Address correspondence to Chad Carlson, Room 1228, HCC12, 530 First Avenue, New York, NY 10016, U.S.A. E-mail: email@example.com
Although many patients with medically refractory focal epilepsy are candidates for resective surgery, patients with multifocal epilepsy and symptomatic generalized epilepsy remain difficult to treat medically and surgically. Corpus callosotomy has been utilized since 1940 for the treatment of seizures, with reports of efficacy in multiple seizure types. Previous studies have demonstrated subsequent lateralization of bilateral/bisynchronous epileptiform activity following callosotomy. To investigate the efficacy of bilateral intracranial electroencephalographic studies immediately following corpus callosotomy, we retrospectively identified 26 patients who underwent corpus callosotomy at our center, 18 of whom had intracranial monitoring following corpus callosotomy. Five of the 18 had focal resections following intracranial electroencephalography (EEG). No patients were seizure free following callosotomy or resection. No differences in postoperative outcomes were seen between patients with intracranial EEG versus those without.
Many patients with treatment-resistant epilepsy (TRE), including symptomatic generalized epilepsy (SGE) and focal epilepsy (FE), are not candidates for resective surgery due to poorly localized/lateralized foci. In these patients, corpus callosotomy (CC) can reduce seizure frequency and intensity in a wide variety of TRE syndromes. Corpus callosotomy was introduced to treat refractory epilepsy in 1940 by Van Wagenen and Herren (VanWagenen & Herren, 1940). The surgical transection of the corpus callosum is often done in stages: transection of the anterior two-thirds followed by a second operation to complete the disconnection. A significant reduction in seizure frequency and morbidity, particularly alleviation of drop attacks, results from both procedures (Asadi-Pooya et al., 2008). Although CC is primarily utilized to treat generalized seizure types, callosal section can unmask FEs with rapid secondary bisynchrony (Fiol et al., 1993; Matsuzaka et al., 1993; Spencer et al., 1993; Matsuzaka et al., 1999; Clarke et al., 2007). Matsuzaka reported 17 cases in which generalized, synchronous slow spike–wave discharges were transformed after CC to unilateral or bilateral independent discharges (Matsuzaka et al., 1999).
At our center, for select patients with TRE with features of both SGE and FE (ictally or interictally), CC was sometimes followed by implantation of bilateral subdural electrodes. This was done to identify a localizable ictal-onset zone unmasked by disconnecting pathways underlying secondary bisynchrony, which could lead to therapeutic focal resection(s). This study explores the efficacy of this approach.
A complete review of the presurgical work-ups, intracranial electroencephalography (ICEEG) findings, surgical interventions, postoperative scalp EEG recordings (when available), and patient outcomes was done to examine the efficacy of this technique in lateralizing ictal foci in patients with poorly lateralized TRE. Twenty-six patients, including 18 with subsequent ICEEG monitoring, were identified following review of all surgeries performed by a single surgeon (W.D.) from 1996 to 2008. A heterogenous approach to the presurgical evaluation and surgical treatment of these patients was noted; three patients had two separate surgical admissions with bilateral ICEEG prior to and following CC. Although the general surgical approach remained consistent, variations in electrode placement based upon preoperative findings were seen. A right hemispheric approach via a vertex craniotomy for transcallosal resection was utilized in all patients. For patients in whom ICEEG was performed, subdural strip and, in some cases, depth electrodes (Ad-Tech, Racine, WI, U.S.A.) were implanted via a vertex craniotomy and/or bilateral burr holes.
Monitoring of the ICEEG was done utilizing a BMSI 5000/6000 system (Nicolet, Madison, WI, U.S.A.) video-EEG system. All ICEEG data were reviewed by an epileptologist for seizures, spontaneous epileptiform discharges, and regions of focal slowing. Statistical analyses were performed utilizing GraphPad Prism version 5.01 for Windows (GraphPad Software, San Diego, CA, U.S.A.).
Seven of 26 patients were female. At the time of CC, the median age was 15.5 years with a median duration of epilepsy of 14.5 years. For all patients, ictal and interictal scalp video-EEG identified either generalized or bilateral multifocal activity, with the exception of one patient with left hemispheric ictal onsets. Magnetic resonance imaging (MRI) was normal in 11 patients. Postoperative changes (i.e., resection cavities) were seen in three patients and nonspecific findings (e.g., atrophy) were seen in three. Three patients had encephalomalacia secondary to old infarcts. Six patients had one of the following findings: previous anterior two-thirds CC, tuberous sclerosis complex, bilateral cortical dysplasia, bilateral temporal atrophy, left hippocampal atrophy, and a subarachnoid cyst. Positron emission tomography (PET) scans were obtained in eight patients: two were normal, one showed hypometabolism consistent with prior resection, one showed mild hypometabolism symmetrically, and the remaining four showed multiple regions of hypometabolism. Magnetoencephalography (MEG) demonstrated bilateral independent epileptiform discharges in two patients; the third was uninterpretable secondary to vagus nerve stimulator (VNS) artifacts. Single-photon emission computed tomography (SPECT) identified bilateral abnormalities in two patients, and in the third demonstrated increased uptake in the right hemisphere. Of the five patients who underwent focal resections, one had PET imaging (normal); neither MEG nor SPECT were obtained.
Figure 1 summarizes the surgical course and outcome of all 26 patients, stratifying by: (1) whether ICEEG was done following callosotomy; (2) whether a focal resection was performed; and (3) the type of CC performed (i.e., complete or partial). The median duration of follow-up was 6.3 years (range 0.9 to 13.1 years). No patients were seizure free. Of the 14 patients with ICEEG following a two-thirds CC, two had focal resections. Of these two patients, one was Engel class II (with subsequent VNS implantation). The second patient died 3 months postoperatively; this was attributed to complications of status epilepticus. For the remaining 12 patients with ICEEG without resection, six had Engel class III outcome and six had an Engel class IV outcome. VNS implantations were performed in one of the patients with class III outcomes and two of the patients with class IV outcomes. Of the four patients without ICEEG following a two-thirds CC, one was class II, one class III following VNS implantation, one was class IV, and one was lost to follow-up.
Of the eight patients who underwent complete CC, four had ICEEG. Three had focal cortical resections (two class III and one lost to follow-up). The remaining patient without a focal resection had a class IV outcome. The four CC patients without ICEEG had class III (one) and class IV (two) outcomes, with one patient lost to follow-up. Postoperative scalp EEG identified lateralized findings in 1 of the 18 patients.
No statistically significant differences (Kruskal-Wallis test) were seen for duration of epilepsy (p = 0.93), age at onset (p = 0.65), or age at the time of surgery (p = 0.90) with regard to surgical outcome grades. Similarly, no significant differences (t test) were seen for duration of epilepsy (p = 0.34), age at onset (p = 0.40), or age at the time of surgery (p = 0.49) between patients who underwent ICEEG and either had or did not have focal resection.
Although a partial or complete hemispheric disconnection with CC may allow for lateralization of seizure foci with rapid secondarily bisynchronous discharges, in this series only 5 (28%) of 18 patients had findings that resulted in focal cortical resections. The absence of focal findings that were amenable to resection in the remaining 13 patients may be secondary to the underlying epilepsy syndrome (i.e., nonfocal, multifocal, and/or SGE), the onset and duration of epilepsy, or other factors. Similarly, a series of 21 patients who underwent CC (median age 18) reported that no patient developed a clear new focality with median follow-up of 2.5 years (Pressler et al., 1999). This high percentage of nonlateralizing or localizing findings suggests that ICEEG should be considered only in patients with documented lateralization on follow-up scalp EEG; it should be avoided in the immediate postcallosotomy period.
In contrast to our findings, a series of children from 5–18 years of age (mean 10 years) was reported in which 9 of 14 patients underwent focal resections following ICEEG (Zupanc et al., 2008). Six (43%) were Engel class I following resection. In another series, 15 patients with infantile spasms (age 5–27 months) developed lateralized hypsarrhythmic EEG patterns after CC (Ono et al., 2008). Postcallosotomy lateralization was associated with an 80% reduction in seizures. Although the reason(s) for the dramatic difference in outcomes is uncertain, the findings in the younger patient cohorts (Ono et al., 2008; Zupanc et al., 2008) suggest better lateralization potential when CC is pursued earlier in the course of a patient’s epilepsy. Alternatively, patients with early onset seizures may have a superior prognosis for lateralization following CC. Although no significant differences were seen in our series with regard to age at onset, age at surgery, and duration of epilepsy, the mean age of onset in the two Engel class II outcomes was 0.88 years versus 3.5 and 2.5 years for Engel class III–IV. Therefore, epilepsy syndromes with onset before age 1 year may have a different underlying etiology than those severe epilepsies that manifest after infancy.
We found no significant differences in outcomes between partial and complete CC, use of ICEEG versus no ICEEG, or for focal resection performed versus only CC (Fisher’s exact test); however, the small sample size limits the power of this study to identify group differences. Despite modest reductions in overall seizure frequency (8 Engel class III and 12 Engel class IV), many families in our study reported improved quality of life as documented in clinical notes. Four patients that had tonic or atonic seizures (i.e., drop seizures) pre-CC had resolution of these seizures post-CC; however, no formal assessment instrument was used, precluding quantification of this outcome.
The retrospective nature of this study—combined with the variations of the preoperative course, operative procedures, and postoperative treatment regimen (e.g., VNS and medication changes)—limits the generalizability of our findings. Potential differences in the underlying etiology/epilepsy syndrome, which may be represented in part by the age at surgery and age at onset, may account for the differences in our outcomes versus other series. However, our findings do not identify significant lateralizing findings in a majority of patients and, therefore, support delaying ICEEG following CC.
We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
None of the authors has any conflict of interest to disclose.