Multimodal intraoperative neuromonitoring in scoliosis surgery: A two-year prospective analysis in a single centre
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.198189
Source of Support: None, Conflict of Interest: None
Aim: The present study is a prospective analysis of neuromonitoring [somatosensory evoked potentials (SSEP) and transcranial motor evoked potentials (TcMEP)] in consecutive scoliosis surgeries done at a tertiary care spine centre.
Keywords: Intraoperative neuromonitoring, scoliosis, somatosensory evoked potentials, transcranial motor evoked potentials
The quest for intraoperative spinal cord monitoring techniques dates back to the early 1960s when Harrington  introduced instrumentation to allow for the correction of spinal column deformities. A retrospective analysis performed by the Scoliosis Research Society in 1974 found that from 1965 to 1971, neurologic complications occurred at a rate of 0.72%, with partial or irreversible injury occurring in 0.65% of the subjects in this patient cohort. Deformity correction in scoliosis surgery is associated with an inherent risk of neurological complications as a result of injury to critical neural structures. A previous combined analysis by the Scoliosis Research Society and the EuroSpine in 1991 reported the results of 51000 surgical cases, and noted an overall injury occurrence of 0.55%. Factors associated with a higher incidence of neurologic injury included combined anterior and posterior repair, neuromuscular scoliosis, and significant kyphosis.,,,, There are various ways by which the spinal cord can get injured. Distraction during deformity correction accounts for the highest risk of spinal cord injury. Direct trauma from surgical manipulation, damage to vascularity during surgical exploration, positional issues, and thermal injury during the use of cautery are some of the contributing factors to intraoperative neurological injuries. The effect of such insults represents a spectrum – at one end, function decreases for the duration of the insult, and at the other end of this spectrum is permanent neurological damage from which recovery is never possible. Between these extremes is a large range over which recovery can occur partially or fully. Intraoperative neuromonitoring (IONM) helps in improving the safety of deformity correction surgery by providing a real-time assessment of neural function.
Somatosensory evoked potential (SSEP) monitoring was widely adopted for IONM in the 1980s. Tamaki and Yamane  and Nash et al., first reported its use in the late 1970s. An extensive review of the outcome for spinal cord monitoring in deformity surgery has been presented by Nuwer  resulting from a survey among the members of the Scoliosis Research Society (SRS). The large survey included 97,586 spinal surgery cases. Of them, 51,263 (53%) were using SSEP for monitoring purposes. Neurological deficits occurred in 0.55% patients, and on analysing the monitored cases, the true positive findings in SSEP were obtained in 0.43% and false negative findings in 0.127% patients. SSEP aids in monitoring only the function of the ascending tracts; however, changes in the motor tracts go undetected.
Merton and Morton  revolutionized spinal cord monitoring in 1980 by demonstrating that single-pulse voltage applied transcranially could result in contralateral motor activity. This was the first time that corticospinal tract monitoring was made possible. Transcranial motor evoked potential (TcMEP), when used in conjunction with SSEP, gave a comprehensive picture of the neurological status. Routine utilization of TcMEP was recommended by Luk et al., following their experience with 30 cases where no false negative results were reported.
When used together, SSEP and TcMEP monitoring allow for a real time assessment of both ascending and descending tracts of the spinal cord, thereby decreasing the likelihood of an iatrogenic neurological injury. Langeloo et al., in their study analysing 145 cases during spinal deformity correction surgery, reported the advantage of using a combination of SSEP and TcMEP, over SSEP alone.
Our study analyses the efficacy of combined SSEP and TcMEP monitoring in patients undergoing scoliosis corrective surgery in a single centre in South India over a period of 2 years.
Approval for this study was obtained from the Institutional review committee. Evaluation of prospectively collected neuromonitoring data from 52 consecutively operated cases of scoliosis was done. Patients diagnosed with scoliosis and operated between August 2013 and July 2015 at Amrita Institute of Medical Sciences, Kochi were selected for the study. The diagnoses of the patients are summarized in [Table 1]. Exclusion criteria were congenital scoliosis, previous spine surgery, and abnormal preoperative neurological status.
The mean age of the patients was 14.6 years (range 7-33 years). Thirty-nine of the patients were females and 13 were males. Demographic details of the cohort are summarized in [Table 2]. Preoperative neurological assessment was done in the outpatient department and recorded, and patients with an intact preoperative neurological status were included in the study. Radiographic data was obtained using X-ray, computed tomography (CT) scan, and magnetic resonance imaging (MRI) scan. Out of the 52 patients, 51 patients underwent posterior scoliosis correction and 1 patient had anterior scoliosis corrective surgery. All surgeries were single-stage procedures.
An established uniform total intravenous anesthesia (TIVA) protocol was followed for all patients, based on the published literature,,, which included induction with propofol and fentanyl, as well as maintenance with dexmedetomidine. Any change in the team of anesthetists and their practices did not influence the outcome of monitoring in any of the cases.
Monitoring for all cases was done by a trained and experienced team of neurophysiologists. The equipment used was Xltek Protektor 4.0 (Natus® Neurology). Recording and analysis was done using EPWorks 4.0 software (Natus Medical Incorporated®). Baseline recordings of SSEP and TcMEP were obtained as soon as the patient was anesthetized, after prone positioning, both intraoperatively and postoperatively.
SSEPs were obtained by stimulating the median and tibial nerves with the stimulus intensity level being. 20–25 mA. Cortical potentials were received via electrodes placed on standard locations over the skull vault,,,, The locations were based upon references from the international literature.
TcMEPs were recorded bilaterally from abductor pollicis brevis, biceps, and deltoid for the upper limb, and tibialis anterior, quadriceps, gastrocnemius, and abductor hallucis for the lower limb. Stimulation was given in the form of a high voltage (300–400 V) stimulus through electrodes placed subcutaneously, overlying the motor cortex regions.
An “alert” was defined as a reduction in the amplitude of at least 50% for SSEP and at least 65% for TcMEP compared to the baseline recordings, and an increase in the latency by more than 10%., In case an alert was sounded, surgery was temporarily halted and intervention was done in the form of initially directing the anesthesiologist to check the body temperature, blood pressure, and oxygen saturation. The technicians checked the integrity of the monitoring circuit. Once signals returned to within normal limits without requiring reversal of any surgical maneuver, rest of the procedure was carried out as planned.
Analysis of the outpatient records and radiographic data was done to determine the eligibility of patients for inclusion in the study. IONM recordings, operative notes, and anesthesia records were then scrutinized. Baseline recordings of SSEP and TcMEP were noted, and intraoperative changes in the amplitude and latency of the potentials were analysed. Surgical and nonsurgical intervention done to reverse the changes in potentials were noted and clinical outcome was assessed postoperatively. Statistical analysis was done to determine the sensitivity and specificity of SSEP and TcMEP monitoring based on correlation with intraoperative events and postoperative clinical outcome. The definitions related to the sensitivity and specificity of the monitoring are along the lines of description by Hilibrand et al., and are detailed in [Table 3].
The cohort in this study consisted of 39 females and 13 males, with a mean age of 14.6 years (range: 7-33 years). Baseline SSEP and MEP values were recordable in all the cases after anaesthesia was given. Out of a total of 52 patients, baseline values of SSEP and TcMEP recorded preoperatively demonstrated a lower baseline value of TcMEP in 4 patients, who had a preoperative diagnosis of neuromuscular scoliosis, as compared to the other groups. However, these 4 patients with neuromuscular scoliosis did not have any preoperative neurological deficits. In addition, none of the other patients had preoperative neurological deficits.
There were 5 cases (9.6%) of adolescent idiopathic scoliosis (AIS), wherein the intraoperative recordings showed a reduction in the amplitude of TcMEP recording to less than 65%. Two of these cases also had a fall in the amplitude of SSEP by more than 50%. However, a time lag of 15 minutes was noticed for the drop to be recorded, compared to the TcMEP reading. Having met the criteria for an alert, surgery was temporarily halted in these cases. In 4 of the cases, the anesthetist checked the vital parameters of the patient. Hypotension was detected and suitable measures were taken to restore hemodynamic stability. This resulted in the evoked potential values returning to within normal limits after a few minutes.
In 1 patient, the TcMEP signal was lost intraoperatively from the lower limbs with normal recording seen in the upper limbs. The vital parameters of the patient were assessed and found to be normal. On checking for the integrity of the circuit, malfunctioning of the leads attached to the lower limbs was discovered. This was set right by the technician and resulted in immediate appearance of normal evoked potential recording from the limb. Hence, no reversal of any surgical maneuver was warranted and the procedure was carried out and completed as planned. None of these cases showed any significant drop in the evoked potential values during the remainder of the surgery. No significant SSEP alterations were noticed during the study. None of the patients had any fresh neurological deficits, therefore, these were classified as true positive cases.
There were no false negative or false positive cases. The rest of the 47 patients (90.38%) in the cohort did not show any changes in the evoked potential values intraoperatively that were significant enough to sound an alert. Postoperatively, their neurological status was intact; therefore, these cases were classified as true negatives, as outlined in [Table 4] and [Table 5].
With the advent of newer instrumentation and techniques that are useful in the correction of spinal cord deformity, there is also an increase in the complexity of disorders in pediatric and adult patients undergoing corrective surgeries. Thus, there is an ever-growing need for newer and improved spinal cord monitoring techniques to reduce the risk of iatrogenic neurological injuries.
The field of IONM has come a long way since 1970s when it was first started. The first modality for intraoperative monitoring was spinal cord evoked potential (SCEP) monitoring, which was introduced in Japan in 1972 by Tamaki et al.,, and Kurokawa. This method tested only the dorsal column integrity and motor pathways could not be tested.
The wake up test, developed by Stagnara and Vauzelle in 1973, helped in identifying motor deficits. It involved allowing the patient to emerge from anesthesia and then assessment of motor function in the upper and lower limbs on command was performed. This test became extremely popular and is still widely used in centers across the world today. Although this test is 100% accurate in detecting gross motor movements when administered properly, the major limitation of this test is that it can detect injury only at a single point in time. A real-time assessment is not possible. The surgeon is, therefore, left with no information as to when exactly, or which maneuver, led to the injury. In addition, there are other inherent problems with the test such as difficulty in waking up the patient, accidental tracheal extubation, ensuring an adequate awakening of the patient to follow commands, and difficulties inherent in the comprehension of commands in an elderly age group or due to other problems. It also results in an increased operating time, as reversal of anesthesia, checking by operating room personnel if the patient is following commands, and re-induction of anaesthesia consumes time. However, the wake-up test can still serve to confirm the presence of an injury to the cord in the presence of significant loss of evoked potential signals. A major study conducted in German spine centres documents the use of the wake-up test as a monitoring modality for scoliosis surgeries. In our study, we did not have to use the wake-up test even once for confirmatory purposes.
The discovery of intraoperative monitoring as being a quick and effective way of assessing injury to the spinal cord led to various evoked potentials monitorings protocols coming into the picture. Our study did not have any case wherein SSEP and TcMEP values were normal intraoperatively but the patient developed postoperative neurological deficits, i.e., a false negative result. Although SSEP monitoring has been shown to have a high false-negative rate,, we were able to get a 100% sensitivity as well as specificity for SSEP. A possible advantage with our study was that it was a single-centre study with uniform protocol followed for all cases. This may not always be possible with multicentric studies. A major limitation of SSEP monitoring is that this modality can only monitor the ascending columns and no inference can be made regarding the integrity of motor nerve roots.
Monitoring of TcMEP began after Mertin and Morton demonstrated that monitoring of corticospinal tracts was also possible by transcranial stimulation leading to motor tract activity. Schwartz et al., have shown that TcMEP is extremely sensitive to spinal cord injury and detects impairment on an average of 5 minutes before the SSEP change appears. We can confirm this in our study wherein two cases had significant TcMEP changes but developed SSEP changes after a lag of 15 minutes, thus indicating that SSEP as a standalone monitoring procedure may not be as sensitive as is clinically required. Lee et al., demonstrated a 5.8% false positive rate. In our study, there were no patients who developed a significant alert intraoperatively that could not be reversed, and the patients awoke without a fresh onset neurological deficit, i.e., we did not find any false positive cases. Downside of TcMEP testing is the inability to use neuromuscular blocking agents, as well as the need for intermittent testing as repetitive movements would hamper surgery.
Through the use of upper limb leads, we were able to identify the isolated malfunction of lower limb leads. Thus, we wish to emphasize the fact that upper limb leads help to serve as controls and can demonstrate the integrity of the circuit. All 5 patients had a return of normal values without requiring surgical intervention and woke up without any neurological deficits.
A pertinent point regarding the importance of preoperative neurological status of the patient with respect to interpreting intraoperative changes in the neurophysiologic parameters was brought out by Pastorelli et al., and Patel et al., The rate of false positives was found to be as high as 13.8% in the former study. The patients in both these studies had an abnormality of the neurological system preoperatively, and the false positive alerts obtained intraoperatively demonstrated that the results must be interpreted with caution in such groups of patients because the neuromonitoring changes seen might likely be a result of the preexisting neurological abnormality.
Although varied reports are available in literature regarding the sensitivity and specificity of SSEP and TcMEP monitoring, multiple studies ,, have shown that the validity and reliability of intraoperative monitoring is increased by the combined use of SSEP and TcMEP recording. Thuet et al., were able to obtain a 99.6% sensitivity whereas Bhagat et al., obtained a sensitivity of 100% and specificity of 99.3% with a multimodal monitoring approach. We have derived a 100% sensitivity and specificity for combined neuromonitoring in our studies, and we concur with the previously published literature that a multimodal IONM gives best results and should be used as a standard line of care in scoliosis corrective surgeries.
Our single center study effectively demonstrates that combined multimodal intraoperative monitoring helps in increasing the safety factor in scoliosis corrective surgeries with a high sensitivity and specificity. Baseline values in neuromuscular scoliosis patients are possibly lower than idiopathic scoliosis patients, and intraoperative variations must be interpreted with caution. Finally, there should be a good collaboration between the spine surgeon, neurologist/neurophysiologist, technicians, and anesthesiologist to accurately interpret the results of IONM recordings, and thus ultimately derive the maximum benefit for the patient.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]