32 Hippocampal Transection
Hippocampal transection was first reported in 2006 as a new surgical treatment for temporal lobe epilepsy.1 Using this method, postoperative verbal memory can be preserved even in patients with left mesial temporal lobe epilepsy without hippocampal atrophy. Verbal memory disturbances after left temporal lobectomy can be observed in pediatric patients as well as in adults.2–4 We applied this new surgical technique to pediatric patients and obtained good seizure outcome and preservation of verbal function. The overall surgical technique and surgical outcomes, including adult and pediatric patients, are presented, and a representative pediatric case is illustrated.
Surgical Rationale and Method
Exposure of the Hippocampus
The hippocampal pathway, closely related with verbal memory, originates in the inferior temporal association cortex and projects directly onto CA1 pyramidal neurons, after running through the perirhinal cortex.5 Therefore, access to the inferior horn via the basal temporal lobe area can compromise this neuronal connection. Zola-Morgan et al demonstrated that bilateral temporal stem can be transected without causing memory disturbance in monkeys, using a task known to be sensitive to human amnesia.6
Based on the previously mentioned neurofunctional data, we place a small corticotomy within 4.5 cm from the temporal tip. After aspirating the gray matter of the superior temporal gyrus along the sylvian fissure, the temporal stem is exposed. If the temporal stem is aspirated just beneath the gray matter of the insula, the temporal horn can be easily opened ( Fig. 32.1A ). The anterolateral part of the ventricle is suctioned as widely as possible and the hippocampal head and amygdala are fully exposed.
Electrocorticography over the Hippocampus and Amygdala
Before starting surgical procedures, specially designed electrodes are placed over the hippocampus and amygdala. Electrocorticography (ECoG) is recorded in many points over the hippocampus. For this purpose, two small strip electrodes with four contacts for exploring the hippocampal body and tail and two square electrodes with four contacts for recording in the hippocampal head and amygdala are applied. Therefore, ECoG is recorded at 12 contacts all over the hippocampus and the amygdala ( Figs. 32.1B,C ).
Hippocampal Transection
The rationale of hippocampal transection is based on the theory of multiple subpial transection (MST) developed by Morrell et al.7 Because the pyramidal cell layer of the hippocampus is within 2 mm from the surface,5 we devised a ring transector 2 mm in diameter ( Fig. 32.2A ). Because the alveus covering the pyramidal layer is a very firm, fibrous tissue, the alveus is sharply cut with microscissors. A 2-mm ring transector is inserted through this slit and the pyramidal layer is transected 4 mm apart using the same distance used for MST. At the bilateral corners of the CA4 portion and near the subiculum, the pyramidal layer becomes deeper and a 4-mm ring transector is used for transection. Toward the posterior portion of the hippocampus, the width of the hippocampus becomes narrower and an oval-formed transector with a 4-mm long diameter is applied for transection of the bilateral corners (Figs. 32.2B).
The transection areas of the pyramidal cell layer are determined based on the results of intraoperative ECoG. After areas with epileptic discharges are transected, ECoG is repeated to detect residual epileptic activity. If residual spikes are found, transection is also performed until prominent spike areas completely disappear ( Fig. 32.2C ).
Seizure Outcome
Between January 2001 and February 2008, we performed hippocampal transection in 45 patients, left side in 23 and right side in 22. The patients consisted of 22 men and 23 women. Patient ages ranged from 2 to 42 years, with a mean of 25 years old. In all patients, preoperative MRI did not demonstrate any sign of atrophy or asymmetry of the hippocampus ( Fig. 32.3 ). In six patients, organic lesions were confirmed in front of the hippocampus. In 12 cases, intra-cranial subdural electrodes were inserted, and laterality of the mesial temporal focus was determined. In the remaining patients, laterality was diagnosed based on scalp electro-encephalography (EEG) with sphenoidal lead, single photon emission computed tomography (SPECT), and neuropsy-chometry data.
Within 2 weeks after surgery, postoperative magnetic resonance imaging (MRI) was examined. Compared with preoperative MRIs, a tract along the sylvian fissure aspirating the gray matter of the superior temporal gyrus was confirmed. However, there was no deformity of the hippocampus after hippocampal transection ( Fig. 32.4 ). In 36 patients who were followed for more than 1 year, 28 (78%) patients were categorized in Engel Class I, 4 (11%) patients were categorized in Class II, 3 patients (8%) were categorized in Class III, and 1 patient (3%) was categorized in Class 4.