1Division of Neurosurgery, Sainte Justine Hospital, Montreal, Quebec, Canada.
2Department of Neurosurgery, UCLA Mattel Children's Hospital, David Geffen School of Medicine at University of California, Los Angeles, California.
J Neurosurg Pediatr. 2020 Jun 1;25(6):567-573. doi: 10.3171/2019.12.PEDS19370. Epub 2020 Feb 21.
Although hemispherectomy is an effective treatment for children with intractable hemispheric epilepsy syndromes, as many as 40% of patients eventually develop seizure recurrence. The causes of seizure recurrence in these patients are incompletely understood. The authors sought to evaluate the efficacy of hemispherectomy at their center and determine whether contralateral MRI abnormalities can predict seizure recurrence.
A retrospective review of consecutive hemispherectomies performed at Miami Children's Hospital between January 2000 and June 2014 was performed. Time-to-event analysis was performed. The "event" was defined as any seizures following resective epilepsy surgery (not including seizures in the first postoperative week and auras). Several preoperative variables were analyzed to determine their suitability to predict seizure recurrence following surgery.
Sixty-nine patients (44 boys) with a mean age of 8.2 ± 5.9 years (range 0.1-20.8 years) underwent 72 hemispherectomies; 67 of these were functional hemispherectomies, while another 5 were completion of a previous functional hemispherectomy (2 completions of functional hemispherectomies, 3 anatomical hemispherectomies). The duration of epilepsy was 5.8 ± 5.5 years with 66 cases (91.7%) having daily seizures. Etiology included stroke (n = 28), malformation of cortical development (n = 11), hemimegalencephaly (n = 11), encephalitis (n = 13), and other (n = 7). Engel class I outcome was achieved in 59 (86%) and 56 (81%) patients at 1 and 2 years of follow-up, respectively. The mean time to seizure recurrence was 33.5 ± 31.1 months. In univariate analyses, the absence of contralateral abnormalities on MRI (HR 4.09, 95% CI 1.41-11.89, p = 0.009) was associated with a longer duration of seizure freedom. The presence of contralateral MRI abnormalities was associated with contralateral ictal seizures on preoperative scalp EEG (p = 0.002). Fifteen patients experienced 20 complications (20/72, 27.8%), including the development of hydrocephalus necessitating CSF diversion in 9 cases (13%), hygroma in 1, hemispheric edema in 1, aseptic meningitis in 2, postoperative hemorrhage in 2, infection in 2, ischemic stroke in 2, and blood transfusion-contracted hepatitis C in 1 case.
Patients with bihemispheric abnormalities, as evidenced by contralateral MRI abnormalities, have a higher risk of earlier seizure recurrence following functional hemispherectomy.
EVD = external ventricular drain; MCD = malformation of cortical development; MEG = magnetoencephalography; PVWM = periventricular white matter; TTE = time-to-event; VPS = ventriculoperitoneal shunt.
尽管大脑半球切除术是治疗顽固性半球性癫痫综合征的有效方法，但多达 40%的患者最终会出现癫痫复发。这些患者癫痫复发的原因尚不完全清楚。作者旨在评估其中心行大脑半球切除术的疗效，并确定对侧 MRI 异常是否可预测癫痫复发。
回顾性分析 2000 年 1 月至 2014 年 6 月期间在迈阿密儿童医院行连续大脑半球切除术的患者。采用时间事件分析。“事件”定义为切除性癫痫手术后的任何癫痫发作（不包括术后第 1 周和先兆的癫痫发作）。分析了几个术前变量，以确定它们是否适合预测手术后癫痫复发。
69 例（44 例男性）患者，平均年龄 8.2±5.9 岁（0.1-20.8 岁），行 72 例大脑半球切除术；其中 67 例为功能性大脑半球切除术，另有 5 例为前次功能性大脑半球切除术的补充切除（2 例功能性大脑半球切除术的补充切除，3 例解剖性大脑半球切除术）。癫痫持续时间为 5.8±5.5 年，66 例（91.7%）患者每日发作。病因包括中风（n=28）、皮质发育畸形（n=11）、半脑巨脑症（n=11）、脑炎（n=13）和其他（n=7）。1 年和 2 年随访时，分别有 59 例（86%）和 56 例（81%）患者获得 Engel Ⅰ级结果。癫痫复发的平均时间为 33.5±31.1 个月。在单因素分析中，对侧 MRI 无异常（HR 4.09，95%CI 1.41-11.89，p=0.009）与更长的无癫痫发作时间相关。对侧 MRI 异常与术前头皮 EEG 上对侧癫痫发作有关（p=0.002）。15 例患者发生 20 例并发症（20/72，27.8%），包括 9 例（13%）需要 CSF 引流的脑积水、1 例血肿、1 例半球水肿、2 例无菌性脑膜炎、2 例术后出血、2 例感染、2 例缺血性中风和 1 例输血相关丙型肝炎。
有证据表明对侧 MRI 异常的双侧半球异常患者，在功能性大脑半球切除术后癫痫复发的风险更高。
EVD = 脑室外引流；MCD = 皮质发育畸形；MEG = 脑磁图；PVWM = 脑室周围白质；TTE = 时间事件；VPS = 脑室-腹腔分流术。
