Corticobulbar Motor Evoked Potentials: A Systematic Review of Technical Variations

Abstract

BACKGROUND AND OBJECTIVES:

Corticobulbar motor evoked potentials (coMEPs) were introduced in 2005 as an alternative to direct stimulation and free running electromyography for continuous monitoring of the facial nerve during skull base surgery. The technique has been expanded for monitoring of multiple cranial nerves (CNs) and adapted across diverse pathologies. Nonetheless, broad adoption has lagged. This study aimed to understand the basis of this disparity by systematically mapping current practices and characterizing the variations in technical implementation and reporting of coMEPs.

METHODS:

We conducted a Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 adherent systematic review of all articles reporting coMEP neuromonitoring published through December 6, 2023. Multiple reviewers screened publications for inclusion. Data were extracted using a standardized template to enable quantitative and qualitative data synthesis. Aggregate data were analyzed to address incidence, reporting variability, and sensitivity/specificity of the method.

RESULTS:

We identified 2711 unique cases using coMEPs, reported across 72 studies (2005-2023). Significant heterogeneity was observed in the technique of coMEPs, including stimulation montages and measurements of accuracy. Overall, 81.9% (59/72) of studies reported clinical outcomes, whereas the remainder were technically focused. coMEPs were often used alongside other neuromonitoring modalities (87.5% of studies, 63/72), but performance was directly compared only 25% of the time (18/72). The facial nerve was the most common single target analyzed (n = 41 publications), followed by the vagus (n = 9). Twenty-two percent of studies analyzed multiple CNs (n = 16). Less than 50% of facial nerve studies, and no vagal nerve study, reported sensitivity or specificity of coMEPs.

CONCLUSION:

coMEPs are an evolving technique in intraoperative neuromonitoring, powerful for their continuous monitoring of the entire motor pathways involving a CN. Comparison of the sensitivity and specificity of coMEPs to other modalities was limited by inconsistent reporting, highlighting the need for further investigation and prospective studies. Standardization of technique may facilitate broader adoption and acceptance of coMEPs.

ABBREVIATIONS:

AI

artificial intelligence

coMEPs

corticobulbar motor evoked potentials

EMG

electromyography

IONM

intraoperative neuromonitoring

MVD

microvascular decompression.

Intraoperative neuromonitoring (IONM) of cranial nerves (CNs) was developed in the 1960s to reduce facial nerve injury during vestibular schwannoma resection.¹ Direct or triggered electromyography (EMG) and free running EMG were the original monitoring modalities introduced. Triggered EMG allows the operator to locate the course of a nerve intraoperatively and assess nerve function on demand; it relies on direct stimulation, thereby providing intermittent feedback (Figure 1). Continuous triggered EMG was subsequently developed to address this limitation, although it still requires exposure of the nerve, which may be challenging in some cases.² Free-running EMG also provides continuous monitoring, however, it relies on irritation of the nerve and is subject to significant artifact. In addition, only select muscles may fire with any of these modalities, with the risk of false negatives.

FIGURE 1.

Differing pathways of EMG and corticobulbar motor evoked potential monitoring. Illustration of the differences in the stimulation location and monitored pathway between coMEPs (blue) and triggered EMG (red). coMEPs monitor the health of the entire neural pathway from motor cortex to neuromuscular junction, whereas triggered EMG monitoring can only distinguish injury distal to the point of stimulation. CN, cranial nerve; coMEPs, corticobulbar motor evoked potentials; EMG, electromyography. © 2026, Used with permission from Julia Wang.

Corticobulbar motor evoked potentials (coMEPs) were introduced in 2005^3,4 as an alternative to direct stimulation and free-running EMGs for facial nerve monitoring during skull base surgery. coMEPs enable continuous monitoring of facial functional integrity from the motor cortex to the neuromuscular junction, without the need for nerve exposure (Figure 1). Similar to transcranial motor evoked potentials (MEPs) for limb monitoring, coMEPs inform surgical decision making, particularly in eloquent areas such as the brainstem.⁵ With free-running EMG, the feedback signal relies on irritation of the nerve. Therefore a severed nerve may be indistinguishable from a healthy nerve, due to the lack of signal produced. With triggered EMG, neural injury can only be detected if stimulating proximal to the site of injury (Figure 2). Therefore, false negatives are a significant issue as the signal from a nerve distal to the site of injury is identical to a healthy nerve. The strength of coMEPs lies in their ability to monitor the entire pathway, allowing them to capture the decrease in nerve signal proportional to the extent of pathway injury at any point.

FIGURE 2.

Comparison of EMG and corticobulbar motor evoked potential signals. Comparative neural signal response tracings between spontaneous EMG, triggered EMG, and corticobulbar motor evoked potentials demonstrating the differential response in the healthy, irritated, injured, and severed nerve. EMG, electromyography; MEP, motor evoked potential. © 2026, Used with permission from Julia Wang.

Since their introduction, coMEPs have been expanded for monitoring of CNs V, VII, IX, X, XI, and XII, although widespread adoption has been limited by surgeon experience and inconsistent outcomes. We hypothesize that these inconsistencies are largely due to variations in technical implementation. This study reviews the current use of coMEPs in skull base surgery, their evolving applications, technical variations, and future potential.

METHODS

Study Selection

We conducted a systematic review using Covidence to include all articles reporting coMEP neuromonitoring published in Pubmed, Embase, and the Web of Science from inception through December 6, 2023 using the following search terms: [corticobulbar OR (cranial AND nerve*) OR (facial AND nerve) OR (oculomotor AND nerve) OR (mandibular AND nerve) OR (vagus AND nerve) OR (accessory AND nerve) OR (spinal AND accessory AND nerve) OR (glossopharyngeal AND nerve)] AND [motor AND evoked AND potential* OR (“motor evoked” AND potential) OR MEP OR cbMEP OR coMEP] AND (operative OR surgery OR neurosurgery). Articles were included in the analysis if they reported any application of coMEP for monitoring of any CN for any cranial or cervical spine pathology and described electrode placement, nerves and muscles studied, and stimulation parameters used. Multiple study designs were permitted, including randomized controlled trials, case-control, cohort, cross-sectional, case series, and case reports. Studies not available in English, animal studies, nonoriginal investigations (reviews, editorials, letters), investigations in which only abstracts were available, and nonpeer reviewed materials were excluded. Each article was manually screened by at least 2 of 4 independent reviewers (C.A.J., R.V.P., M.B., T.M.) in 2 rounds: first by abstracts, then by full text (Figure 3). Discrepancies were resolved by a third reviewer (W.L.B.). This study adhered to the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for systematic reviews. This study was not registered and no protocol was prepared.

FIGURE 3.

Preferred reporting items for systematic reviews and meta-analyses study selection flow diagram. Flow diagram depicting the process of study identification, screening, eligibility assessment, and final inclusion in the qualitative synthesis.

Data Extraction

Our primary outcomes of interest included methodology of implementation as well as sensitivity and specificity of the neuromonitoring modality. We extracted data on study design, country of investigation, pathology, patient age range, and the number of cases using coMEPs. We also recorded the terminology used for coMEPs in each study, as well as the CNs and muscles monitored, and comparative data with other neuromonitoring modalities. Technical data regarding stimulation methods (electrode placement, type of electrodes, whether the electrodes were insulated or noninsulated), stimulation parameters (type, filter settings, pulse duration, train length, interstimulus interval, constant current or constant voltage, and the use of fixed or variable parameters), method of artifact detection, double train application and timing, and equipment used were documented. Anesthetic use, outcome measures, and follow-up time were collected when available, including intraoperative responses, postoperative function, amplitude decrease to deficit, sensitivity, specificity, monitorability rate (defined as proportion of patients in whom an interpretable coMEP signal could be recorded at baseline), and false-positive/negative rates. Studies that did not include clear reporting of CNs monitored or outcome measures were not included in the quantitative synthesis of metrics.

Statistical Analysis

We analyzed aggregate data to assess the incidence, reporting variability, and sensitivity/specificity of coMEPs. Statistical analysis and data visualization were conducted using Excel v16.103.2 (Microsoft Excel) and R version 4.3.2, R studio 2024.12.0 (The R Foundation).

RESULTS

Evolution of coMEPs

Experience with coMEPs has grown in the 2 decades since its introduction for facial nerve monitoring (Figure 4A). Reflective of its origins, the most common coMEP application was in cerebellopontine angle surgeries, primarily for vestibular schwannoma (37/72 studies, 52%), followed by microvascular decompression (MVD) and other skull base pathologies (Figure 4B). Only 4% of studies focused on brainstem pathologies, with the remaining 18% not specified or including other pathologies not otherwise categorized.

FIGURE 4.

Cohort overview. A, Year of publication for included studies on corticobulbar motor evoked potentials. B, Distribution of surgical locations and procedures in which any corticobulbar motor evoked potential monitoring was used. C, Number of patients included per study. D, Frequency heatmap of study location. E, Cranial nerves that were monitored across studies, with percentages reflecting the proportion of included studies monitoring that cranial nerve. Percentages sum to >100, reflecting that some studies included multiple nerves in monitoring. CP, cerebellopontine; MVD, microvascular decompression.

Of 3078 cases using corticobulbar MEPs, reported in 72 studies from 2005 to 2023 (Table 1), at least 2711 were unique, with the remaining 367 cases from groups with multiple publications that could not be verified as distinct. Most studies (91.7%) were case reports or case series (median 26 cases, range 1-367 cases) (Figure 4C), and only one study identified was prospectively designed. coMEP studies largely came from a small group of select countries, led by Japan (25%), followed by China (11.1%), the US (9.7%) and Germany (9.7%) (Figure 4D). A minority of studies (13.9%, n = 10) included pediatric patients (younger than18 years), and only one focused exclusively on this group. The facial nerve was the most commonly analyzed (included in 70.8% of studies, and the exclusive focus of 56.9% of studies), followed by the vagus nerve (included in 27.8% of studies, exclusive focus of 12.5% of studies), with a small number focusing on hypoglossal and glossopharyngeal pathways, indicating incipient experience in the field (Figure 4E). Studies monitored up to 6 CNs concurrently, with the majority monitoring just a single nerve (Figure 5A). Most CNs were monitored by a single muscle, although studies of the facial nerve frequently monitored 2 to 3 muscles, with some studies monitoring up to 5 (Figure 5B).

TABLE 1.

Comprehensive Table of Studies Included in the Systematic Review

Source	Journal	Country of origin	Study design	Pathology	Patient age range (y)	No. of cases involving coMEPs	Cranial nerves used for recording	Muscles studied
Deletis et al6	Clin Neurophysiol	USA	Case control	Brain surgery, NOS	27-68	9	10	Cricothyroid
Fernandez Conejero et al7	Clin Neurophysiol	Spain	Case series	Vestibular schwannoma, posterior fossa meningioma, clivus chordoma, other posterior fossa pathologies	35-61	100	7	Orbicularis oris, orbicularis oculi, mentalis, nasalis
Rogic Vidakovic et al8	Clin Neurophysiol	Croatia	Case control	Left craniotomy for tumor (frontal, temporal, or parietal)	20-68	15	10	Cricothyroid
Kim et al9	Clin Neurophysiol Pract	South Korea	Case series	Cervical spine surgeries	22-79	25	9	Tongue
Bovo et al10	J Neurol Surg A Cent Eur Neurosurg	Switzerland	Case series	Vestibular schwannoma, other CPA pathologies	42-72	33	7	Mentalis, orbicularis oris, nasalis
Tawfik et al11	Ann Otol Rhinol Laryngol	USA	Cohort study	Vestibular schwannoma	38-63	20	7	Orbicularis oris
Tellez et al12	Clin Neurophysiol	USA	Case series	CPA and brainstem pathologies	4-77	41	10	Masseter, facial, pharyngeal, vocal fold, trapezius, tongue
Della Pepa et al13	Neurosurgery	Italy	Case series	CPA extra-axial tumors	—	83	7	Orbicularis oris, orbicularis oculi
Zhang et al14	J Clin Neurosci	China	Case series	Large and giant aneurysms of the anterior circulation	50.0 ± 13.2	21	7	Orbicularis oculi
Wilkinson and Kaufmann15	Can J Neurol Sci	Canada	Case series	Hemifacial spasm or trigeminal neuralgia MVD	—	27	7, 8	Orbicularis oculi, orbicularis oris, mentalis
Hiruta et al16	Clin Neurophysiol	Japan	Case series	CPA tumors	18-80	62	7	Orbicularis oris
Greve et al17	J Clin Monit Comput	Germany	Case series	CPA tumor surgery: vestibula schwannoma, meningioma, other	22-78	79	7	Orbicularis oculi, orbicularis oris, mentalis muscles
Tokimura et al18	Neurosurg Rev	Japan	Case control	Vestibular schwannoma	27-75	20	7	Orbicularis oculi, orbicularis oris
Ling et al19	World Neurosurg	China	Case series	Vestibular schwannoma	22-65	106	7, 5	Masseter, orbicularis oris, orbicularis oculi, mentalis
Fernandez Conejero et al20	Clin Neurophysiol	Spain	Case series	Hemifacial spasm MVD	—	17	7	Orbicularis oris
Fukuda et al21	J Neurol Neurosurg Psychiatry	Japan	Case series	Hemifacial spasm MVD	25-70	25	7	Orbicularis oculi, orbicularis oris, mentalis
Fukuda et al22	World Neurosurg	Japan	Case series	Skull base tumors	50.9 ± 21.8	36	9, 10	Pharyngeal
Ichino et al23	J Anesth	Japan	Case series	Thyroid surgery	56 ± 16	54	10	Laryngeal
Wilkinson et al24	J Neurosurg	Canada	Case series	Hemifacial spasm MVD	18-75	31	7	Orbicularis oculi, orbicularis oris, mentalis
Ito et al25	J Neurosurg	Japan	Case series	Skull base or brainstem tumors	19-69	15	10	Laryngeal
Matthies et al26	Clin Neurol Neurosurg	Germany	Case series	CPA tumors, trigeminal neuralgia	10-76	57	7	Orbicularis oculi, orbicularis oris
Song et al27	Medicine (Baltimore)	China	Case series	CPA tumors	51.4 ± 11.4	97	7	Mentalis
Bozinov et al28	World Neurosurg	Switzerland	Case series	Multiple tumor pathologies	0.5-15	24	7	Orbicularis oculi, orbicularis oris, mentalis
Sarnthein et al29	Clin Neurophysiol	Switzerland	Case series	Vascular and tumor indications	62 ± 19	18	12	Tongue
Martucci et al30	Children (Basel)	Italy	Case control	Thyroid surgery (hemi or total)	3.9	13	10 (recurrent laryngeal)	Laryngeal
Bhimrao et al31	Otol Neurotol	Canada	Case series	Vestibular schwannoma	13-81	367	7	Orbicularis oris
Dong et al4	Clin Neurophysiol	Canada	Case series	CPA tumors, skull base tumors	5-77	76	7	Orbicularis oris
Slotty et al32	J Neurosurg	Japan	Case series	Posterior fossa surgery	50.6 ± 14.1	305	7, 10	Facial muscles (unspecified), pharyngeal
Ilgezdi-Kaya et al33	J Clin Monit Comput	Turkey	Case report	Vagal nerve schwannoma	36	1	10	Vocalis
Tanaka et al34	Cureus	Japan	Case report	Vagal nerve schwannoma	42	1	10	Laryngeal
Fukuda et al35	Surg Neurol Int	Japan	Case series	Skull base tumors	-	10	9,10	Pharyngeal, laryngeal
Greve et al36	Cancers (Basel)	Germany	Case series	Vestibular schwannoma	21.4-83.5	60	7	Orbicularis oculi, orbicularis oris, mentalis
D'Alessandris et al37	Diagnostics (Basel)	Italy	Case series	CPA tumors	>65	35	7	Orbicularis oculi, orbicularis oris
Verst et al38	Acta Neurochir (Wien)	Brazil	Case series	CPA tumors	—	23	7	Orbicularis oculi, mentalis
Fukuda et al39	Neurol Res	Japan	Case series	Skull base tumors	—	37	7	Orbicularis oculi, orbicularis oris
Acioly et al40	Acta Neurochir (Wien)	Germany	Case series	Vestibular schwannoma	15-69	35	7	Orbicularis oculi, orbicularis oris
Goto et al41	Skull Base	Japan	Case series	Skull base tumors, largely vestibular schwannoma	—	26	7	Frontalis, orbicularis oculi, orbicularis oris
Akagami et al3	Neurosurgery	Canada	Case series	Skull base tumors, largely vestibular schwannoma	18-77	71	7	Orbicularis oris
Deletis et al42	Clin Neurophysiol	USA	Case series	Multiple pathologies	21-83	22	10	Laryngeal
Hendriks et al43	Acta Neurochir (Wien)	The Netherlands	Case series	CPA tumor surgery	15-77	39	7,5	Orbicularis oculi, orbicularis oris, nasalis, mentalis, masseter
Fukuda et al44	J Neurol Neurosurg Psychiatry	Japan	Case series	Skull base tumors	—	26	7	Orbicularis oculi, orbicularis oris
Aydinlar et al45	J Clin Neurophysiol	Turkey	Case series	Skull base tumors	4-83	40	7	Masseter, temporalis, orbicularis oculi, nasalis, mentalis
Liu et al46	Chin Med J (Engl)	China	Case series	Large vestibular schwannoma (>4 cm)	38-69	19	7	Orbicularis oculi, orbicularis oris
Acioly et al47	Neurosurgery	Germany	Case series	CPA tumor surgery	5-71	60	7	Orbicularis oculi and oris
Sarnthein et al48	Clin Neurophysiol	Switzerland	Case series	Tumor, vascular	49 ± 21	21	7	Orbicularis oculi, orbicularis oris, nasalis, mentalis
Motoyama et al49	Surg Neurol Int	Japan	Case series	Glossopharyngeal neuralgia MVD	71-79	3	7, 9, 10	Orbicularis oculi, orbicularis oris, pharyngeal, false vocal cord
Wilkinson and Kaufmann50	J Neurosurg	Canada	Case series	Hemifacial spasm or trigeminal neuralgia MVD	—	27	7	Orbicularis oculi, orbicularis oris, mentalis
Kullmann et al51	World Neurosurg	Germany	Case series	Posterior fossa surgery	49 ± 15	63	9, 10, 12	Pharyngeal, tongue
Liu et al52	Clin Neurol Neurosurg	China	Case series	Vestibular schwannoma	19-76	106	7	Orbicularis oculi, orbicularis oris, frontalis
Chen et al53	J Clin Neurophysiol	China	Case series	Skull base or brainstem tumors	—	51	9, 10, 12	Pharyngeal wall, cricothyroid, tongue
Hsu et al54	J Chin Med Assoc	Taiwan	Case series	Hemifacial spasm MVD	28-80	69	7	Orbicularis oculi, mentalis
Sarnthein et al55	Neurosurg Focus	Switzerland	Case series	CPA tumors, trigeminal neuralgia, vascular pathologies	20-84	34	7	Orbicularis oculi, orbicularis oris, nasalis, mentalis
Verst et al56	Acta Neurochir (Wien)	Brazil	Case series	Tumors or cervical/thoracic arthrodesis	21-72	23	7	Mentalis
Fukuda et al57	Acta Neurochir (Wien)	Japan	Case series	Hemifacial spasm MVD	35-73	26	7	Orbicularis oculi, orbicularis oris, mentalis
Bonda et al58	Cureus	USA	Case series	Brain tumors	20-79	7	7	Vocalis, cricothyroid, orbicularis oris, mentalis, tongue
Dalvandi et al59	Eur J Transl Myol	Iran	Case series	CPA tumors	—	25	5, 7, 9, 10, 11	Masseter, frontalis, zygomaticus, buccalis, mentalis, soft palate, vocal cord, trapezius
Xu et al60	Neuro Endocrinol Lett	China	Case series	Acoustics	48 ± 9	29	7	Orbicularis oculi, orbicularis oris, mentalis
Pavoni et al61	J Clin Anesth	Italy	Case series	Thyroid surgery	38 ± 22	20	9	Laryngeal adductors
Mariappan et al62	Journal of Neuroanesthesiology and Critical Care	India	Case report	CPA tumors	45	1	7	Not specified
Smolanka et al63	International Journal of Biology and Biomedical Engineering	Ukraine	Case series	Brainstem tumors and cavernous malformations	—	42	7, 9	Not specified
Hardian et al64	Journal of Neurosurgery	Japan	Case series	Pontine cavernous malformations	18-57	10	7	Frontalis, orbicularis oculi, orbicularis oris
Matsuoka et al65	Journal of Neurosurgery: Spine	Japan	Case series	Spine surgery	43-83	23	7	Orbicularis oris, orbicularis oculi
Romagna et al66	Neurosurgery	Germany	Case series	CPA surgery	34-75	17	5, 7, 9, 10, 11, 12	Vocal cords, soft palate around the uvula, masseter, orbicularis oculi, orbicularis oris, trapezius, tongue
Fukuda et al67	Journal of Neurosurgery	Japan	Case series	Skull base tumors	17-76	22	9, 10	Pharyngeal
Cosetti et al68	Journal of Neurological Surgery, Part B: Skull Base	USA	Case series	CPA tumors	21-73	52	7	Orbicularis oris
Osburn et al69	Neurosurgery	USA	Case report	Hemilingual spasm	52	1	12	Tongue
Kim et al70	Annals of Rehabilitation Medicine	South Korea	Case series	Epidermoid cyst	33-38	2	12	Tongue
Sun et al71	International Journal of Clinical and Experimental Medicine	China	Case series	Tumors	34 ± 5	30	7	Not specified
Taskiran and Küçükyürük72	Istanbul Medical Journal	Turkey	Case series	Skull base tumors	3-72	45	6, 7, 9, 12	Not specified
Kojima et al73	World Neurosurg	Japan	Case series	Extracranial ICA surgery	41-79	8	12	Tongue
Della Pepa et al74	Journal of Neurosurgery	Italy	Case series	Bourth ventricle surgery	35.8 ± 18.8	21	5, 7, 9, 10, 11, 12	Masseter, orbicularis oculi, orbicularis oris, trapezius, tongue, laryngeal
Lapteva et al75	World Neurosurg	Russia	Prospective nonrandomized cohort study	Brainstem and fourth ventricular surgery	42 ± 18	46	9, 10	Soft palate

coMEPs, corticobulbar motor evoked potentials; CPA, cerebellopontine angle; MVD, microvascular decompression; NOS, not otherwise specified.

FIGURE 5.

Cranial nerves and muscles monitored. A, Distribution of concurrent cranial nerves monitored, with monitoring of one cranial nerve being the most frequently studied. B, Number of muscles monitored across the cranial nerves, with only cranial nerve VII being monitored at more than 2 muscles. CNs, cranial nerves.

Correlation of monitoring changes to clinically significant deficits was inconsistently reported. In all, 81.9% of studies (59/72) reported clinical outcomes, of which 48 studies reported some new postoperative deficit. Forty-four studies reported some deficits correlating with monitoring changes, and 27 reported some deficits that did not correlate with monitoring changes. Thirty-three studies reported changes in monitoring that did not result in a new postoperative deficit. Four studies reported no change in intraoperative coMEP corresponding to no postoperative deficits.

Follow-up varied widely, with 77.8% (56 studies) reporting follow-up, ranging from 0 to 30 months (average 6.6 months). Of these, 35 studies included multiple follow-up time points, whereas 18 reported only immediate postoperative follow-up or follow-up of unclear timing, and 3 studies reported just one delayed time point. The second follow-up ranged from discharge or 1 to 2 weeks postoperative up to 30 months.

Signal changes, reported in 26/41 of studies focused on the facial nerve, were measured in different ways across studies. The most common harbinger of facial nerve injury was amplitude reduction (18 studies), with alert criteria ranging from 30% to 75% decrease. Other indicators included increased stimulation thresholds (5 studies) and changes in the final-to-baseline ratio (3 studies). Of those 18 studies that relied on amplitude changes, a majority (61%, n = 11) applied a 50% amplitude decrease threshold to signal postoperative deficit.

Nearly all studies used at least one other modality of facial nerve monitoring (90.2%, 37/41), most of which included facial nerve EMG, in addition to coMEPs (68.3%, 28/41). However, only 24.4% (10/41) of studies used a direct comparator group, most frequently lateral spread response (4/10) or the blink reflex (3/10). Notably, lateral spread response was used only in studies of MVD for hemifacial spasm. Of the 26 studies reporting coMEP changes, 6 compared coMEP changes with traditional IONM changes associated with facial nerve injury.

Specificity or sensitivity was reported in less than half of the monitoring studies focused on the facial nerve (20/41). For orbicularis oris, sensitivity ranged from 53.8% to 100% (median 82.0%, IQR 67.25-90.5)^{3,4,11,13,16,31,47} and specificity ranged from 60% to 89.7% (median 84%, IQR 70.83-88.93).^{3,4,11,13,16,31,40,47} Median orbicularis oculi sensitivity was 41.7% (IQR 40.85-53.6),^13,14,47 with higher specificity (median 93.15%, IQR 90.48-95.5).^13,14,40,47 Mentalis sensitivity (91.3%) and specificity (98.6%) were reported in only one study.²⁷ Median combined sensitivity and specificity for multiple facial muscles (orbicularis oris, oculi, mentalis, and nasalis) was 82.9% (IQR 69.83-97.25) and 82.0% (IQR 78.0-99.0), respectively (Figure 6).^{7,10,28,36-38,45,54,55} Median monitorability was 94% (IQR 87.0-100) across the 26 studies that reported.

FIGURE 6.

Performance of facial nerve monitoring. Sensitivity and specificity of facial nerve monitoring stratified by muscle used for monitoring.

Vagus Nerve Monitoring With Corticobulbar MEPs

Nine studies focused on vagus nerve (CN X) coMEP monitoring, comprising 171 cases. The majority were for intracranial pathologies, predominantly skull base and brainstem conditions. Vagal coMEPs were also commonly applied for extracranial thyroid operations (67/171 cases, 2 studies). Vagal nerve function was assessed by monitoring the vocalis (5 studies) or cricothyroid (3 studies) muscles, or both (1 study), and was not specified in one study. Concurrent monitoring with the laryngeal adductor reflex was reported in 30% of studies.

Anesthetic Approach

Most of the included studies (70.8%, 51/72) report a total intravenous anesthetic approach, predominantly consisting of propofol with additional intravenous narcotic medications. These agents have mild to moderate, dose-dependent effects on MEP monitoring and are viewed as the standard-of-care anesthetic approach during motor monitoring.⁷⁶ Seven studies (9.7%, 7/72), however, report the addition of an inhaled halogenated anesthetic to this regimen,^{14,15,19,24,46,50,60} whereas 2 report administration of nitrous oxide.^41,64 These agents produce marked depression of MEP signals, and thus, their confounding effect must be considered when interpreting the associated studies.⁷⁶ Twelve studies do not report the anesthetic regimen used, and interpretation of these results should be contextualized in the absence of these data.^{6,11,13,26,30,33,37,38,43,58,59,63} Regarding neuromuscular blocking agents, only one study reported the use of paralytic beyond induction,⁷¹ which may potentially confound observed signals.

Technical Variations in coMEP Methodology

Technical practices varied widely, from electrode placement, to electrode type, to stimulation parameters. Stimulating electrodes on the scalp were placed hemispherically (50/72 studies) more than interhemispherically (14/72 studies), with the cathode at Cz and the anode at C3 (left) or C4 (right). Placement was not specified in 8 studies. For studies of the facial nerve specifically, electrode placement was mostly hemispheric (85.3%, 35/41), with 4 studies using interhemispheric placement (9.8%), 1 study using both (2.4%), and 1 study which did not report stimulating montage (2.4%). In studies of the facial nerve, corkscrew electrodes were used most often for stimulation (75.6%, 31/41), with several studies using subdermal needles (9.8%, 4/41), one study using PEG electrodes (2.4%), and 4 studies in which stimulating electrode choice was not reported (9.8%) (Figure 7A). Orbicularis oris and orbicularis oculi were the most commonly selected muscles for monitoring of the facial nerve (Figure 7B).

FIGURE 7.

Specifics of facial nerve stimulation and recording. A, Characteristics of stimulating electrodes and stimulation montage configurations used for facial nerve monitoring across included studies. B, Distribution of facial muscles monitored to assess facial nerve function, shown relative to the total number of muscles recorded simultaneously. Only studies that explicitly specified the recorded muscles were included in this analysis. N/A, not applicable.

For the facial nerve, stimulation was fixed 41.5% (17/41) of the time, variable 53.7% (22/41) of the time, and not reported in 2 studies (4.9%). Reported pulse width ranged from 50 to 500 microseconds, and train lengths varied from 1 to 8 pulses, with 5 pulses being most common (Figure 8A). The stimulation paradigm applied constant voltage in 41/72 studies, and constant current in 23/72 studies overall, and for studies of just the facial nerve, constant voltage was used in 26/41 and constant current in 12/41. The choice of constant current or constant voltage stimulation varied by equipment used (Figure 8B). Inomed (Inomed Inc.) and Nihon Kohden (Nihon Kohden Corporation) equipment were associated with constant current protocols; Cadwell (Cadwell Industries, Inc.), Natus, and Nicolet (Natus Medical Incorporated) equipment were associated with constant voltage; and Axon (Axon Systems) and Medtronic (Medtronic) used both paradigms.

FIGURE 8.

Technical features of facial nerve monitoring. A, Distribution of pulse durations reported across studies and the stimulation parameters used. B, Monitoring equipment used for facial nerve monitoring, stratified by stimulation mode (constant current vs constant voltage). Equipment was grouped by manufacturer, with subsidiaries consolidated under the parent company. C, Proportion of studies using double train stimulation during facial nerve monitoring, along with the distribution of reported intertrain timing intervals.

Artifact detection as a metric of coMEP reliability also varied in practice. Single pulse response was the most common metric of false-negative direct transmission of calvarial electrical stimulation to the peripheral muscles instead of through brain neural pathways. Peripheral response was noted in two-thirds (65.9%, 27/41) of studies of facial nerve monitoring and one-third of studies (31.3%, 5/16) assessing lower CN coMEPs. Strikingly, the method of delivering and measuring the peripheral single pulse response also varied widely across the facial nerve literature, with a double train delivered in only 34.1% (14/41) of studies (not delivered in 24.4% and not reported in 41.5%), and the timing of pulse delivery ranging from 40 ms before to 90 ms after the primary train (Figure 8C). Latency also functioned as a metric of artifact detection, with shorter latencies (<10 ms) being attributed to activation of the facial nerve by extracranial current spread, and was used in another 12.2% of studies (5/41).

DISCUSSION

Corticobulbar MEPs are rapidly broadening in adoption, from their original application for facial nerve monitoring during vestibular schwannoma surgery. However, widespread adoption remains limited by significant variation in methodology and inconsistent correlation with outcome. We present an in-depth review of operative uses, limitations, and opportunities for improvement.

Practice patterns in coMEP monitoring technique varied in almost every detail. This included the type and number of muscles monitored, electrode type and placement, stimulus parameters, and artifact detection techniques. Unsurprisingly, these methodological variations contribute to heterogeneity in results achieved. Less than half of facial nerve studies quantified monitoring accuracy, and those that did showed wide discrepancies due to technical differences. Peripheral response artifacts, typically assessed through the presence or absence of a single pulse response, were inconsistently captured in addition to being delivered and measured in different ways. Latency was also used as an alternative method to identify artifacts of extracranial stimulation. Most of the facial nerve studies used hemispheric scalp electrode placement, which has been demonstrated to be more reliable for CN VII responses. In comparison, studies examining vagal coMEPs favor interhemispheric stimulation over hemispheric stimulation.^34,61 Notably, the choice between constant current and constant voltage stimulation depended on the equipment used, although this did not seem to affect outcomes.

Variability in sensitivity and specificity of different muscles further complicates the reliability of coMEP data.^13,40,47 Previous studies have contended that orbicularis oris recording tends to be more reliable than orbicularis oculi because of reduced susceptibility to peripheral artifact, although we did not find that to be substantiated in our aggregate data.^13,40,47 Moreover, in our review, only 65.9% of facial nerve studies and 31.3% of lower CN studies used single pulse responses to control for peripheral stimulation. Susceptibility to artifact could be mitigated by monitoring multiple muscle groups, which is uncommon across practices because most of the studies focused on facial nerve coMEPs monitored only one (24.4%) or 2 (31.7%) muscles. Monitoring more muscles on a regular basis could improve coMEP reliability.

Reporting of clinical follow-up and outcome, whether immediate or delayed, was also inconsistent. Nearly a quarter of studies reported no follow-up, and 18% lacked outcome measures, often because of their technical focus. This makes it difficult to interpret the long-term predictive value of coMEPs, as immediate postoperative deficits may not reflect permanent damage. Furthermore, although 90.2% of studies of the facial nerve reported using other monitoring modalities alongside coMEPs, less than a quarter of those studies directly compared coMEPs with alternative techniques, leaving a gap in understanding its relative merits. Future studies including immediate and long-term outcomes may better guide the predictive accuracy of coMEPs and improve clinical decision making.

Despite these methodological challenges, coMEPs are associated with improved preservation of facial nerve integrity,⁶⁰ and prediction of outcome.¹³ A meta-analysis of stringently selected studies reported sensitivity (0.81) and specificity (0.94) for coMEPs as a measure of facial nerve function in cerebellopontine angle tumor surgery,⁷⁷ which exceeded the averages found in our review. Importantly, they noted that sensitivity and specificity across studies has improved over time, likely because of refinements in technique. However, some studies suggest that coMEPs may not always influence patient outcomes, despite its accuracy in predicting facial nerve function, potentially due to inconsistencies in how intraoperative data are used by the surgical team.¹¹ Although many studies cite a 50% decrease in amplitude as indicative of potential nerve damage,^{4,11,39,40,77} there is no consensus on a standardized alert threshold. Establishing clear, evidence-based criteria for intraoperative alert will be essential for coMEPs to transition from a purely predictive tool to a preventative one, enabling earlier intervention and potentially improving patient outcomes.

We found that 43% of studies have expanded to include or focus on CNs other than the facial nerve, signaling growing interest in broadening utilization of coMEPs. This trend is reflected in the heterogeneous terminology used to refer to coMEPs, including “PharyngealMEPs”, “facial motor evoked potentials”, and “VagalMEPs”, indicating the diverse ways coMEPs are being applied. In addition to monitoring during tumor resections, coMEPs are increasingly used in MVD for trigeminal neuralgia, hemifacial spasm, and even hemilingual spasm.^21,57,69 Nonetheless, the overall number of cases being monitored with coMEPs remains small, with most studies reporting fewer than 100 cases.

coMEP monitoring remains underused in pediatric populations, with only 13.9% of studies including children. Interestingly, a pediatric-specific study found that coMEP thresholds in children were similar to those reported in adults, suggesting that coMEPs can be safely and effectively used in pediatric patients without modification.²⁸ This represents a significant area of potential growth in coMEP application, given the enrichment of posterior fossa and brainstem pathologies in the pediatric population.

Artificial intelligence (AI) is increasingly embedded in medical research and innovation and is well positioned to augment pattern recognition in IONM, while improving access to and standardization of these tools. Emerging studies demonstrate the feasibility of neural network-based MEP detection and the use of AI to identify clinically meaningful signal features that enhance accuracy and reliability, in addition to furthering research in the field.^78,79 Continued, iterative integration of AI into signal acquisition, quality control, and interpretation has the potential to strengthen performance and accelerate broader adoption of advanced IONM platforms.

We have summarized our findings of the most commonly used approaches and techniques, as well as our recommendations for optimal performance of coMEP monitoring in Table 2. In an effort to increase transparency and consistency, we have created a recommended reporting checklist for future coMEP studies to facilitate standardized high quality data reporting in future publications (Table 3). Continuing to aggregate and analyze our collective experience with this technique is critical to increasing implementation and reliability.

TABLE 2.

Summary of Current Practices and Recommendations for coMEP Monitoring by CN

Topic	Reported in the literature	Most commonly reported	Recommendation
Anesthetic used	TIVA/mixed anesthetic	TIVA	TIVA
CNs monitored	V, VII, IX, X, XI, XII	VII and X	V, VII, IX, X, XI, XII
Muscles monitored	V: Masseter VII: Oculi, nasalis, oris, and mentalis IX: Stylopharyngeus/soft palate X: Vocalis and cricothyroid XI: Trapezius XII: Tongue	V: Masseter VII: Oris and mentalis (most common), variable across studies IX: Stylopharyngeus/soft palate X: Vocalis XII: Tongue	V: Masseter and Temporalis VII: Frontalis, oculi, nasalis, oris, and mentalis IX: Stylopharyngeus/soft palate X: Vocalis/cricothyroid if recording endotracheal tube contraindicated (eg, MRI cases) XI: Trapezius XII: Tongue
Monitorability	V, VII, IX, X, XI and XII	VII, X, and XII	V, VII, IX, X, XI, and XII
Stimulating montage	Interhemispheric Hemispheric	VII: Hemispheric X: Interhemispheric XII: Hemispheric	Hemispheric for all CNs, with option of interhemispheric when monitoring CN X and CN XI
Stimulation parameters (fixed vs variable)	Fixed and variable	No consensus	Variable
Pulse width of stimulation	Constant current: 300-500 Constant voltage: 50-75	Constant current: 500 Constant voltage: 50-75	Constant current: 300-500 Constant voltage: 75
Stimulating electrodes	Corkscrews Subdermal needles Peg electrode	Corkscrews	Corkscrews or subdermal needles
Recording electrodes	Subdermal needles noninsulated or insulated	Noninsulated electrodes (likely due to availability of insulated electrodes)	Insulated (noninsulated electrodes were used in most studies, however use of insulated recording electrodes has been shown to improve signal quality and reliability7,80)
Equipment used (constant current vs constant voltage)	Constant current: Inomed and Nihon Kohden Constant voltage: Cadwell, Natus, and Nicolet Both: Axon and Medtronic	Dependent on equipment available	coMEP monitoring is possible with both CC or CV, however technique must be adjusted to accommodate system's abilities
Timing of double train, if applied	Ranging from 40 ms before stimulating train to 90 ms after stimulating train	No consensus	At least 90 ms after stimulating train to allow for neural recovery from prior stimulation6,81
Alert criteria used and method of signal change (decrease in amplitude vs increase in stimulation threshold)	Amplitude decrease or stimulation threshold increase	Most commonly 50% decrease in amplitude	>50% decrease in amplitude indicating possible weakness >70% decrease in amplitude indicating severe weakness 100% loss of signal indicates paralysis
Timing of outcome assessment	Immediate or “delayed”	No consensus	Posttreatment: immediate, 1-2 wk, 2-3 mo, 6 mo, 1 y
Follow-up duration	No consensus	No consensus	At least 1 y

CC, constant current; CN, cranial nerve; CV, constant voltage; coMEPs, corticobulbar motor evoked potentials; TIVA, total intravenous anesthetic approach.

TABLE 3.

Recommended Reporting Checklist for Future Publications

Recommended reporting checklist
Anesthetic used
Cranial nerve(s) monitored
Muscles monitored
Type of stimulating electrode used, including insulation (yes/no)
Stimulating montage used
Monitorability rate
Equipment used (constant current vs constant voltage)
Stimulation parameters used (fixed vs variable)
Pulse width of stimulation used
Single pulse artifact testing (application of a double train)
Timing of double train, if applied
Alert criteria used and method of signal change measurement (decrease in amplitude vs increase in stimulation threshold)
Clinical outcome
Timing of outcome assessment
Follow-up duration

Limitations

Our study has several limitations, primarily due to the inherent constraints of systematic reviews. Variability in study design, reporting practices, and data quality made direct comparisons difficult. Although we aimed to capture the full scope of coMEP applications, many studies lacked detailed follow-up or quantitative data. Owing to the heterogeneity in the data, we have reported median values and interquartile ranges to reflect observed ranges and distributions. These values do not constitute pooled estimates. Nearly all studies were retrospective in design. Given the lack of prospective data, conclusions regarding impact on outcomes remains limited.

CONCLUSION

coMEP monitoring offers significant potential to improve the continuous monitoring of CN pathways during surgery. Refinement of techniques is needed to achieve improved accuracy and reliability, standardized alert criteria, and enhanced consistency for artifact detection. The field would benefit from standardization of methodologies and more consistent clinical outcome reporting to encourage increased utilization and fully realize the potential of coMEPs in neurosurgery.

Funding

This study did not receive any funding or financial support.

Disclosures

The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.