The current concept of awake brain surgery (ABS) and intraoperative cortical mapping has evolved from the pioneering work of Penfield and Jasper that resulted in the creation of the sensory-motor homunculus.1–3 Their intent was to identify and differentiate indispensable tissue in patients with epilepsy to avoid resections resulting in permanent and functionally significant deficit.4 This idea was later adopted for surgery of tumors, with the aim being to increase the extent of resection (EOR) in eloquent brain areas while preserving neurological function.5
Nowadays, ABS has become a widely adopted and well-established method, with the majority of ABSs being performed for the resection of tumors and the operations being done by neurosurgeons mostly dedicated to tumor surgery. The value of intraoperative electrocorticography (ECoG) to detect afterdischarge potentials (ADP) during cortical mapping seems unclear, and the awareness and discussion of the fact that ADP during surgery have the possibility of causing misguidance by a misinterpretation of the patient’s behavior seems underrepresented.6 There are two reasons for the use of ECoG during the cortical mapping process: 1) to prevent secondary generalization of focal activity, which can lead to an abandonment of ABS; and 2) to detect ADP that may cause a symptomatology that misguides the neurosurgeons’ interpretation of the exact location of critical language areas. Those centers that do not use ECoG for these purposes routinely either do not have the necessary expertise to interpret the results intraoperatively or share the opinion that the necessary setup and time needed outweigh the potential benefit.
In children, the value of using ECoG is even less defined because the experience with ABS in the pediatric population is scarce. In children suffering from seizures, no report on the value of ECoG during ABS has been published so far. Therefore, we analyzed our experience with pediatric ABS with special attention to seizure control and neuropsychological outcome 1 year after the surgery, as well as the value of ECoG during language mapping.
Methods
A retrospective chart review was conducted for all pediatric patients who underwent ABS and language mapping between 2008 and 2019 at a tertiary referral hospital. Demographic and clinical information was obtained from an electronic database including age at ABS, sex, preoperative symptoms, indication for ABS, preoperative structural and functional imaging, results of histopathological workup, surgical reporting including intraoperative events, postoperative complications, as well as seizure and developmental outcomes.
Inclusion criteria were as follows: 1) age at surgery ≤ 18 years; 2) a follow-up of at least 1 year; and 3) complete pre- and postoperative neurological examination.
Presurgical Evaluation
Presurgical workup comprised neurological examination and history, neuropsychological assessment, and EEG. Patients with medically refractory seizures underwent prolonged video-EEG monitoring. In the majority of patients and depending on the etiology, 11C-methionine (MET) and/or 2-[18F]fluoro-2-deoxy-D-glucose (FDG) PET was performed.
Preoperative and postoperative MRI was performed using a 1.5- or 3-T scanner. The structural diagnostic imaging protocol comprised at least a T1-weighted sequence obtained before and after administration of a Gd-based contrast agent, a FLAIR sequence, and a diffusion-weighted sequence in all cases. Based on these sequences, the EOR was classified into gross-total resection (GTR; i.e., EOR ≥ 100%) and subtotal resection (STR; i.e., EOR < 100%).
Patients underwent blood oxygen level–dependent functional MRI (fMRI) prior to surgery to localize and lateralize the functional cortical language areas by using a block design.7 The paradigms used for preoperative mapping included an object-naming paradigm, which was visually presented in all subjects. Curved reformatting to the superimposed structural and fMR images was performed, depicting the relationship between the lesions and the language areas.
Neuropsychological Assessment
A prerequisite for ABS is the patient’s ability to cooperate intraoperatively. Therefore, all children were assessed by a pediatric neuropsychologist and parents were provided with the Child Behavior Checklist (CBCL) to assess behavioral, emotional, and somatic complaints as well as the social skills of their children. If a child was considered a potential candidate, the neuropsychologist prepared him or her in several sessions for the procedure: 1) the procedure was explained to the patient in a child-friendly environment and with the help of suitable illustrations; 2) children visited the operating room and became familiar with the environment; 3) suitable picture boards used during cortical mapping were practiced together; 4) plush toys were selected to be taken into the operating room to increase children’s comfort; and 5) interests of the children were identified to maintain the flow of conversation during surgery. During surgery the patient’s compliance was classified into the following categories: 1) no compliance, necessitating termination of the awake phase; 2) a low level of compliance, allowing communication but preventing standardized cortical mapping; and 3) a high level of compliance.
Neuropsychological testing was performed prior to, 3 months after, and 1 year after ABS to evaluate the level of linguistic development. Therefore, general intellectual ability (GIA) and verbal comprehension (VC) from the Wechsler Intelligence Scale for Children (WISC), as well as verbal learning performance (VLP), immediate free recall (IFR), and delayed free recall (DFR) from the verbal learning and memory test (VLMT), were used.8–10 Results are presented as age-adjusted percentile ranks with mean scores ranging from 16 to 84. Transient (at 3-month control visit) and permanent (at 12-month control visit) neuropsychological deficits were defined as a deterioration from presurgical scores in GIA, VC, VLP, IFR, and DFR.
Technique of ABS
The surgical method of the asleep-awake-asleep technique was similar to that used in adult patients. All operations in this series were performed by one neurosurgeon (T.C.). Anesthesia was maintained by continuous intravenous administration of propofol adapted to the weight of the patient and adjusted to the hemodynamic situation. Generally, a dosage of 0–75 µg/kg/min was used. Analgesia was achieved by continuous administration of remifentanil with a dosage of 0.05–0.10 µg/kg/min. Ventilation was provided through an appropriately sized laryngeal mask. All patients received central intravenous and arterial lines. Patients were positioned in a semilateral position with their head slightly turned to the contralateral side, their ipsilateral shoulder elevated, and their back supported to guarantee patient comfort as well as a free airway and to avoid venous congestion. After placing the skull clamp, the pin and incision sites and relevant skin nerves were infiltrated with a local anesthetic (0.25% bupivacaine). After sterile draping, special attention was given to guarantee a clear field of view and uncovered ears so that unrestricted interaction with the child was possible (Fig. 1A). All surgical procedures were supported by intraoperative neuronavigation (StealthStation S7; Medtronic) to identify lesion margins and potential cortical language areas based on fMRI. Propofol infusion was discontinued during the craniotomy and the laryngeal mask was removed following the opening of the dura mater.
Case 2. Important steps during ABS. Comfortable patient positioning with a clear view to enable interaction with the child (A). Subdural SEs (B) are used to recognize at an early point the seizures and ADP (D) that may be triggered by cortical stimulation (C). Mapping of the stimulated cortex is documented by placing number tags that are logged into the neuronavigation system (E). Language mapping could localize language function in stimulation point 7 (panel E, yellow circle), which was discordant with fMRI (panel E, red spots on fMRI). Figure is available in color online only.
Intraoperative Cortical Mapping
Intraoperative cortical stimulation was performed using an Ojemann Cortical Stimulator (OCS-1; Radionics) (Fig. 1C). Stimulation was biphasic with a stimulation duration of 5 seconds, a pulse width of 500 µsec, and a frequency of 50 Hz. Stimulation intensities were started at a low level and slowly increased in steps of 1 mA, up to a maximum of 15 mA. For language mapping, a neuropsychologist guided the children during the process of naming, counting, and movement. At the same time, the surgeon and the neurologist mapped and marked the relevant brain surface point by point, correlating it with neuronavigation (Fig. 1E). Language testing was performed during the most critical parts of the subsequent resection of tumors or epileptic foci. Picture-naming tasks were performed using direct electrical stimulation to map cortical and subcortical language areas and pathways. A variety of language disorders were observed, including speech arrest, dysarthria, anomia, perseveration, paraphasia, slow speech, and troubled speech initiation. Stimulation-induced seizure activity and the occurrence of ADP was monitored using 2–3 subdural strip electrodes (SEs) (Fig. 1B). In case of seizure activity, the stimulation process was halted and a cold rinse with artificial CSF was applied over the exposed brain. If ADP occurred during stimulation (Fig. 1D), intensities were not increased and the safe stimulation threshold at the specific location was determined. The distances between stimulated area and detected ADP were categorized as follows: 1) close, 2) distant, or 3) close and distant. Distant was defined as not amenable to inclusion into resection or beyond the closest adjacent gyrus.
After completion of the resection, patients were again sedated using propofol and received ventilation with the laryngeal mask until the end of surgery. With the goal being to achieve a GTR of the radiologically defined lesion only, spontaneous extralesional ECoG activity was not used to guide resection of adjacent tissue in these patients.
Statistical Analysis
All data were analyzed using GraphPad Prism software (version 9.1.2). Nonparametric statistics were used to compare pre- and postoperative neurocognitive functional data (Mann-Whitney test, 2-tailed). Statistical significance was defined at an alpha level < 0.05.
Ethics
The study protocol was approved by the institutional review board.
Results
Patient Demographics
In the period extending from January 2008 through December 2019, 11 children underwent ABS at our institution (Table 1). This accounts for 8% of all children who underwent surgery for lesional epilepsy during this period. The median age at the time of surgery was 13 years, with a range of 10–18 years and a male/female ratio of 7:4. All children suffered from seizures as the presenting symptom. Seizures were refractory to therapy in 9/11 (82%) cases, with a seizure load between 1 per month and multiple seizures per day. Otherwise, no neurological deficits were found preoperatively.
Clinical characteristics of 11 children undergoing ABS
| Parameter | No. | % |
|---|---|---|
| Age (yrs) | ||
| Mean | 14 | |
| Median | 13 | |
| Range | 10–18 | |
| Sex | ||
| Male | 7 | 64 |
| Female | 4 | 36 |
| Histology | ||
| Ganglioglioma | 3 | |
| FCD type IIb | 2 | |
| Pilocytic astrocytoma | 1 | |
| Micropolygyria | 1 | |
| LGGNT | 1 | |
| Meningioma | 1 | |
| DNET | 1 | |
| mMCD | 1 | |
| Side of lesion | ||
| Lt hemisphere | 9 | 82 |
| Rt hemisphere | 2 | 18 |
| Main location of lesion | ||
| IFG | 5 | 46 |
| MFG | 3 | 27 |
| STG | 2 | 18 |
| SMG | 1 | 9 |
| fMRI language dominance (n = 9) | ||
| Wernicke lt/rt/bilat | 6:1:2 | 67:11:22 |
| Broca lt/rt/bilat | 6:1:2 | 67:11:22 |
| EOR | ||
| STR | 1 | 9 |
| GTR | 10 | 91 |
IFG = inferior frontal gyrus; LGGNT = low-grade glioneuronal tumor; MFG = middle frontal gyrus; SMG = supramarginal gyrus; STG = superior temporal gyrus.
The indication for ABS was the location of the epileptogenic lesion and/or of the tumor in cortical or subcortical areas, presumably involved in language. This included the frontal lobe in 8/11 (73%) patients, the temporal lobe in 2/11 (18%) patients, and the parietal lobe in 1 patient. Most of the patients (9/11; 82%) received surgery on the left hemisphere, whereas 2/11 (18%) patients underwent surgery on the right hemisphere.
A complete data set was available in 8/11 patients. All were right-handed, and 9/11 patients underwent fMRI prior to surgery. fMRI was not performed in 2 patients because 1) it was not available at the time, or 2) it was aborted due to motion artifacts. In 2 patients, ABS was performed for right hemispheric lesions because fMRI showed bilateral language representations (Fig. 2, Table 2). The mean distance between the lesion and the center of gravity of the nearby fMRI language area was 6.9 ± 7.9 mm, with all lesions but one being localized in the same gyrus.
Case 8. Study of an 11-year-old, right-handed boy who had suffered from 10 to 20 seizures per day since the age of 10 years. During seizures he presented with eye deviation to the right, generalized loss of tone, and twitching of the left upper and lower extremities. Despite several anticonvulsive drugs, there was no reduction in seizure frequency. During prolonged video-EEG monitoring, 146 clinical seizures were recorded within 6 days. MRI revealed cortical thickening along the right intraparietal sulcus merging into its connection with the postcentral sulcus (A, red circle). A methionine PET CT scan showed concordant hypermetabolism (B). fMRI suggested an atypical language lateralization with a bilaterally represented Wernicke area (C, cross) and a predominantly right hemispheric Broca area. Subsequently, depth electrodes were implanted for invasive EEG recording (D). The ictal onset zone and the interictal changes were attributed to the lesion. Indication for a right parietal lesionectomy and ABS was given due to the atypical language representation. Despite the patient’s young age, good preparation by the neuropsychologist enabled excellent cooperation of the patient during the entire operation. GTR (F, resection limits marked in green) was achieved under language testing as shown by postoperative MRI (E, yellow circle); no neurological deficit occurred. Histology revealed the presence of an FCD type IIb. Freedom from seizures was achieved 3 weeks after the surgery. The boy has been seizure free with antiepileptic drugs for more than 24 months. Figure is available in color online only.
Overview of patient characteristics and intraoperative findings during ABS
| Case No. | Age (yrs), Sex | Symptom | Side, Location of Lesion | Language Representation on fMRI | Distance: Lesion to fMRI LA (mm) | fMRI Concordant w/ Language Mapping | EOR | Histology | Patient Compliance | Stimulus-Induced Seizures | Inserted Electrode Type (no. of contacts), Location | ADP Affecting Speech | ADP, Distance to Stimulus | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Broca | Wernicke | |||||||||||||
| 1 | 17, M | MRS | Lt, IFG | NA | NA | NA | NA | GTR | Ganglioglioma | High | No | ECoG NA | ECoG NA | ECoG NA |
| 2 | 11, M | MRS | Lt, MFG | Lt | Lt | 0 | No | GTR | Astrocytoma grade II | High | No | SE (4), PCG; SE (6), IFG | Yes | Close |
| 3 | 17, M | MRS | Lt, STG, MTG, ITG | Bilat (rt > lt) | Lt | 10 | Yes | GTR | Micropolygyria | High | No | SE (4), ITG; SE (4), MTG; GE (20), Wernicke’s area | No | Close, distant |
| 4 | 13, F | Seizure | Lt, IFG, MFG | Lt | Lt | 2 | No | GTR | Meningioma | High | No | SE (6), IFG; SE (4), STG | ADP NA | ADP NA |
| 5 | 13, M | MRS | Lt, IFG | Lt | Lt | 0 | No | GTR | DNET | High | No | SE (4), STG; SE (4), PCG; SE (4), IFG | Yes | Close |
| 6 | 15, M | MRS | Lt, STG | Lt | Bilat (rt = lt) | 16 | Yes | STR | Ganglioglioma | High | No | SE (4), ITG; SE (6), STG; GE (20), Wernicke’s area | Yes | Close |
| 7 | 11, F | Seizure | Rt, MFG, IFG | Bilat (rt > lt) | Rt | 5 | NA | GTR | Ganglioglioma | High | No | SE (6), IFG; SE (4), MFG; SE (4), PCG; GE (20), Broca’s area | No | Close, distant |
| 8 | 11, M | MRS | Rt, SMG | Rt | Bilat (rt = lt) | 23 | NA | GTR | FCD type IIb | High | Yes | GE (10) SMG; GE (10) PCG; DE (4) lesion | ADP NA | ADP NA |
| 9 | 10, F | MRS | Lt, IFG | Lt | Lt | 3 | NA | GTR | LGGNT | Low | No | SE (4), STG; SE (4), PCG; SE (4), IFG | ECoG NA | ECoG NA |
| 10 | 18, F | MRS | Lt, IFG | NA | NA | NA | NA | GTR | FCD type IIb | High | No | SE (4), STG; SE (4), PCG; SE (4), IFG; DE (4), lesion | Yes | Close |
| 11 | 18, M | MRS | Lt, MFG, IFG | Lt | Lt | 3 | Yes | GTR | mMCD | High | Yes | SE (6), PCG; SE (6), MFG | Yes | Close |
DE = depth electrode; GE = grid electrode; ITG = inferior temporal gyrus; LA = language area; MRS = medically refractory seizures; MTG = middle temporal gyrus; NA = not available; PCG = precentral gyrus.
All patients were right-handed, and none had postoperative speech deficits.
Histological workup of the resected specimens revealed ganglioglioma as the most common pathology in 3/11 (27%) patients, followed by focal cortical dysplasia (FCD) type IIb in 2/11 (18%) patients. The histological findings of the other patients included a dysembryoplastic neuroepithelial tumor (DNET), a low-grade glioneuronal tumor, an astrocytoma (WHO grade II), a primarily intraparenchymal meningioma, micropolygyria, and a mild malformation of cortical development (mMCD). GTR was achieved in most patients (10/11; 91%). In the remaining patient STR was the goal because the lesion involved the optic radiation, and hemianopia was considered an unacceptable morbidity in this patient. The mean operative duration was 5 hours (range 3.5–7 hours). No anesthetic complications occurred.
Intraoperative Findings
Results of the intraoperative cortical and subcortical stimulation are summarized in Table 2. The intraoperative compliance of the children was high in 10/11 (91%) and low in 1/11 (9%) cases. The patient with low intraoperative compliance (case 9) was tearful and could not adequately name objects. However, by showing film clips and asking related questions, continuous conversation and monitoring of language function was possible during the resection. Preoperatively the patient had been rated as emotionally unstable, with a marginal depressive mood and anxieties during CBCL assessment. In contrast, the patient in case 8 was highly compliant during ABS but also showed emotional abnormalities. Cortical and subcortical stimulation induced disruption of speech in 6/10 (60%; case 9 excluded) patients. Motor impairments occurred in 3/11 (27%) patients during subcortical mapping of the corticospinal tract. Stimulation-induced seizures occurred in 2/10 (20%) patients, with stimulation intensities of 7.4 and 13.1 mA. All seizures were mild, manageable by irrigation with cold saline, and did not prevent further adequate mapping.
VIDEO 1. Importance of ECoG during language mapping in ABS. © Johannes Herta. Published with permission. Click here to view.
Added value of ABS and language mapping
| Case No. | Detected Functional Boundaries | Added Value |
|---|---|---|
| 1 | Broca’s area | ABS allowed safe corticotomy given that no fMRI data were available that would indicate where corticotomy should not be performed. Posterior resection limit defined by language mapping during surgery. |
| 2 | Broca’s area, CST | Change of surgical strategy. A safe transsulcal approach was adjusted after language mapping because fMRI did show discordant results. |
| 3 | Wernicke’s area | Definition of posterior resection limit during anterior medial temporal lobe resection (concordant w/ fMRI). |
| 4 | None | Resection of an eloquent region shown in fMRI was performed safely because of negative mapping results. |
| 5 | None | Safe resection due to negative mapping & discordant fMRI. |
| 6 | Wernicke’s area | Language mapping confirmed language representation in the MTG as shown in preop fMRI. |
| 7 | None | Negative mapping, added confidence. |
| 8 | None | Negative mapping, added confidence. |
| 9 | None | No mapping possible due to low patient compliance. |
| 10 | Broca’s area, CST | ABS allowed safe corticotomy & resection given that no fMRI data were available that would indicate where corticotomy should not be performed. Detection of resection limits by stimulation. |
| 11 | Broca’s area, CST | Language mapping concordant w/ fMRI allowed a safe & tailored resection of a diffuse lesion (mMCD). |
CST = corticospinal tract.
Case 5. Study of a 13-year-old boy with a DNET in the left inferior frontal gyrus. fMRI (A) revealed pronounced activation during verb generation in the left operculum directly adjacent to and slightly overlapping with the anterior lesion. During intraoperative language mapping, Broca’s area could not be localized. Instead, ADP were elicited with low stimulation intensities of 4 mA that caused slowing of speech and could have been mistakenly interpreted as a language disorder in the absence of ECoG. Resection of the lesion could be performed safely, which is shown in panel B. Figure is available in color online only.
No intraoperative complications occurred.
Functional and Seizure Outcome
None of the children experienced a language disorder or any neurological deficit postoperatively. All patients were seizure free (International League Against Epilepsy class 1a) at the latest follow-up.11 The antiseizure medication was completely discontinued in 4/11 (36%) patients and maintained in 7/11 (64%) patients, respectively. The median follow-up time was 4.3 years, with a range from 1 to 12 years.
Neuropsychological Outcome
Complete neuropsychological assessment was available in 9/11 (82%) patients. Figure 4 shows the results for all patients, which are summarized as line graphs but also displayed on an individual level as heat maps. A total of 2/11 (18%) patients showed GIA values below the age norm (cases 4 and 6) preoperatively, which may explain their poor performance after surgery in all language-specific tests. Many patients (6/11; 55%) showed preoperative deficits in VC. This finding was more pronounced among children who did not have German as their mother tongue (cases 1, 3, 4, and 9). However, because they were born and raised in Austria these children still had sufficient knowledge of the German language to ensure satisfactory language mapping. Comparison of preoperative and 1-year postoperative neuropsychological assessment from the WISC (GIA, p = 0.7472, 95% CI −30 to 45; VC, p = 0.7121, 95% CI −24 to 65) and VLMT (VLP, p = 0.3021, 95% CI −10 to 50; IFR, p = 0.1811, 95% CI 0–37; DFR, p = 0.7171, 95% CI −20 to 40) did not show any significant changes following ABS. Notably, 3 months after surgery a trend toward a slight deterioration that was followed by an overall improvement at 1 year was observed (Fig. 4, right panels). Transient and permanent neuropsychological deterioration of GIA, VC, VLP, IFR, and DFR occurred in 4, 2, 4, 3, and 2 and in 3, 3, 2, 1, and 2 of 9 patients, respectively.
Neuropsychological assessment of children (n = 9) who underwent ABS. Comparison of neuropsychological assessment at 3 different time points: preoperatively (circle), 3 months postoperatively (triangle), and 1 year after ABS (square). The change in the mean percentile ranks of all 9 patients is shown in the line charts (left panels). Although no statistically significant change occurred after ABS, a trend toward an immediate slight deterioration in GIA, VLP, and IFR followed by overall improvement was observed. Heat maps (right panels) illustrate the differences in neuropsychological scores (GIA, VC, VLP, IFR, DFR) on an individual basis (cases 1–11) at the given time points. Results are displayed color-coded in age-adjusted percentile ranks, with normal values ranging from 16 (yellow) to 84 (green) and noticeably low values < 15 (red to orange). Figure is available in color online only.
One patient (case 8) underwent resection of a right-sided FCD type IIb in the supramarginal gyrus. Despite unremarkable results on intraoperative language mapping, the postoperative controls showed an increasing deterioration in auditory memory.
Discussion
Our experience with ABS for mapping of language function in children and adolescents confirms the feasibility and value of this technique in selected cases, even at a young age. Mapping of cortical and subcortical language areas was feasible in 10/11 children. Adequate brain mapping influenced the surgical strategy and allowed targeted resections of tumors or epileptogenic lesions, resulting in termination of the seizure burden in all patients.
We were able to demonstrate that the surveillance by ECoG is not only important in the early detection of stimulation-induced seizure activity,12,13 but also a crucial factor for the correct interpretation of cortical language mapping, given that in 56% of cases ADP occurred simultaneously with a speech disorder during cortical stimulation. With our protocol, no de novo language disorders were observed.
In adult patients, intraoperative stimulation-induced seizures occur frequently, with reported numbers ranging from 7.4% to 21.5%.14–16 In our study motor seizures were induced by cortical stimulation in 2 (20%) patients. In both cases, seizures stopped with cold saline irrigation without the need of pharmacological intervention. Although the anesthetic method used in the reported series is very similar, stimulation settings may vary between centers. In a comparable patient cohort, for instance, a pulse width of 100 µsec instead of 500 µsec and a frequency of 60 Hz instead of 50 Hz were applied.17 Whereas we use stimulation intensities of up to 15 mA in combination with ECoG to detect ADP and seizure activity, other centers advocate the use of lower stimulation intensities without the need for electrocorticographic control.18 In our experience, however, seizures and ADP were induced also at lower stimulation intensities, which may be attributed to the underlying pathologies and history of seizures in all of our patients, but also to incomplete myelination in young patients, making direct comparison difficult.19 Nevertheless, the use of ECoG is simple and may avoid major complications such as brain swelling and hemorrhage after intraoperative generalized seizures. Moreover, postictal drowsiness of the patient can hinder further language mapping and may lead to incomplete resections. ECoG reduces the risk of inducing a seizure by aiding in the determination of ADP thresholds and stopping further stimulation. As seizures evolve from ADP, recruitment can be detected early with ECoG, as seen in both patients suffering from stimulus-induced seizures. In those cases, the tonic phase started approximately 15 seconds after ADP were initiated, leaving enough time for the surgeon to secure the patient’s head in the skull clamp and to start irrigation with cold saline. The detection of ADP not only is of great importance to prevent seizures but also plays a pivotal role in the interpretation of language mapping because it enables a differentiated assessment of the focal stimulus (Video 1). In more than half of our patients, ADP triggered a language disorder during language mapping. This finding is even more important given that we could demonstrate in at least 2 patients that ADP also occurred distant from the stimulating site. Therefore, we believe that without the use of ECoG during cortical language mapping, ADP may have the potential to mislead the surgeon by triggering a speech disorder even though the stimulus was distant to language areas.
Furthermore, our results suggest that the decisive factors for successful ABS in children are as follows: 1) an extensive and child-friendly neuropsychological assessment and careful training of the patients as described previously;20 and 2) a professional interdisciplinary collaboration of neurologists, psychologists, neurosurgeons, and anesthetists before, during, and after the procedure.21,22 This is further supported by the successful outcome of the patients studied, despite frequent language difficulties and occasional intellectual impairment. In a review of 50 pediatric ABS procedures, 2 surgeries needed an interruption of the awake phase due to anxiety and uncooperative behavior of the patient.23 In such situations, we encourage the neuropsychologist to continue the conversation with the child in order to continue monitoring of language function, at least to a limited extent. This approach ensured continued language mapping even in a patient with high levels of anxiety.
Given the small sample size, we could not statistically define neuropsychological parameters that may predict adequate cooperation during surgery. Nevertheless, sufficient understanding of language is a prerequisite for performing ABS. Furthermore, given that depression and anxieties do play an important role in the perception of pain and are associated with a higher failure rate of language mapping in adults, the CBCL is an important tool to test eligibility for ABS.20 The final decision whether language mapping is possible must be made by an experienced pediatric neuropsychologist who evaluates the compliance of the child during several simulation sessions.
Age has long been considered a major limiting factor and ABS has been very restrictively used in pediatric epilepsy centers, leading to only a few reports worldwide.12,17,24–26 The use of ABS initially transitioned from adults to being used in children when Berger reported utilization of hemispheric low-grade glioma surgery in children in 1996. However, in that study Berger concluded that ABS is not feasible in children younger than 12 years of age.24 In 2014, Balogun et al. published the Toronto experience from the Hospital for Sick Children.17 In their cohort of 10 children, 2 were 11 years of age at the time of surgery, comparable with our cohort, in which the youngest child was 10.5 years of age. Experiences are scarce in patients younger than 10 years of age, but successful cases with 8- and 9-year-old children have been reported.12,25,26 We also included 2 patients at the age of 18 years (cases 10 and 11), with lesional epilepsy due to an FCD type IIb in one and an mMCD in the other. Both patients were still managed by pediatricians and pediatric neuropsychologists because their general intellectual performance was low and a child-friendly approach was still needed. Ojemann et al. published the data of 26 children between 4 and 16 years of age who underwent intra- or extraoperative language stimulation and compared them with an adult population.27 These authors could show that the younger the child, the lower the frequency of naming errors that occurred in response to cortical stimulation, which perfectly reflects our results. The individual maturation of the CNS as well as the impact of pathology were subject to speculation as potential influencing factors for language variability.27
Although there are methods as good as or even better than ABS and cortical stimulation to monitor motor function even in the very young,28,29 fMRI and magnetoencephalography as well as Wada testing may complement but not substitute for ABS in monitoring language function.12 The value of ABS to avoid language deficits in general can be discussed under various aspects.24 Particularly in children with a high level of neuroplasticity, the use of ABS must not result in incomplete resections of tumors and/or epileptic foci on the basis of misinterpretation of the intraoperative mapping results.30 Valuing the information gathered during ABS and its thoughtful use is a prerequisite for using this method successfully. Functional imaging, in conjunction with neuronavigation, is already a standard procedure for presurgical planning in children.31 Nevertheless, fMRI has shown high accuracy in language lateralization while performing poorly in precise localization of eloquent language, as supported by our results.32
Further research and technical advances are to be expected in the near future. These include measures to correct for brain shift in neuronavigation, more robust and reproducible results from fMRI and diffusion tensor imaging, as well as the implementation of corticocortical evoked potentials.7,33–38 These techniques may become an alternative to ABS in the future once the level of confidence has increased to the critical point needed. Another valuable alternative to ABS is the implantation of subdural or depth electrodes for extraoperative monitoring, stimulation, and mapping. This method has been proposed in selected patients with epilepsy, is especially useful in children younger than 10 years of age, and is used in various centers.12,39 Regarding the detection of eloquent language cortex, better patient compliance as well as the possibility of longer and repeated testing are clear advantages of this method. Disadvantages are the fixed position of electrodes, the lack of reliable anatomical landmarks, and the variability of language localizations, which are even aggravated by brain pathologies. Therefore, the opinion of several authors is that extraoperative mapping has a good negative predictive value in the determination of eloquent anterior and posterior language cortex.40,41 However, in case of inconsistent results, it cannot replace ABS with intraoperative mapping. Furthermore, sufficient data that correlate the findings of extraoperative mapping with functional language outcome are missing.
Extraoperative mapping results in combination with preoperative functional imaging may determine the surgical strategy in the future and may be the next essential step in the development of brain surgery in eloquent areas. To date, however, ABS still represents the gold standard for resections in eloquent language areas.
Conclusions
Although there is a paucity of evidence regarding the benefits of ABS for mapping language function in children, our results suggest that this surgical technique is of great value in pediatric patients of young age. Our data strongly support the use of neuropsychological examination to assess the feasibility of ABS, especially in neurologically intact pediatric patients who have tumors or epileptogenic foci in language areas. Excellent collaboration of an interdisciplinary team, a well-designed child-appropriate preparation paradigm, and the additional use of ECoG are pivotal factors for successful ABS implementation.
Highlights
Largest series of ABSs in children with epilepsy
Long-term assessment of neuropsychological outcome
High intraoperative cooperation in 91% of patients
Importance of subdural SEs during language mapping
Description of how pediatric patients are prepared for ABS
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Conception and design: Herta. Acquisition of data: Herta, Winter, Pataraia, Czech, Porsche, Leiss, Kasprian. Analysis and interpretation of data: Herta, Porsche, Kasprian. Drafting the article: Dorfer, Herta. Critically revising the article: Dorfer, Feucht, Czech, Slavc, Peyrl, Kasprian, Rössler. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Dorfer. Study supervision: Dorfer.
Supplemental Information
Videos
Video 1. https://vimeo.com/670319188.
References
- 1
Penfield W. The cerebral cortex in man. Arch Neurol Psychiatry. 1938;40(3):417–442.
- 2
Penfield W, Roberts L. Speech and Brain-Mechanisms. Princeton University Press;1959.
- 3
Penfield W, Jasper HH. Epilepsy and the Functional Anatomy of the Human Brain. 1st ed. Little;1954.
- 4↑
Surbeck W, Hildebrandt G, Duffau H. The evolution of brain surgery on awake patients. Acta Neurochir (Wien). 2015;157(1):77–84.
- 5↑
Berger MS, Ojemann GA. Intraoperative brain mapping techniques in neuro-oncology. Stereotact Funct Neurosurg. 1992;58(1-4):153–161.
- 6↑
Duffau H. Contribution of cortical and subcortical electrostimulation in brain glioma surgery: methodological and functional considerations. Neurophysiol Clin. 2007;37(6):373–382.
- 7↑
Bartha-Doering L, Kollndorfer K, Kasprian G, et al. Weaker semantic language lateralization associated with better semantic language performance in healthy right-handed children. Brain Behav. 2018;8(11):e01072.
- 8
Lux S, Helmstaedter C, Elger CE. Normierungsstudie zum Verbalen Lern- und Merkfähigkeitstest (VLMT). Diagnostica. 1999;45(4):205–211.
- 9
Wechsler D. Wechsler. Intelligence Scale for Children. 4th ed. Pearson;2003.
- 10
Wechsler D. Wechsler. Intelligence Scale for Children. 5th ed. Pearson;2014.
- 11↑
Wieser HG, Blume WT, Fish D, et al. ILAE Commission Report. Proposal for a new classification of outcome with respect to epileptic seizures following epilepsy surgery. Epilepsia. 2001;42(2):282–286.
- 12↑
Morshed RA, Young JS, Lee AT, Berger MS, Hervey-Jumper SL. Clinical pearls and methods for intraoperative awake language mapping. Neurosurgery. 2020.
- 13↑
Nossek E, Matot I, Shahar T, et al. Failed awake craniotomy: a retrospective analysis in 424 patients undergoing craniotomy for brain tumor. J Neurosurg. 2013;118(2):243–249.
- 14
Nossek E, Matot I, Shahar T, et al. Intraoperative seizures during awake craniotomy: incidence and consequences: analysis of 477 patients. Neurosurgery. 2013;73(1):135–140.
- 15
Eseonu CI, Rincon-Torroella J, ReFaey K, et al. Awake craniotomy vs craniotomy under general anesthesia for perirolandic gliomas: evaluating perioperative complications and extent of resection. Neurosurgery. 2017;81(3):481–489.
- 16
Pereira LC, Oliveira KM, L’Abbate GL, Sugai R, Ferreira JA, da Motta LA. Outcome of fully awake craniotomy for lesions near the eloquent cortex: analysis of a prospective surgical series of 79 supratentorial primary brain tumors with long follow-up. Acta Neurochir (Wien). 2009;151(10):1215–1230.
- 17↑
Balogun JA, Khan OH, Taylor M, et al. Pediatric awake craniotomy and intra-operative stimulation mapping. J Clin Neurosci. 2014;21(11):1891–1894.
- 18↑
Duffau H, Peggy Gatignol ST, Mandonnet E, Capelle L, Taillandier L. Intraoperative subcortical stimulation mapping of language pathways in a consecutive series of 115 patients with Grade II glioma in the left dominant hemisphere. J Neurosurg. 2008;109(3):461–471.
- 19↑
Ritaccio AL, Brunner P, Schalk G. Electrical stimulation mapping of the brain: basic principles and emerging alternatives. J Clin Neurophysiol. 2018;35(2):86–97.
- 20↑
Santini B, Talacchi A, Casagrande F, et al. Eligibility criteria and psychological profiles in patient candidates for awake craniotomy: a pilot study. J Neurosurg Anesthesiol. 2012;24(3):209–216.
- 21↑
Trevisi G, Roujeau T, Duffau H. Awake surgery for hemispheric low-grade gliomas: oncological, functional and methodological differences between pediatric and adult populations. Childs Nerv Syst. 2016;32(10):1861–1874.
- 22↑
Delion M, Terminassian A, Lehousse T, et al. Specificities of awake craniotomy and brain mapping in children for resection of supratentorial tumors in the language area. World Neurosurg. 2015;84(6):1645–1652.
- 23↑
Lohkamp LN, Mottolese C, Szathmari A, et al. Awake brain surgery in children-review of the literature and state-of-the-art. Childs Nerv Syst. 2019;35(11):2071–2077.
- 24↑
Berger M. The impact of technical adjuncts in the surgical management of cerebral hemispheric low-grade gliomas of childhood. J Neurooncol. 1996;28(2-3):129–155.
- 25↑
Klimek M, Verbrugge SJ, Roubos S, van der Most E, Vincent AJ, Klein J. Awake craniotomy for glioblastoma in a 9-year-old child. Anaesthesia. 2004;59(6):607–609.
- 26↑
Riquin E, Martin P, Duverger P, Menei P, Delion M. A case of awake craniotomy surgery in an 8-year-old girl. Childs Nerv Syst. 2017;33(7):1039–1042.
- 27↑
Ojemann SG, Berger MS, Lettich E, Ojemann GA. Localization of language function in children: results of electrical stimulation mapping. J Neurosurg. 2003;98(3):465–470.
- 28↑
Herta J, Yildiz E, Marhofer D, et al. Feasibility of intraoperative motor evoked potential monitoring during tethered cord surgery in infants younger than 12 months. Childs Nerv Syst. 2022;38(2):397–405.
- 29↑
Murcia D, D’Souza S, Abozeid M, Thompson JA, Djoyum TD, Ormond DR. Investigation of asleep versus awake motor mapping in resective brain surgery. World Neurosurg. 2022;157:e129–e136.
- 30↑
Ho A, Khan Y, Fischberg G, Mahato D. Clinical application of brain plasticity in neurosurgery. World Neurosurg. 2021;146:31–39.
- 31↑
Roland JL, Griffin N, Hacker CD, et al. Resting-state functional magnetic resonance imaging for surgical planning in pediatric patients: a preliminary experience. J Neurosurg Pediatr. 2017;20(6):583–590.
- 32↑
Giussani C, Roux FE, Ojemann J, Sganzerla EP, Pirillo D, Papagno C. Is preoperative functional magnetic resonance imaging reliable for language areas mapping in brain tumor surgery? Review of language functional magnetic resonance imaging and direct cortical stimulation correlation studies. Neurosurgery. 2010;66(1):113–120.
- 33
Prada F, Del Bene M, Mattei L, et al. Preoperative magnetic resonance and intraoperative ultrasound fusion imaging for real-time neuronavigation in brain tumor surgery. Ultraschall Med. 2015;36(2):174–186.
- 34
Silva MA, See AP, Essayed WI, Golby AJ, Tie Y. Challenges and techniques for presurgical brain mapping with functional MRI. Neuroimage Clin. 2017;17:794–803.
- 35
Schmid E, Thomschewski A, Taylor A, et al. Diagnostic accuracy of functional magnetic resonance imaging, Wada test, magnetoencephalography, and functional transcranial Doppler sonography for memory and language outcome after epilepsy surgery: a systematic review. Epilepsia. 2018;59(12):2305–2317.
- 36
Matsumoto R, Nair DR, LaPresto E, et al. Functional connectivity in the human language system: a cortico-cortical evoked potential study. Brain. 2004;127(Pt 10):2316–2330.
- 37
Tamura Y, Ogawa H, Kapeller C, et al. Passive language mapping combining real-time oscillation analysis with cortico-cortical evoked potentials for awake craniotomy. J Neurosurg. 2016;125(6):1580–1588.
- 38
Yamao Y, Suzuki K, Kunieda T, et al. Clinical impact of intraoperative CCEP monitoring in evaluating the dorsal language white matter pathway. Hum Brain Mapp. 2017;38(4):1977–1991.
- 39↑
Wang Y, Fifer MS, Flinker A, et al. Spatial-temporal functional mapping of language at the bedside with electrocorticography. Neurology. 2016;86(13):1181–1189.
- 40↑
Aron O, Jonas J, Colnat-Coulbois S, Maillard L. Language mapping using stereo electroencephalography: a review and expert opinion. Front Hum Neurosci. 2021;15:619521.
- 41↑
Gil Robles S, Gelisse P, Vergani F, et al. Discrepancies between preoperative stereoencephalography language stimulation mapping and intraoperative awake mapping during resection of focal cortical dysplasia in eloquent areas. Stereotact Funct Neurosurg. 2008;86(6):382–390.
