Introduction
Hemispherectomy is a surgical procedure in which an entire hemisphere is resected to treat medically refractory epilepsy. This procedure has largely been supplanted by hemispherotomies, where the pathological hemisphere is disconnected and left in situ instead of being removed. In a hemispherotomy, either more than one lobe or the entire hemisphere is disconnected from the adjoining brain or the opposite hemisphere, respectively, without compromising the blood supply. This shift in surgical technique over the past two decades aims to minimize blood loss, operative morbidity, and mortality. More recently, minimally invasive strategies, such as endoscopic and radiofrequency ablative disconnection techniques, have been introduced to further reduce blood loss and morbidity. The indications for hemispherotomy are determined by the underlying pathology and its extent, in correlation with the patient’s semiology, video electroencephalogram (EEG), neurological status, age, and imaging. Hemispherotomy is an extensive procedure, performed only after comprehensive evaluation and proper counseling with the patient and their caregivers. Such a surgical strategy should be discussed at an epilepsy surgery multi-disciplinary team meeting before these procedures are considered. The following chapter will focus exclusively on the emerging new minimally invasive techniques for hemispherectomy, which include endoscopic hemispherotomy and robotic thermocoagulative hemispherotomy (ROTCH) , as described by the author.
Historical perspective
The term “hemispherectomy” refers to a surgical procedure in which an entire cerebral hemisphere is removed. Friedrich Goltz conceptualized such extensive surgery in 1888 when he first performed it on dogs, demonstrating their survival. Walter Dandy and L’Hermite were the pioneers to perform this surgery on humans, independently describing hemispherectomy for the first time in 1928 to treat diffuse infiltrative glioblastoma multiforme . This procedure remained a rarity, reported in only a few other cases until McKenzie applied it to a patient with epilepsy. Over a decade later, in 1950, Krynauw reported excellent long-term outcomes for seizure control and behavioral improvement in 12 children, thus firmly establishing hemispheric surgery’s role in treating epilepsy. In the ensuing decade, various authors, including Obrador and Laramendi ; Feld ; French and Johnson , Ferey , Mason and Shapiro achieved excellent results with their surgical modifications. By 1961, White had reviewed 269 cases, affirming the procedure’s safety and efficacy. The average follow-up period was 16 months, with some cases monitored for up to 9 years. However, detailed reports were available for only 150 cases. Complications such as dehydration and hydrocephalus were noted, but no incidences of hemosiderosis were reported.
Laine et al. in 1964, followed up with 20 survivors of hemispherectomy and identified a delayed complication, which they termed “intracranial hypotension,” characterized by severe headache, vomiting, hyperthermia, and mental clouding. They raised concerns regarding the vulnerability of the remaining hemisphere. Subsequently, Finally, McKissock reviewed a series of 45 hemispherectomies conducted between 1949 and 1964 and observed hemorrhagic complications at either reoperation or autopsy. The average time to the development of these complications was 4.5 years. In subsequent years, multiple reports described delayed neurological and cognitive deterioration, confirmed at autopsy as superficial siderosis leading to hydrocephalus and even death. The ancient adage “horror vacui”(Nature abhors a vacuum) attributed to Aristotle (350 BC) seemed applicable to the phenomena observed in hollow cranial cavities.
In the following two decades, numerous modifications to the anatomic hemispherectomy were proposed to reduce the incidence of hemosiderosis. It was recommended to occlude the foramen of Monro with a muscle plug and strip the dura to eliminate the subdural space, thereby creating a significant extradural space. Peacock et al. advocated for shunt placement within the cavity. Ignelzi and Bucy, succeeded by others , proposed alternatives like hemidecortication or hemicorticectomy.
Rasmussen introduced the concept of functional hemispherectomy, involving extensive central resection, temporal lobectomy, callosotomy, and frontal and parieto-occipital disconnection. This approach aimed to disconnect rather than remove the hemispheres, despite significant portions still being excised.
Numerous subsequent modifications emerged. Schramm introduced hemispheric deafferentation. Villemure developed peri-insular hemispherotomy, which was followed by variations from Shimizu and Maehara and the trans-sylvian keyhole modification . Mathern revised the lateral hemispherotomy approach by resecting the perisylvian area and performing disconnection through the now-exposed ventricles . Delalande from France introduced parasagittal vertical hemispherotomy (PSVH), achieving a 77% seizure-free outcome in his initial series, with subsequent series reporting up to a 91% seizure-free rate. The advantages of PSVH have been cited as its nonvascular trajectory, superior disconnection of the insula, and reduced chance of incomplete disconnection ( Table 13.1 ). Thus, the trend in procedural evolution favored greater disconnection with less tissue resection.
Technique | EH | ROTCH |
---|---|---|
Indications | Suitable for all types of hemispheric disconnection, particularly easier for atrophic ventricles. | Applicable to all types of hemispheric disconnection, potentially more advantageous for children as the number of trajectories and burns may be reduced. More challenging for atrophic cases due to potential brain shift affecting trajectory and target alignment. |
Duration of surgery | Shorter duration, especially for atrophic hemispheres (1–2 h). | Shorter duration, especially for atrophic hemispheres (1–2 h). |
Learning curve | Significant; requires the surgeon to have substantial experience with endoscopes. | Planning is crucial. As most of the surgery is repetitive, the learning curve is not as steep. |
Technology required | Endoscope with a reliable holder, navigation preferred. | Robotic guidance is mandatory to reduce surgery time. |
Limitations | The temporal stem may be overlooked, especially for nonatrophic pathologies like hemimegalencephaly. | Potential for “skip areas”, hence meticulous planning is critical. |
Repeat Surgery | May be necessary for temporal stem disconnection; not challenging but requires re-exploration of the same opening. | May be necessary if “skip areas” are detected on repeat MRI. Repeat surgery is simpler. |
Minimally invasive | Requires a 3×4 cm flap craniotomy, significantly less invasive than open craniotomies. | “Bloodless” and “ultra-minimally invasive”; requires only single sutures. |
Complications | Fewer than open surgeries, minimized due to less invasiveness and improved visualization with the endoscope. Risks related to hematoma, fevers, etc., still exist (though hematoma has never been encountered). | Considered a blind procedure, hence there’s a potential risk of hematoma; however, meticulous planning greatly minimizes this risk. Nonatrophic cases may require a brief cover with mannitol and steroids due to brain swelling. Fever risk is very low. |
Outcomes | Excellent | Excellent |
More recently, Chandra et al. developed an endoscopic-assisted hemispherotomy (EH), a minimally invasive procedure using endoscopic and robotic assistance through a small incision (3×2 cm) achieving seizure freedom in up to 80% of cases. This technique was further described by Sood and Wagner . The advancements in surgical safety, neuroimaging, and functional assessment have solidified hemispherotomy as a viable treatment for intractable hemispheric epilepsy.
Endoscopic-assisted techniques have gained popularity in recent years Endoscopic-assisted techniques have gained popularity in recent years. Since then, this technique has been the preferred approach at our institute, with its success attributed to the following:
- 1.
Enhanced endoscopes and camera systems that facilitate deep surgical procedures while preserving critical structures and minimizing invasiveness.
- 2.
Significant reduction in blood loss and prevention of hypothermia are crucial, especially since most patients are children.
- 3.
The procedure is performed in-depth as a disconnection technique, which is easily executed by angling the endoscope, as will be described subsequently.
Moreover, Chandra et al. have recently introduced a novel “bloodless” technique for hemispheric disconnection using robotic-guided radiofrequency ablation.
Pathology
Typical pathologies leading to hemispheric involvement include drug-resistant epilepsy due to neonatal injury, vascular insults (resulting in hemispheric infarct and porencephaly), hemimegalencephaly, hemispheric cortical dysplasia, Sturge–Weber syndrome, and Rasmussen’s encephalopathy . We can categorize the pathologies into those presenting with (1) an acute onset of massive insult followed by a fixed neurological deficit, such as infantile hemiplegia resulting from trauma or vascular insults; (2) conditions involving malformations of cortical development (MCD), which range in severity from focal cortical dysplasias to extensive hemispheric cortical dysplasia, exemplified by hemimegalencephaly . In these patients, neurological manifestations are often delayed or progressive. This group also encompasses those with Sturge–Weber syndrome (encephalotrigeminal angiomatosis) , characterized by the typical facial port-wine stain (nevus flammeus) and pial angiomatosis, leading to intractable epilepsy, impaired cognitive development, and hemiparesis; and (3) a cohort of patients who experience acquired progressive neurological deterioration with medically refractory epilepsy, such as those with Rasmussen’s encephalopathy, first described by Rasmussen et al. in 1958. This condition, of unknown etiology, involves chronic childhood encephalitis and results in progressive neurological decline alongside medically refractory epilepsy. This perplexing disease, which has been linked to both infectious and autoimmune causes, typically begins at 4–5 years of age, most commonly presenting with Epilepsia partialis continua. Although the onset may be delayed, the progression is invariably relentless once it begins.
Patients most suited for hemispheric surgery typically exhibit contralateral hemiparesis with a “helper hand” (i.e., intact proximal tone due to innervation from the ipsilateral motor cortex and poor distal fine finger movements due to denervation from the damaged contralateral cortex) and independent ambulation, as gait is not affected by distal motor group denervation. They may also present with some degree of hemianopsia. However, these are merely general guidelines. Hemispheric surgery may be considered for any patient with medically refractory epilepsy arising from an entire single hemisphere with demonstrated pathology. Surgeons wishing to adopt minimally invasive techniques, such as endoscopic and the more recently introduced radiofrequency ablative methods, should first receive training from experienced practitioners. It is also advisable to initially apply these techniques in atrophic cases, given the lower degree of difficulty.
Role of functional localization
Investigations utilizing functional localization play a limited role for patients undergoing hemispheric surgery. However, they can be used to assess the residual functions of the affected hemisphere, particularly when surgery is considered for older children and adults . Thus, evaluating motor and language functions in the affected hemisphere may be of benefit, especially for adults. Functional MRI (fMRI) is currently the technique of choice in most centers, although the intracarotid amobarbital (WADA) test is also employed in a few centers. Language function is usually transferred to the unaffected hemisphere, and most studies have not shown any worsening of language abilities following surgery, with only a few case reports suggesting otherwise . Similarly, motor function deterioration may occur in one-third of cases, particularly in patients who did not exhibit preoperative distal upper limb weakness, such as a compromised pincer grip.
Electrophysiologic studies
It is essential to perform both interictal and ictal electroencephalograms (EEGs) as an integral part of the diagnostic workup. There should be complete concordance between clinical localization and radiological findings before considering hemispheric surgery. Interictal surface EEG often shows asymmetric tracings with abnormal slowing, low background voltage, and multifocal independent sharp waves and spikes over the affected hemisphere . Characteristic EEG abnormalities may be present; for example, in hemimegalencephaly, the EEG may demonstrate hemihypsarrhythmia .
It is also crucial to determine whether there are abnormalities in the nonaffected hemisphere. Bilateral independent epileptogenic foci are associated with a poor prognosis for seizure control . Generally, favorable outcomes from hemispheric surgery are linked to interictal EEG suppression over the abnormal hemisphere, the absence of contralateral slowing, a lack of generalized discharges, and the absence of bilateral independent spiking . Developmental anomalies, such as hemimegalencephaly, more commonly exhibit bilateral EEG abnormalities (up to 75%) compared to acquired lesions, such as those occurring postinfarct. The distinction between these two types of lesions contributes to the better outcomes observed in the latter. Thus, the presence of bilateral EEG abnormalities alone does not preclude the consideration for hemispherectomy . In general, EEG findings correlate well with structural imaging abnormalities. For hemispherotomy, precise lateralization, rather than intra-hemispheric localization, is most critical. In cases lacking radiographic lateralization, both surface EEG and, when necessary, invasive EEG monitoring play a critical role in both lateralization and localization.
Neuroimaging
MRI plays an important role in the preoperative evaluation of hemispherotomy candidates. The utility of a CT scan is negligible or limited (e.g., for identifying calcification in Sturge–Weber syndrome). A 3 Tesla MRI, if available, is preferred as it provides excellent visualization of subtle structural abnormalities across the cortical surface and the subcortical structures. It also offers a high-resolution assessment of the unaffected hemisphere. MRI is instrumental in delineating the severity of disease in patients with vascular or posttraumatic injuries, leading to atrophy and porencephalic cysts. Cerebral atrophy is characterized by the loss of gray and white matter and an enlarged ventricle, depending on the severity and chronicity of the pathology. In Sturge–Weber syndrome, CT and plain X-rays may show “tram-track” hemispheric calcification, while gadolinium-enhanced MRI demonstrates pial angiomatosis . In hemimegalencephaly, structural imaging reveals a marked enlargement of the affected hemisphere with abnormal thickening of the cortical mantle . High-resolution structural MRI has revolutionized the diagnosis of MCD . A 7 Tesla MRI can provide high-resolution structural sequences such as magnetization-prepared two rapid acquisition gradient echoes (MP2RAGE), fluid and white matter suppression MP2RAGE (FLAWS), and susceptibility-weighted imaging, which are crucial for advancing the diagnosis of MCD in clinical practice. Additionally, there is the future potential for postprocessing of multiparametric ultra-high-resolution and quantitative data to enable automated detection of MCD via machine learning . MRI may be helpful in the early diagnosis of Rasmussen encephalitis , but a definitive diagnosis requires the appropriate clinical picture along with a biopsy demonstrating perivascular lymphocytic cuffing and gliosis . Detailed morphometric studies of MRI can reveal complex changes occurring in the unaffected hemisphere. A recent study showed an increase in cortical volume but reduced subcortical gray matter in the unaffected hemisphere in patients with Rasmussen’s encephalopathy compared to controls, perhaps as compensation for the affected side’s cortical atrophy . Functional imaging is useful for evaluating the patient. Positron emission tomography (15O-PET) can be beneficial in tracking disease progression in conditions such as hemimegalencephaly , Sturge–Weber syndrome and Rasmussen syndrome . It commonly reveals diffuse unilateral interictal hypometabolism in the affected hemisphere.
One study demonstrated increased glutamate concentrations in the affected hemisphere, measured by MR spectroscopy in young children with Sturge–Weber syndrome, correlating with the severity and frequency of seizures . Such findings may indicate severe epilepsy, suggesting the need for early surgical resection . Single-photon emission computed tomography (SPECT) scanning, providing cerebral blood flow information, may show hypoperfusion in the affected hemisphere . More recently, studies have demonstrated the lateralizing value of magnetoencephalography (MEG) for hemispherotomy patients . MEG is also beneficial for detecting epileptogenic areas in porencephalic cysts . In summary, functional studies are most important in cases without lateralizing findings on structural imaging studies. Furthermore, they provide the clinician with sufficient confidence regarding the functional adequacy of the unaffected hemisphere before performing a major procedure like hemispherotomy.
From the perspective of performing endoscopic hemispherotomies, pathologies may be divided into atrophic (postinfarct, late Rasmussen syndrome) and nonatrophic (hemimegalencephaly, early Rasmussen, hemispheric cortical dysplasias) types. For early-stage learners, it is advisable to address atrophic pathologies first.
Indications
The indications for hemispherectomy include damage to one hemisphere accompanied by medically intractable epilepsy and associated neurological deficits. Common etiologies include extensive hemispheric dysplasia, hemimegalencephaly, Sturge–Weber disease, Rasmussen encephalitis, and perinatal infarction. Broadly, one group presents with maximal hemispheric deficits, contralateral hemiplegia, and hemianopsia, usually due to a fixed perinatal insult. Another group presents with intractable epilepsy accompanied by progressive neurological deterioration due to an evolving pathology, such as Rasmussen encephalitis. Another method of classification divides the pathologies into atrophic (porencephalic cysts, postinfarct, “burnt out” Rasmussen’s) and nonatrophic (early Rasmussen’s, MCD, hemimegalencephaly). This classification is also useful for grading the difficulty or ease of surgery.
It is crucial to understand that not all hemispherotomy procedures are of equal complexity. Hemispherotomy for atrophic pathologies is generally straightforward and can be completed quickly. In contrast, surgery for conditions like hemimegalencephaly can be quite challenging, especially in young children (less than 1 year of age), where the anatomy may be distorted, tissues fragile, and hemostasis difficult to achieve.
Hemispherotomy should not be considered if preoperative evaluation fails to confirm that ictal activity is confined to the affected hemisphere. To contemplate hemispherotomy, a combination of medically intractable epilepsy, lateralizing semiology, hemispheric involvement on high-quality MRI, and localization with video-EEG are required. Furthermore, hemispherotomy should not be pursued if less extensive surgery (such as corpus callosotomy or focal or multilobar resection) could be effective.
Seizure semiology
Candidates for hemispherotomy may experience a variety of seizure types and severities, including drop attacks, focal motor seizures, complex partial seizures, epilepsia partialis continua, and generalized seizures. These seizures are often frequent and severe, incapacitating patients in their daily activities, with individuals typically experiencing 10 to 200 seizures per day. Approximately 80% of patients predominantly exhibit a partial motor seizure pattern clinically. Epilepsia partialis continua is characteristic of chronic encephalitis, such as Rasmussen’s, while infantile spasms are associated with various etiologies. A patient must indeed have medically intractable seizures to be considered for hemispherotomy. It is the neurologist’s responsibility to ensure that the patient has received appropriate antiepileptic drugs in the correct dosages and combinations for a sufficient period. In cases of pathologies like MCD, early surgery is advisable due to the high likelihood of antiepileptic drug failure. Additionally, a trial of immunoglobulin therapy, plasmapheresis, and steroids may be considered for Rasmussen’s syndrome.
Neurologic examination
In addition to medically refractory epilepsy, patients often exhibit other symptoms of unilateral hemispheric damage, such as impairment of cortical sensory modalities, hemianopsia, hemiplegia, and spasticity. Many can walk, with arm movement typically more impaired than leg movement. Fine motor control, like the pincer grasp, is often absent, but gross motor control, such as handgrip, may be retained. Hemispherectomy usually does not create new motor deficits in these cases. Occasionally, patients may experience immediate postoperative hypotonia and loss of voluntary movement, which can last from a few days to weeks. However, potential postoperative worsening of weakness or even hemiplegia should be included in the informed consent process. Completion of hemianopsia postoperatively is inevitable and should also be communicated during the consent process. Nevertheless, this should not serve as a contraindication for hemispheric surgery.
Language and motor function evaluation
Preoperative language function depends on whether the dominant hemisphere is affected and the extent of disease progression. The younger the age at which the injury occurs, the more likely it is that language functions will transfer to the healthy hemisphere. Therefore, the potential for postoperative language deficits hinges on the timing of this language function transfer, and an incomplete transfer is a relative contraindication to hemispherectomy. Older children with dominant hemisphere involvement and retained language function should undergo an assessment for lateralization of language, such as a WADA test or fMRI. The “worst-case scenario” involves the late manifestation of progressive disease, like Rasmussen encephalitis, in the dominant hemisphere, which may lead to severe language deficits after a dominant hemispherectomy.
Similarly, an fMRI is very useful for understanding the residual motor function in the affected hemisphere. Often, there is a transfer of motor function to the healthy side. If significant motor function remains on the affected side, informed consent must be obtained for the potential loss of this function.
Electroencephalography and imaging findings
Concordance between EEG abnormalities and structural anomalies observed in preoperative imaging studies is common. Recording EEG abnormalities from the unaffected hemisphere is not unusual, whether they are secondary to or independent of the abnormalities in the diseased hemisphere. Bilateral abnormal EEG findings do not constitute an absolute contraindication for hemispherectomy, as they may represent secondary epileptogenesis that could resolve after surgery . A recent multicentric study involving 29 centers and 1,257 patients demonstrated bilateral independent/synchronous spikes in 27% of cases . However, this finding did not seem to significantly influence the prognosis or outcome.
Etiology, natural history, and timing of surgery
The timing of surgery may be influenced by the following aspects of the etiology and natural history: (1) whether the condition is congenital or acquired, (2) whether it is strictly unilateral or potentially bilateral, and (3) whether it is static or progressive. Congenital pathologies such as large porencephaly resulting from in utero or perinatal insult and Sturge–Weber syndrome generally have a better seizure prognosis than conditions like hemimegalencephaly or hemispheric dysplasia. Acquired unilateral pathologies, such as Rasmussen’s syndrome, have a more favorable prognosis compared to infectious processes, which may involve both hemispheres.
The timing of surgery is determined by considering the severity of epilepsy, the patient’s age, the natural history of the disease, and the adequacy of therapeutic trials with anticonvulsant medications. Performing hemispherotomy before the second or third year of life carries a lower risk of increasing deficits, making it ideal for children who present early for diagnosis and evaluation. In cases of later-onset conditions, such as Rasmussen encephalitis in older children, the timing remains controversial. The complete transfer of language and motor functions to the healthy hemisphere is less likely in older children. However, medically refractory epilepsy can lead to neuropsychological decline, which may necessitate early surgery. Accumulating evidence suggests that seizures, more than the pathological substrate, may significantly impede cognitive and psychosocial development . In certain conditions, such as congenital brain malformations leading to catastrophic infantile epilepsy, very early surgery may be required. Based on our experience, hemispherectomy can be safely performed at the age of 4–6 months. A recent multicentric study showed a 66% Class I outcome in children undergoing hemispherotomy at less than 3 months of age, with over 25% developing hydrocephalus . Another large multicentric study indicated a worse prognosis for patients less than 3 months of age at the onset of seizures, with generalized seizures, nonstroke substrates, and a history of previous resective surgery.
Presurgical work up
The preoperative work-up should include both interictal and ictal video EEG, as well as MRI (1.5/3.0 Tesla as per the epilepsy protocol). More advanced investigations like ictal SPECT (single photon emission computed tomography), fluorodeoxyglucose PET (FDG-PET), and MEG are optional . However, most centers prefer to perform at least one functional investigation such as SPECT or FDG-PET, as it provides information about the metabolic status of the affected lobe. fMRI for language and memory usually cannot be performed in children but may be performed in patients who are cooperative, to assess the function of the affected lobe. It is important to discuss the patient in a comprehensive epilepsy case conference before deciding on surgery. Preoperative counseling and informed consent are critical, as the patient’s attendants/caregivers must fully understand the possible necessity of postoperative ventilation, a prolonged hospital stay, and extensive rehabilitation. From a practical radiological perspective, we classified the pathologies broadly into atrophic and nonatrophic pathologies. The criteria used for the atrophic group included widening of the cortical sulci compared to the normal side, especially around the sylvian sulcus, dilation of the ipsilateral lateral ventricle and temporal horn, and presence of large porencephalic cysts due to postinfarct sequelae. Pathologies like postinfarct porencephaly and some cases of advanced Rasmussen’s disease were included in atrophic pathologies. In contrast, hemimegalencephaly, hemispheric cortical dysplasia, Rasmussen’s disease without cortical atrophy, and Sturge–Weber disease were included under the nonatrophic group. Such a classification was useful for us in the initial stages as the atrophic cases allowed a smoother learning curve than the nonatrophic cases for an endoscopic procedure.
Goals of surgery
The primary purpose of hemispherotomy is to achieve seizure control through complete disconnection of the epileptogenic abnormal hemisphere from the healthy hemisphere. In anatomical hemispherectomies, this was accomplished by complete excision of the affected hemisphere. In functional hemispherectomies, the emphasis is on complete disconnection rather than resection, although some tissue is resected in each of its variations. More recent techniques of hemispheric disconnection focus primarily on complete disconnection with significantly less or no resection . The second primary goal is to improve psychosocial and cognitive development. Adequate seizure control can be expected to lead to better psychosocial and cognitive development and an improved quality of life .
Surgical approaches
Currently, there are five cardinal accepted techniques of hemispherotomies, which include:
- 1.
Modified functional hemispherectomy ;
- 2.
Peri-insular hemispherotomy ;
- 3.
Trans-sylvian hemispherotomy ;
- 4.
PSVH ;
- 5.
Endoscopic hemispherotomy.
Recently, a sixth category has been added, which includes:
- 6.
Robotic Thermocoagulative hemispherotomy (ROTCH).
Here, only 5 and 6 will be described.
Endoscopic hemispherotomy
EH has emerged as a viable option for performing hemispheric disconnection . Following an initial technical description by the author , the procedure has become the surgery of choice at our institute. Following our report on this technique, it is now being performed at other institutes as well . We feel that, in the current scenario, EH forms a logical approach for performing hemispheric disconnection for the following reasons:
- •
Recent developments in endoscope technology, particularly the introduction of HD camera systems, 3D endoscopes, and compatible endoscopic instruments, for example, suction devices, bipolar systems, etc., optimize the visualization and performance of endoscopic procedures.
- •
Most hemispherotomy patients are children. It is essential to reduce the size of the cranial opening, minimize blood loss, and preserve body temperature. Hence, it is vital to develop minimally invasive techniques that achieve these objectives.
- •
Hemispherotomy is a “disconnective procedure.” It does not involve the excision of large masses. Hence, there is no reason why it cannot be transformed into an endoscopic procedure.
Principles of endoscopic-assisted hemispherotomy
A cadaveric study in 2010 introduced the concept of possible endoscope use in hemispherotomy . It described the introduction of two endoscopes, one through a frontal (for performing frontal disconnection) and the other through a parietal burr hole (for completing posterior disconnection). Although it was a significant study, its principles could not be translated into clinical use due to the challenges of hemostasis and optimal disconnection through burr holes. If the endoscopic procedure is to be developed for hemispherotomy, certain cardinal principles must be followed:
- •
CSF cisternal and ventricular access for hemispheric disconnection: The approach must be through preformed CSF spaces, either through cisterns or ventricles, to achieve adequate visualization through the endoscope.
- •
Use of bipolar and suction: There must be adequate control of bleeding through simultaneous use of suction and bipolar diathermy, requiring minimal space and design modifications for these instruments.
- •
Expanding cone principle of minimal access neurosurgery: The focus of the approach should follow that of an expanding cone ensuring the access area at the target is much larger than the entry point. For interhemispheric transcallosal hemispherotomy, the point of entry should have access from genu to splenium ( Fig. 13.1 ).
Figure 13.1
The figure illustrates the advantage of endoscopic access through an anterior cranial opening. Lines “a” and “b” are parallel lines drawn along the genu and splenium, respectively. The point of entry of the endoscope € is situated between these two lines to allow optimal surgical visualization. The cone of visualization forms a triangle joining the anterior and posterior limits of surgical exposure (genu and splenium, respectively) with the entry point in the cranium. The base of this triangle (3) now represents the entire extent of the surgical area. As shown here, an anterior approach reduces the size of the surgical area and avoids bridging veins. The distance from the splenium in this approach, compared to that in Fig. 13.2 , is overcome with the use of an endoscope. Neuronavigation is used to decide the final cranial opening site to avoid any bridging veins.
- •
Anterior location of cranial opening: The endoscope should be placed much more anteriorly than described in Delalande’s parasagittal technique . As shown in Fig. 13.1 , the entry point of the scope should be equidistant from the anterior (genu of the corpus callosum, CC) and the posterior limits of surgical exposure (splenium), decreasing the chance of encountering significant bridging veins. Fig. 13.2 shows the approach using the craniotomy suggested typically for parasagittal vertical hemispherotomy. It may be seen that the anterior and posterior surgical limits (a & b respectively) are widely separated. This would make the endoscopic approach through this corridor more difficult.
Figure 13.2
This figure summarizes the main steps of endoscopic hemispherotomy, consisting of corpus callosotomy, anterior disconnection, middle disconnection, and posterior disconnection.
- •
Brain laxity following CSF cisternal/ventricular drainage: The brain can become significantly lax once CSF spaces are drained, creating significant operating space as emphasized by surgical experts like Yasargil . Following CSF drainage, sufficient space is available to perform corpus callosotomy followed by hemispheric disconnection. Such equations surprisingly did not change much even in complex pathologies like hemimegalencephaly. Performing hemispherectomy through the ventricular corridor helped us preserve the anatomical perspective.
- •
Use of rigid endoscope holder: The authors wish to emphasize that this aspect of surgical nuance is crucial and should not be compromised. Unlike endoscopic pituitary surgery, where free-hand endoscopic assistance may be preferred, endoscopic hemispherotomy is a complex procedure that requires the surgeon to navigate through critical structures. Therefore, it is essential to stabilize the endoscope with an optimal holding device. The author prefers to use a robotic device (ROSA Medtech, Montpellier, France) which provides haptic feedback; that is, during movements, the weight of the endoscope is supported by the robotic device, thus reducing surgeon fatigue. This is akin to a modern operating microscope, where releasing the magnetic locks does not result in the microscope’s weight falling onto the surgeon’s hands. In the absence of a robotic system, other high-quality holders are also recommendable. The authors suggest that the holders be of optimal quality to allow for the best surgical performance (suggested holders include the pneumatic endoscope holder, for example, the Point Setter holding system from Karl Storz, GmbH & Co. KG, Tuttlingen, Germany).
- •
Use of customized instruments: The author prefers customized extra-long suction tubes with detachable tapering tips. Additionally, having an efficient bipolar system is crucial. The author opts for variable impedance, low-temperature, and nonstick bipolar systems with long forceps (suggested options include Sutter Medizintechnik or molecular resonance-based systems like Vesalius from Medilife technologies). Single-channel endoscopic bipolar systems are also beneficial for achieving hemostasis in depth. The author wants to highlight that variable impedance systems are more of a necessity than a luxury, as the surgeon will encounter tissues with highly varying densities. For instance, the frontal disconnection involves normal brain parenchymal density, whereas the basal nuclei encountered during the middle disconnection have a much higher density, requiring the bipolar cautery to be set at a higher level. Concurrently, controlling bleeding from the lenticulostriate arteries would necessitate lower impedance.
Surgical technique
Under general anesthesia, the patient is positioned supine (see Figs. 13.1–13.4 ) . Planning the site of the cranial opening is crucial, with the optimal trajectory planned using neuronavigation. As previously mentioned, ( Fig. 13.1 ), the trajectory for the approach to endoscopic hemispherotomy is much more anterior than that proposed for transcortical vertical hemispherotomy. An anterior trajectory not only avoids bridging veins but also reduces the size of the footplate at the target ( Fig. 13.1 , line 3 ). A transverse 5 cm incision was made anterior to the coronal suture, and a bone flap measuring 4×3 cm was raised on one side of the midline ( Fig. 13.3 ). Neuronavigation (with gadolinium-enhanced MRI) was used in all cases to mark the exact site of the craniotomy, which was especially useful for avoiding bridging veins. The dura was opened in a C-shaped manner, with the base over the sagittal sinus. The ipsilateral hemisphere was retracted using a brain retractor connected to the Leyla arm. It is important to note that after the dural opening, the brain is usually up to the surface, making the prospect of proceeding with endoscopic surgery seemingly daunting for beginners. However, by deliberately releasing CSF from the interhemispheric fissure, the brain gradually relaxes, resembling the opening of a book. We advise early learners to involve colleagues more adept with endoscopic techniques in the initial stages, especially if the epilepsy surgeon is not familiar with the endoscope. The author prefers to perform the entire procedure (right after dural opening) using a rigid 0-degree, 30 cm long, 10 mm thick endoscope (Karl Storz, GmbH & Co. KG, Tuttlingen, Germany). Sometimes, a 30-degree scope may be required while performing the middle disconnection. The author prefers to stabilize the endoscope with a robotic device (ROSA Medtech, Montpellier, France), which also doubles as a neuronavigation device. The robotic device provides haptic feedback, preventing any sudden unwanted movements and reducing surgeon fatigue while maneuvering the scope. However, if a robotic device is not available, the surgeon should use a high-quality rigid holder as previously mentioned. The first step in surgery is to use the endoscope to inspect under the dura to ensure no bridging veins are present. This is done as soon as the brain is gently retracted by a few millimeters to one side. If any veins are encountered, they are coagulated and cut if they are anterior to the cranial opening. For veins posterior to the cranial opening, they are carefully mobilized and preserved. The anterior placement of the cranial opening, as compared to the traditional PSVH approach, usually prevents encountering any bridging veins. Additionally, the use of preoperative neuronavigation allows the surgeon to plan the cranial opening to avoid bridging veins. Furthermore, retraction in these surgeries is typically no more than 3 cm, with more retraction at depth, causing minimal stretch on the veins and adhering to the principle of minimally invasive surgery, that is, “expanding cone with base inferiorly.” After retracting the hemisphere to one side, the interhemispheric cistern is opened, and the CC along with the anterior cerebral arteries (ACAs) are carefully exposed. The surgery is then carried out in three basic steps ( Figs. 13.2–13.4 ) : (1) complete corpus callosotomy, (2) anterior and middle disconnection, and (3) posterior disconnection.
