Basic Principles and Recent Trends of Transcranial Motor Evoked Potentials in Intraoperative Neurophysiologic Monitoring

Abstract

Transcranial motor evoked potentials (TcMEPs), which are muscle action potentials elicited by transcranial brain stimulation, have been the most popular method for the last decade to monitor the functional integrity of the motor system during surgery. It was originally difficult to record reliable and reproducible potentials under general anesthesia, especially when inhalation-based anesthetic agents that suppressed the firing of anterior horn neurons were used. Advances in anesthesia, including the introduction of intravenous anesthetic agents, and progress in stimulation techniques, including the use of pulse trains, improved the reliability and reproducibility of TcMEP responses. However, TcMEPs are much smaller in amplitude compared with compound muscle action potentials evoked by maximal peripheral nerve stimulation, and vary from one trial to another in clinical practice, suggesting that only a limited number of spinal motor neurons innervating the target muscle are excited in anesthetized patients. Therefore, reliable interpretation of the critical changes in TcMEPs remains difficult and controversial. Additionally, false negative cases have been occasionally encountered. Recently, several facilitative techniques using central or peripheral stimuli, preceding transcranial electrical stimulation, have been employed to achieve sufficient depolarization of motor neurons and augment TcMEP responses. These techniques might have potentials to improve the reliability of intraoperative motor pathway monitoring using TcMEPs.

Introduction

It is essential to monitor motor function during spinal surgery that may damage the spinal cord or the spinal nerve roots. Transcranial motor evoked potentials (TcMEPs), which are muscle action potentials elicited by transcranial brain stimulation, have been the most popular method for the last decade to monitor the functional integrity of the motor system during surgery. In 1980, Merton and Morton discovered that it was possible to stimulate the human brain through the intact scalp using a high-voltage single electrical stimulus, and that TcMEPs could be recorded from the limb muscles.^1,2) Thereafter, this method was used on anesthetized patients in the operating theater to monitor motor function during neurosurgical operations on the spinal cord.³⁾ However, it was difficult to record reliable and reproducible potentials under general anesthesia, especially when inhalation-based anesthetic agents that suppressed the firing of anterior horn neurons were used.⁴⁾ Furthermore, a single electrical stimulus applied over the skull could not generate multiple descending volleys, which are required to generate TcMEPs, due to the suppression of motor cortical excitability under general anesthesia.⁵⁾ Advances in anesthesia, including the introduction of intravenous anesthetic agents,^6,7) and progress in stimulation techniques, including the use of pulse trains,^8–10) improved the reliability and reproducibility of TcMEP responses.

Methodology

I. Anesthesia

Inhalational anesthetics such as isoflurane and nitrous oxide attenuate the amplitudes of TcMEPs, because of the suppression of the excitability of cortical and spinal motor neurons and interference with synapse transmission.^4,11,12) Propofol causes less suppression of the excitability of motor neurons than inhalational agents.^13–15) Consequently, total intravenous anesthesia using propofol and opioid is widely recommended as optimal.¹⁶⁾ Other favorable intravenous agents include ketamine, etomidate, and benzodiazepines.^16,17–20) Neuromuscular blockade is often omitted after intubation.^19,21,22) Otherwise, muscle relaxant can be administered with a constant infusion according to the amplitude of muscle responses to peripheral nerve stimulation (“train of four” technique).^22,23)

II. Stimulation

The electrode placement is on the skull based on the international 10–20 electroencephalograph (EEG) system.²¹⁾ We prefer the method of Matsuda and Shimazu, with the electrode symmetrically on the skull 5 cm outside and 2 cm forward of Cz.^23,24) Although EEG cup electrodes or needle electrodes may be used, cork screw-like needle electrodes are preferable because of their secure placement and low impedance.^16,21) While the cathode becomes the stimulating electrode with increasing intensity of the current, anodal stimuli evoke potentials more efficiently than cathodal stimuli.^16,21,25) Constant-current stimulators are better than constant-voltage stimulators because current delivered to the brain does not depend on the impedance of stimulating electrodes especially when impedance changes during surgery.⁵⁾

A short train of stimuli is preferable in anesthetized patients, while there remain controversies regarding the optimal parameters of the short train of stimuli including the number of pulses in the train, individual pulse duration, inter-pulse interval, and train repetition rate. We currently used a train of five biphasic stimuli with 0.5-ms in duration (two phases of 0.25 ms in each stimulus) and an inter-pulse interval of 2 ms.

III. Recording

TcMEPs can be recorded either with surface or needle electrodes. Needle electrodes inserted into the belly muscle yield greater amplitude due to their low impedance.⁵⁾ Averaging of several trials is not always required because of high signal-to-noise ratios of TcMEPs.¹⁶⁾ When choosing the muscles to record from for monitoring the functional integrity of corticospinal tract, small muscles in hands and feet should be included due to the rich corticospinal tract innervation.⁵⁾

Safety Issues

Safety concerns with intraoperative neurophysiologic monitoring using transcranial electrical stimulation include brain damage, scalp burns, seizure, bite injury, cardiac arrhythmias, and accidental injury resulting from patient movement.²⁶⁾ Continuous direct brain stimulation over a period of a few seconds with a frequency of 50–60 Hz was reported to easily induce seizures.^27,28) In spite of rarity of seizure related to transcranial electrical stimulation, epilepsy should be included in the contraindication of transcranial electrical stimulation.^5,26) The vigorous contractions and twitches of proximal muscle groups after transcranial electrical stimulation interfere with surgery and put patients at risk of spinal cord injury, and spinal nerve root, eye, tongue, and lip injuries.²⁹⁾ However, accidental injury resulting from patient movement can be avoided by brief surgical pauses (a few seconds) for monitoring of TcMEPs, coordinated between surgical and electrophysiological teams.³⁰⁾ Bite blocks are recommended to avoid tongue and lip lacerations.

Interpretations

In spite of the introduction of intravenous anesthetic agents and stimulation techniques using pulse trains,^6–10) TcMEPs are much smaller in amplitude compared with compound muscle action potentials evoked by maximal peripheral nerve stimulation, and vary from one trial to another in clinical practice, suggesting that only a limited number of spinal motor neurons innervating the target muscle are excited by the currently used transcranial stimulation techniques in anesthetized patients.³¹⁾ It is thought that TcMEPs reflect the activity of only 1.8–8.9% of the motor neuron pool innervating the target muscle during surgery,^32–35) particularly as Taniguchi et al. demonstrated that the amplitude of TcMEPs corresponded to about 5% of the amplitude of compound muscle action potentials evoked by maximal peripheral nerve stimulation.³⁶⁾ The smaller amplitude of TcMEP may also be due to the desynchronization of the descending volley, which may occur and lead to a decrease in amplitude because of “phase cancellation” phenomena.³⁷⁾ We previously examined the proportion of recruited motor neurons by multipulse transcranial electrical stimulation after eliminating the desynchronization of the descending volley using Magistris’s technique,³⁷⁾ and demonstrated that only 20% of motor neurons innervating the target muscle are recruited during TcMEP monitoring under general anesthesia.³⁸⁾

Regarding the fluctuation of TcMEPs, Kajiyama et al. reported that CMAPs vary from trial to trial even under partial neuromuscular blockade and under the strictly controlled low-dose propofol anesthesia.³⁹⁾ Similarly, it was reported that TcMEP responses degrade or fade over the duration of a surgery although the mechanism could not be explained.⁴⁰⁾ Therefore, reliable interpretation of the critical changes in TcMEPs remains difficult and controversial, including 50%,^41,42) 70%,⁴³⁾ 80%,^44,45) or 100%^46,47) attenuation of amplitude. TcMEPs are demonstrated to be very sensitive to ischemic and compressive insults to the spinal cord although the disappearance of TcMEPs does not necessarily reflect a motor deficit.^48,49) Additionally, in our clinical experience, we have occasionally encountered false negative cases in which patients have suffered from focal post-operative segmental motor weakness mostly due to single nerve root injury, despite no significant change in TcMEP activity during surgery.²³⁾ In such cases, TcMEPs may not have been reliable monitors of activity in motor units damaged intraoperatively, because of radicular overlap and different dominancy of each nerve root innervating the recorded muscle.^50,51)

Recent Developments

To reduce the false negative cases, more spinal motor neurons innervating the target muscle should be recruited to augment TcMEP responses. Almost two decades passed without any significant developments in anesthetic technique except for the introduction of the short-acting muscle relaxant rocuronium. Sugammadex to reverse rocuronium was demonstrated to facilitate motor evoked potentials monitoring during spinal surgery.⁵²⁾

Recently, several facilitative techniques using central or peripheral stimuli (conditioning stimulation), preceding transcranial electrical stimulation, have been employed to achieve sufficient depolarization of motor neurons and augment TcMEP responses during surgery.^29,53–57) According to Journée et al., conditioning stimulation can be classified into two categories: (1) heteronymous stimulation in which conditioning stimuli are applied at a different site from a test stimulus and (2) homonymous stimulation in which both conditioning and test stimuli are applied at the same site.⁵⁷⁾ One homonymous conditioning stimulation technique previously reported is recurrent pulse trains at low frequency (2–5 Hz), which was demonstrated to progressively facilitate TcMEP responses.^21,58–60) Another homonymous conditioning is double-train stimulation developed by Journée et al.⁵⁴⁾ They demonstrated that double-train stimulation elicited a marked facilitation of TcMEPs when the inter-train interval (ITI) was short (10 ms ≤ ITI ≤ 40 ms) or long (ITI ≥ 0.1 s). Taking these previous homonymous conditioning techniques into account, we systematically investigated the optimal setting of multiple transcranial electrical pulse trains (Fig. 1), so-called multi-train stimulation (MTS) to enhance TcMEP responses, in which a pulse train was delivered repeatedly at repetitive rates of 2 Hz, 5 Hz, and 10 Hz (ITI; 0.5 s, 0.2 s, and 0.1 s, respectively).⁶¹⁾ The amplitudes of TcMEPs increased with the number of train stimuli, and the strongest augmentation of TcMEPs was observed at a repetition rate of 5 Hz (ITI; 0.2 s). In addition, MTS significantly reduced the trial-to-trial variability of TcMEPs (Fig. 2), and enabled to obtain the stable responses throughout a surgery (Fig. 3), indicating that MTS could overcome “anesthetic fade.”

Fig. 1.

A schematic of the multi-train stimulation (MTS) technique, in which a pulse train is delivered repeatedly at constant repetitive rates (e.g., 2 Hz, 5 Hz, and 10 Hz). A pulse train consists of five biphasic stimuli with 0.5-ms in duration (two phases of 0.25 ms in each stimulus) and an inter-pulse interval of 2 ms. Transcranial motor evoked potentials are recorded from a pair of needle electrodes inserted in the muscle belly of the abductor hallucis (AH).

Fig. 2.

The within-patient variability of the amplitude of transcranial motor evoked potentials recorded from the abductor halluces muscle, which is assessed with the coefficient of variation (CV: standard deviation/mean). There is a statistically significant difference in CV between single-train stimulation (STS) and multi-train stimulation (MTS) (*p = 0.026, Mann-Whitney U test).

Fig. 3.

The trial-to-trial variability of transcranial motor evoked potentials recorded from the abductor halluces muscle. Multi-train stimulation (MTS) can reduce the trial-to-trial variability, and enables to obtain more stable responses throughout a surgery when compared to single-train stimulation (STS).

Conclusion

Intraoperative neurophysiologic monitoring has been indispensable with the recent advancement of surgical techniques for the treatment of more complicated spinal diseases which might damage neural tissues. Although the development of commercial equipment for intraoperative monitoring enabled reliable and reproducible evoked potentials, there have been some reports of false negative cases. Further advances in the TcMEP monitoring technique are required to enable it to be a good predictor of the patient’s motor function during spinal surgery.