Intraoperative neuromonitoring of spinal cord function is an essential component of spinal deformity correction surgery [1]. Motor and sensory functions of the spinal cord are assessed intraoperatively using transcranial motor evoked potential (TcMEP) and somatosensory evoked potential (SSEP), respectively [2,3].

Amplitude-based alarm criteria for TcMEPs are routinely used to detect physiological loss of motor function. A reduction in amplitude of >80% from baseline and/or a threshold elevation of ≥100 V is considered a reliable indicator of impending motor pathway compromise, particularly when consistent across multiple trials and muscle groups. However, TcMEPs are highly sensitive to various anesthetic, physiological, and technical factors, such as hypotension, hypothermia, changes in anesthesia depth, and neuromuscular blockade. Therefore, interpretation of amplitude changes must account for these potential confounders before attributing them to surgical insult. When TcMEP amplitude loss is confirmed to be of surgical origin, immediate corrective action is warranted to potentially reverse neurological dysfunction within the operative window [4]. Intraoperatively, a decrease in TcMEP amplitude may occur either abruptly or gradually over the course of the surgical procedure. If corrective measures are initiated promptly following the detection of physiological compromise, functions may be restored either during the operation (rescue) or, in some cases, after a delay of days to weeks. Upon a surgical alert, the surgical team halts further intervention, examines the surgical field, and promptly reviews the preceding surgical steps to identify potential causes of spinal cord compromise. In some cases, a specific maneuver or step responsible for the functional deterioration can be identified; in others, no clear cause of cord compression may be apparent. The surgical response may include spinal stabilization, reduction of correction, removal of rods or screws, or even aborting the procedure with plans to return at a later date. The specific action taken depends on whether TcMEP changes are reversible and the overall surgical context [5].

In the event of non-reversal of TcMEP changes, localizing the site of neural compromise becomes a critical objective. A recent modification of the descending neurogenic evoked potential (DNEP) technique—known as dynamic spinal cord mapping (DSCM)—has been proposed to address this issue [6]. DSCM involves the use of epidurally placed catheter electrodes to stimulate the spinal cord at various levels, with evoked responses recorded from a peripheral nerve in the lower limb to help identify the level of cord compression or dysfunction. It is pertinent to mention here that both DNEP and its modification, DSCM, essentially assess the sensory pathway of the spinal cord, and sensory dysfunction is an unreliable surrogate for motor pathway compromise. As such, they offer only indirect assessments of motor pathway integrity. Moreover, epidural catheter placement may not be feasible or appropriate in all surgical contexts. The reliance on sensory conduction as a surrogate for motor function is particularly problematic, as sensory pathways can remain intact even in the presence of significant motor deficits. This limitation can lead to diagnostic uncertainty, reducing the utility of these methods in intraoperative decision-making [7–9]. Trans-spinal myogenic evoked potentials (TsMEP) address these shortcomings by providing a direct evaluation of motor tract function without requiring additional instrumentation.

Through a case of kyphoscoliosis, we describe the applicability and benefits of a novel technique of transpinal stimulation of spinal cord motor fibers at different levels with recording of myogenic evoked potentials (TsMEPs) to localize the site of motor pathway compromise. TsMEPs represent a novel intraoperative technique involving direct electrical stimulation of the spinal cord via preplaced pedicle screws, with myogenic responses recorded from peripheral muscles. Unlike conventional pedicle screw stimulation, which primarily assesses screw placement accuracy via triggered electromyography (EMG) and reflects sensory nerve root proximity or breach, TsMEPs are designed to evaluate the functional integrity of spinal motor pathways. The key distinctions lie in the stimulation protocol (train of pulses vs. single pulse), the target (central motor tracts vs. peripheral nerves), and the outcome measure (compound muscle action potentials vs. threshold response or EMG firing). TsMEP thus provides a unique modality for intraoperative localization of motor deficits, particularly when transcranial motor evoked potentials (TcMEPs) are lost or inconclusive. Notably, TsMEPs can be elicited from multiple vertebral levels intraoperatively. By assessing the presence or absence of evoked responses at each level, the technique enables electrophysiological localization of the site of motor pathway disruption. This capability is particularly valuable when TcMEPs are lost and localization is critical to guide surgical decision-making.

A 14-year-old male with congenital left convex thoracolumbar kyphoscoliosis (Cobb angle, 83.1°), neurologically intact on admission, was scheduled for deformity correction and posterior spinal fusion.

Intraoperative neuromonitoring was performed as follows. To minimize the impact of anesthetic variables on MEP recordings, the standard protocol of total intravenous anesthesia was employed throughout the procedure, with continuous monitoring of the depth of anesthesia using the bispectral index (BIS). Neuromuscular blocking agents were avoided after induction to preserve neuromuscular transmission. Control recordings were obtained from the bilateral brachioradialis (BR), and motor function was monitored using electrodes placed in the bilateral rectus abdominis (RA), vastus lateralis, tibialis anterior, extensor hallucis longus, and abductor hallucis. Stimulation was delivered using a train of biphasic square-wave pulses, each with a duration of 75 microseconds, in fast-charge mode [10–12]. Baseline responses were recorded after surgical exposure and before the placement of pedicle screws.

Fig. 1 shows a time stack of TcMEP recordings at various stages of the surgery. Following pedicle screw placement on both the concave and convex sides, neuromonitoring responses were recorded. During the cantilever reduction on the convex side, a progressive decline in the TcMEP amplitude was noted. Initially, this reduction was seen in both control and monitored muscles (Stack B of Fig. 1) and was attributed to increased depth of anesthesia due to the addition of the inhalational agent isoflurane (BIS, approximately 24). Throughout the surgery, episodes of signal loss in both control and monitored muscles were found to correlate with deeper anesthesia levels (BIS, approximately 24) and were reversible upon lightening the anesthesia (BIS approximately 40), confirming the anesthetic sensitivity of TcMEP responses. These measures ensured that the recorded MEP responses reflected true motor pathway conduction, minimizing the confounding influence of pharmacological suppression. Following the reduction in the depth of anesthesia (BIS, approximately 40), TcMEP signals recovered in the control muscles only, but not in the monitored muscles (Stack C of Fig. 1). After consultation with the anesthesia and surgical teams, and ruling out confounding factors, a surgical alert was sounded. The reduction maneuver was reversed, and a temporary rod was placed on the concave side for stabilization. The MEPs in the monitored muscles showed recovery in 5–20 minutes (Stack D of Fig. 1), following which a slow step-wise correction was attempted with more frequent MEP monitoring to ensure neurological safety. Again, a decrease in MEPs, predominantly on the right side, was observed (Fig. 2A). The procedure was then concluded with partial recovery of signals noted at the time of skin closure (Fig. 2B). A few hours after extubation, the power in the lower limbs was 5/5 with no neurological deficit.

On the 3rd postoperative day, the patient developed a complete neurological deficit in both lower limbs over a few hours. Re-exploration was performed on the next day. Intraoperative MEPs demonstrated preserved responses in the control muscles but not in the lower limb muscles and RA (Fig. 3).

Based on computed tomography imaging and the absence of response from RA, the level of spinal cord compression was inferred to be above T9. To determine the level of compression electrophysiologically, a technique involving direct stimulation of the cord via pedicle screws was employed.

The temporary rod was removed. A separate set of similar screws were tested ex vivo for electrical conductivity. The stimulation set up was adjusted to deliver a train of 8 biphasic square-wave pulses, each with a duration of 0.5 msec and an interstimulus interval of 3 msec, in current mode. Two needle electrodes were placed on the lowest screws on either side. Stimulation was initiated at 15 mA and gradually increased. At the D8–D9 pedicle level, evoked responses were observed in all monitored muscles, except the control muscle (Fig. 4A). The stimulation threshold was 30 mA, with optimal near-suprathreshold responses recorded at 40 mA. However, a similar response was not observed at the D3–D4 level (Fig. 4B).

The screw on the right side at the D7 level was removed, and no medial breach of the pedicle was identified. Decompression in the form of laminectomy was performed at D7, and both pedicles were drilled. TcMEP and trans-spinal stimulation through pedicle screws at the D3–D4 level were then attempted. Following the correction, TcMEPs continued to show no evoked response in the monitored muscles, although responses remained present in the control muscles. However, trans-spinal stimulation through pedicles elicited responses at a higher intensity of 60 mA, with a gradual decrease in stimulation threshold over the following 45 minutes. As shown in Fig. 5, robust responses were recorded in the monitored muscles on the left side compared to the right side via pedicle screw stimulation at D3–D4. The reappearance of TsMEPs was considered as a positive sign of improving motor pathway function, and surgical closure was completed. Immediately after extubation following the second-stage surgery, the patient demonstrated no motor power in the lower limb.

On further evaluation, a gradual improvement in motor power in both lower limbs was observed. By postoperative day 9 following the second surgery, muscle power had improved to 2/5. The patient subsequently underwent a third procedure, during which an in-situ fusion was performed and the temporary rod was replaced with final rods. At the time of discharge, motor power in both limbs had improved to 3/5. Upon outpatient follow-up 1 month after the index surgery, power had further improved to 4/5 bilaterally.

In the event of a neurophysiological alarm and subsequent surgical alert, the surgical team must make an on-table decision regarding the appropriate course of action. These may range from stabilization of the spine to removal of all implants and instrumentation. TsMEP offers a valuable modality for localizing the site of motor pathway dysfunction within the spinal cord, thereby enabling more informed and targeted surgical decision-making. Additionally, TsMEPs may serve as a useful tool for monitoring the effects of corrective measures taken after the loss of TcMEP signals, allowing for real-time assessment of motor function recovery and aiding in the localization of spinal cord compromise.

This technique is a modification of an earlier technique first described in 1988, which involved direct stimulation of the spinal cord using stimulating electrodes placed in the spinous processes and recording of evoked potentials from peripheral nerves [10]. Although initially intended to monitor motor tract function, subsequent studies showed that evoked potentials in peripheral nerves were predominantly sensory in origin [7,8]. A more recent modification of this technique involves direct stimulation of the spinal cord via epidurally placed catheters, enabling localization of spinal cord dysfunction [6]. TsMEP employs a train of pulses sufficient to activate lower motor neurons, making it similar to the stimulation protocol used in transcranial MEPs. The responses recorded following trans-spinal stimulation were confirmed to be motor in nature based on several electrophysiological characteristics. First, the evoked waveforms exhibited multiphasic morphology with expected latencies (20–35 ms) from target muscle groups, consistent with direct activation of motor pathways with a single train of stimulation. Second, the responses were reproducible across multiple trials and were abolished under deep anesthesia or neuromuscular blockade, supporting their origin from motor neuron activation. Third, the latencies of the responses corresponded to expected conduction times from the spinal stimulation site to the respective limb muscles, clearly distinguishing them from sensory or reflex-mediated signals. Collectively, these features affirm the myogenic nature of the responses elicited through TsMEP. These TsMEP responses were rapidly elicited intraoperatively following stimulation, allowing for near-immediate interpretation and correlation with surgical events. This responsiveness facilitated a timely localization of motor pathway compromise and enabled the assessment of functional recovery following decompression. Although not instantaneous, the latency (20–35 ms) and high reproducibility of TsMEP responses underscore their practical utility for real-time intraoperative decision-making.

Quantitative impedance measurements of the screw pathway were not recorded in this case; however, ex vivo testing was performed to confirm the electrical conductivity of the screws before their intraoperative use for stimulation. While pedicle screws are routinely used in triggered EMG to assess screw placement accuracy, their application for delivering multi-pulse stimulation to evaluate central motor pathways—as implemented in TsMEP—is novel. To the best of our knowledge, no prior published reports have described the use of pedicle screw stimulation in this manner to localize motor tract dysfunction following TcMEP loss. This highlights the innovative nature of our approach and its potential to expand the application of standard spinal instrumentation in functional neuromonitoring.

In the present case, stimulation was delivered via preplaced screws in the pedicle; however, this technique can be adapted to stimulate the cord using ball-tip monopolar probes inserted into the pedicle tract in cases where removed screws are unavailable. Additionally, similar spinal cord stimulation may be performed through adjacent spinous processes, offering flexibility in situations where pedicle access is limited or contraindicated.

While TsMEP offers a novel and practical approach for assessing motor pathway integrity, certain limitations must be acknowledged. The current technique necessitates temporary removal of the connecting rod between pedicle screws to prevent spread of the current via the rod, which may not be feasible in all intraoperative scenarios, especially in cases involving unstable deformities. Furthermore, the lack of insulation on standard pedicle screws can result in diffuse current dispersion, potentially reducing stimulation specificity and introducing electrical artifacts. Signal interference from adjacent hardware or poor contact at the screw-skin interface may also affect response quality. These challenges highlight the need for future development of insulated or selectively conductive screw systems tailored for neuromonitoring applications. Until such modifications are available, meticulous technique and cautious interpretation are essential to ensure the reliability of TsMEP recording.

However, with the development of rods coated with insulated material—without compromising their mechanical properties—this technique could potentially be applied without the need to remove the rod during stimulation. In the current approach, electric current disperses broadly from one screw to another across the entire exposed surface of the screws. The design of pedicle screws with selective insulation, leaving only specific parts conductive, would allow for more targeted and controlled current delivery. Furthermore, adapting this technique for use in non-deformity cases and cervical spine procedures could broaden its applicability, provided that anatomical considerations and hardware access allow for safe stimulation. Combining TsMEP with peripheral nerve recording strategies, as employed in DNEP, may enable its use even under neuromuscular blockade, thereby enhancing its clinical versatility. With continued advancements, TsMEP holds promise both as a rescue modality and as a stand-alone technique for intraoperative motor pathway monitoring.

Unlike the recently described technique of DSCM, TsMEP does not require the placement of additional catheter electrodes in the epidural space and enables direct assessment of the integrity of the motor tracts.

To conclude, trans-spinal motor evoked potential is a promising technique that warrants further exploration, standardization, and use in situations of non-reversal of TcMEP signals during surgery or even as a stand-alone technique for monitoring of spinal cord motor function.