Surgical Outcome Of MRI-negative Refractory Extratemporal Lobe Epilepsy

Objective:It is estimated that about 30%of patients with epilepsy will be refractory to medical therapy.Epilepsy surgery is an important therapeutic option in these patients.Of these,the majority are for seizures arising in the medial temporal lobe.The extratemporal epilepsies constitute perhaps 35%of all refractory epilepsy.Patients with normal MRI findings and extratemporal lobe epilepsy are very rare and have less favorable outcomes.Most studies in the world investigating the outcomes of patients with normai MRI findings who underwent(nonlesional)extratemporal epilepsy surgery are confined to a highly select group of patients with limited follow-up.An increasing proportion of patients in recent years who undergo invasive epilepsy surgery evaluations are MRI-negative extratemporal lobe epilepsy(MNETLE)patients for the development of SEEG.The aim of this study is to determine outcome of resective epilepsy surgery in MNETLE patients who underwent invasive evaluations and to determine factors governing outcome.This study aims to analyze the factors associated with the prognosis of surgery and to improve the outcome of MNETLE surgery.We retrospectively reviewed a cohort of patients who underwent epilepsy surgery for MNETLE.We analyzed electro-clinical,brain imaging,and intracranial EEG(iEEG)features to determine outcome and predictors of outcome.Methods:We retrospectively studied 28 consecutive patients(13 females,15 male)admitted to the Epilepsy Monitoring Unit at University Hospital Case Medical Center from August 2006 to November 2015,with completely negative epilepsy surgery protocol 3 Tesla MRI studies.MRI-negative defined as epilepsy protocol 3.0T MRI scan without an epileptogenic lesion.Extratemporal epilepsy was defined as epilepsy where the seizure onset zone,as determined by intracranial EEG onset and/or successful resection,was completely outside the anatomical boundaries of the temporal lobe.Four patients(2 hemispherectomy,1 occipital-temporal disconnection and 1 occipital-parietal-temporal resection)were included because their intracranial EEG onsets were clearly outside the temporal lobe,but underwent wider resections to maximize chances of seizure freedom.Patients with prior resective epilepsy surgeries in outside hospitals,and those with subtle MRI findings such as blurring of grey-white matter junctions were excluded.Electro-clinical,pathological and surgical data were evaluated.All patients had prolonged surface video-EEG monitoring using the 10-10 International Electrode System and subsequent intracranial EEG.In surface EEG,a regional pattern was defined as interictal epileptiform discharges(IEDs)or EEG seizure onsets restricted to one chain,or had amplitude in one chain that was at least twice as high as the adjacent bipolar electrode chain.A lateralized pattern was determined if both ipsilateral electrode chains were involved and the contralateral hemisphere pattern was<50%in amplitude.A generalized pattern was identified defined if the initial EEG change was bi-hemispheric and diffuse.Non regional included lateralized and generalized patterns.In intracranial EEG,regional epileptiform discharges were considered if they were confined to a single sub-lobar region.High Frequency Oscillations(HFOs)were defined as a rhythmical fast activity with frequency higher than 80Hz in invasive EEG recording.Subdural grids(SDG)and/or stereotactic(SEEG)electrodes were implanted to:(1)to more precisely localize the ictal onset zone,and(2)to perform functional mapping for eloquent areas.Intraoperative electrocorticography(ECoG)was used in some cases to further avoid eloquent cortex resection.Their positioning and number are designed according to the prior hypotheses to address the following issues:(1)Demonstrating that brain regions suspected of being involved in seizure onset and early propagation show the expected ictal pattern.(2)Considering the possibility that this pattern might in fact reflect the propagation of an ictal discharge generated elsewhere.(3)Delineating the border of the epileptogenic zone as precisely as possible,to perform the minimum cortical resection.(4)Assessing whether the complete removal of the epileptogenic zone will be possible or not.For this,the eloquent areas that are of interest,relative to the hypothetical epileptogenic zone and with respect to the possible boundaries of the planned resection,need to be investigated.(5)Evaluating the precise relationships between an anatomical lesion(when present)and the epileptogenic zone.For this,whenever possible,the epileptogenicity of the lesion itself needs to be investigated,and in any cases of the surrounding cortex,the number of the"lesional"electrodes to use depending on the morphology,extent,and anatomical location of the lesional process.Epilepsy surgery outcome,assessed at last follow-up visit,was classified using the International League Against Epilepsy(ILAE)scoring system.In our study,only ILAE class 1 was classified as seizure freedom and ILAE class 1-3 were considered as good outcome.Focal cortical dysplasia(FCD)was classified according to the ILAE classification.Statistical analysis was performed using SPSS statistics 22(IBM).Wilcoxon rank-sum test was used to analyze type of iEEG and surgery,and pathology findings.Fisher exact test was used to analyze sex,seizure frequency,seizure type,irritative zone,High Frequency Oscillations(HFOs),EEG onset zone,lateralization and localization of operation.We used an Analysis of Variance(ANOVA)for age at surgery,age at seizure onset,and duration of epilepsy.Statistical significance was defined as P<0.05.Additional multivariate Cox proportional hazard regression model was used.Results were considered statistically significant at the 5%level.Kaplan-Meier survival analysis was done to calculate the probability of seizure freedom.Statistical significance was tested using the log-rank test and comparison of 95%confidence intervals(CIs).Results:28 patients were included in this study.15(53.6%)were adult(≥18 years of age),13(46.4%)were children and 15(53.6%)were male.The mean age at epilepsy onset was 9.06 years(range 0.1-29 years),and at surgery was 16.75 years(range 0.9-58 years).Mean duration of epilepsy was 7.66 years(range 0.1-41 years),and 10 patients(35.7%)had left-sided surgery.25 patients underwent interictal positron emission tomography(PET),and 13 patients had interictal and ictal single-photo emission computed tomography(SPECT).Complete baseline and post-op neuropsychological evaluations were done in most patients.No significant association was found with outcome after surgery.All cases had at least 6 months of follow up(mean 32[6-1 13]months).IEDs in surface EEG were found in all;patients,19(67.9%)of those patients had regional IEDs.17(60.7%)patients had regional EEG onset in the surface EEG.In 2(3.6%)patients,surface EEG onset was obscured by artifact.HFOs in intracranial EEG were’ found in 13(54.2%)patients.6(21.4%)patients had auras(3 visual aura,2 somatosensory aura and 1 unspecific aura),18(64.3%)had complex partial seizures,and 9 had(31.0%)versive seizures.10(37.5%)patients had secondary generalized tonic-clonic seizures(GTCS).Seizure semiology had lateralizing value in 16(57.1%).24(85.7%)patients underwent an intracranial EEG evaluation.9(32.1%)patients had stereoelectroencephalography(SEEG),7(25%)subdural grids(SDG)and 8(28.6%)both,SEEG and SDG.HFOs in intracranial EEG were found in 13(54.2%)patients Epileptogenic zone was studied in 28 patients;in 18(64.3%)it was lateralized to the right hemisphere.The most frequent location of the epileptogenic zone was the frontal lobe.The following types of resections were performed:frontal(16 patients),occipital(4 patients),parietal(2 patients),parieto-occipital(1 patient),hemispheric(2 patients),occipital-parietal-temporal(2 patients),insula(1 patient).18(64.3%)patients underwent a cortical resection(corticectomy),7(25%)lobectomy,2(7.1%)hemispherectomy,1(3.5%)disconnection.The mean follow up was 32[6-113]months.At the time of latest follow-up,13(46.4%)patients were seizure-free(ILAE1)and 18(64.3%)had a good(ILAE 1,2,3)outcome.The most frequent pathological findings were focal cortical dysplasia(FCD)in 21 patients.Most of them were FCD ILAE type Ⅱ(13).In univariate analysis with a 2 tailed Fisher exact test,regional surface IEDs(P=0.016),regional surface EEG onset(P=0.024)and regional intracranial(P=0.036)EEG onset were significantly associated with seizure freedom(Table 3).A single type of intracranial evaluation(SEEG or SDG)was found to have a predictive value for better outcome than those with both types of intracranial evaluations(P=0.034)(Table 3).No significant relationship was found between clinical and pathological variables and operation type with surgical outcome.Kaplan-Meier survival analysis was done to estimate the probability of seizure freedom predictors after surgery.Only regional surface IEDs and>120 electrode contact studies were statistically significant.Multivariate analysis(Multivariate Cox proportional hazard regression model)was used to determine independent variables that can be used as predictors of recurrence(or seizure freedom).Only undergoing evaluations with<120 electrode contacts was found to have significant value predicting seizure recurrence(HR=4.283,CI=1.342-13.676).Surgical complications were observed in 5 patients.One complication(asymptomatic hematoma)was associated with depth electrodes implantation in the frontal lobe.New neurologic deficits developed after surgery in 4 patients(14.3%);hemianopia in one,upper limb paresthesias in one and hemiparesis in two(after hemispherectomy).Conclusion:As perhaps can be expected,poorer surgical outcomes have more commonly been reported in MNETLE than in MRI positive ETLE.Invasive EEG is a powerful tool in the pre-surgical evaluation of patients with MNETLE.Invasive EEG implantation that include the irritative zone and EEG onset zone as indicated by surface EEG,as well as wider brain coverage predict seizure freedom,contingent upon a sound anatomo-electro-clinical hypothesis for implantation.