Kuhl D E, Engel J, Phelps M E, Selin C
Ann Neurol. 1980 Oct;8(4):348-60. doi: 10.1002/ana.410080403.
Seventeen patients with partial epilepsy had electroencephalographic (EEG) monitoring concurrent with cerebral positron emission computed tomography (PECT) after 18F-fluorodeoxyglucose (18FDG) and 13N-ammonia (13NH3) were given intravenously as indicators of local cerebral glucose utilization (LCMRglc) and relative perfusion, respectively. In 12 of 15 patients who had unilateral or focal electrical abnormalities, interictal 18FDG scan patterns clearly showed localized regions of decreased (14 to 58%) LCMRglc that correlated anatomically with the eventual EEG localization. These hypometabolic zones appeared normal on x-ray computed tomography in all but 3 patients and were unchanged on 18FDG scans repeated on different days. In 5 of 6 patients who underwent anterior temporal lobectomy the interictal 18FDG scan correctly detected the pathologically confirmed lesion as a hypometabolic zone, and removal of the lesion site resulted in marked clinical improvement. In contrast, the ictal 18FDG scan patterns clearly showed foci of increased (82 to 130%) LCMRglc that correlated temporally and anatomically with ictal EEG spike foci and were within the zones of interictal hypometabolism (three studies in 2 patients). 13NH3 distributions paralleled 18FDG increases and decreases in abnormal zones, but 13NH3 differences were of lesser magnitude. When the relationship of 13NH3 uptake to local blood flow found in dog brain was applied as a correction to the patients' 13NH3 scan data, local alterations in perfusion and glucose utilization were usually matched in both the interictal and the ictal state. We conclude that the interictal 18FDG-PECT scan is useful in aiding localization of the dysfunctional cerebral zone most likely to be responsible for seizures in patients being considered for anterior temporal lobectomy. With further development, emission computed tomography may help in better categorizing the various forms of the disorder and in elucidating the basic mechanisms of epilepsy in humans.
17例部分性癫痫患者在静脉注射18F-氟脱氧葡萄糖（18FDG）和13N-氨（13NH3）后，分别作为局部脑葡萄糖利用率（LCMRglc）和相对灌注的指标，同时进行脑电图（EEG）监测和脑正电子发射计算机断层扫描（PECT）。在15例有单侧或局灶性电异常的患者中，12例的发作间期18FDG扫描模式清楚地显示出局部LCMRglc降低区域（14%至58%），其在解剖学上与最终的EEG定位相关。除3例患者外，这些低代谢区在X线计算机断层扫描上显示正常，且在不同日期重复进行的18FDG扫描中无变化。在6例接受前颞叶切除术的患者中，5例的发作间期18FDG扫描正确地将病理证实的病变检测为低代谢区，切除病变部位后临床症状明显改善。相比之下，发作期18FDG扫描模式清楚地显示出LCMRglc增加区域（82%至130%），其在时间和解剖学上与发作期EEG棘波灶相关，且位于发作间期低代谢区内（2例患者的3项研究）。13NH3分布与异常区域内18FDG的增加和减少平行，但13NH3的差异幅度较小。当将在犬脑中发现的13NH3摄取与局部血流的关系应用于校正患者的13NH3扫描数据时，灌注和葡萄糖利用的局部改变在发作间期和发作期通常是匹配的。我们得出结论，发作间期18FDG-PECT扫描有助于对考虑进行前颞叶切除术的患者中最可能导致癫痫发作的功能失调脑区进行定位。随着进一步发展，发射计算机断层扫描可能有助于更好地对该疾病的各种形式进行分类，并阐明人类癫痫的基本机制。
