Image-Guided Epilepsy Surgery
Over the past several years, neurosurgery has witnessed a resurgence of interest in stereotactic and navigational techniques. There are several reasons for this renewed interest. The most important is the recent technical progress in brain imaging, specifically with regard to cross-sectional imaging such as computed tomography (CT) and high-resolution magnetic resonance imaging (MRI).1 These imaging modalities have replaced the use of ventriculography and standardized brain atlases and allow direct, patient-specific anatomic targeting of brain structures. Along with advances in the development of compact, high-speed microprocessors, this has allowed for the commercial development of stereotactic navigation systems. Today, these systems are widely available and commonplace in most neurosurgical practices. The recent trend toward minimally invasive surgical procedures has also done its part to create interest in surgical stereotaxis. Stereotactic navigation allows for minimization of scalp incisions, craniotomies, and brain resections. This has the potential to make neurosurgical procedures more economical with shorter hospital stays and a more rapid return to preoperative lifestyle.1–2
Like other disciplines of neurosurgery, epilepsy surgery has benefited from the development and refinement of stereotactic navigation. Typical uses of intracranial stereotaxy in epilepsy surgery have included minimizing the invasiveness of diagnostic and resective procedures, defining trajectories to deep-seated cerebral structures, defining a navigational plan for the resection of small, subcortical lesions, and confirmation of resective boundaries in lesional resection. This chapter discusses some of these applications and describes a novel use of stereotaxy to assist in the perioperative evaluation of epilepsy surgery candidates.
Lesional Epilepsy Surgery
Stereotactic navigational techniques for lesional epilepsy surgery utilize the same principles already refined for image-guided resection of other radiographic lesions such as tumors and vascular malformations. In these cases, the navigational system can be used to plan the incision and craniotomy, localize the lesion, and confirm the extent of resective boundaries. Once again, navigational systems can be useful in localizing small, subcortical epileptogenic lesions such as cavernous malformations and low-grade tumors that may not produce grossly visible changes on the brain surface. This allows the surgeon to create a direct and minimally disruptive pathway to the lesion and surrounding epileptogenic zone. For more extensive resections, such as those performed in patients with cortical dysplasia, the primary utility is in defining the boundaries of lesional resection, which are often defined on fluid attenuated inversion recovery (FLAIR) and T1-weighted MRI. Such lesions often have a normal gross appearance, and intraoperative electrocorticography (ECoG) can yield variable results. This makes it difficult to assess the optimal extent of resection. Image guidance aids in defining the resection as long as the surgeon understands the relationship between radiographic changes and pathological substrate. The actual epileptogenic zone, which is defined as the area of cortex indispensable for the generation of clinical seizures, may be more or less extensive than those radiographic changes seen on MRI. The ability to define the “lesion” using stereotactic guidance is the first step toward a comprehensive operative plan that also includes metabolic, functional, and electrographic data. If ECoG is employed for research or clinical decision making, the navigation system can be used to demonstrate electrode position for the benefit of electroencephalographers and other ancillary staff in the operating room.3
Selective Amygdalohippocampectomy
Stereotactic techniques have been adapted to guide the selective resection of mesial temporal structures while sparing the temporal neocortex for the surgical treatment of mesial temporal lobe epilepsy. Such selective resection may confer protection to neuropsychological function, especially when surgery involves the dominant hemisphere.4 Using image guidance, such a resection can be accomplished while minimizing neocortical disruption and the size of the skin incision and craniotomy. Without the benefit of image guidance, it would be necessary to employ a craniotomy large enough to expose recognized surgical landmarks such as the Sylvian fissure. Furthermore, it would be difficult to place the cortical incision in an optimal location for the most efficient resection of the mesial temporal structures. After patient positioning and registration, navigation systems allow the surgeon to determine the relationships of various intracranial structures with scalp position in order to optimize location of the skin incision. We utilize the navigation system both to place the skin incision in the same coronal plane as the junction of the amygdala and hippocampus and to guide the transection of the temporal stem white matter to the inferior (temporal) horn. By entering the ventricle at this location, we are able to minimize the size of the cortical opening and the extent of retraction needed to perform the amygdalar, hippocampal, and parahippocampal resections. Typically, a 1.5 to 2 cm cortical incision is large enough to allow resection of these structures. Other approaches to the mesotemporal structures have been described. 5 Navigational tools have been also used to assess the extent of hippocampal resection during selective amygdalohippocampectomy.6
Callosotomy
The use of navigational tools has also facilitated the safe and accurate performance of callosal disconnection. Although many procedures have evolved to replace callosotomy, this technique remains useful in a small subset of patients with severe, medically intractable epilepsy. The goal of surgery is typically disruption of the rostral two-thirds of the corpus callosum while sparing the more caudally located association fibers. Although this is a discrete, easily localized structure, the interhemispheric approach can be complicated by the presence of midline, cortical draining veins and their interference with retraction. In an attempt to limit such retraction, craniotomies and approaches that are located anterior to the coronal suture have been employed.3