Somatosensory-Evoked Potentials

Early attempts at monitoring relied solely on recording SSEPs. Reports of false-negative outcomes when using only SSEP monitoring illustrated the need for multimodality monitoring.^6–8 SSEPs remain the standard technique for intraoperative monitoring. Newer techniques such as EMG and MEP have been developed and are used as an adjunct to SSEP monitoring.

Motor-Evoked Potentials

Because of the well-reported limitations of SSEP monitoring, the recording of MEPs has been advocated. Myogenic transcranial MEP offers the advantage of assessing the entire length of the corticospinal tract from the cortex to the distal extremity beyond the neuromuscular junction. The most significant issue when recording MEPs is the anesthetic regimen: total intravenous anesthetic without neuromuscular blockade needs to be utilized to provide the optimal environment for recording MEPs. This requires experience on the part of the surgical and anesthetic teams. For example, surgical exposure may be somewhat more challenging and the depth of anesthesia may require closer vigilance.

Electromyography

Spontaneous electromyographic recordings provide real-time data that are sensitive to surgical manipulation or compression.² Myotomes are selected for recording based on the operative level and the nerve roots most at risk. Burst or train activity is considered significant and is thought to represent ongoing compression or stretch. Spontaneous EMG is sensitive but not specific.⁹ As well as recording spontaneous electromyographic activity, a number of other applications have used electromyographic recordings to determine the proximity of nerve roots. Direct nerve root testing with EMG recording aids in the dissection of nerve roots off intradural tumors especially in the conus region. Triggered EMG has been used to aid in placement of pedicle screws.

Evidence for Somatosensory-Evoked Potentials

Most of the earlier literature predominantly evaluated the role of SSEPs alone in spinal surgery. As far back as 1982, Grundy and colleagues¹¹ reported a study showing that a wake-up test was not necessary provided SSEPs were monitored and stable. The authors prospectively studied the effects of moderate hypotension on 24 patients undergoing spinal fusion with Harrington rod instrumentation. Five of the 24 patients had alterations in their SSEPs and required an intraoperative wake-up test, all of which had normal results.

Epstein and coworkers¹² evaluated the role of SSEPs in cervical surgery by comparing the outcomes of patients who were monitored and operated on over a 3-year period (1989–1991) with those who were not monitored and had been operated on between 1985 and 1989. No instances of quadriplegia in the 100 patients who were monitored versus eight in the 218 who were not monitored were reported. The authors conclude that the reduction of neurologic deficit was attributed in part to early SSEP detection of vascular or mechanical compromise and to the immediate alteration of anesthetic or surgical technique in response to SSEP changes. Kombos and coworkers¹³ report similar findings in a prospective evaluation of the impact of SSEP monitoring during anterior cervical surgery. In a prospective study of 100 patients, they deduce that SSEP monitoring was easy to perform and helped to increase the safety during anterior cervical surgery. Monitoring of both cortical and subcortical sites for SSEP responses has been shown to increase the reliability of SSEPs during spinal surgery.¹⁴

Evidence for Motor-Evoked Potentials

Concerns over false-negative results when using SSEPs alone have led to the proposal of alternative strategies, with either monitoring of MEPs alone or used in combination with other modalities.⁶ Hilibrand and investigators¹⁵ analyzed the data of 427 patients who underwent anterior or posterior cervical spine surgery over a 2-year period. Twelve of their patients had loss of amplitude of MEPs, of which 10 had complete reversal of the loss after prompt intraoperative intervention and the remaining 2 had a new postoperative deficit. The sensitivity and specificity for MEPs was 100% in their series, whereas SSEP had only a sensitivity of 25%, although it was 100% specific. The authors conclude that transcranial MEPs appeared to be superior to conventional SSEPs for identifying evolving motor tract injury during cervical spine surgery.

Difficulties in obtaining and maintaining MEP responses with transcranial stimulation has led some authors to propose direct spinal cord stimulation to obtain neurogenic MEPs in certain pathologies.¹⁶ Komanetsky and colleagues¹⁷ compared two methods of stimulation when obtaining neurogenic MEPs. They prospectively compared spinous process stimulation with percutaneous stimulation in obtaining neurogenic MEPs in 184 patients. Both methods were found to be sensitive to neurologic deficit. When responses were obtained, the percutaneous method was found to be sufficiently reliable to obviate the need for the spinous process method.

Numerous studies have addressed the difficulties in obtaining reliable predictors for postoperative C5 palsy.¹⁸ Tanaka and colleagues¹⁸ evaluated the usefulness of transcranial MEPs for prediction of the occurrence of postoperative C5 palsy after cervical laminoplasty. They prospectively evaluated 62 consecutive patients, three of which developed postoperative transient C5 palsy. No critical decrease in amplitude occurred in any of the 62 patients. Because of this, the authors conclude that postoperative C5 palsy after cervical laminoplasty was not associated with an intraoperative injury (Table 7-1).

Evidence for Electromyography

There are fewer Level I or II evidence articles in the literature that validate the use of intraoperative EMG monitoring in spinal surgery. Dimopoulos and investigators¹⁹ conducted a prospective randomized trial to correlate the findings of intraoperative EMG with immediate postoperative pain in patients undergoing lumbar microdiscectomy (Level II). They found no correlation between intraoperative electromyographic findings and postoperative pain.

Krassioukov and coauthors²⁰ examined the neurologic outcomes of 61 patients, most of whom were treated for spinal/spinal cord tumours (61%) or adult tethered cord syndrome (25%). Patients underwent multimodal neurophysiologic monitoring with EMG monitoring of the lower-limb muscles, external anal sphincter (EAS), external urethral sphincter (EUS), and lower-limb SSEPs. Spontaneous electromyographic activity was observed in the lower-extremity muscles and/or EAS and EUS in 51 cases (84%). In addition to spontaneously recorded electromyographic activity, electrically evoked EMG activity was also used as an intraoperative adjunct. The presence of electrically evoked EMG activity in structures encountered duringmicrodissection altered the plan of treatment in 24 cases (42%).

Similar findings were reported by Paradiso and colleagues,¹ who examined the use of intraoperative monitoring in tethered cord syndrome. Posterior tibial nerve SSEPs were found to have high specificity, but low sensitivity, for predicting new neurologic deficits. In contrast, continuous EMG showed high sensitivity and low specificity. Evoked EMG accurately identified functional neural tissue.

Multimodality Monitoring

To overcome the limitations of individual monitoring modalities, many teams have explored the value of multimodal monitoring to obtain a more robust real-time assessment of intraoperative neural function. Pelosi and researchers²¹ investigated the combined monitoring of MEPs and SSEPs in 126 spinal operations. Combined monitoring was successfully achieved in 104 operations; it was possible only to monitor a single modality in 18 patients (16 SSEPs, 2 MEPs). No response to either modality could be recorded in two patients. The authors report that combined monitoring was superior to single-modality techniques both for increasing the number of patients in whom satisfactory monitoring could be achieved and for improving the sensitivity and predictivity of monitoring.

Gunnarsson and coauthors⁹ correlate the clinical and electrophysiologic findings in a prospective, consecutive series of 213 patients. Intraoperative electromyographic activation had a sensitivity of 100% and a specificity of 23.7% for the detection of a new postoperative neurologic deficit. SSEPs had a sensitivity of 28.6% and specificity of 94.7%. The authors conclude that combined intraoperative neurophysiologic monitoring with EMG and SSEP is helpful for predicting and possibly preventing neurologic injury during thoracolumbar spine surgery.

Costa and colleagues’²² prospective observational study reports on the use of combined multimodal monitoring in a cohort of patients undergoing a variety of spinal procedures. Combined SSEP and MEP monitoring was successfully obtained in 38 of 52 patients (73%), whereas MEPs from at least 1 target muscle were obtained in 12 patients (23%); both SSEPs and MEPs were absent in 2 patients (3.8%). Significant intraoperative changes occurred in one or both modalities in five patients, two of which were transient, whereas three had persistent changes associated with new deficits or worsening of the preexisting neurologic disability. The authors suggested that intraoperative combined monitoring is a safe, reliable, and sensitive method to detect and reduce neurologic injury to the spinal cord. However, there was no comparative group that did not use neuromonitoring.

After comparing transcranial MEPs and SSEP monitoring in a large cohort of patients who underwent cervical spine surgery, Hilibrand and investigators¹⁵ highlight the fact that, although SSEPs were specific, they remain relatively insensitive. In addition, they highlight the need for multimodal monitoring in cervical spine surgery.

Case Example

A 64-year-old man with known ankylosing spondylitis presented with a 6-month history of progressive gait difficulties and impairment of upper-limb fine motor function. His clinical examination confirmed that he had cervical myelopathy. Preoperative imaging (Fig. 7-1 A and C) showed that he had marked cord compression on the T2-weighted magnetic resonance imaging. He was electively booked for surgery where he underwent a multilevel cervical laminectomy and instrumented fusion. During the decompression stage of the operation, the surgeon was alerted that the patient had developed spontaneous electromyographic train activity in an upper-limb muscle (see Fig. 7-2A). Shortly after this, the patient’s SSEP responses in all four limbs deteriorated (see Fig. 7-2B). A diagnosis of a presumptive cord injury was made and treatment was instituted with blood pressure elevation and administration of steroids. The patient’s SSEP traces in his upper limbs began to recover toward the end of the case. Postoperative imaging showed new high signal changes in the spinal cord (Fig. 7-1B and D). Immediately after surgery, the patient was quadriparetic, legs worse than arms, but his deficit had improved substantially by the time of discharge. The immediate diagnosis and institution of treatment potentially prevented a complete injury and a suboptimal clinical outcome for this patient.

FIGURE 7-1 Preoperative and postoperative T2 magnetic resonance imaging (MRI) and lateral C-spine radiographs. Preoperative T2 MRI (A) showing marked cord compression and (B) postoperative T2 MRI showing new high-signal change in the cord substance. C, D, Preoperative and postoperative plain lateral radiographs.

FIGURE 7-2 Multimodality intraoperative recordings of electromyographic (EMG) and somatosensory-evoked potential (SSEP) responses. Intraoperative electromyography showing the run of spontaneous EMG activity before the SSEP deterioration (A). The patient’s SSEP tracing is shown at various time points. At 14.56 (arrow), the patient had complete loss of his SSEP responses. Treatment was instituted immediately, and by 15.11, there had begun to be some recovery of the SSEP response.

Anesthetic Regimen for Monitoring

Some of the best evidence available in the intraoperative monitoring literature pertains to the optimal anesthetic regimen to use. In 1997, a randomized trial of 20 patients was undertaken to compare the effects of ketamine with those of fentanyl (both combined with midazolam) on cortical SSEP monitoring during major spinal surgery.²³ Cortical SSEP latencies were not significantly affected in either group. The authors conclude that both ketamine and fentanyl allowed the recording of reliable cortical SSEPs, but a longer delay for voluntary postoperative motor assessment was observed in the ketamine group (Level II).

Laureau and colleagues²⁴ report a prospective, randomized trial comparing the effects of propofol and midazolam in 2 groups of 15 patients undergoing surgery for idiopathic scoliosis. The amplitude of the cortical SSEP responses decreased after induction in both groups; in the midazolam-treated group, the amplitudes were smaller. The most significant finding in the study was that both propofol and midazolam seemed to be acceptable hypnotic agents for total intravenous anesthesia during intraoperative monitoring in the surgical treatment of scoliosis. This study has limitations in that it is difficult to correlate postoperative neurologic deficits and intraoperative monitoring data because no SSEP changes occurred and also because cortical SSEPs were recorded only unilaterally (Level II).

The effects of low-dose propofol as a supplement to ketamine-based anesthesia during intraoperative mo-nitoring of MEPs has been reported by Kawaguchi and coauthors.²⁵ Intraoperative monitoring of MEPs was performed in 58 patients who underwent elective spinal surgery. Anesthesia was maintained with nitrous oxide-fentanyl-ketamine with or without low-dose propofol. The authors found that low-dose propofol could be effectively used as a supplement to ketamine-based anesthesia during intraoperative monitoring of myogenic MEPs as long as a train of pulses was used for transcranial stimulation. Addition of propofol significantly reduced the ketamine-induced psychedelic effects.

In 2001, Samra and colleagues’²⁶ study was published comparing remifentanil- and fentanyl-based anesthesia for intraoperative monitoring of SSEPs with special attention to the speed of recovery from the anesthetic. They studied 41 patients who were randomized into 2 groups to receive either remifentanil- or fentanyl-based anesthesia. The authors report that both remifentanil and fentanyl infusion allowed satisfactory monitoring of SSEPs. Remifentanil infusion offers the advantage of quicker recovery from anesthesia and less variability of SSEP morphology.

Ku and coworkers²⁷ have reported the effect of sevoflurane/nitrous oxide versus propofol anesthesia on SSEP monitoring during scoliosis surgery. They randomized 20 patients into 2 groups to receive either sevoflurane or propofol. Changes in anesthetic concentrations produced little effect on the latency of SSEP, but the effect on the variability of amplitude was significant. The authors conclude that sevoflurane produced a faster decrease and recovery of SSEP amplitude, as well as a better conscious state on waking, than did propofol (Level II).

A number of studies have compared the effects of propofol versus isoflurane on monitoring SSEPs and MEPs intraoperatively.^28–30 Clapcich and investigators²⁹ randomized 12 patients into 4 groups receiving either variable doses of isoflurane-nitrous oxide combinations or a propofol combination. Chen²⁸ randomized a group of 35 patients into 2 groups to receive either an isoflurane or a propofol infusion. Liu and coauthors³⁰ reported a randomized trial of 60 patients who were randomized to receive similar types of infusions. All of the groups reported that both propofol and isoflurane decreased SSEP amplitude while the anesthetic took effect. Propofol caused less suppression of the cortical SSEP responses with better preservation of SSEP amplitude. Chen²⁸ reports that MEPs were recordable in all patients receiving propofol, but in only 50% of patients receiving isoflurane²⁸ (Level II).

In 2006, Lo and coauthors’³¹ study was published comparing a desflurane-nitrous oxide combination with propofol total intravenous anesthetic regimen. They prospectively randomized 20 patients undergoing scoliosis correction surgery into 2 groups. Reproducible MEP responses were always obtained throughout the procedure. The authors conclude that both desflurane and total intravenous anesthetic with propofol allow successful monitoring. They suggest using abductor hallucis muscle for recording MEPs when using desflurane (Level II).

Limitations of Current Evidence

All of the articles discussed in the previous section are deemed Level II evidence. No Level I evidence exists to support the use of intraoperative monitoring. Moreover, many of the class II studies are limited by relatively small sample sizes. The lack of validated quantitative methods to assess intraoperative electromyographic changes also poses a limitation in evaluating the literature for this neurophysiologic modality.

With the advancement and development of new anesthetic agents, the question as to which anesthetic regimen is optimal for intraoperative monitoring is likely to be asked many times in the future. The answering of such a question in the presence of clinical equipoise could potentially be deciphered by performing a prospective, randomized, blinded, controlled trial (PRCT).

What is much more challenging, however, is to try and establish a similar level of evidence to support the use of intraoperative neurophysiologic monitoring. The difficulty is that trials on this subject do not lend themselves well to the PRCT type model. Devising any such outcome study, which would include blinding the surgeon, would be ethically unacceptable to those who routinely use intraoperative monitoring.

It appears likely, therefore, that the answers of future questions on whom should be monitored and how they should be monitored will more likely come from well-documented, nonrandomized, prospectively collected studies on large series of patients rather than other types of study design.

Which Spinal Patients Should Be Monitored?

Currently, no universal agreement exists as to which patients should undergo intraoperative monitoring during spinal surgery. The published literature would indicate that patients with a preexisting cord deficit have a greater risk for having a major neurophysiologic alert and are at greater risk for an intraoperative injury.^{2,32, 33} Given that these patients appear to be at greater risk, it would seem prudent to monitor them.

MEPs are reliable in predicting clinical motor outcome, and their use has altered the surgical approach; for example, gross total resections of intramedullary tumors are more readily attempted as long as MEP data indicate the intact functional integrity of the corticospinal tract.³⁴

The applicability and effectiveness of intraoperative monitoring increase with its use and familiarity. Therefore, we recommend that it be used for all spine surgery cases where neurologic compromise is a potential risk. The most effective application of these techniques requires a team-based approach involving the surgical, anesthetic, and monitoring personnel.

What Modalities Are Optimal to Monitor?

Which set of tests is optimal to monitor needs to be individually tailored for each surgery. The choice of what modalities to use is dependent on the preference of the surgeon, the nature and localization of the pathology involved, and the technical feasibility of the modality in the specific context. A good way to select which tests to use is to choose a combination of modalities most specific to the neural tissue at risk during the procedure—MEPs for anterior cord, SSEPs for posterior cord, and EMGs for roots at risk, respectively.

The combined electrophysiologic exploration of motor and sensory potentials has proved to be the most useful tool for monitoring patients during spinal surgery. Combined MEPs and SSEPs (as well as EMG) provide independent and complementary information, and improve spinal cord monitoring. Good practices such as being consistent in its use and keeping a good and constant communication between the monitoring and surgical team are key for its success.