1McConnell Brain Imaging Center, Montreal Neurological Institute, McGill University, Montréal.
2Philips Canada, MRI Research Department, Montréal; and Departments of.
J Neurosurg Pediatr. 2020 Dec 25;27(3):243-252. doi: 10.3171/2020.7.PEDS20397. Print 2021 Mar 1.
The authors sought to assess the utility of arterial spin labeling (ASL) perfusion 3T-MRI for the presurgical evaluation of poorly defined focal epilepsy in pediatric patients.
Pseudocontinuous ASL perfusion 3T-MRI was performed in 25 consecutive children with poorly defined focal epilepsy. ASL perfusion abnormalities were detected qualitatively by visual inspection and quantitatively by calculating asymmetry index (AI) maps and significant z-score cluster maps based on successfully operated cases. ASL results were prospectively compared to scalp EEG, structural 3T-MRI, FDG-PET, ictal/interictal SPECT, magnetoencephalography (MEG), and intracranial recording results, as well as the final surgically proven epileptogenic zone (EZ) in operated patients who had at least 1 year of good (Engel class I/II) seizure outcome and positive histopathology results.
Qualitative ASL perfusion abnormalities were found in 17/25 cases (68%), specifically in 17/20 MRI-positive cases (85.0%) and in none of the 5 MRI-negative cases. ASL was concordant with localizing scalp EEG findings in 66.7%, structural 3T-MRI in 90%, FDG-PET in 75%, ictal/interictal SPECT in 62.5%, and MEG in 75% of cases, and with intracranial recording results in 40% of cases. Eleven patients underwent surgery; in all 11 cases the EZ was surgically proven by positive histopathology results and the patient having at least 1 year of good seizure outcome. ASL results were concordant with this final surgically proven EZ in 10/11 cases (sensitivity 91%, specificity 50%). All 10 ASL-positive patients who underwent surgery had positive surgical pathology results and good long-term postsurgical seizure outcome at a mean follow-up of 39 months. Retrospective quantitative analysis based on significant z-score clusters found 1 true-positive result that was missed by qualitative analysis and 3 additional false-positive results (sensitivity 100%, specificity 23%).
ASL supports the hypothesis regarding the EZ in poorly defined focal epilepsy cases in children. Due to its convenience and noninvasive nature, the authors recommend that ASL be added routinely to the presurgical MRI evaluation of epilepsy. Future optimized quantitative methods may improve the diagnostic yield of this technique.
作者旨在评估动脉自旋标记（ASL）灌注 3T-MRI 在前瞻性评估儿童局灶性癫痫中的应用价值。
对 25 例局灶性癫痫定义不明确的儿童进行伪连续 ASL 灌注 3T-MRI 检查。通过视觉检查定性检测 ASL 灌注异常，并通过计算基于成功手术病例的不对称指数（AI）图和显著 z 分数聚类图进行定量检测。将 ASL 结果与头皮脑电图、结构 3T-MRI、FDG-PET、发作间/发作期 SPECT、脑磁图（MEG）以及颅内记录结果进行前瞻性比较，并与接受手术且至少有 1 年良好（Engel 分级 I/II）发作结局和阳性组织病理学结果的手术患者的最终手术证明致痫区（EZ）进行比较。
25 例患者中 17 例（68%）发现 ASL 灌注异常，20 例 MRI 阳性病例中有 17 例（85.0%），5 例 MRI 阴性病例中无一例发现异常。ASL 与 66.7%的局部头皮脑电图发现、90%的结构 3T-MRI、75%的 FDG-PET、62.5%的发作间/发作期 SPECT 和 75%的 MEG 结果一致，与 40%的颅内记录结果一致。11 例患者接受手术治疗；在所有 11 例患者中，最终手术证明 EZ 阳性，组织病理学结果阳性，且患者至少有 1 年良好的发作结局。ASL 结果与这 11 例患者最终手术证明的 EZ 结果一致（敏感性 91%，特异性 50%）。10 例接受手术的 ASL 阳性患者均有阳性手术病理结果，且在平均 39 个月的随访中术后癫痫发作得到良好控制。基于显著 z 分数聚类的回顾性定量分析发现，1 例定性分析漏诊的真阳性结果和 3 例假阳性结果（敏感性 100%，特异性 23%）。
ASL 支持儿童局灶性癫痫患者 EZ 的假说。由于其方便和无创性，作者建议将 ASL 常规添加到癫痫的术前 MRI 评估中。未来优化的定量方法可能会提高该技术的诊断效果。
