Intra‐Operative Neurophysiological Monitoring in Patients with Thoracic Spinal Stenosis: Its Feasibility and High‐Risk Factors for New Neurological Deficit

Authors: Wanyou Liu, Yinkun Li, Junyin Qiu, Benlong Shi, Zhen Liu, Xu Sun, Zezhang Zhu, Yong Qiu

Abstract

Objective

Considering the high risk of postoperative neurological complications for patients with thoracic spinal stenosis (TSS), intra‐operative neurophysiological monitoring (IONM) has been used for detecting possible iatrogenic injury timely. However, the IONM waveforms are often unreliable. This article is designed to determine the test performance of somatosensory evoked potentials (SEP) and motor evoked potentials (MEP) during surgical thoracic decompression in patients with TSS, and to investigate the risk factors associated with deteriorated neurologic function at immediate postoperation.

Methods

Patients undergoing posterior spinal fusion from February 2009 to December 2020 were retrospectively reviewed. Patients were divided into the deteriorated neurologic function (DNF) group and the improved/intact neurological function (INF) group based on the postoperative neurological status. Demographic parameters such as gender, age, height, weight, etiology and IONM data were compared between groups. Demographics and IONM data between DNF and INF groups were compared by independent t or nonparametric tests. The incidence of abnormal SEP was analyzed by Chi‐square test.

Results

A total of 108 patients (63 males, 45 females) with an average age of 53.5 ± 14.0 years were included. The SEP and MEP records were available in 94 and 98 patients, with the overall success rates being 87.0% and 90.7%, respectively. The sensibilities and specificities were 100% and 88.2% for SEP, 100% and 98.8% for MEP, respectively. There were 17 patients in DNF group and 91 patients in INF group. High weight (79.1 ± 14.6 vs 69.7 ± 15.7 kg, P = 0.024), high inter‐side difference of MEP amplitude (899.1 ± 997.5 vs 492.3 ± 512.4 μV, P = 0.013) and high incidence of abnormal SEP (94.1% vs 64.8%, P = 0.024) were observed in the DNF group. Fourteen (82.4%) patients in the DNF group showed improvement in neurological status during follow‐up.

Conclusions

The overall success rates were 87.0% for SEP and 90.7% for MEP in patients with TSS.

Introduction

Thoracic spinal stenosis (TSS) is a relatively rare degenerative disease which is characterized by a reduction in the volume of the thoracic spinal canal. Therefore, the classical symptom was paresthesia in the chest and abdomen and incontinence and lower limb hemiplegia on account of compression of the spinal cord or nerve roots. Though posterior spinal decompression and fusion has been regarded as effective treatment for patients with TSS, the outcome tends to be poor for its tenuous blood supply, natural kyphosis and shortening pedicles. ^1^ , ^2^ Postoperative neurological complications are reported with an incidence of 14.5% to 33%, ^3^ , ^4^ , ^5^ , ^6^ remaining a problem for worldwide spine surgeons.

In order to timely detect possible iatrogenic injury and maximally guarantee the safety of decompression surgery, the intra‐operative neurophysiological monitoring (IONM) technique mainly including somatosensory evoked potentials (SEP) and motor evoked potentials (MEP) have been frequently emphasized. The IONM has been routinely applied during the surgery of TSS. Unfortunately, the IONM results are often unreliable. The incidence of false–negative IONM in detecting postoperative neurological complications was reported to be more than 0.36%, which could lead to disastrous outcomes for the patients. ^7^ , ^8^ Therefore, this retrospective study was (i) to determine the test performance of SEP and MEP during surgical thoracic decompression in patients with TSS; and (ii) to investigate the risk factors associated with deteriorated neurologic function at immediate postoperation in this cohort.

Materials and Methods

Subjects

TSS patients undergoing decompression surgery from February 2009 to December 2020 in our center were retrospectively reviewed. The following inclusion criteria were (i) patients underwent posterior spinal decompression and fusion surgery; (ii) with preoperative thoracic CT and MRI demonstrating symptomatic TSS; (iii) with muscle strength examination of bilateral lower extremities at preoperation, postoperation and follow‐up; and (iv) with intact IONM data. The exclusion criteria (i) patients combined with cervical/lumbar stenosis; and (ii) undergoing anterior or anterior+posterior approach. Based on the postoperative neurological status, patients with deteriorated neurologic function were divided into the DNF group, and those with improved/intact neurological function were into the INF group. This study was approved by the Clinical Research Ethics Committee of Nanjing Drum Tower Hospital (No. 2021‐398‐01).

Anesthetics and Muscle Relaxants

The total intravenous anesthesia was performed in all patients, of which the protocols were in accordance with previous studies. ^9^ , ^10^ Anesthesia was successively induced midazolam (0.06 mg/kg), propofol (2–3 mg/kg), cisatracurium (0.2 mg/kg) and fentanyl (3 μg/kg). No muscle relaxant or inhalation anesthesia was used after the patient was induced. Anesthesia was maintained with a propofol (80–120 μg·kg^−1^·min^−1^) infusion, supplemented with remifentanil (0.2–1 μg·kg^−1^·min^−1^) and dexmedetomidine (0.2 μg·kg^−1^·h^−1^) infusion. The bispectral index ranged from 40 to 60.

IONM Protocol

All stimulation and recording trials of SEP and MEP were performed using the commercially available neurophysiological monitoring workstations (Protektor, XLTEK, Oakville, Canada; NIM‐ECLIPSE, Medtronic, Minneapolis, MN, USA; Fig. 1). The IONM protocol were performed as ^9^ (i) SEP: bilateral posterior tibial was stimulated at 2–3 cm above ankle, and recording point was placed at Cz with reference to Fz. The parameters of SEP square waves were 0.2 ms duration, 2.1 Hz frequency, 20mA‐40mA intensity, 100 ms time base, 30–1000 Hz filter bandwidth; and (ii) MEP: MEP was elicited with a brief anodal pulse train over motor cortex regions C3‐C4, and recording point was placed at bilateral abductor hallucis muscles, tibialis anterior and abductor pollicis brevis of the upper extremity, respectively. The parameters of MEP stimulation were 6–9 pulses, 50 μs stimulus duration, 2–4 ms inter‐stimulus interval, 200–500 V stimulus intensity, 100 ms time base and 10–1000 Hz filter band pass. The P37 and N50 latency, amplitude were recorded for SEP, and the latency and amplitude were recorded for MEP.

IONM alerts would be reported if fulfilling any of the following (i) SEP latency increased >10%; (ii) SEP amplitude decreased >50%; or (iii) MEP amplitude decreased >80% on one or both sides. The abnormal SEP was defined as previous in ^10^ (i) absent SEP waveforms, unilateral or bilateral; and (ii) prolonged unilateral latency, or bilateral latencies normalized with body height and >2.5 standard deviation (SD) over mean values calculated from normal control with reference to the equation by Chen et al. ^11^ (P37 latency = 0.277*height − 7.144, SD = 1.071); and (3) asymmetrical inter‐side difference of latency or amplitude>2.5 SD of normal control.

Parameters

Demographic parameters such as gender, age, height, weight and etiology were assessed for each patient. The SEP (P37, N50) and MEP latency, SEP and MEP amplitude were recorded as IONM data.

Statistical Analysis

Data were statistically analyzed with the SPSS software 22.0 (SPSS, Chicago, IL, USA). Patients' demographics were shown with mean ± SD. Demographics and IONM data between DNF group and INF groups were compared by independent t test if fulfilling the Normality test or nonparametric test if not. The incidence of abnormal SEP was analyzed by a chi‐square test. Statistically significant difference was defined as P < 0.05.

Results

General Data

A total of 108 patients with TSS (63 males, 45 females) were included in the study. The average values of age, height and weight were 53.5 ± 14.0 years (range 13–82 years), 163.0 ± 12.1cm (range 130–178 cm) and 74.9 ± 14.1kg (range 35–152kg), respectively. The etiologies of the 108 patients were as ossification of the posterior longitudinal ligaments (OPLL, n = 17), ossification of the ligamentum flavum (OLF, n = 57), OPLL + OLF (n = 15), thoracic disk herniation (TDH, n = 19). The postoperative changes in target muscle strength of lower limbs were shown in Table 1.

IONM Performance

The SEP and MEP records were available in 94 and 98 patients, with the overall success rates being 87.0% and 90.7%, respectively. IONM alerts were reported in 27 patients for SEP and 18 patients for MEP. The text performance of SEP and MEP monitoring was summarized in Table 2. Additionally, a total of 75 (69.4%) abnormal SEP were observed in this cohort.

Comparison between DNF Group and INF Group

A total of 17 patients were divided into DNF group and the other 91 patients were into INF group, respectively. The gender, age, height, weight and IONM data were compared between two groups, which showed high weight (79.1 ± 14.6 vs 69.7 ± 15.7 kg, P = 0.024), high inter‐side difference of MEP amplitude (899.1 ± 997.5 vs 492.3 ± 512.4 μV, P = 0.013) and high incidence of abnormal SEP (94.1% vs 64.8%, P = 0.024) in DNF group (Table 3, Fig. 2). Fourteen (82.4%) patients in DNF group showed improvement in neurological status during the follow‐up period, of which the details were showed in Table 4.

Discussion

IONM Performance in TSS Patients

In the current, the sensibilities and specificities were 100% and 88.2% for SEP, 100% and 98.8% for MEP, respectively. Hence, the IONM results seemed to be quite reliable for those patients with successful monitoring waveforms. Nowadays, it has been well recognized that the IONM performance is closely associated with the etiologies, ^9^ , ^10^ , ^12^ , ^13^ , ^14^ and TSS is an obvious obstacle for stable and reliable IONM results. ^15^ , ^16^ Melachuri et al. ^15^ retrospectively included 771 patients who underwent spinal thoracic fusion with SEP monitoring, of whom nine (1.2%) patients lost SEP responses at baseline. In the Prospective Multicenter Study of the Monitoring Committee of the Japanese Society for Spine Surgery and Related Research, Kobayashi et al. ^16^ reported 4% patients with no detectable waveform from all muscles during posterior decompression and fusion surgery for thoracic ossification of the posterior longitudinal ligament. In the current study, a total of 108 TSS patients were analyzed, of whom only 94 (87.0%) SEP baselines and 98 (90.7%) MEP baselines were successfully obtained. Therefore, multiple IONM monitoring at least including SEP, MEP and electromyogram should be together applied in patients with TSS. In addition, several other IONM models such as spinal cord evoked potentials and descending neurogenic evoked potentials were also recommended if possible.

Inter‐side Difference of IONM

The inter‐side difference of both SEP and MEP were demonstrated to be quite common in patients with various etiologies, indicating its possible physiological mechanism in each cohort. ^11^ , ^17^ Chen et al. ^11^ compared the inter‐side difference between scoliosis patients and normal control, and drew the conclusion that the inter‐side difference of both P37 latency (1.1 ± 1.6 vs 0.2 ± 0.3ms) and N45 latency (1.2 ± 1.6 vs 0.2 ± 0.3ms) were significantly higher in scoliosis patients. Recently, the abnormal SEP, as an important factor of inter‐side difference, was proposed as a critical indicator for alterations in somatosensory pathways. Chen et al. ^11^ found the rate of abnormal SEP was higher in patients with adolescent idiopathic scoliosis (32.6%, 15/46) than in those patients with congenital scoliosis (12.1%, 4/33), implying that the sensory deficit in patients with adolescent idiopathic scoliosis could be linked to an etiologic factor besides secondary factors. Shi et al. ^9^ reported that the rate of abnormal SEP in patients with asymptomatic Chiari malformation‐associated scoliosis was about 31.7%. In the current study, significantly high rate of abnormal SEP (69.4%, 75/108) was obtained in patients with TSS, showing obvious sensory deficit in this cohort. In addition, motor deficit was found in 88.9% (192/216) target muscles of lower extremities in our results. We believed that the significant sensory and motor deficit in this cohort should be responsible for the relatively low successful rate of IONM monitoring. The direct compression of spinal cord including the posterior column pathways and motor pathways functions, a more tenuous blood supply predisposing the cord to injury with manipulation of TSS were assumed as the reasons for failed IONM monitoring in these cases.

IONM Alerts and Neurologic Function

Surgical decompression is clearly the most effective method for the treatment of TSS, which is unfortunately associated with a high risk of deteriorated neurologic function at immediate postoperation. ^3^ , ^4^ , ^5^ , ^6^ Young and Baron ^3^ reported 14.5% TSS patients suffered from acute neurologic deterioration after surgical treatment. Takahata et al. ^4^ tracked 30 TSS patients treated with posterior approach, and found 10 (33.3%) patients with neurological deterioration immediately after the surgery. In our study, though postoperative improved/intact neurological function was observed in 82 (75.9%) patients, worsened neurological function was still found in 17 (15.7%) patients. To the best of our knowledge, the high risk of deteriorated neurologic function should be attributed to the unique anatomical factors of thoracic spine. ^3^ , ^14^ The thoracic spinal canal is relatively narrow because of natural kyphosis and shortening pedicles of thoracic vertebrae. Additionally, poor blood supply in this region lead to high risk of ischemia–reperfusion injury. Nevertheless, IONM alerts were reported in all the 17 patients with worse neurological function in our study, which highly proved the necessary of IONM monitoring in TSS‐related surgery.

Risk Factors for Deteriorated Neurologic Function

Several possible indicators for iatrogenic neurological deficit after spinal surgery have been illustrated in the literature. ^4^ , ^6^ , ^18^ , ^19^ Wang et al. ^20^ reported BMI was usually higher in patients with thoracic compression myelopathy compared with cervical compression myelopathy. The possible reason was assumed as the increased loading of the spine accelerates proliferation and degradation, which also contributed to the development of TSS. Meanwhile, patients with higher weight tended to show severe neurological atrophy because of little exercise and long bed time. In the current study, the comparison between DNF and INF groups showed that high weight, high inter‐side difference of MEP amplitude and high incidence of abnormal SEP were high‐risk factors for deteriorated neurologic function at immediate post‐operation in TSS patients. Therefore, more frequent monitoring should be performed in TSS patients, especially in those with the above high‐risk factors. It should be mentioned that 14 (82.4%) patients in DNF group showed improvement in neurological status during the follow‐up period. Hence, promising prognosis could be expected in most DNF patients with timely and appropriate managements.

Strengths and Limitations

The current study first, investigated the feasibility and performance of IONM in TSS patients based on a relatively large sample size, demonstrating that high weight, high inter‐side difference of MEP amplitude and high incidence of abnormal SEP were high‐risk factors for new neurological deficit. However, there were still several potential limitations in the current study. The retrospective nature might lead to uncertain bias of the results. Though total intravenous anesthesia was performed in all patients, IONM results would still be impacted by the individual details of the anesthetic scheme. The difference of IONM between patients with different etiologies were not compared. In addition, though improvement in neurological status was obtained during follow‐up period, longitudinal postoperative follow‐up was still necessary for evaluating long‐term outcomes of neurological status in DNF patients.

Conclusions

IONM could be effectively and reliably applied during the posterior spinal decompression surgery in patients with TSS, and the overall success rates are 87.0% for SEP and 90.7% for MEP. TSS patients with high weight, high inter‐side difference of MEP amplitude and high incidence of abnormal SEP have a higher risk of deteriorated neurological function at immediate postoperation.

Author Contributions

Conception and design of the Wanyou Liu, Benlong Shi, Zezhang Zhu, Yong Qiu; acquisition, analysis and interpretation of Wanyou Liu, Yinkun Li, Junyin Qiu, Benlong Shi, Zhen Liu, Xu Sun, Zezhang Zhu, Yong Qiu; drafting the article and revising it critically for important intellectual Wanyou Liu, Yinkun Li, Junyin Qiu, Benlong Shi, Zezhang Zhu, Yong Qiu; all authors have given approval to the final version of the manuscript to be published.

Conflict of Interest

All authors have no conflict of interest to declare. This study was approved by the ethics review board of the affiliated Drum Tower Hospital of Nanjing University (No. 2021–398‐01).