INTRODUCTION
Intramedullary (IM) spinal cord tumors are rare diseases; however, they can lead to neurological deficits and mortality, significantly affecting the patient’s quality of life [1,2]. The primary goal of surgery for IM tumors is twofold: removing the tumor to alleviate symptoms and prevent recurrence, and preserving normal neural structures to avoid postoperative neurological deterioration [3,4]. Nevertheless, surgical resection of IM tumors poses significant challenges due to the spinal cord’s small diameter and dense population of essential neural fibers, increasing the risk of intraoperative injury [5].
Intraoperative neurophysiological monitoring (IONM) is therefore widely utilized during IM tumor surgeries to minimize neural injury while maximizing tumor resection [6]. Motorevoked potentials (MEPs) and direct wave (D-wave) monitoring represent key modalities of IONM [7,8]. MEP evaluates the functional integrity of motor pathways in real-time through stimulation of the motor cortex, yet it can be influenced by anesthetics and neuromuscular blockers and may transiently disappear during surgery [7,9]. Conversely, D-wave monitoring involves recording corticospinal tract signals directly from subdural or epidural electrodes following transcranial stimulation, providing more stable signals less affected by anesthetics [10-13]. Importantly, D-wave signals remain detectable even after the loss of MEP, serving as a complementary and reliable indicator for evaluating corticospinal tract integrity during critical surgical moments [11,14].
However, current literature lacks sufficient data on surgical outcomes and neural protection associated with D-wave monitoring in situations when MEP signals are lost. Thus, the present study aims to assess the utility of D-wave-guided surgery in IM tumor resection specifically after the loss of MEP signals, focusing on its role in preventing additional neurological deficits and enabling effective tumor removal.
MATERIALS AND METHODS
1. Study Design and Patient Selection
From January 2012 to May 2025, all patients who underwent tumor resection while receiving multimodal IONM during surgery for IM tumors at a single tertiary referral center were retrospectively analyzed. D-wave monitoring was first implemented at our institution in 2015, with most cases utilizing this method performed after 2021. Thus, patients who underwent IM tumor resection with D-wave monitoring were designated as the study group, while patients who underwent IM tumor resection without D-wave monitoring were classified as the historical comparison group. The common inclusion criteria for both groups were: (1) patients aged 18 years or older, and (2) those who experienced intraoperative MEP loss. Both groups shared an exclusion criterion of a postoperative follow-up duration shorter than 6 months. The diagnoses of patients included in the D-wave-guided surgery group were ependymoma (EPN) and cavernous angioma (CA). For an appropriate comparative analysis, patients diagnosed with tumors other than EPN, subependymoma (subEPN), or CA were additionally excluded from the historical comparison group (Fig. 1). Our institutional review board waived the requirement for informed consent and approved the study protocol and chart review (approval No. 2409-124-1574). All investigations were conducted in accordance with our institutional review board of guidelines and regulations.
2. Data Collection
The data collected included patient age at the time of surgery, sex, preoperative neurological status (categorized by motor deficits, sensory abnormalities, gait disturbance, and bowel and bladder dysfunctions), preoperative and postoperative 1 month, 3 months, 6 months, 12 months, and the last follow-up modified McCormick (MMC) scale [15,16], with functional independence defined as an MMC scale of 1 or 2, pathological diagnosis, preoperative neck disability index (NDI) [17] or Oswestry Disability Index (ODI) [18], modified Japanese Orthopaedic Association (mJOA) score [19], EuroQoL 5-dimension 5-level (EQ-5D-5L) index [20], IONM changes, postoperative motor status, complications, and follow-up duration. The tumor location was identified preoperatively using T1-weighted, T2-weighted, and contrast-enhanced T1-weighted magnetic resonance (MR) images, while the extent of resection was assessed based on intraoperative findings and immediate postoperative MR images. Tumor height was measured using sagittal T2-weighted and contrastenhanced T1-weighted MR images. Follow-up MR images were also collected to evaluate tumor recurrence.
3. Surgical Methods
All patients underwent surgery via a posterior approach. For D-wave monitoring, following laminectomy, a 1×4 strip electrode was gently inserted into the caudal epidural space of the surgical field (Fig. 2). The electrode was placed in the epidural space to avoid direct injury to the spinal cord, and to minimize interference with surgical instrument manipulation, it was positioned only on the caudal side. Durotomy and arachnoidotomy were performed, followed by median myelotomy or posterolateral myelotomy depending on the tumor’s location. Tumor resection was started with intermittent monitoring of MEP and continuous monitoring of somatosensory-evoked potential (SSEP). To prevent subtle vibrations of the spinal cord, D-wave monitoring was performed using intermittent monitoring rather than continuous monitoring. During tumor resection, neurological function monitoring was primarily based on MEP. As MEP began to deteriorate, D-wave monitoring was performed more frequently. In cases where MEP was lost, D-wave monitoring was relied upon to monitor neurological function while continuing tumor resection. After tumor resection, the pia, arachnoid, and dura were reconstructed primarily. Before fully closing the dura, the 1×4 strip electrode was removed, and the epidural dead space was eliminated by filling the intradural cavity with plas-ma solution or normal saline, creating a tamponade effect. Patients who did not undergo D-wave monitoring underwent the same procedures, except for the insertion and removal of the 1×4 strip electrode.
4. Anesthesia Protocol
Induction of general anesthesia was performed with continuous infusion of propofol (effective site concentration 4 μg/mL) and remifentanil (effective site concentration 4 ng/mL). Maintenance of anesthesia was achieved with continuous infusion of propofol (3–5 μg/mL) and remifentanil (3–6 ng/mL) using a target-controlled infusion pump (Orchestra, Fresenius, Germany). Muscle relaxants were administered only at the onset of anesthesia induction for endotracheal intubation, with no additional neuromuscular blocking agents given thereafter. During surgery, mean blood pressure was continuously monitored through an arterial pressure measuring device and kept within ±20% of preoperative value. Body temperature was continuously monitored using an esophageal stethoscope with a thermistor. A forced air warming device was used to maintain normothermia.
5. Intraoperative Neurophysiological Monitoring
IONM was conducted using the Eclipse workstation (Axon System, Hauppauge, USA). MEP was elicited through transcranial electrical stimulation, with electrodes inserted at the C3´ and C4´ positions, which are located 1–2 cm anterior to C3 and C4 of the international 10–20 system. The stimulation intensity ranged between 350–500 V with a train of 5–7 pulses. The pulse duration ranged from 50–75 μsec, and the interstimulus interval was 0.2 msec (500 pulses per second). Monitoring was performed on the deltoid (D), biceps (B), triceps (T), thenar muscle (TH), vastus medialis, tibialis anterior (TA), gastrocnemius, abductor hallucis (AH), and sphincter muscles. Low-frequency filtering was set at 30 Hz, high-frequency filtering was utilized at 1,500 Hz, and a notch filter was applied at 60 Hz. Baseline MEPs were obtained approximately 30 minutes to 1 hour after a single dose of muscle relaxant was administered during intubation. No additional muscle relaxants were administered thereafter. Real-time free-running electromyography was also monitored concurrently. For SSEP, surface stimulation electrodes were positioned on the median nerve at the wrist area for upper limb monitoring, and the posterior tibial nerve at the ankle for lower limb monitoring. The stimulation intensity was adjusted to 18–25 mA for the median nerve and 28–35 mA for the posterior tibial nerve, with a stimulation frequency of about 2.8 Hz for both nerves. Recording needle electrodes for SSEP were placed at positions C3 and C4 (according to the international 10–20 system) for upper extremity measurements, and at Fpz-Cz (according to the international 10–20 system) for lower extremity measurements, without the use of additional SSEP filtering. D-wave monitoring was performed for IM spinal cord tumors located at or above the T11 level. The recording electrode, a 1×4 strip electrode, was placed in the distal epidural space of the surgical field and used for intermittent monitoring via transcranial electrical stimulation, in the same manner as MEP. The stimulation intensity was set between 350–500 V with a train of 3–5 pulses, and filtering was conducted from 20 Hz to 3,000 Hz. No averaging was performed, and recording was conducted through channels 1–2 and 1–3.
6. Alarm Criteria
The alarm criteria for MEP were defined as a decrease in amplitude to 0% (all or none methods) in any kind of muscle. This threshold was selected based on prior studies suggesting that the complete loss of MEP is more associated with long-term postoperative neurological deficits compared to partial reduction [21,22]. In addition, this study aimed to investigate the supplementary role of D-wave monitoring specifically in cases where MEPs are completely lost. For SSEP, the alarm criteria were defined as a delay in latency of more than 10% or a decrease in amplitude of more than 50% [23]. The alarm criteria for D-wave were defined as a decrease in amplitude of less than 50% [10]. During surgery, if MEP, SSEP, or D-wave decreased to the alarm criteria, mean blood pressure was increased to enhance spinal blood perfusion, the surgical field was irrigated with warm saline, and body temperature was increased using air warmer and warm blanket. Additionally, 5 mg of dexamethasone was injected based on the surgeon’s preference. These interventions aimed to facilitate the recovery of decreased MEP, SSEP, and D-wave.
7. Outcome Measurement
The patients were divided into 2 groups: those who underwent D-wave-guided surgery following MEP loss and those who underwent non–D-wave-guided surgery following MEP loss, and a comparison was made between the groups. Functional improvement was defined as any improvement in the MMC scale at the last follow-up compared to the preoperative MMC scale. To assess the longitudinal changes in MMC, we compared MMC scales at each follow-up time point. Subgroup analyses were performed according to tumor pathology and the presence or absence of D-wave attenuation to evaluate differences in MMC trends. Additionally, extent of resection, complications, tumor recurrence, and the need for secondary surgery were compared between groups. Finally, changes in IONM findings were analyzed in relation to preoperative and postoperative motor grades, and the degree of motor recovery at the last follow-up was assessed.
8. Statistical Analysis
Descriptive statistical analysis was performed. The comparison of categorical variables was conducted using the chi-square test or Fisher exact test, while the comparison of continuous variables was performed using the Mann-Whitney U-test. Categorical variables are presented as number (%), and continuous variables are presented as median (interquartile range). All statistical analyses were conducted using IBM SPSS Statistics ver. 29.0 (IBM Co., USA), and a p-value <0.05 was considered statistically significant.
RESULTS
1. Baseline Characteristics
The median age at surgery was slightly younger in the D-waveguided surgery group compared to the non–D-wave-guided surgery group (39.0 years vs. 44.5 years), but the difference was not statistically significant (p=0.156). Regarding preoperative symptoms, pain was significantly more common in the non-D-wave-guided surgery group (82.1%) compared to the Dwave-guided surgery group (44.4%, p=0.041). Other preoperative symptoms including motor deficits, sensory abnormalities, gait disturbance, and bowel and bladder dysfunctions did not significantly differ between groups. Preoperative functional metrics, including NDI or ODI, mJOA score, EQ-5D-5L index, were comparable between the 2 groups. Additionally, there was no significant difference between groups regarding the proportion of patients with preoperative functional independence. Tumors were predominantly located at the cervical level in both groups, with similar distributions observed for cervicothoracic and thoracic locations. Pathologically, the majority of tumors in both groups were EPNs (all World Health Organization grade 2), with no significant difference in the distribution of tumor pathology between the groups. The median tumor height was numerically greater in the D-wave-guided surgery group (38.0 mm vs. 27.9 mm), but this difference was not statistically significant (p=0.336) (Table 1).
2. Clinical Outcomes
At the last follow-up, functional improvement was significantly higher in the D-wave-guided surgery group (66.7%) compared to the non–D-wave-guided surgery group (17.9%, p=0.011). Among patients with EPN, the rate of functional improvement remained significantly higher in the D-wave-guided surgery group compared to the non–D-wave-guided surgery group (74.4% vs. 9.1%, p=0.003). In contrast, patients with CA revealed no significant differences between the 2 groups. However, it is important to note that the median follow-up duration was significantly shorter in the D-wave-guided surgery group (12.6 months) than in the non–D-wave-guided surgery group (36.0 months, p=0.001) (Table 2). The changes of MMC scale were analyzed over time between groups (Fig. 3). Although no statistically significant differences were observed at most time points in the overall cohort, patients in the D-wave-guided surgery group showed a trend toward better MMC scale at the last follow-up. Among patients with EPN, significantly better MMC scale was observed in the D-wave-guided surgery group at postoperative 3 months and postoperative 6 months, with a favorable trend thereafter. In patients with CA or in subgroup analyses stratified by the degree of D-wave amplitude reduction, no significant differences were observed, though patients with preserved D-wave amplitude (amplitude ≥50% of baseline; D ≥50%) tended to maintain more stable or improved MMC scale. Gross total resection was achieved in all patients of the D-wave-guided surgery group and in most patients of the non– D-wave-guided surgery group (96.4%, p=1.000). Complication rates were low and comparable between groups. Notably, although the strip electrode for D-wave monitoring was placed in the caudal epidural space, no patients in the D-wave-guided surgery group experienced epidural hematoma after surgery. Recurrence or regrowth occurred only in the non–D-waveguided group (7.1%, p=1.000) (Table 2).
3. IONM Changes and Postoperative Motor Status
Among patients undergoing D-wave-guided surgery, the D-wave amplitude was not attenuated in 5 patients (55.6%), attenuated but remained ≥50% of baseline in 2 patients (22.2%), and attenuated to <50% of baseline in 2 patients (22.2%). Preoperative motor status, indicated by motor grade ≤3, did not differ significantly between the 2 groups (11.1% vs. 3.6%, p=0.432). Postoperative motor grade ≤3, evaluated immediately after surgery, was observed in 66.7% of patients in the D-waveguided surgery group versus 39.3% in the non–D-wave-guided surgery group, although this difference did not reach statistical significance (p=0.251). At the last follow-up, the majority of patients in both groups were self-ambulatory without external aids, with no significant difference between groups (88.9% vs. 89.3%, p=1.000) (Table 3).
4. Representative Case
A 39-year-old male patient was diagnosed with an IM tumor extending from C2 to C5, as confirmed by MR images (Fig. 4A). The tumor was resected using multimodal IONM including D-wave monitoring. After midline myelotomy, tumor dissection proceeded along lateral and ventral margins (Fig. 4B and C). During resection, significant decreases in bilateral MEP amplitudes occurred (right TA: 20%, right T, D, AH: 5%–10%; left D, T, TA, AH: 20%, left TH: 50%; left B: loss; Fig. 4D). As MEP amplitudes declined, D-wave monitoring frequency increased, initially decreasing to 40%, transiently recovering to 80%, then dropping to 20% until resection completion (Fig. 4E). The tumor was completely removed, with pathology confirming EPN. Postoperative motor deterioration occurred but resolved fully by 12.6 months.
DISCUSSION
In this study, we evaluated the efficacy of D-wave-guided surgery in IM spinal cord tumor resection when MEP loss occurred, by comparing the outcomes with those of non–D-wave-guided surgery. Although the follow-up period in the D-wave-guided surgery group was relatively shorter compared to the non–Dwave-guided surgery group, the rate of functional improvement at the last follow-up was higher in the D-wave-guided surgery group. Additionally, there were no significant differences between the 2 groups in terms of complications, extent of resection, or recurrence rates.
In spinal tumor surgery, IONM is widely used to minimize neurological deficits and maximize tumor resection [24,25]. MEP is one method of monitoring the motor pathway, and intraoperative MEP attenuation is highly associated with motor deterioration immediately after surgery [26,27]. D-wave is monitored through the direct activation of corticospinal tract fibers and serves as an independent prognostic factor for long-term functional outcomes [12,28]. Kimchi et al. [29] presented the sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) of D-wave in predicting postoperative functional outcomes at 3 time points: 1 day after surgery, 6 weeks after surgery, and at the last follow-up, for surgeries on IM spinal cord tumors (sensitivity 40%, 33%, 100%; specificity 100%, 83%, 90%; NPV 70%, 71%, 100%; PPV 100%, 50%, 50%, respectively). MEP loss can occur during IM spinal cord tumor surgery, and since MEP loss precedes D-wave attenuation, D-wave can serve as a guide for tumor resection in situations where MEP loss is present [14]. In this study, patients who underwent D-wave monitoring completely resected the remaining tumor through D-wave-guided surgery following MEP loss. These patients experienced shortterm worsening of motor function immediately after surgery due to MEP loss, but showed a tendency for gradual recovery over time. Although some patients experienced a decrease in D-wave amplitude, the rate of functional improvement at the last follow-up was higher in the D-wave-guided surgery group compared to the non–D-wave-guided surgery group. Rather than resecting the tumor without monitoring for additional neural damage after MEP loss, using D-wave monitoring to assess the integrity of the corticospinal tract while removing the tumor may help functional improvement.
In addition to utilizing D-wave monitoring to guide resection after MEP loss, efforts should also be directed toward preventing MEP loss during surgery. Common causes of intraoperative MEP deterioration include direct neural injury, excessive traction or compression, hypotension, reduced spinal cord perfusion, and hypothermia [30,31]. To mitigate these risks, it is essential to avoid excessive manipulation of neural tissue, maintain adequate mean arterial pressure, ensure sufficient spinal cord perfusion, and regulate intraoperative body temperature [31,32]. Specific procedural considerations may also contribute to the prevention of MEP loss depending on tumor pathology. In EPN, strict adherence to midline myelotomy along the spinal cord’s median septum is essential [33]. Furthermore, when dissecting the ventral aspect of the tumor where a feeder vessel is present, meticulous preservation of the normal spinal cord tissue and vasculature is of particular importance [34]. In cases of CA, spinal cord injury can be minimized by careful dissection that preserves the gliotic rim [35,36]. When performing posterolateral myelotomy, particular attention should be paid to the corticospinal tract anterior to the dorsal rootlets [37]. For ventrally located lesions, it is critical to avoid injury to the anterior spinal artery [38].
D-wave is robust under general anesthesia and unaffected by muscle relaxants, offering the advantage of monitoring the integrity of the corticospinal tract. Unlike MEP, which can suddenly decrease, the D-wave decreases gradually. Therefore, if a decrease in the D-wave is detected, it offers the advantage of being able to identify the cause of the decrease and make the necessary corrections [39]. However, D-wave monitoring has several limitations, including the inability to provide information on sensory conduction, the requirement that monitoring can only be performed at levels T10–11 and above, the lack of specific muscle group information, potential difficulty in electrode placement due to scarring or spinal cord swelling, and the possibility of inaccurate information if the electrode position shifts [10,11,40]. In this study, considering the risk of inducing subtle vibrations in the spinal cord during delicate tumor dissection when performing continuous monitoring, intermittent monitoring of the D-wave was used instead. Tumor resection was initially guided by MEP, and when MEP loss occurred, D-wave intermittent monitoring was performed more frequently to guide the resection of the remaining tumor. D-wave monitoring is typically performed by positioning a catheter-type electrode in the rostrally and caudally subdural or epidural space at the surgical site. Placing the electrode in the subdural space offers the advantage of obtaining a larger amplitude with relatively lower impedance; however, electrode insertion can be difficult due to spinal cord swelling or scar tissue [41,42]. In our country, catheter-type electrodes for D-wave monitoring are not available; therefore, in this study, a strip electrode was used as an alternative. The strip electrode was gently inserted into the epidural space to minimize the risk of spinal cord injury and was positioned on the caudal side of the surgical field to minimize interference with surgical instrument manipulation. Compared to conventional D-wave monitoring techniques, we do not believe that the use of a 1×4 strip electrode in our study offers a distinct advantage or disadvantage in terms of signal gain or technical characteristics. However, this method enabled effective acquisition of orthograde signals even with monitoring limited to the caudal side, and may serve as a viable alternative in cases where traditional catheter-type electrodes are not available. We believe this represents a practical advantage in terms of clinical accessibility. An appropriate D-wave amplitude was obtained in all patients who underwent D-wave monitoring. D-wave attenuation of less than 50% of baseline was observed in 2 patients. However, since the electrode was positioned only on the caudal side, it cannot be excluded that D-wave attenuation may have been a false alarm due to factors such as cerebrospinal fluid drainage. Despite the D-wave attenuation, the tumor was completely resected based on the clear margins of the EPN observed under the microscope, and no tumor was left behind. Postoperatively, there was temporary worsening of motor function, but recovery was achieved. Furthermore, although a larger strip electrode was used in the epidural space compared to a catheter-type electrode, no patients developed postoperative epidural hematoma.
This study, being a retrospective analysis conducted at a single institution, has several limitations. First, the small sample size limits the generalizability of the findings. Second, D-wave monitoring was introduced relatively later, resulting in differences in follow-up periods between the 2 groups, thereby posing the potential for bias related to historical comparison and limiting the interpretation of the results. Nonetheless, to the best of the authors’ knowledge, this is the only study to report outcomes using D-wave monitoring with a strip electrode. It provides important evidence for evaluating the practicality and efficacy of D-wave monitoring with a strip electrode, highlighting the need for larger-scale prospective studies to validate and further investigate these findings.
CONCLUSION
In patients undergoing IM spinal cord tumor resection, Dwave-guided surgery enabled continued tumor removal even after MEP loss, with favorable functional outcomes and no increase in complication or recurrence rates. Although immediate postoperative motor deficits were more common in the Dwave-guided surgery group, long-term recovery was comparable. D-wave monitoring thus offers a practical and safe method for preserving neural integrity in high-risk situations, particularly when standard MEP is lost. These findings support its role as a complementary modality in IONM and warrant further validation through larger prospective studies.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Print
