Surgery for epilepsy

32.1

Introduction

Epilepsy has been recognized as a serious medical condition for thousands of years . Estimates of its prevalence range from 0.5% to 1%; thus 30–80 million people suffer from epilepsy. Despite this, medical therapies fail in up to 30% of patients. On the other hand, surgical therapies in selected patients whose seizures that cannot be controlled with medications have a 50%–75% cure rate . The success rate for surgery depends on the ability to resect the neural structures producing seizures without causing significant functional damage. This is in part determined by patient selection and preoperative evaluation; however, intraoperative evaluations can provide additional information and are the oldest techniques established for this purpose.

This chapter will first review the important aspects of the presurgical evaluation and then the various operative considerations in patients undergoing epilepsy surgery. This will be followed by a detailed description of the history and current practice of intraoperative electrocorticography (ECOG) used to define the area of resection. We will also discuss the effect of anesthesia and electrode selection on corticography.

32.2

Preoperative evaluation

The preoperative evaluation, which is critical to successful epilepsy surgery , is designed to answer the following four questions: (1) Does the patient really have epilepsy?, (2) Is the epilepsy refractory to optimal medical therapy?, (3) What is the anatomy of the epileptogenic zone?, and (4) Is the anatomy and the physiology of the epileptogenic zone such that it can either be resected or disconnected from the rest of the brain without causing unacceptable loss of function?

The first question is generally answered using continuous video and electroencephalography (EEG) recordings to record a number of the patient’s characteristic seizures. This provides an objective description of the seizure and the accompanying EEG changes. It is the requisite for diagnosing the patient’s epilepsy and optimizing medical therapy. If medical therapy fails, the next step toward surgical therapy is defining the epileptogenic zone. The epileptogenic zone is the area from which seizures are likely to arise. Inclusion of a given region of cortex in the epileptogenic zone is a statistical concept based on the probability that it contains networks that could produce a seizure in a clinically relevant time period. Estimating this probability is a clinical decision based on both evidence of anatomic and neurophysiologic abnormalities. Anatomic evidence of abnormal network connectivity is most traditionally provided by high-quality MRI with special attention to the hippocampus, although newer techniques using dynamic changes in MRI signals are promising. Other anatomic techniques such as positron emission computed tomography and single-photon emission computed tomography can provide metabolic and blood flow images that could also suggest abnormal underlying networks. Neurophysiologic techniques such as scalp and intracranial EEG and magnetoencephalography can help define the epileptic zone prior to surgery and intraoperative neurophysiologic techniques discussed in Section 32.4 can be used to refine the preoperative definition of the epileptogenic zone.

The preoperative evaluation also includes testing to evaluate the probability that resecting the epileptogenic zone would lead to a functional deficit. These include routine MRI and functional MRI as well as neuropsychological testing and the Wada test.

32.3

Surgical options

Understanding the different surgical options is critical to high-quality intraoperative neurophysiology. The surgical options for treating epilepsy have expanded over time. The earliest medical writings in the Smith Papyrus ( https://ceb.nlm.nih.gov/proj/ttp/flash/smith/smith.html ) suggested a relation between brain injury and seizures. Hippocrates and the Greek and Roman physicians reinforced that connection ; however, it was Hughlings Jackson in the 1880s who systematically studied the relationship between seizures and structural lesions that set the stage for epilepsy surgery in man. In 1886 Victor Horsley , in collaboration with Jackson and the physiologist David Ferrier, operated on a patient with Jacksonian march epilepsy using intraoperative electrical stimulation as a guide. They removed scar tissue from the brain and the patient’s seizures improved. After this, epilepsy surgery grew quickly. Fedor Krause had operated on over 400 patients by 1931 . The modern era of preoperative evaluation and intraoperative corticography began with Penfield and Jasper who used a combination of electrical stimulation to localize cortical functions and ECOG to localize tissue that was potentially epileptogenic. Jasper suggested that spikes were the only form of electrical activity that characterized epileptogenic cortex and made the seminal observation that the location of epileptic activity rather than detailed waveforms modifies the clinical manifestations of a seizure. Bailey and Gibbs recorded spikes most frequently near the temporal pole and used depth electrodes to acutely determine whether spikes were superficial or deep. They extended the surgical resection to include most of the spikes. Although their epilepsy cure rate was not high, it was higher in patients in whom no spikes were recorded after the resection. The subsequent studies of Green and Scheetz and later Falconer et al. relied more on preoperative recordings and intraoperative stimulation than intraoperative corticography. These authors felt that regions where thiopental did not induce fast activity were a better indicator of epileptic cortex than spikes. They generally did not tailor the resection if most of the spikes were in the area of planned resection.

Over the next 60 years, building on this background, the surgical procedures were gradually refined A recent meta-analysis by Liu et al. reviewed the outcomes of surgical (mostly temporal lobectomy) versus medical therapy, concluding that the relative risk of seizure freedom was 3.72 (95% CI 2.75–5.03, P <.0001) with surgery compared to medical therapy only. This large experience with epilepsy surgery has led to a number of important considerations regarding surgical therapy.

Three limitations in the use of surgical resection are important. First, some regions that appeared to be the primary, or only, onset region for seizures were not resectable because of the significant risk for loss of important neurological function. These are considered to be so-called eloquent areas of cortex and include language, vision, and primary movement areas. The second observation was that even if a primary onset area was definitively identified, the prospect of seizure freedom with resection was still not 100%. In fact, the best results obtained to date in cases of mesial temporal sclerosis or cortical dysplasia approach 90%, but not 100% . Moreover, even if seizure freedom is achieved over the course of the first year, eventual failure with recurrence of seizures over the following 5–10 years occurs in another 10% of patients . The third finding, despite the widespread use for many years of intracranial electrode recordings for up to 2–3 weeks at a time and intraoperative recordings of epileptogenic spikes, is that the epileptogenic zone can include diffuse regions of cortex or several onset areas, so that a single region of resection would never likely be very successful.

It is helpful to revisit the rationale for surgical treatment of epilepsy in light of these three observations. While it might be expected that eliminating all seizures is the only goal of epilepsy surgery, there is a broader way of evaluating the goals of surgical intervention. Surgery is only considered once the patient has not achieved adequate clinical benefit from multiple antiepileptic medications. It is generally the most invasive treatment on the spectrum of potential treatments for epilepsy. Although it is hoped that surgery can eliminate seizures, it is still considered justifiable if the outcome is a significant reduction in seizure severity and frequency. Depending on the overall seizure burden and the quality of life that the patient has, surgery can still be considered a valid therapy when it can achieve less than Engel Class 1 outcomes (free of disabling seizures). As such, several different surgical options have become available ranging from stimulation of the vagus nerve, stimulation of the anterior thalamus, closed-loop recording of the seizure focus and then stimulating within or near that region, and a variety of resection procedures including corticectomy, topectomy, lesionectomy, lobectomy, hippocampectomy and amygdalectomy, corpus callosotomy, and subpial transections. Several of these surgical options will be briefly described next.