Value of Evoked Potential Changes Associated with Neck Extension Prior to Cervical Spine Surgery
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.360905
Source of Support: None, Conflict of Interest: None
Keywords: Cervical spine surgery, multimodal intraoperative monitoring, neck extension, neurological deficits
Intraoperative neuromonitoring (IONM), as the standard operation in spinal deformity surgery, is an effective means to reduce nerve injury. Combining somatosensory evoked potentials (SSEPs), transcranial motor evoked potentials (TcMEPs), and free-run electromyography (EMG), multimodal intraoperative monitoring (MIOM) can fully assess the integrity of the anterior as well as the posterior spinal cord, and help surgeons intervene to reverse the immediate cause of intraoperative spinal cord injury in cervical spine surgery.,, Previous studies have confirmed some risk factors of nerve injury in cervical spine surgery, including positioning, preoperative cervical spinal stenosis, operation technique, and vascular injury due to hypotension or arterial interruption., Positioning of the neck is one of the most common risk factors of nerve injury in cervical spine surgery, except for surgical factors. However, the value of MIOM remains controversial before cervical spine surgery.,,,,,,
In this study, we aimed to analyze the EP changes associated with neck extension, including intubation and positioning of the neck, and explore the value of MIOM in detecting spinal cord injury and reducing permanent nerve injury caused by neck extension during the pre-surgical step of cervical spine surgery.
Data of 191 patients undergoing cervical spine surgery under MIOM were collected at our medical center from June 2014 to June 2020. There were 122 males and 69 females. The mean age was 55 years (range, 15–86 years). The Japanese Orthopaedic Association (JOA) scores were 11.85 ± 3.72. In total, 127 (66.49%) patients had an anterior surgical approach, 55 (28.88%) had a posterior approach, and only 9 (4.71%) patients had a combined approach. The main surgical indications included cervical myelopathy, radiculopathy, trauma, tumor, and infection [Table 1]. Appropriate institutional ethics committee was obtained.
The procedure was performed under total intravenous anesthesia (TIVA). The administration dose of propofol (1 to 2 mg/kg), fentanyl (3 to 5 μg/kg), cisatracurium (0.1 to 0.12 mg/kg) were given for anesthesia induction. Propofol (3 to 5 μg/ml) and remifentanil (0.05 to 20 μg/kg/min) were used for target-controlled infusion (TCI) to maintain the depth of anesthesia after intubation. Intraoperative inhaled anesthetics and bolus dose of muscle relaxants were avoided.
Somatosensory evoked potentials (SSEPs)
For the upper limbs, C3' and C4' (2 cm posterior to C3 and C4) were applied for recording locations with subdermal needle electrodes according to the International 10–20 System of Electrode Placement. Stimulation to the upper extremities was delivered through the median or ulnar nerve at the wrist. For the lower extremities, Cz or Cz' (2 cm posterior to Cz) was used as the recording location. The reference electrode was placed at Fz. The posterior tibial nerve is usually stimulated at the ankle. The ground electrode was placed at the deltoid muscle. Parameter setting: rate of 2.1 to 4.7 Hz; current intensity: 15–25 mA for upper limbs and 25–40 mA for lower limbs with constant current stimulation, duration of 300 μs. Timebase: 50 ms for upper limbs and 100 ms for lower limbs, 100–200 average. Band pass filters were 30 to 500 Hz.
Transcranial motor evoked potentials (TcMEPs)
Repetitive stimulation was performed using corkscrew electrodes, which were applied at C3 (anode) and C4 (cathode) locations. The recording electrodes were inserted into the muscle with paired subdermal needle electrodes. The recording target muscles were deltoid, biceps, brachioradialis, abductor pollicis brevis, tibialis anterior, and adductor hallucis. Parameter setting: Stimulation intensities of 200 to 400 V constant voltage stimulation were used for transcranial stimulation, duration of 500 μs. Stimuli were delivered in trains of seven with a frequency of 500 Hz.
Free-run electromyography (EMG)
A continuous free-run EMG activity was recorded from the muscles that were recorded in the TcMEPs.
The level of residual neuromuscular blockade was assessed by recording nerve-conducted responses from the abductor pollicis brevis to unilateral train-of-four (TOF) electrical stimulation applied either to the median nerve. Upon recording 3/4 (TOF) responses, repeat TcMEPs were obtained.
The data of MIOM were recorded by the same intraoperative neuromonitoring device (Endeaver CR IOM16, USA), including SSEPs, TcMEPs, free-run EMG, and TOF. The baseline of SSEPs and TcMEPs was established after intubation but before positioning. Alert criteria: SSEPs amplitude attenuated more than 50% and/or latency increased more than 10% compared with baseline values; TcMEPs used the standard “all or none” of amplitude; and the free-run EMG was the presence of significant EMG activity.
The preoperative neurological function of all 191 patients was evaluated using the JOA.
Magnetic resonance imaging (MRI) and X-ray of the cervical spine were collected from 191 patients. Cervical spinal cord compression was determined by MRI in sagittal and axial planes. The Pavlov values of each segment from C3 to C7 were measured by lateral cervical radiographs.
A true positive (TP) was defined as the presence of EP changes followed by a new postoperative neurological deficit, or presence of EP changes and then recovery after the intervention, but lack of a new postoperative neurological deficit. Conversely, a true negative (TN) was defined as the absence of EP changes and lack of a new postoperative neurological deficit. A false positive (FP) was defined as the presence of EP changes that were not followed by a new postoperative neurological deficit. A false negative (FN) was defined as the absence of EP changes followed by a new postoperative neurological deficit.
Analysis of related factors between the group with EP changes and the group with no EP changes associated with neck extension were analyzed using non-parametric Kruskal–Wallis' test. And P < 0.05 was considered statistically significant. All analyses were performed using SPSS 25.0.
A total of 10 patients had EP changes. The changes in five patients (2.62%) were alerts combining SSEPs and TcMEPs caused by neck extension during intubation or positioning. Five patients were diagnosed with cervical myelopathy (1/80), Ossification of the posterior longitudinal ligament (OPLL) (1/12), cervical tumor (1/26), cervical trauma (1/28), and infection (1/6) [Table 1].
Not all patients in the said five cases were repositioned after observing abnormal EP signals. In case no. 5, no repositioning was conducted; however, the EP signals recovered after decompression and there were no new postoperative neurological deficits. In case no. 1 [Figure 1] and [Figure 2], the EP signals gradually recovered after repositioning and no new postoperative neurological deficits were observed. However, the EP signals in case no. 2, no. 3, and no. 4 had not yet recovered after repositioning. In case no. 2, the procedure went on but new postoperative neurological deficits were observed so that the second stage of posterior decompression was conducted and the patient's nerve function recovered at discharge. The procedure also went on in case no. 3, and new postoperative neurological deficits were observed but the patient has not yet recovered. In case no. 4, the procedure was terminated due to a positive result of the wake-up test, and the patient's neurological function recovered within 6 months of follow-up. No FPs or FNs were documented [Table 2].
By analyzing and comparing the difference between the two groups, it was found that the average preoperative JOA scores of the group with EP changes were lower than those of the group with no EP changes (P = 0.037 < 0.05). There was no statistical significance in gender, average age, mean Pavlov ratio, and the minimum Palov ratio between the two groups (P > 0.05) [Table 3].
No significant EMG activity was observed during the pre-surgical step in all cases included in this study.
In cervical spine surgery, the position of the neck before cervical spine surgery is a high-risk step causing spinal cord injury. Some investigators have found that in cervical spine surgery, the position of the neck had more EP changes and could cause severe nerve injury than the operation itself. MIOM is sensitive and specific in detecting impending neurological injury during neck extension in cervical spine surgery. In this study, we found that MIOM could identify spinal cord injury associated with neck extension before cervical spine surgery. Active and effective interventions could prevent or reduce permanent postoperative neurological deficits. Severe spinal cord compression may be a risk factor for EP changes.,,,
The diameters of the cervical canal may be reduced significantly during neck extension.,, EP changes might occur in patients with previous cervical spine pathology because the spinal cord was compressed, aggravated by neck extension in the stenotic spinal canal., However, in this study, no statistical significance was identified in the mean Pavlov ratio and minimum Palov ratio between the two groups. It meant that the stenotic bony spinal canal was not an important risk factor of nerve injury caused by neck extension during intubation or positioning. Besides, Li et al. have found that a diagnosis of cervical spondylotic myelopathy increased the risk of an intraoperative neurophysiological alert. Four of the five positive cases were diagnosed with myelopathy. Moreover, patients with a pre-existing cord compromise had a higher risk of intraoperative neurophysiological deterioration. The preoperative computed tomography (CT)/MR images [Figure 3] showed that all the five positive cases had spinal cord compression, and the mean JOA scores were 9.40 ± 2.58 (<11.00), indicating the presence of severe myelopathy. A greater degree of preoperative compression to the cord led to a higher risk of neurological complications in patients. So, patients with severe myelopathy or preoperative spinal cord compression must be focused on monitoring the neurological injury. In addition, according to our results and experience, no EP alert was detected in anterior cervical discectomy and fusion (ACDF) with radiculopathy during positioning. For ACDF without cervical myelopathy or preoperative spinal cord compression, many investigators believed that monitoring was not necessary.,,,, Therefore, we suggested MIOM during intubation and positioning before cervical spine surgery should be used for cervical myelopathy or preoperative severe spinal cord compression, and it was not necessary for ACDF with radiculopathy routinely.
It had been proved that some reversible operations could cause injury to the spinal cord, including positioning (neck positioning, taping shoulder), surgical operations (stretch, implant), and non-surgical factors (hypotension, etc.). Re-positioning of the neck may reverse EP changes in potential and prevent permanent neurological deficits.,, The principle for the management of EP changes associated with neck extension was as follows: pause surgery, reposition, take steps to enhance perfusion of the spinal cord, elevate and maintain mean arterial blood pressure between 85 and 95 mmHg, administer methylprednisolone, and wait 20 min to see if EP changes recover or not. If partial or total recovery of EP signals is observed, continue the surgery; if no recovery of EP signals is observed and the cervical spine is unstable, including spinal instability due to trauma, continue the surgery; while the spine is stable, a wake-up test should be considered.,, However, the efficacy of these practices was never examined, and not all surgeons intervene according to the above principles once encountering the EP alert, instead they might choose other interventions based on their understanding of the causes of EP changes and experience with related cases. In case no. 5, the cervical spinal cord was compressed by a pre-existing tuberculous abscess, and the EP changes were observed when positioning the neck extension. No repositioning was conducted and the procedure was continued. Fortunately, EP signals were completely recovered after the removal of the abscess, and no new postoperative neurological deficits were observed. It was the cold abscess of tuberculosis on the ventral side of the spinal cord that caused EP changes because of the neck extension that could result in the aggravation of the spinal cord compression from it. There was no guarantee that EP signals would recover in time by repositioning after EP changes because the compression of the spinal cord was still there. Therefore, whether to reposition after EP changes should be decided according to the specific conditions of patients. It needs to be emphasized that the EP recovery could be delayed and even no recovery until the end of surgery. In case no. 1, EP signals did not recover immediately after repositioning but recovered before decompression, 1 h after the beginning surgery, and no new postoperative neurological deficits were observed. EP signals in case no. 2 had not yet recovered after repositioning, and even after continued anterior decompression so that the second stage of posterior decompression was conducted and the patient's nerve function recovered when discharged. So far, it remains to be further studied whether it is necessary to wait for EP signal recovery to continue surgery for patients with a stable cervical spine. Furthermore, both the patient and surgeon should be aware that the procedure might be forced to terminate before completion. It can be helpful for guiding decision-making at the time of EP alert that preoperative discussion with the patient about the significance of EP alerts and the emergency plans for impending spinal cord injury. Meanwhile, we should obtain informed consent from the patient.
In the five positive cases, intervention measures were taken after EP changes, except for case no. 3, who had a permanent preoperative neurological deficit, the neural function of the other four cases recovered during or after the surgery. Case no. 3, undergoing a posterior–anterior surgery, could be completely relieved of the compression theoretically; however, the postoperative recovery of neural function was not ideal. We speculated that it might be attributed to severer spinal stenosis, prolonged spinal compression, and myelopathy. Appel et al. stated that compared with the single approach, the recovery rate of EP signals after repositioning the neck was lower in the group with the combined approach because of the fact that the group with the combined approach surgery included patients with severer cervical stenosis and spinal cord compression. In contrast, a “double hit” mechanism might occur during the procedure, that is, when the spinal cord was disturbed during positioning, it became more sensitive to subsequent operations. However, because EP signals of case no. 3 had not recovered after repositioning, we had no proof of that.
At present, there are few studies on principles and procedures dealing with EP changes caused by neck positioning, and this should be a focus of future studies about this topic. Nevertheless, it can always help prevent or reduce the permanent postoperative neurological deficit by taking effective interventions when abnormal EP changes occur, regardless of whether following the conventional principles or not.
The multimodal intraoperative monitoring could identify spinal cord injury associated with neck extension before cervical spine surgery. Active and effective interventions could prevent or reduce permanent postoperative neurological deficits. Severe spinal cord compression may be a risk factor for EP changes.
Financial support and sponsorship
This work was supported by the National Natural Science Foundation of China (81801224), Guangzhou Municipal Science and Technology (201707010089), and Guangzhou Municipal Science and Technology (2017A030310219).
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]