intraoperative neurophysiological monitoring during

clinical articleJ neurosurg Pediatr 17:147–155, 2016

Surgical treatment of pediatric cranial base tumors such as craniopharyngiomas, chordomas, angiofi-bromas, pituitary adenomas, and Rathke’s cleft cysts

has been evolving from conventional open skull base ap-proaches to novel, less invasive techniques like endoscop-ic endonasal surgery (EES).5,18,19,32,41 For properly selected tumors, EES offers several advantages over traditional

methods, including the sparing of disfiguring facial inci-sions and craniotomy. EES allows the surgeons to access the entire ventral skull base, from the crista galli to the upper cervical spine, with minimal postoperative compli-cations.15–17 The morbidity and mortality after pediatric skull base tumor resection has been reduced significantly with EES.5 However, during any cranial base surgery,

aBBreViatiOnS BAEP = brainstem auditory evoked potential; CMAP = compound muscle action potential; CN = cranial nerve; EEA = endoscopic endonasal approach; EES = endoscopic endonasal surgery; EMG = electromyography; ICA = internal carotid artery; IONM = intraoperative neurophysiological monitoring; JNA = juvenile nasopharyngeal angiofibroma; SSEP = somatosensory evoked potential.SuBmitted July 29, 2014. accePted July 6, 2015.include when citing Published online October 30, 2015; DOI: 10.3171/2015.7.PEDS14403.

Intraoperative neurophysiological monitoring during endoscopic endonasal surgery for pediatric skull base tumorscheran elangovan, mBBS,1 Supriya Palwinder Singh, mBBS,1 Paul gardner, md,1 carl Snyderman, md,1,3 elizabeth c. tyler-Kabara, md, Phd,1 miguel habeych, md, mPh,1 donald crammond, Phd,1 Jeffrey Balzer, Phd,1,4 and Parthasarathy d. thirumala, md, mS1,2

Departments of 1Neurological Surgery, 2Neurology, and 3Otolaryngology, University of Pittsburgh Medical Center; and 4Department of Neuroscience, University of Pittsburgh, Pennsylvania

OBJectiVe The aim of this study was to evaluate the value of intraoperative neurophysiological monitoring (IONM) using electromyography (EMG), brainstem auditory evoked potentials (BAEPs), and somatosensory evoked potentials (SSEPs) to predict and/or prevent postoperative neurological deficits in pediatric patients undergoing endoscopic endonasal surgery (EES) for skull base tumors.methOdS All consecutive pediatric patients with skull base tumors who underwent EES with at least 1 modality of IONM (BAEP, SSEP, and/ or EMG) at our institution between 1999 and 2013 were retrospectively reviewed. Staged procedures and repeat procedures were identified and analyzed separately. To evaluate the diagnostic accuracy of significant free-run EMG activity, the prevalence of cranial nerve (CN) deficits and the sensitivity, specificity, and positive and negative predictive values were calculated.reSultS A total of 129 patients underwent 159 procedures; 6 patients had a total of 9 CN deficits. The incidences of CN deficits based on the total number of nerves monitored in the groups with and without significant free-run EMG activity were 9% and 1.5%, respectively. The incidences of CN deficits in the groups with 1 staged and more than 1 staged EES were 1.5% and 29%, respectively. The sensitivity, specificity, and negative predictive values (with 95% confidence intervals) of significant EMG to detect CN deficits in repeat procedures were 0.55 (0.22–0.84), 0.86 (0.79–0.9), and 0.97 (0.92–0.99), respectively. Two patients had significant changes in their BAEPs that were reversible with an increase in mean arterial pressure.cOncluSiOnS IONM can be applied effectively and reliably during EES in children. EMG monitoring is specific for detecting CN deficits and can be an effective guide for dissecting these procedures. Triggered EMG should be elicited intraoperatively to check the integrity of the CNs during and after tumor resection. Given the anatomical complexity of pediatric EES and the unique challenges encountered, multimodal IONM can be a valuable adjunct to these procedures.http://thejns.org/doi/abs/10.3171/2015.7.PEDS14403KeY wOrdS intraoperative; neurophysiological monitoring; pediatric; endoscopic endonasal surgery; skull base; technique

©AANS, 2016 J neurosurg Pediatr Volume 17 • February 2016 147

Unauthenticated | Downloaded 02/28/22 05:11 PM UTC

c. elangovan et al.

J neurosurg Pediatr Volume 17 • February 2016148

there is potential risk to neurovascular structures such as the internal carotid artery (ICA), anterior cerebral arteries, and cranial nerves (CNs) that cause temporary or perma-nent neurological deficits.36 Neurological complications, including hemiparesis and cranial nerve palsies, can be predicted and prevented by utilizing real-time continuous intraoperative neurophysiological monitoring (IONM).25–

27,35–39 IONM using somatosensory evoked potentials (SSEPs),2 brainstem auditory evoked potentials (BAEPs), and electromyography (both free-run electromyography [EMG] and triggered EMG)26,27 have been valuable in conventional skull base surgeries. Similarly, the SSEP and BAEP values and free-run EMG of the CNs during EES for skull base tumors in adults have been previously de-scribed.35–39 However, to our knowledge, sufficient data to establish the value of IONM in pediatric EES have not been published.

EES for pediatric skull base tumors poses unique chal-lenges, including anatomical limitations, when compared with adults. For example, the size of the nasal aperture and limited pneumatization of the sinuses is likely to limit dissection in patients younger than 2 years.33 Staged and repeat procedures are performed for bulky tumors20,23 and frequently recurring lesions like craniopharyngiomas, pi-tuitary tumors, and chordomas. During staged procedures, the anatomical distortion caused by primary dissection can increase the risk of neurovascular injury in later stages. Similarly, there is an increased risk of neurovascular in-jury during repeat procedures because of scar tissue from previous surgery.14 In the event of a vessel injury or highly vascular tumors, blood loss is poorly tolerated in children in comparison with adults.42 Even with these significant risks, EES is associated with better quality of life,1 which can have lasting importance in the pediatric population.

The aim of this study is to investigate the value of sig-nificant changes in IONM for predicting and/or prevent-ing postoperative neurological deficit following EES for pediatric skull base tumors. This information will be very useful for optimizing the utilization of various intraopera-tive monitoring modalities in order to reduce the morbid-ity and mortality associated with EES for pediatric skull base tumors.

methodsWe retrospectively reviewed all consecutive EES pro-

cedures performed in children at our institution between 1999 and 2013. In total, 129 patients younger than 18 years who underwent 159 procedures with IONM were identi-fied. The inclusion criteria for the study were patients who underwent EES with IONM using at least 1 modal-ity (BAEP, SSEP, and/or EMG); 2 patients who did not have documented postoperative neurological status were excluded. This study was approved by the local institu-tional review board for retrospective review of the clini-cal outcomes. The subsequent description of the neuro-physiological monitoring techniques has been discussed in detail in earlier publications,35–39 and a brief overview is given below.

The monitoring protocol was decided preoperatively based on the neurovascular structures at risk. Since the lo-

cations of the tumors in patients differed, different proto-cols were tailored for individual patients. EMG was used based on the surgeon’s impression of the potential involve-ment during exposure or resection.

neurophysiological monitoringPhysician oversight and interpretation were performed

using a combined on-site and remote model used by the University of Pittsburgh Medical Center. In all cases, a board-certified neurophysiologist was on-site and imme-diately available for interpretation and consultation, and physician (neurologist) oversight, supervision, and in-terpretation were performed in person or remotely. The overseeing physician provided supervision to 4 to 5 cases simultaneously on average, with a maximum of 8 cases. No special adaptations were required for pediatric cases.

BaePBaseline BAEP levels were obtained by stimulating

both ears independently: left ear and right ear. Recordings were made continuously throughout the procedure by de-livering a click stimulus to 1 ear, either the left ear or right ear, at an 85-dB hearing level at a stimulus rate of 17.5 Hz. White noise was applied to the contralateral ear at a 65-dB hearing level. The observation interval was 12 msec. Recording channels included Cz-A1, Cz-A2, and Cz-Cv2. Amplifier bypass was 100 Hz to 1 KHz for all channels. Baseline BAEP responses were obtained after the initia-tion of anesthesia, positioning of the patient, and in some cases before any major manipulations.

SSePAfter the induction of anesthesia and positioning of the

patient, baseline SSEP responses were established, except in cases of basilar invagination or other severe brainstem or cervicomedullary compression, in which case baseline SSEP values were obtained prior to positioning. Upper- and lower-extremity SSEP responses were continuously obtained throughout the procedure. Subdermal needle electrode pairs were used for stimulating the median or ulnar nerves bilaterally for the upper-extremity SSEPs and the tibial and peroneal nerves bilaterally for the lower-ex-tremity SSEPs. Recordings were obtained from the scalp, cervical region, and Erb’s point with subdermal elec-trodes. All electrodes were placed per the international 10-20 system. P4/Fz and P3/Fz scalp electrodes were used to record cortical potentials for upper-extremity SSEPs. An extra Pz/Fz channel was used for lower-extremity cor-tical SSEPs. A cervical electrode was localized at the C7 spinous process or mastoid and referenced to the scalp electrode Fz. Erb’s point recordings were obtained us-ing EPs and EPd electrodes that were placed close to the brachial plexus. Band-pass filters were set from 10 to 250 Hz for cortical recordings and 30 to 1000 Hz for cervi-cal and Erb’s point recordings. The stimulation frequency was 2.33 to 2.45 Hz with a duration of 0.2 to 0.3 msec. The averages were computed for either 128 or 256 trials, depending on the signal quality.

emg monitoringContinuous free-run EMG activity was recorded using


neurophysiological monitoring of pediatric skull base tumors

J neurosurg Pediatr Volume 17 • February 2016 149

pairs of subdermal needle electrodes that were placed 1 cm apart in or near the muscle groups innervated by a CN. The sensitivity, time base, and bandwidth were established at 50 mV/division, 100 msec/division, and 3 Hz to 1 KHz for recording the responses. The details of the monitored CNs and muscle groups are shown in Table 1. Triggered EMG was recorded from the appropriate CNs (Table 1) during the procedure using the same electrodes placed for free-run EMG recording. The CNs were stimulated using a constant-voltage monopolar stimulator with current in-tensity ranging from 0.2 to 2 V based on the amplitude of the response. The lowest current stimulation, which was used to obtain a response, was determined as the thresh-old. The monopolar stimulator was introduced separately outside the endoscope. A return electrode was placed to field. The sensitivity, time base, and bandwidth were es-tablished at 50 mV/division, 5 msec/division, and 3 Hz to 1 KHz for recording the triggered EMG responses.

alarm criteriaFor BAEPs, persistent decreases in amplitude of more

than 50% of wave V and/or a persistent absolute latency increase of the peak of wave V ≥ 0.5 msec were consid-ered clinically significant. Changes in more than 2 consec-utive averaged trials were considered “persistent changes.” For SSEPs, a 10% increase in latency or 50% decrease in amplitude relative to baseline was considered clinically significant.31,35–37

Changes in more than 2 consecutive averaged tri-als were considered persistent changes. The absence of free-run EMG activity was considered baseline in each case. The detection of nerve manipulation, compression, stretch, and/or permanent injury was based on changes from the baseline recordings. Significant free-run EMG activity, when present for prolonged periods of time (≥ 100 msec) from a CN, was reported to the surgeons as 1 alert and was also recorded in the patient’s records.38,39 Trig-gered EMG responses were obtained upon request using a constant-current stimulator. All negative and positive responses were communicated to the surgeons. When ap-propriate, the lowest possible current required to stimulate a nerve was obtained. Furthermore, we also attempted to obtain triggered EMG responses after tumor resection was completed.

medical records reviewThe medical records of the 129 patients were reviewed

to determine if any new neurological deficits were pres-ent after the surgical procedure. The medical records were reviewed independently without knowledge of the IONM changes. Any new postoperative motor or sensory deficits were considered to be due to iatrogenic injuries. Deficits were classified as transient or permanent. Transient deficits had to have documented evidence in the medical records of complete improvement to baseline. Permanent deficits were defined as those that did not improve to baseline in subsequent follow-up visits (> 1 month).

data analysisThe means and standard deviations were calculated for

the continuous variables, and ratios were calculated for categorical variables. We calculated the percentages to determine the distributions of the types of procedures, as well the postoperative pathological diagnoses. For the data analysis, the number of nerves monitored for each patient was calculated by considering unilateral as “1” and bilat-eral as “2” nerves monitored. Depending on the presence or absence of significant free-run EMG activity, data were divided into 2 groups and analyzed. Staged procedures and repeat procedures were identified and analyzed separately to determine the incidence of significant activity and CN deficits. To evaluate the diagnostic accuracy of significant free-run EMG activity, we calculated the prevalence of CN deficits and the sensitivity, specificity, and positive and negative predictive values. We further calculated the like-lihood ratios in order to compare the probability of obtain-ing significant free-run EMG activity if the patient had a deficit to the probability of obtaining a significant free-run EMG activity if the patient was healthy. To evaluate the measure of uncertainty, we calculated the 95% confidence intervals.

resultsdemographic data

The total number of children who underwent at least 1 IONM modality was 129, of whom 74% were male and 26% were female. Angiofibroma, pituitary adenoma, chordoma, and craniopharyngioma were the most com-mon diagnoses. The most common surgical approaches were transsellar (32%) and transclival (29%), followed by transpterygoid (11%) and others. There were 16 planned staged procedures (in 14 patients) and 10 repeat proce-dures. Two patients underwent a repeat staged procedure. A detailed description of the demographic information is shown in Table 2.

clinical OutcomesThe incidence of postoperative CN deficits was 9

(2.8%) of 321 CNs monitored. Of the 6 patients who had 9 CN deficits, 3 had a diagnosis of chordoma and 3 had ju-venile nasopharyngeal angiofibroma (JNA). Four CN defi-cits were transient (1.2%) and 5 were permanent (1.5%). Two patients underwent staged procedures, 1 patient un-derwent a repeat procedure, and 1 patient underwent a

taBle 1. monitored cns and muscle groups

CN Monitored Muscle Group

Occulomotor nerve Medial rectus muscleTrochlear nerve Superior oblique muscleAbducent nerve Lateral rectus muscleTrigeminal nerve MasseterFacial nerve Orbicularis oris, orbicularis oculi, &

mentalis (ipsilateral)Motor component of the glosso

pharyngeal nerveSoft palate (after intubation)

Recurrent laryngeal component of the vagus nerve

Cricothyroid muscle

Spinal accessory nerve Trapezius muscleHypoglossal nerve Tongue muscles


c. elangovan et al.


staged procedure followed by a repeat procedure at a later date. The details of the clinical deficits are shown in Table 3. No patient experienced quadriparesis, hemiparesis, or death after the procedure. There were no new postopera-tive vision deficits.

Free-Run EMG Activity and Neurological DeficitsFree-run EMG monitoring was performed in 62 (39%)

patients. The total number of CNs monitored in these cases was 321. Significant free-run EMG activity was observed in 55 (17%) nerves, and 266 (83%) nerves did not have sig-nificant free-run EMG activity. Five CN deficits were ob-served among the CNs with significant free-run EMG ac-tivity; 3 deficits were permanent and 2 were transient. Four CN deficits were observed among CNs without significant free-run EMG activity. All 4 of these deficits were in CN VI; 2 were transient and 2 were permanent. Seven CN defi-cits were observed in staged or revision EES, whereas 2 CN deficits were observed after the first stage of the proce-dure. The overall incidence of CN deficits was 2.8%. The incidence of CN deficits, when significant free-run EMG activity was observed, was 9% (3.6% were transient deficits and 5.4% were permanent deficits), and the incidence of CN deficits in the group without significant free-run EMG activity was 1.5% (0.75% were transient deficits and 0.75% were permanent deficits). The incidence of CN deficits in the group with only 1 staged EES was 1.5%, and the inci-dence of CN deficits in the group with a revision or staged procedure was 27%. An overview of significant EMG ac-tivity and CN deficits is detailed in Table 4. In addition, we evaluated significant free-run EMG activity in patients who underwent staged or repeat procedures and those with nerve deficits (Table 5).

Statistical Analysis of Significant Free-Run EMG ActivityThe prevalence of CN deficits in staged procedures was

4.2%, which is higher than the prevalence of 2.8% in the first staged procedure for all CNs monitored (p = 0.35). Free-run EMG activity had a high specificity and negative predictive value for all procedures, and higher specific-ity for repeat procedures (Table 6). The sensitivity of this study was low for all procedures. The likelihood of having a significant EMG activity was significantly higher in pa-tients undergoing a staged or repeat procedure.

Triggered EMGs were recorded in 7 patients and 12 cranial nerves; we were able to obtain compound muscle action potential (CMAP) responses from all 12 CNs (Fig. 1). No deficit was noted in those CNs without a significant change after tumor resection.

BaeP and SSePBAEP monitoring was performed in 16 (10%) patients,

of whom 2 patients showed significant BAEP changes in-traoperatively. One patient underwent a primarily trans-clival resection of JNA, and the other was a JNA that had a

taBle 2. demographic data

Variable No. (%)

No. of procedures 159No. of patients 129 Male 91 (71) Female 38 (29) Transclival 46 (29) Transsellar 51 (32) Transplanum 11 (7) Transpterygoid 17 (11) Transcavernous 1 (0.6) Other (>1 approach) 33 (21) 1–5 25 (19) 5–10 26 (20) 10–18 78 (61) Craniopharyngioma 14 (11) Chordoma 19 (15) Pituitary tumor 21 (16) Rathke’s cyst 9 (7) Angiofibroma 22 (17) Meningocele/encephalocele 6 (5) CSF leak 5 (4) Dermoid/epidermoid 4 (3) Other 29 (22)Staged procedure 16 (10)Repeat procedure 10 (6)Monitoring SSEP 156 (98) BAEP 16 (10) Free-run EMG 62 (39) EMG* 7 (11)

* Among 62 patients who underwent freerun EMG monitoring.

TABLE 3. Clinical deficits in individual patients

Age (yrs) Sex Diagnosis Op

CN Monitored

Side Monitored

Free-Run EMG Performed? Nerve Deficit (type)

No. of Stages/Repeat

10 M JNA TC III, VI Lt Yes, CN III; no, CN VI Lt CN III, lt CN VI (both transient) 3 stages13 M JNA TP VI Bilat Yes, lt CN VI Lt CN VI (permanent) 5 stages/1 repeat10 F Chordoma TC IX Rt Yes Rt CN IX (permanent) 1 stage11 M Chordoma TP XII Bilat Yes, lt CN XII Lt CN XII (transient) 1 stage11 M JNA TC V, VI Bilat Yes, rt CN V; no, CN VI Rt CN VIII (permanent); rt CN VI (transient) 2 stages10 M Chordoma TC VI Bilat No Bilat VI (permanent) 1 repeat

TC = transclival; TP = transpterygoidal.




primarily transpterygoid route. We observed a significant decrease in the amplitude of the wave V of the BAEPs during tumor resection. These changes were transient and improved with an increase in mean arterial pressure. No postoperative hearing deficits or weakness was observed. No SSEP changes were noted in any of the patients. No postoperative sensory or motor deficits were seen.

case exampleA 7-year-old boy presented with nasopharyngeal ob-

struction, headache, and diplopia from left CN VI palsy. MRI revealed a large clival lesion, consistent with chor-doma, involving the left parapharyngeal and petrous ICA (Fig. 2). A left cervical incision was made to isolate the ICA for proximal control. Intraoperatively, the left petrous ICA, just distal to foramen lacerum, was injured with a cutting rongeur. The site of bleeding was packed, and a vascular clip was placed on the cervical ICA with no change in SSEPs over 15 minutes. After removal of the packing, there was brisk back-bleeding from the distal stump of the injured ICA but still no change in the base-line SSEPs. Consequently, the ICA was sacrificed using endoscopic bipolar electrocautery.

The patient remained hemodynamically and electro-physiologically stable, and a decision was made to pro-

ceed with resection. Following resection, under continued SSEP monitoring, the patient was taken for endovascular evaluation and permanent occlusion of the ICA. Angiog-raphy showed an ipsilateral persistent trigeminal artery that filled the middle cerebral artery, and there was rapid cross-fill across the anterior communicating artery. As predicted by the IONM, the patient awoke without deficit. Postoperative MRI showed no diffusion-weighted imag-ing restriction and complete tumor removal, which was safely facilitated by IONM (Figs. 3 and 4).

discussionThe wide variation in pathology in this series is shown

in Table 2. The most common pathologies were craniopha-ryngioma, chordoma, JNA, and Rathke’s cleft cyst. While invasive chordomas could potentially damage the CNs in the foramina, Rathke’s cyst and craniopharyngioma (be-ing noninvasive) have less propensity to do so. However, nerve compression could occur. Similarly, intracranial extension of the JNAs could cause compressive lesions. The neurovascular structures that were at risk varied in individual cases depending on the stage of tumor.

emg monitoringOur results show that free-run EMG monitoring has

TABLE 4. Incidence of significant free-run EMG activity in the monitored CNs and nerve deficits

CN

No. of Monitored

CNs

Significant Free-Run EMG Activity No Significant Free-Run EMG Activity

No. (%)No. of Nerve

DeficitsNo. of Permanent

Deficits No. (%)No. of Nerve

DeficitsNo. of Permanent

Deficits

III 54 14 (26) 1 0 40 (74) 0 0IV 45 4 (9) 0 0 41 (91) 0 0V 11 4 (36) 1 1 7 (64) 0 0VI 87 5 (5.7) 1 1 82 (94.3) 4 2VII 16 3 (19) 0 0 13 (81) 0 0IX 29 9 (31) 1 1 20 (69) 0 0X 30 9 (30) 0 0 21 (70) 0 0XI 10 2 (20) 0 0 8 (80) 0 0XII 39 5 (13) 1 0 34 (87) 0 0Total 321 55 (17) 5 3 266 (83) 4 2

TABLE 5. Incidence of significant free-run EMG activity in the staged procedure and CN deficits

CN

No. of CNs Monitored Significant Free-Run EMG Activity CN Deficits, ≥2 Stages/Repeat

Total 1 Stage ≥2 Stages/Repeat, n No. 1 Stage, n (%) ≥2 Stages/Repeat, n (%) 1 Stage, n ≥2 Stages/Repeat, n

III 54 35 19 14 10 (71) 4 (29) 0 1IV 45 34 11 4 3 (75) 1 (25) 0 0V 11 1 10 4 0 4 (100) 0 1VI 87 38 49 5 3 (60) 2 (40) 0 5VII 16 7 9 3 1 (33) 2 (67) 0 0IX 29 13 16 9 4 (44) 5 (56) 1 0X 30 13 17 9 7 (78) 2 (22) 0 0XI 10 2 8 2 0 2 (100) 0 0XII 39 13 26 5 2 (40) 3 (60) 1 0Total 321 156 165 55 30 (55) 25 (45) 2 7


c. elangovan et al.


high specificity and a negative predictive value for detect-ing CN deficits. This allows the surgeon to feel relatively confident during tumor removal that there is no CN in-jury if there is no significant EMG activity. However, if a nerve is transected abruptly, there will be only brief or no EMG activity.12 We also observed that free-run EMG has a lower sensitivity for detecting cranial deficits dur-ing EES. Free-run EMG activity is observed secondary to the activation of a CN due to manipulation during tumor dissection, bipolar use, as well as other maneuvers during the surgery.38,39 Free-run EMG activity has been classified as bursts, spikes, and neurotonic discharges based on the amplitude and frequency of the discharges.21,22 It has been observed that manipulation of the CN is the primary rea-son for bursts and spike discharges with no correlation to nerve injury.25,27,28 In our practice, we use free-run EMG activity as a guide to alert the surgeon to the proximity of the nerve, wherein the surgeon understands the limita-tions and risks where a positive test indicates that he might be very close to a nerve. Our alarm criteria for significant free-run EMG activity allows for a careful dissection of the tumor, thereby reducing the incidence of neurological deficits.

In our study, the incidence of CN deficits in the group with significant free-run EMG activity was higher when compared with the group without significant activity. This is expected given the relationships between nerve dissec-tion/manipulation and free-run EMG and supports the con-cept that increased activity can be associated with injury. We also observed that children who had recurrent tumors or underwent staged procedures had a higher likelihood of significant free-run EMG activity, as well as a correspond-ing increase in the prevalence of CN deficits. Staging is used more often in pediatric cases since blood loss is poor-ly tolerated in children with bulky or vascular tumors. Re-current tumors require dissection of the fibrotic areas that are in close proximity to critical structures. There are scar tissues and adhesions from previous surgery, which makes tumor dissection challenging and in turn increases the risk of CN deficit. Neurotonic discharges, which are the high-frequency, high-amplitude discharges observed mostly in facial nerve monitoring, are a precursor of postoperative nerve injury.21,25 We did not observe neurotonic discharges uniformly in all instances of CN free-run EMG monitor-ing. Since we recorded all changes in significant free-run EMG activities, this could have lowered the sensitivity rate during the analysis.

Triggered EMG responses are obtained secondary to electrical stimulation of the CN in the tumor field during dissection, which produces CMAPs with specific laten-

TABLE 6. Prevalence, sensitivity, specificity, predictive value, and likelihood ratios of significant EMG activity to detect CN deficits

Significant Free-Run EMG Activity All Procedures ≥2 Stages

No. of CNs 321 165No. of deficits 9 7Prevalence 2.8% 4.2%Sensitivity (95% CI) 0.55 (0.22–0.84) 0.42 (0.11–0.79)Specificity (95% CI) 0.83 (0.79–0.87) 0.86 (0.79–0.9)Positive predictive value

(95% CI)0.08 (0.03–0.2) 0.12 (0.03–0.32)

Negative predictive value (95% CI)

0.98 (0.95–0.99) 0.97 (0.92–0.99)

Likelihood ratio (95% CI) 0.02 (0.00–0.04) 3.07 (1.2–7.8)

Fig. 1. left: MR image obtained in a patient with JNA extending into the cavernous sinus. right: Triggered EMG was performed to identify right CNs III, IV, and VI.

Fig. 2. Preoperative T2weighted axial MR images showing a large chordoma in a 7yearold patient adjacent to the left parapharyngeal ICA (left) and extensive involvement of the left petrous ICA (right).




cies that identify the CN during a procedure.27,29 Triggered EMG can be used as a mapping tool to identify the CNs along their tracts during the procedure. In our series, not all patients received triggered EMG monitoring; the adop-tion of triggered EMG coincided with increased surgeon experience that facilitated more aggressive resection. We believe adequate mapping of a CN during dissection and obtaining a triggered EMG CMAP response from a proximal site after complete resection provides valuable information regarding the integrity of the nerve. Based on this study and published suggestions, CN free-run EMG and triggered EMG monitoring, especially during staged procedures and for recurrent tumors, might be valuable for reducing morbidity by identifying the nerve during tumor dissection.11,12 Communicating free-run EMG ac-tivity to the surgeon plays a key role in complex decision making during tumor dissection. The predictability of the increased risk of clinically significant injury by free-run EMG and the lack of deficits in those few cases with docu-mented triggered EMG strongly support the role of these modalities during EES.

We report 9 CN deficits of 321 monitored CNs in chil-dren. Four CN deficits were transient and 5 CN deficits were permanent, which were seen in children with JNA and chordoma. Studies on the endoscopic and conventional resection of JNA have reported CN deficits in 0.6% to 15% of patients.6,9 Studies show that CN deficits were higher in the conventional resection of JNA, as well in patients who underwent surgery for recurrence (43%).24 The wide variation in results was secondary to the degree of tumor resection, tumor size, and intracranial extension.6,9,10,13,24 Our reported CN deficits of CNs III, V, and VI after JNA resection in children are similar to previously reported se-ries, especially given tumor size and the completeness of the resections. An even wider range in CN deficits after chordoma resection has been reported as 8% to 70%.28,30,40 Since traditional skull base surgery used for resection of JNA and chordomas includes lateral and midline ap-proaches where neurovascular structures are encountered before the tumor, the incidence of postoperative deficits may be higher in those patient groups. In our series with pediatric patients, the use of free-run EMG and triggered EMG contributed to the identification of the CNs during tumor dissection. This may not only improve the ability to preserve the nerves, but also help increase the degree of

resection by increasing surgeon confidence in nerve loca-tion and involvement relative to the tumor. Free-run EMG and triggered EMG have been shown to be effective in the past.26,27,35–39 However, a control group was not used during this study, as it is difficult to set up a control group in a pediatric patient population.

SSeP and BaePJNAs are vascular tumors whose resection could be

complicated by vascular injury and intraoperative hemor-rhage that lead to hypovolemia.13 Chordomas, which are tu-mors of the clival region, can compress the brainstem, and dissection at this location can lead to changes in BAEPs as observed in our study. Chordoma resections that involve the brainstem had previously reported motor weakness secondary to brainstem infarction.8,28,30 In the series by Brockmeyer et al. involving 55 patients, they reported 2 patients with hemiparesis and 4 patients with permanent CN deficits. 4 In a series involving 26 patients, Teo et al. reported 1 patient with quadriparesis and 12 CN deficits.34 We observed significant changes in BAEPs, which were reversed with an increase in mean arterial pressure. No patient who underwent EES for resection of pediatric skull base tumors experienced quadriparesis, hemiparesis, or death after the procedure. BAEPs are sensitive to changes in stretch on the auditory nerve, and hypoperfusion along the brainstem lemniscal pathways. BAEP changes can be measured by changes in latency and amplitude of wave-forms, though wave V is the most prominent.31 A larger decrease in brainstem blood flow causes changes in the amplitudes of SSEP and BAEP.3 Monitoring SSEP and BAEP for tumors in proximity to the brainstem can pro-vide a multimodal approach that protects the brainstem’s sensory and auditory pathways since the blood supply to the brainstem consists of single end arteries, and a sudden change in 1 pathway may not be reflective of changes in the other pathway.

SSEP plays a critical role in the resection of tumors with a high risk of ICA injury such as chordomas.7 In the absence of prior balloon test occlusion, intraoperative SSEP becomes the surrogate for the physiological impact of arterial injury or occlusion. This is illustrated by the case example, whereby the decision to completely occlude the injured ICA both intraoperatively and postoperatively

Fig. 3. Postoperative angiogram showing a persistent trigeminal artery filling the ICA and middle cerebral artery (lateral view) (left), as well as brisk cross-fill via the anterior communicating artery (anteroposterior view) (right). Fig. 4. Postoperative T2weighted MR image showing the complete re

section of an extensive skull base chordoma (left), which was adjacent to the now occluded left ICA (right).


c. elangovan et al.


was made based solely on SSEP data. In addition, the sta-bility of intraoperative potentials allowed for the removal of the remainder of tumor, thereby significantly impacting outcome.

This is a retrospective analysis of the results, which limits our capacity to collect long-term information and is therefore a limitation. Also, our series includes patients with pituitary adenomas, Rathke’s cysts, and CSF leak re-pairs, which are low risk for carotid artery injury. SSEP monitoring was used in these patients early on during the development of EES as a standard protocol. Our review of a previous large series40 did indicate that the changes in mean arterial blood pressure and anesthetic protocols did have an effect on SSEP, which were more common during the early part of the procedure. This was attributed to the learning effect of the surgical and neuroanesthesia team. This could have introduced potential selection bias in our study. But, in the current study, we did not find any neurovascular deficits in the patients without monitoring.

Deficits were determined by a chart review. This meth-od has its pros (a relatively inexpensive modality for re-searching rich, readily accessible, and existing data) and cons (incomplete documentation, unrecorded/unrecover-able documents, and problematic verification of informa-tion). The best possible outcome would be obtained by do-ing a prospective study.

conclusionsIONM can be applied effectively and reliably during

pediatric EES. EMG monitoring is specific for detecting CN deficits and can be an effective guide to dissection in these procedures. Triggered EMG can be elicited intra-operatively to check the integrity of the CNs during and after tumor resection. Given the anatomical complexity of pediatric EES and the unique challenges encountered, multimodal IONM can be a valuable adjunct to these pro-cedures.

references 1. Abergel A, Cavel O, Margalit N, Fliss DM, Gil Z: Compari-

son of quality of life after transnasal endoscopic vs open skull base tumor resection. Arch Otolaryngol Head Neck Surg 138:142–147, 2012

2. Bejjani GK, Nora PC, Vera PL, Broemling L, Sekhar LN: The predictive value of intraoperative somatosensory evoked potential monitoring: review of 244 procedures. Neurosur-gery 43:491–500, 1998

3. Branston NM, Symon L, Crockard HA, Pasztor E: Relation-ship between the cortical evoked potential and local cortical blood flow following acute middle cerebral artery occlusion in the baboon. Exp Neurol 45:195–208, 1974

4. Brockmeyer D, Gruber DP, Haller J, Shelton C, Walker ML: Pediatric skull base surgery. 2. Experience and outcomes in 55 patients. Pediatr Neurosurg 38:9–15, 2003

5. Chivukula S, Koutourousiou M, Snyderman CH, Fernandez-Miranda JC, Gardner PA, Tyler-Kabara EC: Endoscopic endonasal skull base surgery in the pediatric population. J Neurosurg Pediatr 11:227–241, 2013

6. Cloutier T, Pons Y, Blancal JP, Sauvaget E, Kania R, Bres-son D, et al: Juvenile nasopharyngeal angiofibroma: does the external approach still make sense? Otolaryngol Head Neck Surg 147:958–963, 2012

7. Gardner PA, Tormenti MJ, Pant H, Fernandez-Miranda JC,

Snyderman CH, Horowitz MB: Carotid artery injury during endoscopic endonasal skull base surgery: incidence and out-comes. Neurosurgery 73(2 Suppl Operative):ons261–270, 2013

8. Gay E, Sekhar LN, Rubinstein E, Wright DC, Sen C, Janecka IP, et al: Chordomas and chondrosarcomas of the cranial base: results and follow-up of 60 patients. Neurosurgery 36:887–897, 1995

9. Godoy MD, Bezerra TF, Pinna FdeR, Voegels RL: Complica-tions in the endoscopic and endoscopic-assisted treatment of juvenile nasopharyngeal angiofibroma with intracranial extension. Braz J Otorhinolaryngol 80:120–125, 2014

10. Hackman T, Snyderman CH, Carrau R, Vescan A, Kassam A: Juvenile nasopharyngeal angiofibroma: The expanded endo-nasal approach. Am J Rhinol Allergy 23:95–99, 2009

11. Harner SG, Daube JR, Ebersold MJ, Beatty CW: Improved preservation of facial nerve function with use of electrical monitoring during removal of acoustic neuromas. Mayo Clin Proc 62:92–102, 1987

12. Harper CM: Intraoperative cranial nerve monitoring. Muscle Nerve 29:339–351, 2004

13. Huang Y, Liu Z, Wang J, Sun X, Yang L, Wang D: Surgi-cal management of juvenile nasopharyngeal angiofibroma: analysis of 162 cases from 1995 to 2012. Laryngoscope 124:1942–1946, 2014

14. Hwang JM, Kim YH, Kim JW, Kim DG, Jung HW, Chung YS: Feasibility of endoscopic endonasal approach for recur-rent pituitary adenomas after microscopic trans-sphenoidal approach. J Korean Neurosurg Soc 54:317–322, 2013

15. Kassam A, Snyderman CH, Mintz A, Gardner P, Carrau RL: Expanded endonasal approach: the rostrocaudal axis. Part I. Crista galli to the sella turcica. Neurosurg Focus 19(1):E3, 2005

16. Kassam A, Snyderman CH, Mintz A, Gardner P, Carrau RL: Expanded endonasal approach: the rostrocaudal axis. Part II. Posterior clinoids to the foramen magnum. Neurosurg Focus 19(1):E4, 2005

17. Kassam AB, Snyderman C, Gardner P, Carrau R, Spiro R: The expanded endonasal approach: a fully endoscopic trans-nasal approach and resection of the odontoid process: techni-cal case report. Neurosurgery 57(1 Suppl):E213, 2005

18. Komotar RJ, Roguski M, Bruce JN: Surgical management of craniopharyngiomas. J Neurooncol 92:283–296, 2009

19. Laufer I, Anand VK, Schwartz TH: Endoscopic, endonasal extended transsphenoidal, transplanum transtuberculum approach for resection of suprasellar lesions. J Neurosurg 106:400–406, 2007

20. Pirris SM, Pollack IF, Snyderman CH, Carrau RL, Spiro RM, Tyler-Kabara E, et al: Corridor surgery: the current paradigm for skull base surgery. Childs Nerv Syst 23:377–384, 2007

21. Prass RL, Kinney SE, Hardy RW Jr, Hahn JF, Lüders H: Acoustic (loudspeaker) facial EMG monitoring: II. Use of evoked EMG activity during acoustic neuroma resection. Otolaryngol Head Neck Surg 97:541–551, 1987

22. Prass RL, Lüders H: Acoustic (loudspeaker) facial electro-myographic monitoring: Part 1. Evoked electromyographic activity during acoustic neuroma resection. Neurosurgery 19:392–400, 1986

23. Prosser JD, Vender JR, Alleyne CH, Solares CA: Expanded endoscopic endonasal approaches to skull base meningiomas. J Neurol Surg B Skull Base 73:147–156, 2012

24. Pryor SG, Moore EJ, Kasperbauer JL: Endoscopic versus tra-ditional approaches for excision of juvenile nasopharyngeal angiofibroma. Laryngoscope 115:1201–1207, 2005

25. Romstöck J, Strauss C, Fahlbusch R: Continuous electromy-ography monitoring of motor cranial nerves during cerebel-lopontine angle surgery. J Neurosurg 93:586–593, 2000

26. Schlake HP, Goldbrunner R, Siebert M, Behr R, Roosen K: Intra-operative electromyographic monitoring of extra-ocular




motor nerves (Nn. III, VI) in skull base surgery. Acta Neuro-chir (Wien) 143:251–261, 2001

27. Schlake HP, Goldbrunner RH, Milewski C, Krauss J, Traut-ner H, Behr R, et al: Intra-operative electromyographic moni-toring of the lower cranial motor nerves (LCN IX-XII) in skull base surgery. Clin Neurol Neurosurg 103:72–82, 2001

28. Sekhar LN, Pranatartiharan R, Chanda A, Wright DC: Chor-domas and chondrosarcomas of the skull base: results and complications of surgical management. Neurosurg Focus 10(3):E2, 2001

29. Sekiya T, Hatayama T, Iwabuchi T, Maeda S: Intraoperative recordings of evoked extraocular muscle activities to moni-tor ocular motor nerve function. Neurosurgery 32:227–235, 1993

30. Sen C, Triana AI, Berglind N, Godbold J, Shrivastava RK: Clival chordomas: clinical management, results, and compli-cations in 71 patients. J Neurosurg 113:1059–1071, 2010

31. Shah A, Nikonow T, Thirumala P, Hirsch B, Chang Y, Gard-ner P, et al: Hearing outcomes following microvascular de-compression for hemifacial spasm. Clin Neurol Neurosurg 114:673–677, 2012

32. Stippler M, Gardner PA, Snyderman CH, Carrau RL, Preve-dello DM, Kassam AB: Endoscopic endonasal approach for clival chordomas. Neurosurgery 64:268–278, 2009

33. Tatreau JR, Patel MR, Shah RN, McKinney KA, Wheless SA, Senior BA, et al: Anatomical considerations for endo-scopic endonasal skull base surgery in pediatric patients. Laryngoscope 120:1730–1737, 2010

34. Teo C, Dornhoffer J, Hanna E, Bower C: Application of skull base techniques to pediatric neurosurgery. Childs Nerv Syst 15:103–109, 1999

35. Thirumala P, Lai D, Engh J, Habeych M, Crammond D, Balzer J: predictive value of somatosensory evoked potential monitoring during resection of intraparenchymal and intra-ventricular tumors using an endoscopic port. J Clin Neurol 9:244–251, 2013

36. Thirumala PD, Kassasm AB, Habeych M, Wichman K, Chang YF, Gardner P, et al: Somatosensory evoked potential monitoring during endoscopic endonasal approach to skull base surgery: analysis of observed changes. Neurosurgery 69 (1 Suppl Operative):ons64–ons76, 2011

37. Thirumala PD, Kodavatiganti HS, Habeych M, Wichman K, Chang YF, Gardner P, et al: Value of multimodality monitor-ing using brainstem auditory evoked potentials and somato-sensory evoked potentials in endoscopic endonasal surgery. Neurol Res 35:622–630, 2013

38. Thirumala PD, Mohanraj SK, Habeych M, Wichman K, Chang YF, Gardner P, et al: Value of free-run electromyo-graphic monitoring of extraocular cranial nerves during ex-panded endonasal surgery (EES) of the skull base. J Neurol Surg Rep 74:43–50, 2013

39. Thirumala PD, Mohanraj SK, Habeych M, Wichman K, Chang YF, Gardner P, et al: Value of free-run electromyo-graphic monitoring of lower cranial nerves in endoscopic endonasal approach to skull base surgeries. J Neurol Surg B Skull Base 73:236–244, 2012

40. Tzortzidis F, Elahi F, Wright D, Natarajan SK, Sekhar LN: Patient outcome at long-term follow-up after aggressive mi-crosurgical resection of cranial base chordomas. Neurosur-gery 59:230–237, 2006

41. Venkataramana NK, Anantheswar YN: Pediatric anterior skull base tumors: Our experience and review of literature. J Pediatr Neurosci 5:1–11, 2010

42. Winter V: Blood loss in pediatric neurosurgery. J Neurosurg Nurs 2:31–39, 1970

disclosureThe authors report no conflict of interest concerning the materi-als or methods used in this study or the findings specified in this paper.

author contributionsConception and design: Thirumala, Elangovan, Gardner, Snyder-man, Tyler-Kabara, Habeych, Crammond, Balzer. Acquisition of data: Thirumala, Gardner, Snyderman, Tyler-Kabara, Habeych, Crammond, Balzer. Analysis and interpretation of data: Thirum-ala, Elangovan, Singh. Drafting the article: Elangovan. Critically revising the article: Elangovan, Singh. Reviewed submitted ver-sion of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Thirumala. Statistical analy-sis: Thirumala, Elangovan. Administrative/technical/material sup-port: Thirumala, Gardner, Snyderman, Tyler-Kabara, Habeych, Crammond, Balzer. Study supervision: Thirumala.

correspondenceParthasarathy D. Thirumala, Center for Clinical Neurophysiology, Department of Neurological Surgery, UPMC Presbyterian, Ste. B-400, 200 Lothrop St., Pittsburgh, PA 15213. email: [email protected].


intraoperative neurophysiological monitoring during

Documents