Introduction
Children with drug refractory focal epilepsy are candidates for neurosurgical intervention if the epileptogenic zone (EZ) can be identified and disrupted through resection, disconnection, ablation, or neuromodulation . In some instances, the EZ may be restricted to a distinct region of the brain, such as with a developmental tumor, but in other cases may involve an entire lobe, multiple lobes, or an entire hemisphere of the brain. The aim of surgery in such cases is to prevent the propagation of seizure activity from the EZ to the remainder of the brain whilst minimizing the neurological and neuropsychological sequelae associated with the intervention . Identification and accurate delineation of the EZ through seizure semiology, electroencephalography (EEG), and imaging modalities (both structural and functional) can be particularly challenging especially when large volumes of brain are involved or when there is rapid propagation between the brain regions . Subsequently, invasive EEG exploration is frequently required to help delineate the seizure onset zone and allow the mapping of critical neurological functions that may be within or adjacent to putative resection margins .
Intuitively, larger resection volumes have the greatest chance of achieving seizure freedom but also carry increased risks of neurological and neuropsychological deficits especially when performed in the language-dominant hemisphere . The age of the patient, developmental status including cognitive and behavioral factors and underlying pathology are all key determinants for the choice of the operative technique . Lobar and multilobar resection and disconnection procedures, as well as combinations thereof, can be safely undertaken when neurological function is already lost . In some instances, an expected neurological dysfunction may be acceptable to the patient or carers, such as mild or transient deficits, or when the function has bilateral cortical representation . Common lobar and multilobar epilepsies in children include temporal (TLE) , insular-opercular (IOE) , frontal (FLE) , occipital (OLE) , parietal (PLE) lobe epilepsies and combinations thereof such as temporal-plus (TPE) , which includes temporo-parietal, temporo-occipital, temporo-insular, temporo-frontal, and temporo-occipito-parietal (TOP) aka posterior quadrant epilepsy . Multilobar disconnection has been favored over resection procedures due to the potential for shorter operative times, reduced blood loss and lower incidences of superficial hemosiderosis, subdural hemorrhages and traction-related complications secondary to brain slump, but concerns surrounding incomplete disconnection may lead to lower seizure remission rates .
In this chapter, we will discuss the operative techniques, seizure remission rates, associated complications and, where evidence exists, the relative merits of resection versus disconnection techniques for lobar and multilobar epilepsy. IOE has already been described in the insular epilepsy chapter. Hemispheric disconnection, corpus callosotomy, and neuromodulation are also discussed in other chapters and will not be further addressed in this chapter.
Lobar epilepsy surgery
Temporal lobar epilepsy
Compared to adults, isolated TLE is less common in pediatric epilepsies which frequently result from structural developmental aberrations that are associated with more severe and frequent seizures. Despite this, TLE still represents the most common single-lobe epilepsy in contemporary pediatric operative series, representing two-thirds of all cases. Focal cortical dysplasia (FCD), perinatal or postnatal ischemia, neocortical gliosis (most commonly following trauma) and glioneuronal tumors are amongst the most common histopathological substrates, with over half of these patients also having concomitant “dual” finding of mesial temporal sclerosis (MTS) .
Surgery for mesial TLE is discussed in more detail in the temporal lobe epilepsy chapter, but it has been proposed that some patients fail to achieve long-term seizure remission due to the presence of temporal-plus (T+E) or pseudo-temporal lobe epilepsy (PTLE) . T+E is where EZ is primarily within the temporal lobe but also extends beyond, whilst in PTLE the EZ is within a neighboring anatomical structure and then rapidly propagates to the temporal lobe along white matter fiber tracts ( Fig. 14.1 ). In both cases, patients invariably demonstrate semiological features attributable to the temporal lobe, usually have a normal MRI or show findings of MTS and have ictal and interictal scalp EEG consistent with TLE. Of those with T+E, the brain regions into which the seizure-onset zone extended into included the anterior insula (35%), temporo-parieto-occipital junction (25%), fronto-parietal operculum (20%), orbitofrontal cortex (10%), and cingulum (10%) .
Common regions associated in temporal-plus and pseudo-temporal epilepsy along with white matter fiber tracts implicated in rapid propagation to the temporal lobe. The uncinate fasciculus connects the orbitofrontal cortices (blue) to the anterior temporal lobe and mesial temporal structures including the amygdala (cyan) and hippocampus (yellow). This has a role in social and emotional processing. The cingulate cortex (red) lies immediately above the corpus callosum. The cingulum bundle comprises association fibers that lie deep in the cingulate cortex and is continuous with the parahippocampal gyrus posteriorly providing connectivity between the frontal, parietal, and temporal lobes. The cingulum bundle is part of the Papez circuit and hence has a critical role in emotional processing. The inferior longitudinal fasciculus lies within the white matter of the basal temporal lobe below the temporal horn of the lateral ventricle and connects the temporal pole to the occipital lobe (lingual gyrus shown in green). It has a role in object recognition and discrimination as well as providing parallel connectivity as part of the ventral (semantic) language stream. The arcuate fasciculus and third part of the superior longitudinal fasciculus (SLFIII) connect the dorsolateral frontal lobe to the posterior aspects of the superior and middle frontal gyri. This connection is continuous via the arcuate fasciculus and discontinuous via the inferior parietal lobule (Geschwind’s area) in the case of the SLFIII. The inferior parietal lobule comprises the supramarginal gyrus (pink) and the angular gyrus (brown). These fiber tracts subserve functional roles in the phonemic aspects of language.
Surgical techniques for temporal lobar epilepsy
Surgical treatment options for TLE include a standard anteromesial temporal lobe resection (ATLR) , selective amygdalohippocampectomy (SAH), laser interstitial thermal therapy, and radiofrequency thermocoagulation . The lesser-known procedure of temporal disconnection has also been described as an alternative to ATLR . These results have also not been verified by other centers but may provide a valid alternative to temporal lobectomy. One limitation of temporal disconnection compared to ATLR, however, is that the retained tissue cannot be sent for histological analysis ( Fig. 14.2 ).
Temporal disconnection
Temporal disconnection surgery involves a lateral neocortical disconnection through the posterior temporal white matter extending from the superior temporal sulcus to the temporal horn of the lateral ventricle . The lateral neocortical entry through the superior temporal sulcus prevents damage to the optic radiation whilst ensuring disconnection of the inferior longitudinal fasciculus and U-fibers which connect the temporal pole and basal temporal regions to the occipital lobe . Within the atrium of the lateral ventricle, the hippocampal tail along with the fimbria/fornix can be identified and transected allowing visualization of the quadrigeminal cistern. Anteriorly the lateral approach to the temporal horn of the lateral ventricle through the superior temporal sulcus allows for the temporal stem to be disconnected by connecting the apex of the inferior limiting sulcus of the insula to the temporal horn of the lateral. The disconnection is then carried anteriorly through the superior aspect of the amygdala to the limen insulae allowing separation from the basal ganglia above and fronto-temporal connections through the uncinate fasciculus whilst sparing the projections of the inferior fronto-occipital fasciculus.
Anterior temporal lobe resection
In the majority of pediatric series, an ATLR is carried out in accordance with the technique described by Spencer et al. . In the language-dominant hemisphere, the lateral neocortical resection is restricted to 3.5 cm from the temporal pole, whilst in the nondominant hemisphere this may be extended to 5 cm. In cases of lobar temporal epilepsy, the lateral neocortical resection may be extended more posteriorly, but this is likely to result in a contralateral superior temporal hemianopsia and phonemic language deficits in the language-dominant hemisphere. Numerous STL approaches have been described, namely, the transsylvian , transcortical, subtemporal, and supracerebellar transtentorial. Each of these approaches was designed to reduce the volume of temporal neocortical resection required when resecting the mesial temporal structures to minimize postoperative neuropsychological sequelae. Given that isolated MTS is less common in children with epilepsy it follows that ATLR is associated with an improved seizure remission rate compared to STL . Neuropsychological assessments have revealed no change in the overall IQ values between the approaches in children, although older age at the time of surgery, irrespective of the technique, has been identified as a risk factor for cognitive decline overall .
Surgical outcomes for temporal lobar epilepsy
To date, only a single case series has reported on temporal disconnection surgery as an alternative to temporal lobe resection. This series reported on 47 patients (27 right) with nonlesional temporal lobe epilepsy of whom one-third underwent presurgical SEEG evaluation to confirm the seizure-onset zone. Aided by neuronavigation, temporal disconnection resulted in seizure remission in 85% (40/47), defined as an Engel class 1 outcome at 2 years. The series was comprised predominantly of young adults but the technique may be more applicable to children with temporal lobar epilepsy where the pathological substrate is more likely to be a malformation of cortical development and hence have a more diffuse EZ. The most common complication reported was a contralateral homonymous superior temporal quadrantanopia in 50% (23/47). The severity of the visual field defects and implications for driving based on binocular Estermann perimetry, however, were not reported. Nevertheless, the rates are higher than would be expected from contemporary series reporting standard ATLR where driving eligibility ranges from 70% to 95%, especially when tractrography of the optic radiation is used to guide the resection intraoperatively . Other reported complications include mild hemiparesis in 2% (1/47), mild facial paresis in 2% (1/47), basal ganglia ischemia in 4% (2/47), entrapment of the temporal horn in 4% (2/47), and superficial wound infection in 2% (1/47). As expected from patients undergoing dominant temporal lobe surgery, the disconnection procedure resulted in a reduction in verbal memory by 13% (6/47) which is in keeping with large series reporting on both ATLR and SAH .
A large multicenter series of 168 patients including both pediatric and adult patients (age range 3–59 years) demonstrated that in those with SEEG-proven T+E the probability of achieving seizure remission following was 14.8% compared to 74.5% with unilateral mesial TLE at 10 year follow-up . Of note, 50% of the T+E cases demonstrated MTS on MRI with a trend toward a posttraumatic or infectious etiology compared to a history of febrile seizures. Based on this, it is vital to differentiate patients with T+E from mesial TLE as those with T+E are more likely to benefit from multilobar resections rather than ATLR alone .
Frontal lobar epilepsy
FLE is the second most common cause of drug-resistance epilepsy in contemporary pediatric series, accounting for 20%–30% of cases . Seizure remission rates are generally less favorable than TLE due to the lack of anatomical boundaries and the underlying histological substrates . In up to half of cases, no epileptogenic lesion is identified on MRI and even in cases where a focal abnormality is identified, authors still recommend SEEG sampling prior to definitive resection as the EZ has been found to be widespread in one-third of cases and aids in differentiation of bifrontal onset from rapid propagation to the contralateral frontal lobe .
Surgical techniques for frontal lobar epilepsy
Surgical options for FLE include lesionectomy, sublobar, or lobar resection and sublobar or lobar disconnection . Histological evaluation of resected tissue reveals that FCD is the most common atiology accounting for 60% of cases . FCD type 2b was significantly more common than FCD type 1b, which is differentiated by the presence of dysmorphic neurons in the former and by more diffuse distribution in the latter . Other pathological findings include nonspecific findings, such as gliosis, as well as encephalitis, tuberous sclerosis, and encephalomalacia secondary to trauma or ischemia
Frontal lobar resection
Frontal lobectomies are commonly extended back to the precentral sulcus to avoid contralateral motor weakness or paralysis. If the EZ extends into the precentral gyrus then this may also be included in the resection if it involves the inferior (face motor) region as this has bilateral cortical innervation and is likely to result in only a transient weakness. Involvement of the superior precentral gyrus or paracentral lobule may require adjunctive neuromodulation techniques such as responsive neural stimulation (RNS) . In addition to the lateral neocortex (superior, middle, and inferior frontal gyri), the anterior cingulate, orbitofrontal gyri, and subcallosal (SC) area are extensively resected to avoid residual pathological tissue . The mesial aspect of the frontal lobe consists of the superior frontal gyrus (SFG), the cingulate cortex (Cing), SC area, and gyrus rectus (GR). The posterior aspect of the mesial surface of the SFG is known as the supplementary motor area (SMA) and demarcated from the paracentral lobule posteriorly by the paracentral sulcus. The white matter of the frontal lobe is made up of association fibers, the largest of which are the SLF (segments 1–3), commissural fibers of the corpus callosum and projection fibers containing fronto-fugal fibers that project to the brainstem and cerebellum. Other important projection fibers are the UF connecting the basal frontal and anterior temporal lobes as well as the IFOF between the frontal and parieto-occipital cortices. The more recently described frontal tract of Aslant projects from the region of the mesial SFG immediately anterior to the SMA to the pars opercularis of the inferior frontal gyrus and has been implicated in verbal fluency in the language-dominant hemisphere .
Frontal lobectomy is commonly performed through a large fronto-temporal craniotomy exposing the interhemispheric sulcus medially, the sylvian fissure inferiorly and the central sulcus posteriorly . The first stage of the resection involves an interhemispheric dissection to identify the callosomarginal artery within the cingulate sulcus and the pericallosal arteries within the callosal sulcus overlying the corpus callosum. In cases where the interhemispheric dissection is difficult to perform, such as in the presence of large bridging veins, the approach should be performed through the substance of the SFG. The lateral neocortex can then be safely removed once these arteries have been protected thereby transecting the SLF1 (deep to the SFG and superior to the cingulate sulcus), SLF 2, and 3 (within the white matter deep to the middle and inferior frontal gyri, respectively) as well as the cingulum bundle deep to the cingulate cortex. In the dominant hemisphere the frontal operculum is spared from the resection to prevent an expressive speech deficit. The next stage of the resection focuses on the basal frontal lobe incorporating the orbitofrontal gyri and GR. It is important to remain in front of the anterior perforated substance to prevent damage to lenticulostriate perforators supplying the ventral basal ganglia structures as well as remaining subpial to avoid damage to the olfactory bulb and nerve. Finally, as the resection progresses the GR is removed through continued subpial aspiration extending superiorly along the mesial surface to incorporate the SC area.
Frontal disconnection
Frontal lobe disconnection procedures are undertaken with a corticotomy anterior to the precentral sulcus extending inferiorly to the pars opercularis of the inferior frontal gyrus ( Fig. 14.3 ). The corticotomy is then deepened inferiorly through the frontal operculum to sylvian fissure and more superiorly to enter the frontal horn of the lateral ventricle. The superior-most aspect of the dissection is continued through the centrum semiovale to posterior aspect of the supplementary motor cortex remaining anterior to the corticospinal tract and paracentral lobule on the medial surface. From within the frontal horn of the lateral ventricle an inside-out corpus callosotomy is performed through the mid-body with identification of the pericallosal artery signifying complete disconnection. The pericallosal arteries are then traced anteriorly to the anterior communicating complex allowing disconnection of the genu and rostrum. A fronto-basal disconnection is then completed by extending the dissection through the SC gyrus, GR and along the orbitofrontal cortices to the sylvian fissure. The olfactory nerve overlies the olfactory sulcus between the GR and medial orbitofrontal gyrus at the base of the brain and is a helpful landmark to ensure complete basal disconnection. The edge of the sphenoid wing can be used to demarcate the posterior extent of the fronto-basal disconnection avoiding damage to the ventral diencephalic structures. Where language function is being preserved the dorsolateral disconnection line is carried along the inferior frontal sulcus instead of the sylvian fissure sparing the frontal operculum of the dominant hemisphere. In the case of a disconnection procedure, care is taken to preserve arteries and large draining veins to prevent postoperative brain tissue infarction and subsequent swelling ( Table 14.1 ).
| Study | Type | Demographics | Structures resected or disconnected as % of series | Seizure remission rates (% achieving Engel 1) | Neuropsychological outcomes | Complications | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Age | Laterality | Neurological deficits | Hydrocephalus | Infection | Morbidity | Mortality | |||||
| Retrospective—single center | 0.5–16.6 years | Resection laterality not specified | Frontal disconnection or resection 33% (15/46) |
| Not reported | Transient hemiplegia 2% (1/46) | 0% | Chest infection 2% (1/46) | Overall morbidity 9% (4/46) including CSF leak 2% (1/46) and symptomatic anemia requiring transfusion 2% (1/46) | 0% | |
| Retrospective—single center | 10.65 (range 0.2–18) years | Right 83% (5/6) |
| Overall Seizure-free: 66% (4/6) at 1 year | Cognitive decline 17% (1/6) | New neurological deficit 33% (2/6) | 0% | 0% | Overall morbidity 50% (3/6) | 0% | |
| Retrospective—single center | Mean 10 (+/- 4.9 SD) years | Resection laterality not specified |
| Overall Seizure-free: Frontal lobe resection 63% with median 8.1 year follow-up | Not reported | Not reported | |||||
| Retrospective—single center | 23.4 (+/- 12.3 SD)` years | Right 62% (26/42) |
|
| Seizure freedom resulted in improved full-scale IQ and verbal IQ but no difference in performance IQ or working memory scores. | Not reported | |||||
Surgical outcomes for frontal lobar epilepsy
There are no comparative studies in the literature on surgical outcomes between frontal lobar disconnection and resection procedures. In contrast to temporal lobe surgery, frontal lobe resections tend to be more extensive and are associated with poorer seizure remission rates. This is most likely due to the lack of anatomical boundaries within the frontal lobe and due to FCD being relatively more common. Reported long-term seizure-free outcomes vary widely between studies ranging from 11% to 63% . Due to the lack of homogeneity between patient selection, surgical procedure, underlying histological diagnosis, and presence of epileptogenic lesions on MRI and presurgical SEEG it is not possible to compare the reported series. A large series comprising 75 patients reported seizure remission rates of 63% at a median follow-up of 8.1 years with over half of these discontinuing antiepileptic drugs . An identifiable lesion was present on MRI in all but one of these patients, indicating that this was a major bias for patient selection in this series. Benign tumors also conveyed improved postsurgical outcomes compared to FCD. Other prognostic indicators included focal epileptic discharges in EEG, lesions distant from eloquent cortex and completeness of lesion resection on postoperative MRI. Similarly, a series in which SEEG exploration was performed in all patients in order to guide tailored resection reported durable seizure remission rates of 56% and 53% at 2 and 5 years, respectively . SEEG allowed more conservative resection volumes with almost half of patients receiving only a prefrontal resection and very few underdoing a complete (motor-sparing) frontal lobectomy. Those undergoing prefrontal resection achieved seizure remission of 53% (11/19) compared to 12% (2/5) following motor-sparing frontal lobectomy indicating a more extensive EZ in the latter group that may be multilobar or extend into the motor cortex.
Regarding frontal lobe disconnection procedures, a small series of 6 patients undergoing predominantly nondominant frontal lobe disconnections achieved a seizure remission rate of 66% (4/6) in comparison to a slightly larger series of 13 patients where a seizure remission rate of 31% (4/13) was reported at 1-year follow-up . These series are too small to draw inferences between resective frontal lobectomy and frontal lobe disconnections. Few studies reported postoperative surgical outcomes, but frontal lobe disconnection was associated with transient hemiplegia in a small proportion of patients, presumably due to edema extending to the motor cortex or corticospinal tracts . Fewer studies still, report neuropsychological outcomes following extensive frontal lobe resections and disconnection procedures, but the attainment of seizure freedom does appear to convey improvement in full-scale and verbal IQ .
Multilobar epilepsy
Surgical approaches for T+E are tailored based on the preoperative imaging investigations as well as the results of any invasive EEG sampling. These include cases where the temporal lobe is the primary seizure-onset zone but with extension into nearby anatomical structures. TOP aka posterior quadrant, temporo-occipital, temporo-parietal, fronto-temporal, and temporo-insular seizure onset may necessitate resection or disconnection of the entire lobe (lobar) or confined to regions therein (sublobar) as part of a tailored resection.
Surgical techniques for multilobar epilepsy
Temporo-occipito-parietal disconnection surgery
A traditional TOP disconnection begins with an extended ATLR or temporal disconnection, as described above ( Fig. 14.4 ). The atrium of the lateral ventricle is then entered from laterally through a corticotomy in the posterior aspect of the superior temporal gyrus ensuring that the vein of Labbe is preserved. The cortical resection through the posterior superior temporal gyrus is then curved around the posterior aspect of the sylvian fissure through the supramarginal gyrus extending to the postcentral sulcus thereby transecting the SLF II and III fibers and the AF that course from the dorsolateral frontal lobe through the inferior parietal lobule to the posterior aspects of the superior and middle temporal gyri. The cortical resection is then extended superiorly to the vertex utilizing the postcentral sulcus as the anterior border and to the falx with an anterior-to-posterior inclination as the dissection proceeds medially . Similarly, partial TOPs have also been described when the parietal limb of the lateral cortical resection does not reach the postcentral sulcus, therefore, sparing variable amounts of the supramarginal gyrus . Within the atrium of the lateral ventricle, the splenium and isthmus of the corpus callosum are then transected utilizing the falco-tentorial edge as a landmark of complete disconnection. White fiber dissection studies have also shown that a significant proportion of commissural parietal fibers pass through the isthmus of the corpus callosum and hence recommend including this within the disconnection . The tail of the hippocampus and fornix can be identified anteriorly and inferiorly to the splenium and disconnected at their junction if the hippocampus had not been resected previously as part of the ATLR.
Related posts:
Evidence in pediatric epilepsy surgery
Controversies in the timing of pediatric epilepsy surgery: is earlier better?
Epilepsy in eloquent cortex: resection versus responsive neurostimulation
Lesional epilepsy: lesionectomy versus ECoG-guided resection
Temporal lobe epilepsy with preserved function: multiple hippocampal transection versus neuromodulation (deep brain stimulation, responsive neurostimulation)
Corpus callosotomy: anterior two-thirds (two-stage) versus complete (one-stage)
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
