| Objectives:The goals of this study were to assess the diagnostic value of MRI, VEEG,and MEG and combination of those tests in the pre-surgical evaluation of epilepsy, and to analysis the sensitivities and specificities and predictive values of post-surgical outcome of those tests. Methods:134 patients, which have been diagnosed as epilepsy on clinical ground and EEG and as candidates for operation, were undergone presurgical evaluation which including seizure semiology, MEG, MRI, VEEG and ECoG interoperation and subsequent surgery. The coherence of localization between pre- and inter-operation is observed. Patients were followed up and the post-surgical outcomes were assessed. By chi-square test, the accuracy and predictive value of noninvasive presurgical investigations are evaluated. Logistic regression analysis was applied to assess the relationships between pre-surgical predictors and post-surgical outcome.Results:1. One hundred thirty-four patients have undergone operation, and ratio of male/female are 91/43, and ratio of cryptogenic/symptomatic are 58/76. Mean age at surgery are 21.4 (range, 5-52yr), mean duration of epilepsy are 8.45yr (range, 0.5-40yr). Those were grouped into mesial temporal lobe epilepsy (MTLE),neocortical temporal lobe epilepsy and extratemporal epilepsy, and the numbers of patients in the three groups are 72,18 and 44, respectively. There are no differences among of three groups with respect to sex of patients, age at surgery, the duration of epilepsy. Mean period of follow-up are 0.89(range, 0.3-2.25yr), 97 patients achieved post-surgical seizure-controlled effectively. 2. 56 of 72 mesial temporal lobe epilepsy (MTLE) patients showed hippocampus and/or amygdala sclerosis on MRI, and 40 patients of the group had only this abnormality on MRI. There were 30 of the 40 patients with simple hippocampus/amygdala sclerosis on MRI who had concordant lateralized MRI and interictal-EEG and 26 (86.7％) of them acquired post-operative remission. However, only 4 of the other 10 patients with no concordant lateralization achieved seizure-relief. MEG is more sensitive to neocortical spike than the spike of mesial-basal temporal, its concordant rates with strip electrode laid on necrotic region and with depth electrode placed in mesial-basal of temporal lobe were 70.9％ and 34.7%, respectively.3 The number of patients who had bilateral epileptogenic zone determined by interictal VEEG, MEG, and ECoG were 23, 9, 9, respectively; and the number of patients whose origin of epileptiform discharges could not to be distinguished from frontal or temporal lobe was 27, 9, and 6, respectively. Chi-square test presented significant difference (P＜0.05), and suggested that MEG and ECoG are superior to VEEG in identification of the origin of epileptiform discharges come from unilateral or bilateral hemisphere and from frontal or temporal lobe. 4. The concordant rates of MRI, interictal VEEG and MEG with ECoG were 69.4％,77.6％, 70.9％ respectively. Chi-square test revealed that no statistically significant differences were found, combined with the post-surgical outcome, anyone of these noninvasive presurgical investigations alone does not has sufficient predictive value regard to post-surgical seizure free(P＞0.05). Only when the three investigations had concordant results, the concordant rate with ECoG was the highest one as 95.7%,and the post-surgical seizure free might be positively predicted (P＜0.01). 5. The sensitivities of those pre-surgical tests were between 59.8-84.5％, MRI had the most scope, and the specificities of those tests were between 35.1-66.7%, EEG of sphenoidal electrodes achieved the highest. 6. Logistic regression analyses revealed that three factors, as followed, were related with seizure controlling after operation(P＜0.05): clinical manifestation, the position of epileptogenic zone, the number of spike focus in ECoG interoperation. Conclusions:1. MRI has high sensitivity and credibility in diagnosis hippocampus/amygdala sclerosis in M...
|
PDF Full Text Request