Intraoperative Mapping and Monitoring for Cortical Resection

Introduction

Resective surgery for epilepsy requires an intimate knowledge of the functional and eloquent areas of cortex. Reliance on classic anatomic descriptions of language and motor localization, however, is insufficient, given the variable nature of eloquent cortex. Functional neuroimaging as well as neuronavigation have improved cortical localization on preoperative work-up. Although these modalities provide a rough map of the cortical areas involved in language and motor function, they lack the spatial resolution required for functional preservation. Furthermore, cortical reorganization due to slow-growing lesions or malformations adds another layer of complexity to understanding functional organization. Therefore it is useful to perform intraoperative cortical and subcortical mapping to identify functionally active language and motor areas of cortex. Direct cortical stimulation and mapping is the existing “gold standard” for the maximum resection around a tumor or seizure focus, even in cases where the normal cortical architecture may be disrupted. ^,

Penfield and Boldrey described the initial techniques for direct cortical stimulation for mapping of the central sulcus. Since then, with further technical refinements, direct cortical stimulation has become the gold standard for providing real-time representation of motor and language areas. This chapter describes the proper steps to intraoperative cortical and subcortical mapping of motor, sensory, and language areas.

Anatomic Considerations

A comprehensive understanding of anatomy is critical to successful mapping and surgical resection. The central sulcus and sylvian fissure are consistent landmarks that are visible on the convexity of the brain. The central sulcus separates the precental from the postcentral gyrus. The precentral and postcentral gyri are connected inferiorly by a gyrus known as the inferior pli de passage, rolandic operculum, or subcentral gyrus. The motor and sensory cortex located in the precentral and postcentral gyri, respectively, are both organized with the foot and leg representation in the medial aspect of the cortex, whereas the arm, hand, head, neck, and tongue areas are located laterally ( Fig. 92.1 ). The central sulcus has a sinusoidal contour with three major folds; the middle fold is a posteriorly facing convexity that represents the hand area in the motor and sensory areas—the pli de passage moyen. This sinusoidal shape has been recognized as the omega sign on MR images, with the middle being the hand area. The primary somatosensory cortex is the postcentral gyrus, demarcated as Brodmann areas 3, 1, and 2 from an anterior to posterior direction, and mediates two-point discrimination and proprioception of the corresponding area of the body. Injury to the somatosensory representation of the hands or feet affects dexterity or ambulation, whereas injuries to the representative areas of the face and tongue have fewer clinical consequences. Furthermore, in the nondominant motor cortex, as facial movement receives bilateral innervation, resection of this area typically does not cause significant deficit beyond drooping of the contralateral corner of the mouth, except for dominant hemisphere resection of mouth and laryngeal cortex, which can be devastating to speech.

In terms of neurovascular structures in the rolandic region, important structures are the precentral sulcus artery, central artery, vein of Trolard, and perisylvian cortical arteries and veins. All efforts should be undertaken to ensure that any vessel of significant size is not sacrificed, as loss of these vessels would cause significant loss of function and postoperative morbidity. This is particularly important during resections around the operculum, as the arterial supply to the motor cortex originates from the sylvian fissure, and en passage vessels must be preserved during such resections.

The left hemisphere is language dominant in 90% to 95% of patients, with the remaining 5% to 10% of the population having either bilateral or right lateralizing language function. ^, Factors that contribute to atypical language dominance include left handedness and early life insults to or developmental anomalies of the left hemisphere. With the aid of cortical stimulation, a better understanding of the neurobiology of language has been attained. Cortical stimulation not only is useful in confirmation and better characterization of canonical anterior and posterior language sites, but also allows for the identification of other sites implicated in the function of language. ^,

Indications

Preoperative evaluation of a patient should be used to determine whether the tumor or the target area of epilepsy resection is in a non-eloquent, near-eloquent, or presumably eloquent location. This type of topographic nomenclature is derived based on the surgeon’s knowledge of functional neuroanatomy and has previously been described in the literature. This includes the primary sensorimotor cortex, posterior portion of the superior temporal gyrus and the inferior parietal lobule (Wernicke area), inferior posterior dominant frontal lobe (Broca area), lateral premotor cortex, mid and posterior fusiform gyrus, calcarine cortex (visual cortex), basal ganglia, internal capsule, and thalamus, and the white matter connectivity of each. Involvement of any of these areas of brain in epilepsy or a lesion should be considered eloquent. If the lesion approaches but does not involve eloquent cortex, it is considered near eloquent. If it is in a separate anatomic region altogether, it is considered non-eloquent. This classification system has great implications for the operative strategy chosen. Lesions of non-eloquent brain generally do not require further consideration for more invasive means of cortical mapping, unless specific functions such as music or arithmetic are considered. However, lesions that are in near-eloquent or eloquent cortex may be candidates for intraoperative cortical stimulation. If intraoperative cortical mapping is required for language, the patient must be able to reliably participate in intraoperative monitoring while awake, including positioning and answering pertinent questions without major dysphasia.

Preoperative Functional Imaging

Functional imaging can be a useful noninvasive first step in determining the localization of lesions in eloquent cortex. However, although functional MRI (fMRI) has largely replaced Wada testing for language lateralization, there are major limitations to relying on fMRI alone for localization, with a wide range of sensitivity and specificity in published studies. More recently, new functional neuroimaging methods—magnetoencephalography (MEG) and transcranial magnetic stimulation (TMS)—have shown the capacity to localize motor cortex with an accuracy of within 1 cm, as compared with direct cortical stimulation. Preoperative functional neuroimaging has two major advantages. First, chronic lesions may cause migration of functional activity to adjacent cortex from that of classical anatomic localization. Therefore an understanding of where the new “eloquent” cortex has been displaced in relation to the lesion may refine surgical approach. Second, functional neuroimaging may give the surgeon information about the most hazardous regions of the tumor with regard to perioperative morbidity. This allows the surgeon not only to plan for the maximal safe resection, but also to be able to better counsel patients regarding surgical risks, goals, and adjuvant treatment options.

Preoperative Preparations