the epileptogenic zone. The outcome measure for phase I is complete cessation of seizures. The predictor variable for phase II is ATL. The outcome measures for phase II are seizure cessation, quality of life, neuropsychological function, psychiatric status, vocational and educational achievement, and patient satisfaction. Quality of life, neuropsychological function, and psychiatric status will be assessed by previously standardized and validated tests. All outcome measures will be repeated yearly on each subject until the end of the study; the enrollment period will be four years so that all patients will have a minimum of 1 year follow-up. Seventy-five percent of subjects will be followed for more two years. Controls will be patients with temporal lobe epilepsy confirmed by the algorithm who defer surgery or are ultimately not candidates for resective surgery (i.e., bitemporal seizure onset) and are followed for a minimum of 1 year. Statistical analysis of phase I will include logistic regression analysis, proportional hazards model for time to first seizure, and a step-wise procedure. Phase II statistical analysis will include as appropriate a categorical format (chi square or Fisher's Exact test), repeated measures analysis, interaction assessment, and ANOVA with covariance adjustment.