JNNP Online First, published on April 9, 2014 as 10.1136/jnnp-2013-307069 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp-2013-307069 on 9 April 2014. Downloaded from Review Advances in epilepsy surgery Mark Nowell,1,2,3 Anna Miserocchi,1,3 Andrew W McEvoy,1,3 John S Duncan1,2 1Department of Clinical and ABSTRACT deficit zones. The notion of a single discrete area of Experimental Epilepsy, UCL This review summarises exciting recent and forthcoming EZ is attractive in its simplicity, although the high Institute of Neurology, London, UK advances that will impact on the surgical management failure rate of resective surgery runs counter to 2MRI Unit, Epilepsy Society, of epilepsy in the near future. This does not cover the this. Rather, it is possible that in at least some cases Chalfont St Peter, UK current accepted diagnostic methodologies or surgical there are several structures involved in epileptogen- 3 Department of Neurosurgery, treatments that are routinely practiced today. The esis, and a more comprehensive notion of the EZ National Hospital for content of this review was derived from a PubMed needs to be considered. Neurology and Neurosurgery, ‘ ’ London, UK literature search, using the key words Epilepsy Surgery , The characteristic electrophysiological pattern of ‘Neuromodulation’, ‘Neuroablation’, ‘Advances’, the EZ is the presence of high-frequency oscilla- Correspondence to between 2010 and November 2013. tions or ‘rapid discharges’. Bartolomei in 2008 Mark Nowell, introduced the concept of the ‘epileptogenic index Department of Clinical and ’ Experimental Epilepsy, UCL (EI) , a novel quantitative measure that characterises Institute of Neurology, Box 29, INTRODUCTION the epileptogenicity of brain structures recorded National Hospital for Despite advances in imaging and the accumulation with depth electrodes.3 The EI is based on spectral Neurology and Neurosurgery, of neurological and surgical experience, the out- and temporal factors, with statistically high values Queen Square, London, corresponding to structures involved early in the WC1N 3BG, UK; comes for seizure freedom in epilepsy surgery have [email protected] not changed significantly over the last 20 years. ictal process. They found that their measure of EI Currently, 20–40% of patients with epilepsy are effectively distinguished between mesial temporal Received 21 October 2013 considered refractory to medical treatment.1 Less lobe epilepsy (MTLE) and lateral neocortical epi- Revised 29 January 2014 than 50% of these are candidates for focal resective lepsy. Furthermore, in MTLE they found a statistic- Accepted 18 March 2014 surgery, with rates of long-term seizure freedom ally significant correlation between the duration of ranging from 30% to 60% depending on the oper- epilepsy and the number of structures disclosing ation.2 Some argue that this apparent lack of pro- high EI values, suggesting that MTLE is a gradually gress is a reflection of a lowered threshold to offer evolving process which progresses over time. surgery, and that with continued refinement of Visualisation of the EI, in 3D space as part of an techniques, increasingly challenging cases are being integrated multimodal model, is the obvious next taken on. However, there is widespread agreement step. This would provide an alternative to the elec- that there remains great potential to improve non- trical source imaging, but also provide a more pharmacological management, to achieve either detailed, contoured brain map of seizure likelihood, better seizure control or complete seizure freedom. that could be used by the surgeon to stratify the There are three broad directions in which the risk-benefit ratio of cortical resection. Work is et al next major advances may occur. First, there is the already underway to make this a reality. David http://jnnp.bmj.com/ continued refinement of the current methodology. in Grenoble, report the use of statistical parametric An improved hypothesis for the epileptogenic zone mapping to visualise a quantification of the seizure (EZ), based on advanced presurgical evaluation, onset zone.4 This has been applied to case studies including intracranial EEG, and thus, better patient of insular epilepsy, and also group studies on selection for cortical resections. This is probably MTLE. This technique offers much promise in most relevant to sufferers of non-lesional extratem- future research and clinical practice, with obvious poral epilepsy, and is most likely to lead to applications in future guidance of stereoEEG on September 26, 2021 by guest. Protected copyright. improvements in the rates of seizure freedom. (SEEG) implantations, and in the delineation of Next, there are improved surgical methods for epileptogenic networks over time. achieving a precisely targeted cortical or subcortical resection. These can be grouped as neuroablative Advances in imaging techniques, and include disconnection of propaga- The rates of seizure freedom following resective tion pathways, and destruction of epileptogenic surgery in sufferers of non-lesional extratemporal foci. Neuroablation may be applied in the epilepsy remain poor.2 It is clear that better patient treatment of focal and generalised epilepsy. Third, selection is required in these cases, with improved neuromodulation may take a broader role, with the methods for imaging the EZ and guiding the possibility of improving quality of life and being a implantation of intracranial EEG. For some time useful palliation. This is most applicable to patients there has been interest in unmasking previously who are currently not candidates for resective occult structural lesions, using non-routine mag- epilepsy surgery, because their epilepsy arises from netic resonance (MR) sequences and voxel-based To cite: Nowell M, eloquent cortex, is multifocal or generalised. morphometric analyses.5 There are also develop- Miserocchi A, McEvoy AW, mental techniques, such as EEG-fMRI, which et al. J Neurol Neurosurg fi Psychiatry Published Online REFINEMENT OF CURRENT METHODOLOGY require re nement and further evaluation in clinical First: [please include Day The epileptogenic index practice.6 Month Year] doi:10.1136/ The purpose of presurgical evaluation is to define 3D multimodality imaging is the simultaneous jnnp-2013-307069 the EZ, and to define the surrounding functional display of different structural and functional NowellCopyright M, et al. J NeurolArticle Neurosurg author Psychiatry (or2014; their0:1 –employer)7. doi:10.1136/jnnp-2013-307069 2014. Produced by BMJ Publishing Group Ltd under licence. 1 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp-2013-307069 on 9 April 2014. Downloaded from Review datasets, tailored to individual patients. The feasibility of this MR-guided focused ultrasound additional tool in a busy epilepsy surgery practice has been Magnetic resonance-guided focused ultrasound surgery demonstrated,7 and a prospective study is currently underway to (MRgFUS) is an accurate method of delivering high doses of validate the usefulness of this in presurgical evaluation and sur- transcranial ultrasound energy to a discrete intracranial focal gical management. Ultimately, improvements in the surgical out- point.10 The major historical barrier to this method was the comes in this patient group will depend on better imaging, need to create a craniectomy window prior to treatment, to including visualisation of electrical abnormalities in 3D, reliable prevent the ‘defocusing’ effect of the skull. However, recent imaging integration and robust planning and implementation of advances in phased array transducer technology have overcome intracranial EEG (figure 1). this defocusing effect, so that the treatment can be administered in a ‘closed’ method without the need for conventional surgery. The MRgFUS consists of a clinical 3 T MRI, with a transcra- NEUROABLATION nial hemispheric array transducer that has 1024 ultrasound ele- There is an increasing trend in all forms of surgery towards min- ments. The patient’s head is fixed to the system in a stereotactic imally invasive techniques. This is most pertinent for neurosur- frame and the transducer is filled with degassed water to allow gery, which often requires access to deep parts of the brain. ultrasound waves to propagate toward the patient’s head. Accurate navigation to these areas without the need for signifi- Treatment planning is based on MRI, and MR thermometry is cant brain retraction has been solved by the application of used for target verification during the procedure. The treatment stereotaxis. However, there remains the problem of how to can be administered on an outpatient basis, and treatment effect ‘execute’ the surgery once instruments have been safely navi- can be monitored by postoperative MRI (figure 2). gated to their targets. This problem is best framed in the MRgFUS has previously been used to execute a selective context of epilepsy surgery with cortical resection following medial thalamotomy in the treatment of chronic neuropathic SEEG. However, it is also possible to consider patients with hip- pain.11 A US food and drug administration (FDA)-approved pocampal sclerosis (HS), where even ‘selective’ amygdalohippo- phase I trial using MRgFUS thalamotomy in the treatment of campectomies carry the risk of new cognitive deficit. essential tremor has just been completed, and showed clinical There are a number of interesting alternatives to ‘execute’ improvements in 15 patients.12 There are plans for further trials lesioning at the site of the