F.S.S. Leijten

Interictal magnetoencephalography and the irritative zone in the electrocorticogram

Magnetoencephalography (MEG) is considered a useful tool for planning electrode placement for chronic intracranial subdural electrocorticography (ECoG) in candidates for epilepsy surgery or even as a substitute for ECoG. MEG recordings are usually interictal and therefore, at best, reflect the interictal ECoG. To estimate the clinical value of MEG, it is important to know how well interictal MEG reflects interictal activity in the ECoG. From 1998 to 2008, 38 candidates for ECoG underwent a 151-channel MEG recording and 3D magnetic resonance imaging as a part of their presurgical evaluation. Interictal MEG spikes were identified, clustered, averaged and modelled using the multiple signal classification algorithm and co-registered to magnetic resonance imaging. ECoG was continuously recorded with electrode grids and strips for approximately 1 week. In a representative sample of awake interictal ECoG, interictal spikes were identified and averaged. The different spikes were characterized and quantified using a combined amplitude and synchronous surface-area measure. The ECoG spikes were ranked according to this measure and plotted on the magnetic resonance imaging surface rendering. Interictal spikes in MEG and ECoG were allocated to a predefined anatomical brain region and an association analysis was performed. All interictal MEG spikes were associated with an interictal ECoG spike. Overall, 56% of all interictal ECoG spikes had an interictal MEG counterpart. The association between the two was >or=90% in the interhemispheric and frontal orbital region, approximately 75% in the superior frontal, central and lateral temporal regions, but only approximately 25% in the mesial temporal region. MEG is a reliable indicator of the presence of interictal ECoG spikes and can be used to plan intracranial electrode placements. However, a substantial number of interictal ECoG spikes are not detected by MEG, and therefore MEG cannot be considered a substitute for ECoG.

Contralateral MRI abnormalities affect seizure and cognitive outcome after hemispherectomy

Objective: To explore whether EEG and MRI abnormalities in the “healthy” hemisphere influence seizure and cognitive outcome after functional hemispherectomy. Methods: This is a retrospective consecutive cohort study of 43 children who underwent functional hemispherectomy between 1994 and 2008. Results of preoperative EEG recordings were reviewed for the existence of (inter)ictal epileptic or background abnormalities in the contralateral hemisphere. Preoperative MRIs were reexamined for the existence of unequivocal contralateral abnormalities. Postoperative seizure status was assessed, and of 34 children, IQ or mental developmental index (MDI) scores were obtained preoperatively and postoperatively. Seizure freedom was defined as Engel 1A. Contralateral EEG and MRI abnormalities were studied in relation to seizure and cognitive outcome. Results: Thirty-three children achieved seizure freedom (77%). Of the 11 patients with contralateral MRI abnormalities, only 45% were seizure free, compared with 88% of the 32 patients without contralateral MRI lesions (p = 0.030). Children with contralateral MRI abnormalities more often were severely retarded after surgery (MDI/IQ <55; 90% vs 42%, p = 0.030). Postoperative MDI/IQ scores improved in none of the children with, but in 38% of those without contralateral MRI abnormalities (p = 0.034). Contralateral epileptic or background EEG abnormalities did not affect seizure outcome or postoperative cognitive performance. Four of 6 children with bilateral epileptic encephalopathy reached seizure freedom. Conclusion: Unambiguous contralateral MRI abnormalities are significantly associated with seizure recurrence, severe mental delay, and lack of cognitive improvement and may be considered a relative contraindication for hemispherectomy. Contralateral EEG abnormalities do not negatively influence postsurgical outcome.

Epilepsy surgery in tuberous sclerosis: The Dutch experience

INTRODUCTION Epilepsy associated with tuberous sclerosis complex (TSC) is drug resistant in more than half of the patients. Epilepsy surgery may be an alternative treatment option, if the epileptogenic tuber can be identified reliably and if seizure reduction is not at the expense of cognitive or other functions. We report the pre-surgical identification of the epileptogenic tuber and post-surgical outcome of patients with TSC in The Netherlands. METHODS Twenty-five patients underwent the pre-surgical evaluation of the Dutch Comprehensive Epilepsy Surgery Programme, including a detailed seizure history, interictal and ictal video EEG registrations, 3D FLAIR MRI scans and neuropsychological testing. Suitability of the candidates was decided in consensus. Seizure outcome, scored with the Engel classification, and cognition were reassessed at fixed post-surgery intervals. RESULTS Epilepsy surgery was performed in six patients. At follow-up, four patients had Engel classification 1, two had classification 4. Improved development and behaviour was perceived by the parents of two patients. Epilepsy surgery was not performed in 19 patients because seizures were not captured, ictal onset zones could not be localised or were multiple, interictal EEG, video EEG and MEG results were not concordant, or seizure burden had diminished during decision making. A higher cognition index was found in the surgical patients compared to the non-surgical candidates. CONCLUSIONS Epilepsy surgery can be performed safely and successfully in patients in whom semiology, interictal EEG, ictal EEG, MEG and the location of tubers are concordant. In other cases the risk of surgery should be weighed against the chance of seizure relief and in case of children subsequent impact on neurodevelopment.

Stimulation of the phrenic nerve as a complication of vagus nerve pacing in a patient with epilepsy

We report a complication of chronic vagus nerve stimulation, a new treatment for therapy-resistant epilepsy.1-3 Case report. A 42-year-old man had had between 1 and 12 seizures a month, with clustering, since he was 22. He had tried various antiepileptic drugs without success and had given up his work as a teacher. He was considered ineligible for epilepsy surgery because of adverse MRI findings and ictal EEG findings of multilobar, left-sided seizure onset. A vagus nerve stimulator (Neurocybernetics Prosthesis Model 100 and Bipolar Stimulation Leads Model 300 Series, Cyberonics, Webster, TX) was implanted in February 1997. At the time, he was taking phenytoin, vigabatrin, and topiramate. The left vagus nerve was identified between the carotid artery and the internal jugular vein, through a medial incision next to the left sternocleidomastoid muscle. Two helical electrodes …

Individualized prediction of seizure relapse and outcomes following antiepileptic drug withdrawal after pediatric epilepsy surgery

The objective of this study was to create a clinically useful tool for individualized prediction of seizure outcomes following antiepileptic drug withdrawal after pediatric epilepsy surgery. We used data from the European retrospective TimeToStop study, which included 766 children from 15 centers, to perform a proportional hazard regression analysis. The 2 outcome measures were seizure recurrence and seizure freedom in the last year of follow‐up. Prognostic factors were identified through systematic review of the literature. The strongest predictors for each outcome were selected through backward selection, after which nomograms were created. The final models included 3 to 5 factors per model. Discrimination in terms of adjusted concordance statistic was 0.68 (95% confidence interval [CI] 0.67‐0.69) for predicting seizure recurrence and 0.73 (95% CI 0.72‐0.75) for predicting eventual seizure freedom. An online prediction tool is provided on www.epilepsypredictiontools.info/ttswithdrawal. The presented models can improve counseling of patients and parents regarding postoperative antiepileptic drug policies, by estimating individualized risks of seizure recurrence and eventual outcome.

Activation time - a powerful constraint in noninvasive functional source imaging

Theinverseproblemof imagingbioelectromagneticsources is ill-posed. In orderto getsolutionsthatarereliablea priori constraintshave to beapplied. In orderto have any diagnosticvaluethesesolutionsshould allow aninterpretation in termsof theunderlyingelectrophysiology, andshouldbe verifiableby invasivemeasurements. An a priori constraintthathasfounda widespreadapplication in both the fields of EEG and ECG is restrictingthe assumedsourcesto a singlecurrent dipole. In somecases thedipole modelindeedleadsto an’image’ thatcanbeinterpretedin termsof (pathological)electrophysiology. Examplesarethelocalizationof anectopicfocusin electrocar diologyandof anepilepticfocusin electroencephalography. Although in many casesa current dipole can adequately model the observed bioelectromagneticfields, its application is not always valid. This is, e.g., the casewhen the actualsourcesinvolve nearsimultaneousactivationof large areasof cardiacor cortical tissue.In suchcasesthesource is distributed, anda singledipolemodel mayleadto a misinterpretation of theunderlyingelectrophysiology. Sometimesinvasive measure mentsare available that can be usedto verify the resultsof noninvasive modeling, e.g. through endocardialcathetermeasureme ntsor epicardial or epicorticalmeasurementsperformedduring surgery. However, it should berealizedthat in many casesinvasive measurementsareremotemeasuremen tsaswell, andadirectinterpretationwill notalwaysallow to eitherdiscardor accept a computeddipoleposition. A more generalconstraint, allowing both focal and distributedsources,canbe formulatedwhenthe geometry on which thesourcesresideis known. This geometry(theepiandendocardialsurfacesor thecorticalsurface)canbeextractedfrom MRI. The inverseproblem thusformulatedis linear, but ill-posed. In thecaseof thetheinverseEEGproblem, given the poor conductivity of the skull, it is severely ill-posed. For the MEG, the null-spacerelatedto the insensiti vity to radialsourcesposesproblems.Standar d treatment of suchproblems involves applying Tikhonov regularization,wheremathematical constraintson thesolutions areformulated. The leadsto solutionsthat areeithervery smoothandreducedin amplitude, or areunreliable (if not enough regularization is applied). Thesepropertieshamper both the direct electrophysiological interpretationand the validationthrough invasivemeasurements. In order to do betterconstraints basedon electrophysiology can be applied. A model that has beensuccessfullyapplied in the ECG inverse problem is basedon the known amplitude andshapeof the ventricular transmembraneaction potential[1]. Assumingtheseto beknown, for theQRS complex of theECGtheinverseproblemcanbeformulated in termsof theactivation timeontheventricular surface. Solutions basedon activation time have a direct electrophysiological interpretation, and can and have beenvalidated throughinvasivemeasureme nts[2]. In this paper , theapplicationof anadaptationof theactivation time approachto a particular EEG/MEGinverseproblemis presented. Thesettingis thecharacterizationof interictal spikes in the simultaneously measure d EEG/MEGof patientshaving temporal lobe epilepsy. It is assumedthat when suchspikes occur larger areasof the temporal lobe become active, and that single dipole models cannot adequately describethis process.Therefore,distributedsource solutions arecomputed,basedon a descriptionof the cortical geometry. Standardzero order Tikhonov (minimum norm) and activation time solutionsare consider ed. On someof thesepatients, epilepsysurgery is performed,during which epicortical measurementsare taken. Theseare usedto validate thecomputedsolutions,by comparingmeasuredepicorticogramsto thosesimulatedon basisof the EEG/MEGinversesolutions.Finally, anattemptis madeto interpret thesolutionsin termsof electrophysiology. Simulationstudiespresentedearlierhaveshown thefeasibilityof thisapproach[3, 4].