Intraoperative Electrophysiology in Children – Single Institute Experience of 96 Examinations
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.284352
Source of Support: None, Conflict of Interest: None
Keywords: Electrophysiology, intraoperative, neurosurgery, pediatricKey Message: The following article describes importance of intra operative electrophysiological examinations (IEE) in children based on an experience of 96 patients. The authors have recommended that IEE should be still performed even though the success rates might be lower than in adults due to incomplete myelination.
Neurosurgical interventions at eloquent sites are high-risk operations at every age. Injuring these neurological structures can lead to permanent deficit and deterioration of the quality of life. Technical innovations can help to extend the surgical boundaries,, and modern imaging with neuronavigation can visualize anatomical and functional territories. However, significant differences may occur in the location of certain eloquent regions,, and slowly progressing pathologies; brain plasticity can also alter the normal functional anatomy., Checking the functional integrity of the structures is possible with the application of intraoperative neuromonitoring methods. At the same time, with all these devices the expectations have changed. Nowadays, maximal safe resection is expected from a neurosurgical operation, and surgical resection has become the first choice even in high risk diseases such as intramedullary spinal cord tumors. The positive prognostic effect of radical excision is also well known in other cerebral and spinal cord pathologies.,,,,,,
Specific pathologies tend to occur in childhood. Certain primary central nervous system tumors may involve eloquent regions such as the brain stem and spinal cord. The number of resective epilepsy surgeries is also high, and these interventions may also involve speech or motor areas of the brain. Operation of the tethered spinal cord mainly occurs in childhood. Although the mapping and monitoring techniques are well standardized in adult patients, due to developmental and maturation processes their application in childhood could be limited as certain modalities such as d-wave recording or speech mapping need a mature nervous system., The aim of the present study was to evaluate the feasibility of different intraoperative electrophysiological techniques in childhood.
Between September 2012 and June 2016 we performed 96 neurosurgical operations on 95 patients under 14 years of age (mean: 5.7 years, 0.3–14 years, male/female ratio: 37/58) using intraoperative monitoring and mapping modalities.
During the surgeries, we conducted complex examinations tailored to the actual pathology and localization with an Inomed ISIS IOM Portable device (Inomed, Germany). This equipment can simultaneously record 16 bipolar and eight monopolar channels, and has the capability of stimulating in seven channels, including high-frequency direct cortical stimulation, direct nerve stimulation, bipolar stimulation for somatosensory evoked potentials (SSEP), and motor evoked potentials (MEP) examinations. We used standard preset scenarios for each specific pathology and localization, with minor fine tuning when necessary.
We used direct cortical mapping for lesions involving the central region. During this procedure, we stimulated the exposed brain surface with a bipolar stimulator. We applied biphasic stimulus using a pulse width of 1000 μs, which was reduced to 500 μs when epileptic seizures occurred, with an amplitude of 4–12 mA at a frequency of 50 Hz. We detected compound muscle action potentials on the contralateral side. We regularly used eight channels, including the facial muscles and the upper and lower extremity muscle groups. We stimulated for approximately 3 s to have an effective response. When the stimulation evoked epileptic seizures, we used ice cold irrigation fluid to cool the brain surface, as any agent depressing the cortical activity could jeopardize the feasibility of further examinations. To localize the central region we also recorded SSEP, detecting the phase reversal phenomenon. We stimulated the contralateral median nerve and detected the SSEP waves directly from the brain surface with a strip electrode with four or eight contacts. By detecting the reversal of the SSEP wave phase, we could determine the localization of pre and postcentral gyrus.
In our practice, we performed white matter mapping with similar parameters such as the motor cortex mapping, however, we used a handpiece with a smaller tip distance (2 mm versus 5 mm) we tended to go up to higher amplitudes.
For mapping the nucleus of the facial nerve, we used high-frequency stimulation with 50 Hz, the pulse width of 1000 μs, and amplitude up to 4.9 mA, with a small tip distance handpiece.
Sensory functions were monitored with SSEP. By stimulating peripheral nerve in the upper and lower extremities (tibial and median nerve), we recorded the activity of the sensory cortex at the sites of C3'-C4' for the upper and Cz-Fz for the lower extremity according to the international 10–20 electroencephalography (EEG) system. The stimulation frequency was 4.7 Hz and the applied pulse width was 200 μs. However, as this method gives information regarding the dorsal column, during intramedullary surgeries, the dorsal myelotomy and retraction can eliminate SSEP waves at the very beginning of the surgery. Thus, SSEP does not have a major role in determining the extent of safe resection. Motor pathways were monitored by stimulating the brain at the C3-C4 or C1-C2 a site according to the international 10–20 EEG system. We could evoke motor activity detected as compound muscle action potentials in the extremities. We stimulated up to 220 mA with a pulse width of 1000 μs. We applied a five train of stimuli with alternating polarity at 4 ms interstimuli interval. The sustained evoked potential refers to the intact motor pathway, whereas the vanishing waves to postoperative paresis.,
Detecting the d-wave gives predictable information regarding the motor pathway. The electric activity is recorded above the spinal cord epidurally after transcranial stimulation. If the amplitude of the d-wave falls below 50%, a permanent postoperative deficit occurs. When it stays above 50% of the baseline, recovery can be expected, even when short-term deterioration is experienced after surgery. The applied parameters are similar to MEP stimulation, but we used only a single, 500 μs stimulus. For reliable results, four or five responses were averaged.
Peripheral nerves, such as the cranial nerves, were monitored reliably during the surgeries of tethered spinal cord. The method is similar to cranial nerve mapping/monitoring. We placed needle electrodes in the corresponding muscle group and applied direct stimulus on the site where we expected the nerve fibers passing. We used a single bipolar stimulus with a pulse width of 200 μs up to 10 mA. Compound muscle action potentials were recorded by the needle electrodes and free-run electromyography (EMG) was used to detect mechanical traction of the nerve.
During the resection of pathologies involving the fibers of the cauda equina or sacral regions, vegetative functions need to be monitored. In our practice, we used direct stimulation and detection of compound muscle action potential from the external anal sphincter combined with bulbocavernosus reflex (BCR). The latter includes peripheral stimulation of the pudendal nerve and detection of the response at the external anal sphincter. We used a single stimulus on the pudendal nerve at the site of the genitals and bipolar stimulation with a pulse width of 200 μs up to 30 mA. This response involves a polysynaptic reflex arc, which correlates with vegetative functions.
A special setup was used during selective dorsal rhizotomy (SDR). This surgery can relieve the symptoms of lower limb spasticity in patients with cerebral palsy. We prefer a monosegmental approach to minimize invasivity and postoperative recovery interval.,, We performed a laminectomy at the level of the conus, usually at Th.XII-L.I. level. We exposed the conus medullaris and the fibers constituting the cauda equina. Separating the motor fibers can be aided by free run EMG as mechanical manipulation of motor fibers evokes significant EMG response while the manipulation of sensory fibers does not. We divided the roots distal form S.2., including the vegetative fibers, by anatomical characteristics. The remaining fibers from L.1. to S.1. were then divided by segments, and each segment was dissected to three subdivision. Each of these sensory fascicles were stimulated determining the motor response threshold at 1 Hz and 100 μs pulse width. When we reached the threshold, we applied a 50-Hz tetanic stimulation lasting 1 s and detected the compound muscle action potentials in six channels in both lower extremities (m. adductor, m. quadriceps femoris, m. biceps femoris, m. tibialis anterior, m. peroneus longus, and m. triceps surae). The responses were analyzed and stratified in five groups based on the strength of the result. We cut in each segment on both sides, i.e., the two more active parts out of the three.
All the surgeries were carried out under general anesthesia with total intravenous anesthesia using propofol and fentanyl. We administered short-acting muscle relaxant for intubation. The relaxants were completely washed out by the time of the electrophysiological monitoring. The index of consciousness was monitored with IoC View™ monitoring system (Morpheus Medical, Barcelona, Spain) or BIS™ Brain Monitoring System, (Covidien/Medtronic USA).
The aim of our study was to define the feasibility of the different electrophysiological modalities. During surgeries, we usually use a combination of these methods. The results are summarized in [Table 1], and the characteristics of the pathologies are presented in [Table 2].
Somatosensory evoked potentials
SSEP was performed in 14 cases. The mean age at the time of surgery was 7.6 years (2.1–13.8 years). We recorded SSEP once for phase reversal detection during central region tumor resection, seven times during intramedullary tumor resection, in five cases during orthopedic spine surgery, and in one occasion during lumbar intradural pathology. We found reproducible waves in seven cases. The average stimulation amplitude of the peripheral nerve was 19.3 mA (9.3–29.3 mA).
Motor evoked potentials
We applied MEP recording in 22 cases. The mean age of patients at the time of the surgery was 7.1 years (1–14 years). The pathologies included one optic glioma involving the corticospinal tract, five brainstem gliomas, eight intramedullary spine tumor, five orthopedic spine reconstructions, two lumbar intradural pathologies, and one thoracal extramedullar tumor resection. Reproducible MEP responses were detected in 19 cases, with a mean stimulation amplitude of 122 mA (80–170 mA).
D-wave detection was done in four cases. All pathologies were cervical or thoracal intramedullary lesions. The mean age of the patients at the time of the surgery was six years (0.8–11.2 years). Out of the four cases, we found reproducible d-wave in only one case, with stimulation amplitude of 160 mA.
Cranial nerve mapping (trigeminal and facial nerve)
We used cranial nerve mapping in one case during the surgery of a grade II meningioma at the petroclival region. The patient was eight years old. We could elicit reproducible waves at 1.3 mA.
Peripheral nerve mapping
Peripheral nerve monitoring was done in 34 cases. All of them were during the surgery of the tethered cord. The mean age of the patients at the time of the surgery was 3.5 years (0.3–13.5 years). Stable responses were obtained in 30 cases at a mean stimulation amplitude of 1.3 mA (0.8–2.5 mA). Out of the four unsuccessful cases, on three occasions we performed direct stimulation of the tethering bundle at the standard amplitude, and as there was no response, we found it safe to be transected. There was only one case where we could not detect response up to 10 mA stimulation amplitude.
Motor cortex mapping
We performed direct motor cortex mapping in three cases, in one case we combined this with phase reversal detection. The mean age of the patients at the time of the surgery was 8.1 years (6.4–9.2 years). Pathologies were tumors adjacent to the central region. In all cases, direct cortical stimulation was successful with a mean stimulation amplitude of 9.4 mA (8–12 mA).
White matter mapping
We had three pediatric cases where we used white matter stimulation. The mean age of the patients at the time of the surgery was 4.8 years (2–9.6 years). The pathologies included one thalamic tumor, one deconnective surgery for epilepsy, and one optic glioma involving the corticospinal tract. We could detect reproducible waves in two cases with a 9 mA mean stimulation amplitude (6–12 mA).
Selective dorsal rhizotomy protocol
During our study period, we operated on 33 spastic patients. The mean age of the patients at the time of the surgery was 6.9 years (2.9–12.7 years). All the patients had cerebral palsy. All of them underwent single-level selective dorsal rhizotomy. In all cases, separation of the motor fibers based on the irritation responses on the free-run EMG was successful. The grading of the spastic response was done at 330 location (five levels in each patients per sides, due to the limited testing of the L.1. fibers in this level, we arbitrarily transected 50% of the sensory bundle). The average minimum threshold amplitude was 2.1 mA (1–7.6 mA) and the average maximum threshold amplitude was 12.6 mA (3–29.7 mA). The hardware limit of the stimulation was 30 mA. In three patients in the case of 22 bundles, we could not detect response at the upper hardware limit. In these cases, we arbitrarily chose the fibers to be transected.
This method was used in eight cases, all of them were detethering surgery. The mean age of the patients at the time of the surgery was 3.4 years (0.6–8.7 years). Out of the eight cases, we could detect BCR waves in five cases; the mean stimulation amplitude was 24.2 mA (20–30 mA).
Facial nucleus mapping
During the surgery of six brainstem pathologies, we performed facial nucleus mapping to determine the safe entry location. In five cases, we could detect responses at a mean amplitude of 2.4 mA (0.9–4.9 mA). The mean age of the patients at the time of the surgery was 8.5 years (2.9–14 years). We were not able to detect responses in the case of the eldest patient. Based on our data, we found that the motor nuclei of the facial nerve could be mapped reliably.
Safety of neurosurgical operations is utmost of importance. Several technical innovations are now available to make such surgeries safer such as functional imaging, tractography, fiber tracking, or neuronavigation. Intraoperative imaging is also widely available. However, parallel to extending the surgical resection, the postoperative adverse events must be minimized. Real-time functional feedback regarding the neurological structures that are being operated on can be obtained by intraoperative electrophysiological monitoring. Adding this to the complex diagnostic and monitoring scenario makes extended resection safer.
However, there are several factors that distinguish pediatric electrophysiology from adult cases. Maturation and developmental processes may limit certain monitoring possibilities. In addition, there are a number of pathologies that tend to appear in childhood. The incidence of intrinsic central nervous system tumors peaks at the ages of 3 and 14 years. Gliomas involving the posterior fossa and the brainstem are also more often in childhood. Developmental anomalies affecting the brain, spinal cord, or spine are often diagnosed in childhood; sometimes they can even be diagnosed in utero or at the moment of the delivery. Safe and radical solutions for these cases are as important as in the older ages.
We reviewed the feasibility of electrophysiological examinations in 96 pediatric operations. We also applied a combination of different mapping and monitoring methods tailored to the actual pathology.
Motor cortex mapping was employed three times, and we found that this is a very reliable method in childhood as well as in the adult population. Although there are controversial data in the literature about the possibility of white matter mapping, we found that, as in adult population, it is also possible in pediatic population. Our success rate was 66% implying that this method can be used reasonably during the surgery involving motor pathways. The failure could be in connection with the immature myelination process, however, in our cases, successful mapping was available at the age of 2.0 and 9.6 years. The failed attempt was at the age of 2.9 years. We also need to mention that younger patients needed higher amplitude of stimulation. More data is needed to confirm the possible connection between the success and the myelination process.
Facial nerve mapping was applied in only one case with success. We also tried to stimulate the facial nerve nucleus to map the floor of the fourth ventricle. This method is useful to determine the entry point during the surgery of brainstem pathologies when the usual landmarks cannot be identified because of the distorted anatomy. In 83% of the cases, we could detect reliable response. The failure was observed in the case of the eldest patient. Hence, we consider that the reason for it is not related to age.
MEP recording was used during the surgery on spinal or brainstem pathologies. We found a success rate of 86%. However, this method has its limitations. It cannot predict if the paresis is permanent and has no mapping value, neither in brainstem nor in spinal cord pathologies. The examination must be repeated at short intervals to make it safe. However, it can prolong the surgery, as during stimulation the resection has to be paused. Among our patients, we did not find any relation between the failure of MEP detection and the preoperative deficit as two of them were neurologically intact and the third case had severe neurological symptoms. Although the eldest of these children was 2.9 years old (the other two were one year and 2.7 years old), the small number of cases does not make it possible to gain statistically significant data regarding the correlation of failure and patient's age.
We observed a low success rate in the case of SSEP (50%). The observed failure did not depend on the preoperative neurological deficit. However, the missing SSEP response did not influence the surgery as in our practice this modality is used as an auxilliary method, both in intracranial (combined with direct cortical mapping) and spinal (combined with MEP and d-wave detection) procedures. Especially in intramedullary pathologies, dorsal myelotomy can deteriorate the SSEP responses even when they are present at the beginning of the surgery, thus, MEP and d-wave were the methods we relied on during these cases. MEP alone cannot predict if a potential postoperative paresis will be transient or permanent. However, the decrement of d-wave amplitude below 50% of the baseline refers to permanent deficit. Thus, the standard method should be d-wave recording to predict the motor outcome combined with MEP.,,
We conducted d-wave detection in four cases. There was only one patient where we could detect reproducible waveform, a 10-month-old child. The reason might be that the detection of d-wave is susceptible to many factors. As described before, the maturation process of the spinal cord has a significant influence on the possibility of d-wave detection., In addition, to detect the wave, we need the synchronous activity of the corticospinal tract. The number of activated fibers, good preoperative functions, and mature nervous system with complete myelinization has a positive correlation with the possibility of d-wave detection. Upper levels are better as the exiting fibers decrease the number of activated fibers at the lower regions. At the lower thoracal level, or the level of the conus, d-wave recording cannot be used, as there is no spinal cord at the site of distal recording. Many of these factors are infavorable for recording among children.
Peripheral nerve stimulation was one of the easiest and most reliable method in our practice. The success rate was 88%, but taking into consideration that three out of four failures can be referred as negative mapping (accepting that no response with standard stimulation parameters can be considered and there are no active fibers in the examined bundle), we can say that in 33 out of 34 cases this technique helped the surgery.
The success rate of BCR detection was 63%. In cases where we could not detect BCR response, all patients had some degree of urine retention, which might be correlating with the missing BCR response.
During SDR surgeries, we used a standardized protocol. In our database, we found that with the described parameters we could detect responses in all patients. However, in three patients, there were certain bundles where we could not reach the threshold amplitude; however, it is more likely to be caused by the hardware limit of the stimulation amplitude at a maximum of 30 mA.
In our practice, intraoperative monitoring (where the applied modalities were successfully recorded) has been valuable in defining the safe boundaries of surgical resection, either by sustained stable responses or by warning signs during examinations. Although we use it routinely during surgeries of eloquent regions and we had no control groups to assess the differences in terms of outcome, we do believe that this tool helps minimize the risk of surgery related deterioration.
Based on our data, we can suggest that all intraoperative monitoring techniques can be used in childhood. Certain modalities, especially those that deal with peripheral nerves (cranial nerve monitoring, peripheral nerve monitoring, SDR protocol), offer high success rates even at a very young age; however, we have also demonstrated that even the modalities that require mature nervous system could be feasible. Thus, we strongly recommend using all the necessary and available electrophysiological methods at any age, keeping in mind that some of them may have a lower success rate compared to adults.
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2]