Assessment of the Postoperative Progression in Patients With Intramedullary Spinal Cord Tumors Based on the Severity of Preoperative Gait Disturbance

Abstract

Study Design

Retrospective comparative study.

Objectives

This study aimed to determine whether the severity of preoperative gait disturbance remains after surgical resection in patients with intramedullary spinal cord tumors (IMSCTs), and to identify any factors influencing poor improvement in postoperative gait disturbance.

Methods

The study included a total of 128 patients with IMSCTs requiring surgical excision between 2006 and 2019. Based on the degree of preoperative gait disturbance assessed by the modified McCormick scale (MMCS) grade, patients were categorized into Mild (I-II) and Severe (III-V) groups. The mean postoperative follow-up period was 55.5 ± 34.3 months, and demographic and surgical characteristics were compared between the two groups.

Results

Significant differences were observed in age at surgery, tumor location, tumor size, estimated blood loss, intraoperative motor-evoked potential disappearance, extent of resection, and tumor histopathology between the Mild and Severe groups. In the Mild group, at the final follow-up, only 7.3% of patients experienced improvement, 56.0% showed no changes, and 36.7% experienced deterioration. Conversely, in the Severe group, 26.3% of patients experienced improvement, 31.6% showed no changes, and 42.1% experienced deterioration. Tumor location and age at surgery were identified as factors correlated with poor improvement in postoperative gait disturbance in the Mild group.

Conclusions

Irrespective of the preoperative gait disturbance degree, approximately 40% of patients with IMSCTs experienced deterioration in gait after tumor resection. For preoperative MMCS grade I-II cases, older age at surgery and thoracic IMSCTs would be important factors associated with poor improvement in postoperative gait disturbance.

Keywords

intramedullary spinal cord tumors modified McCormick scale gait disturbance tumor resection Japan orthopaedic association score

Introduction

Intramedullary spinal cord tumors (IMSCTs) are relatively rare, with an incidence of approximately 1.1 cases per 100 000 people, constituting 2-4% of all neoplastic lesions in the central nervous system.^1-3 The majority of these tumors necessitate treatment due to the manifestation of sensory and/or motor disorders, including pain, numbness, muscle weakness in the extremities and trunk, and gait disturbances. Specifically, the onset of gait disturbance and the progression of lower limb paralysis in the preoperative period significantly diminish the patient’s quality of life, representing a serious condition.^4-7 Despite surgical resection being the primary treatment for IMSCTs, concerns arise due to the tumor’s origin from the intramedullary spinal cord, raising apprehensions about the surgical procedure potentially invading the normal spinal cord and exacerbating postoperative neurological symptoms.^5,8,9 The decision on whether to perform surgical resection when preoperative neurological symptoms are mild or more severe, such as the presence of gait disturbance, remains inconclusive and requires careful consideration.

Recent studies have reported surgical outcomes for each histological type of IMSCTs and the prognostic factors influencing them, highlighting tumor histology as the most critical predictor of neurological outcomes after surgical resection.^10-16 However, these reports have predominantly focused on individual histological types, with only a limited number providing a comprehensive analysis of all histological types together. In addition, they did not specifically address preoperative degree of gait disturbance and solely evaluated the involvement of preoperative neurological status. Therefore, evaluating the postoperative progression in patients with IMSCTs, particularly those with mild or severe preoperative gait disturbance, and comprehending the factors influencing the exacerbation of postoperative gait disturbance, are crucial for confirming the advantages of surgical procedures in this patient population.

The objective of this study was to examine the surgical and clinical outcomes following tumor resection in patients with IMSCTs, and to evaluate whether the severity of preoperative gait disturbance remains postoperatively.

Materials and Methods

Study Design, Participant Demographics, and Characteristics

This retrospective comparative study included consecutive 128 patients with primary IMSCTs at the cervical level (central region between C1 and C7) or thoracic area (T1 to T12), requiring surgical excision at our institution between 2006 and 2019. The inclusion criteria were as follows: 1) patients diagnosed with IMSCTs who underwent tumor resection without facetectomy or fixation with spinal instrumentation, 2) those without neuromuscular disorders, trauma, congenital vertebral deformity, or other pathological conditions, 3) those without a previous history of spinal surgery, and 4) those with a minimum of 3-year follow-up after tumor resection. Patients who underwent postoperative radiotherapy/chemotherapy or had a dumbbell-shaped tumor, intradural extramedullary spinal cord tumor, or extradural spinal cord tumors, such as metastatic tumors, were excluded. All surgeries for the resection of IMSCTs were conducted by the same skilled orthopedic spine surgical team. In our institute, for cases of IMSCTs confined within five vertebral bodies, tumor removal is performed only by decompression without spinal fixation. As all eligible patients with IMSCTs met these conditions, no cases were treated with spinal fusion. Intraoperative motor evoked potentials (MEP) were necessarily performed at the time of tumor removal for all patients with IMSCTs in the study, irrespective of tumor localization or size. Data on demographic characteristics (age at surgery, sex), imaging findings (tumor location, size [sagittal, axial view max]), and surgical characteristics (surgical time, estimated blood loss [EBL], approach for tumor resection, intraoperative MEP disappearance, resection, and histopathology of tumor) were retrospectively obtained by reviewing patients’ medical records and magnetic resonance imaging (MRI) findings.

This study received approval from the Ethics and Institutional Review Board at Keio University School of Medicine (Approval number: #20110142), and all participants provided their informed consent for inclusion in the study before undergoing tumor resection.

MRI Results

Preoperative MRI was conducted to validate the extent of tumor location and size. The maximum size of each IMSCT was examined on sagittal or axial views on T2-weighted MRI. An orthopedic spine surgeon not directly involved in the surgeries measured the tumor size at three different time points. The average of these measurements was used as the tumor size. The intraobserver reliabilities, examined using the intraclass correlation coefficient (ICC) (1, 1) and ICC (1, 3) formulas, were 0.86 and 0.95, respectively. To examine interobserver error, 100 randomly chosen axial MRI findings were independently evaluated by three orthopedic spine surgeons not directly involved in the surgeries. The interobserver reliabilities, tested using the ICC (2, 1) and ICC (2, 3) formulas, were 0.85 and 0.94, respectively. Postoperatively, MRI was performed to confirm the extent of tumor resection, categorized into two types according to the operative records: gross total resection (GTR) and partial resection (PR). The standard definition of GTR was employed, indicating complete removal of the tumor, confirmed by the absence of microscopic remnants and no neoplasm at the tumor’s origin site. Conversely, PR was defined when a small tumor fragment or an extraforaminal or paravertebral portion remained unexcised, based on the documented removal of 50% to 99% of the tumor, as confirmed by intraoperative ultrasonography or postoperative MRI findings.¹⁷ An independent radiologist from our institution reviewed the MRI images each time to evaluate the extent of tumor resection.

Assessment of Neurological Function and Clinical Outcomes

All patients with IMSCTs in this study were evaluated for the degree of gait disturbance at “pre-operation,” “immediately after surgery,” and “the final follow-up” using the modified McCormick scale (MMCS; grade I, normal gait; grade II, mild gait disturbance not requiring support; grade III, gait disturbance requiring support; grade IV, with assistance; and grade V, wheelchair needed).^18,19 Additionally, to determine whether the degree of preoperative gait disturbance can affect clinical outcomes after surgery, the participants were divided into two groups according to preoperative MMCS. Patients with preoperative MMCS grade I-II were included in the Mild group, while patients with preoperative MMCS grades III-V were assigned to the Severe group. The outcome variables were compared between the two groups, and changes in postoperative MMCS were analyzed.

Statistical Analysis

All data were presented as mean ± standard error of mean, and categorical variables were expressed as percentages. Demographic and surgical characteristics data based on the degree of preoperative gait disturbance were compared at “preoperation,” “immediately after surgery,” and “the final follow-up” using the Kruskal–Wallis test, followed by the Dunn test for multiple comparisons. The Mild and Severe groups were assessed using the Mann–Whitney U test or Fisher’s exact test, respectively. A P-value of <0.05, <0.01, or <0.001 was considered statistically significant. Additionally, univariate analysis was conducted on all variables to identify potential factors influencing improvement in postoperative gait disturbance in the Mild group. Variables with a P-value of < 0.25 in the univariate analysis were included in the multiple logistic regression analysis. P-values of < 0.05 with a 95% confidence interval were considered statistically significant in the univariate and multivariate logistic regression analyses. All statistical analyses were performed using Statistical Package for the Social Sciences software version 29.0 (IBM Corp., Armonk, NY, USA).

Results

This study comprised a cohort of 71 men and 57 women, with a mean age at the time of surgery of 47.4 ± 16.4 years. IMSCTs were localized in the cervical spine in 71 patients or thoracic in 57. The mean tumor size measured 28.0 ± 29.5 mm (sagittal max) and 9.3 ± 4.4 mm (axial max). The mean surgical time was 351.9 ± 133.1 min (range 142-1041 min), and the mean EBL was 95.7 ± 104.6 mL (range 5-623 mL). All surgeries were conducted using the posterior approach, with dorsal median sulcus (DMS) in 106 cases or dorsal root entry zone (DREZ) in 22 cases. Intraoperative motor-evoked potential (MEP) disappearance was observed in 44 out of 128 cases. Although GTR was the intended outcome for all study participants, it was successfully achieved in 107 patients, while 21 patients underwent PR. Tumor histopathology types included ependymoma in 54 patients, cavernous hemangioma in 31, low-grade astrocytoma in 20, hemangioblastoma in 15, lipoma in 3, myxopapillary ependymoma in 2, ganglioglioma in 2, and subependymoma in 1. Fourteen patients experienced postoperative complications, with one case (0.8%; low-grade astrocytoma) presenting cerebrospinal fluid leakage, which improved with conservative treatment. Another patient with low-grade astrocytoma (0.8%) developed a surgical site infection, leading to additional surgery (incision and thorough cleaning of the wounds) and the administration of antibiotics to address the local infection. Tumor recurrence and regrowth necessitating additional tumor resection occurred in 12 patients (9.4%; including 4 cases of ependymoma, 4 cases of low-grade astrocytoma, 1 case of cavernous hemangioma, 1 case of lipoma, 1 case of ganglioglioma, and 1 case of subependymoma). The mean postoperative follow-up period was 55.5 ± 34.3 months. Preoperatively, 53 patients presented with MMCS grade I, 56 with grade II, 17 with grade III, 1 with grade IV, and 1 with grade V. Immediately after surgery, 30 patients presented with MMCS grade I, 50 with grade II, 35 with grade III, 7 with grade IV, and 6 with grade V. At the final follow-up, 35 patients presented with MMCS grade I, 60 with grade II, 20 with grade III, 10 with grade IV, and 3 with grade V (Table 1).

Table 1.

Demographic and Clinical Characteristics of all Patients With IMSCTs.

Case	n = 128
Age at surgery (y/o)	47.4 ± 16.5
Sex, no. (%)
Men	71 cases (55.5)
Women	57 cases (44.5)
Location, no. (%)
Cervical	71 cases (55.5)
Thoracic	57 cases (44.5)
Tumor size (mm)
Sagittal max	28.0 ± 29.5
Axial max	9.3 ± 4.4
Surgical time (min)	351.9 ± 133.1
EBL (mL)	95.7 ± 104.6
Approach for tumor resection, no. (%)
DMS	106 cases (82.8)
DREZ	22 cases (17.2)
Intraoperative MEP disappearance, no. (%)
(+)	44 cases (34.4)
(−)	84 cases (65.6)
Resection, no. (%)
GTR	107 cases (83.6)
PR	21 cases (16.4)
Histopathology of tumor, no. (%)
Ependymoma	54 cases (42.2)
Cavernous hemangioma	31 cases (24.2)
Low grade astrocytoma	20 cases (15.6)
Hemangioblastoma	15 cases (11.7)
Lipoma	3 cases (2.3)
Myxopapillary ependymoma	2 cases (1.6)
Ganglioglioma	2 cases (1.6)
Subependymoma	1 case (0.8)
MMCS (Preoperation), no. (%)
I	53 cases (41.4)
II	56 cases (43.7)
III	17 cases (13.3)
IV	1 case (0.8)
V	1 case (0.8)
MMCS (Immediately after surgery), no. (%)
I	30 cases (23.4)
II	50 cases (39.1)
III	35 cases (27.3)
IV	7 cases (5.5)
V	6 cases (4.7)
MMCS (Final follow-up), no. (%)
I	35 cases (27.3)
II	60 cases (47.0)
III	20 cases (15.6)
IV	10 cases (7.8)
V	3 cases (2.3)
Follow-up period after surgery (month)	55.5 ± 34.3

Values indicate mean ± standard deviation. IMSCTs, intramedullary spinal cord tumors; MMCS, modified McCormick scale; EBL, estimated blood loss; DMS, dorsal median sulcus; DREZ, dorsal root entry zone; MEP, motor-evoked potential; GTR, Gross total resection; PR, partial resection.

In the Severe group, IMSCTs were more commonly found in the thoracic spine (78.9%). Meanwhile, tumors were frequently located in the cervical spine in the Mild group (61.5%) (P = 0.002). The mean age at surgery was significantly older (P = 0.004), and tumor size was larger (sagittal max, P = 0.042; axial max, P = 0.016) in the Severe group compared to the Mild group. Furthermore, the Severe group exhibited a significantly greater volume of EBL (P = 0.008) and more cases of intraoperative MEP disappearance (P = 0.001). In the Mild group, 89.0% of the patients were achieved GTR, compared to 47.4% in the Severe group (P < 0.001). Regarding tumor histopathology, low-grade astrocytoma was significantly more common in the Severe group, while ependymoma was most common in the Mild group (P = 0.006). However, there were no significant group differences in terms of sex (P = 0.807), surgical time (P = 0.750), approach for tumor resection (P = 0.861), and postoperative follow-up period (P = 0.523) (Table 2).

Table 2.

Comparison of Demographic and Surgical Data According to the Degree of Preoperative Gait Disturbance in Patients With IMSCTs.

	Mild Group (Pre-Operative MMCS = I or II) n = 109	Severe Group (Pre-Operative MMCS = III-V) n = 19	P Value
Age at surgery (y/o)	45.8 ± 16.1	56.8 ± 186.4	0.004^**
Sex, no. (%)			0.807
Men	61 cases (56.0)	10 cases (52.6)
Women	48 cases (44.0)	9 cases (47.4)
Location, no. (%)			0.002^**
Cervical	67 cases (61.5)	4 cases (21.1)
Thoracic	42 cases (38.5)	15 cases (78.9)
Tumor size (mm)
Sagittal max	25.4 ± 27.9	44.0 ± 34.4	0.042^*
Axial max	9.0 ± 4.4	12.1 ± 4.6	0.016^*
Surgical time (min)	352.1 ± 137.2	350.7 ± 111.3	0.750
EBL (mL)	83.3± 89.5	163.3 ± 150.4	0.008^**
Approach for tumor resection, no. (%)
DMS	90 cases (82.6)	16 cases (84.2)	0.861
DREZ	19 cases (17.4)	3 cases (15.8)	0.861
Intraoperative MEP disappearance, no. (%)
(+)	31 cases (28.4)	13 cases (68.4)	0.001^**
(−)	78 cases (71.6)	6 cases (31.6)	0.001^**
Resection, no. (%)
GTR	97 cases (89.0)	9 cases (47.4)	<0.001^***
PR	12 cases (11.0)	10 cases (52.6)	<0.001^***
Histopathology of tumor, no. (%)			0.006^**
Ependymoma	52 cases (47.7)	2 cases (10.5)
Cavernous hemangioma	25 cases (22.9)	6 cases (31.6)
Low grade astrocytoma	12 cases (11.0)	8 cases (42.0)
Hemangioblastoma	14 cases (12.8)	1 case (5.3)
Lipoma	2 cases (1.9)	1 case (5.3)
Myxopapillary ependymoma	1 case (0.9)	1 case (5.3)
Ganglioglioma	2 cases (1.9)	0 case (0)
Subependymoma	1 case (0.9)	0 case (0)
Follow-up period after surgery (month)	56.4 ± 34.5	50.3 ± 33.1	0.523

Values indicate mean ± standard deviation, ^*P < 0.05, ^** < 0.01, ^*** < 0.001, statistically significant difference. IMSCTs, intramedullary spinal cord tumors; MMCS, modified McCormick scale; EBL, estimated blood loss; DMS, dorsal median sulcus; DREZ, dorsal root entry zone; MEP, motor-evoked potential; GTR, Gross total resection; PR, partial resection.

Regarding changes in postoperative gait status according to each preoperative MMCS grade, 38 out of 56 patients with preoperative grade II, and 3 of 17 patients with preoperative grade III showed almost no gait disturbance (grade I or II) immediately after surgery. In contrast, 4 out of 53 patients with preoperative grade I, 18 of 56 patients with preoperative grade II, and 14 of 17 patients with preoperative grade III experienced worse gait disturbances (grades III–V). At the final follow-up, 42 out of 56 patients with preoperative grade II improved to grade I or II, while only 4 of 17 patients with preoperative grade III showed similar results. Based on the aforementioned results, changes in MMCS grade were further examined in the Mild and Severe groups. In the Mild group, improvement was observed in only 1 out of 109 patients. However, there was no change in 66 patients, and 42 patients experienced deterioration immediately after surgery. At the final follow-up, 8 patients experienced improvement, 61 patients did not show changes, and 40 patients experienced deterioration. Meanwhile, in the Severe group, improvement was observed in 3 out of 19 patients. However, there was no change in 8 patients, and 8 experienced deterioration immediately after surgery. Furthermore, at the final follow-up, 5 patients experienced improvement, 6 did not show changes, and 8 patients experienced deterioration (Table 3).

Table 3.

Summary of Changes in MMCS Grade Immediately After Surgery and the Final Follow-Up in Patients With IMSCTs.

	MMCS Grade (Immediately After Surgery)							MMCS Grade (Final Follow-Up)
		I	II	III	IV	V	Total	I	II	III	IV	V	Total
MMCS grade (Preoperation)	I	29	20	4	0	0	53	27	22	3	1	0	53
	II	1	37	15	1	2	56	8	34	11	2	1	56
	III	0	3	6	5	3	17	0	4	5	6	2	17
	IV	0	0	0	1	0	1	0	0	0	1	0	1
	V	0	0	0	0	1	1	0	0	1	0	0	1
	Total	30	60	25	7	6	128	35	60	20	10	3	128
All of the cases, n = 128
Improvement 4 cases (3.1%)								Improvement 13 cases (10.2%)
No change 74 cases (57.8%)								No change 67 cases (52.3%)
Deterioration 50 cases (39.1%)								Deterioration 48 cases (37.5%)
Mild group (Preoperative MMCS grade = I or II), n = 109
Improvement 1 cases (0.9%)								Improvement 8 cases (7.3%)
No change 66 cases (60.6%)								No change 61 cases (56.0%)
Deterioration 42 cases (38.5%)								Deterioration 40 cases (36.7%)
Severe group (Preoperative MMCS grade = III-V), n = 19
Improvement 3 cases (15.8%)								Improvement 5 cases (26.3%)
No change 8 cases (42.1%)								No change 6 cases (31.6%)
Deterioration 8 cases (42.1%)								Deterioration 8 cases (42.1%)

MMCS, modified McCormick scale, IMSCTs, intramedullary spinal cord tumors.

In the Mild group, the univariate analysis indicated that patients with MMCS grades III–V at the final follow-up were significantly older at the time of surgery (P = 0.044), had a higher proportion of localization in the thoracic spinal cord (P = 0.001), exhibited a larger tumor size on MRI sagittal view (P = 0.044), or experienced a longer surgical time (P = 0.010) compared to those with MMCS grades I–II at the final follow-up. The histopathology of the tumor also significantly differed between the two groups (P = 0.036). Conversely, sex (P = 0.630), tumor size (axial max, P = 0.113), volume of EBL (P = 0.070), approach for tumor resection (P = 0.925), intraoperative MEP disappearance (P = 0.282), and resection rate (P = 0.401) did not show significant differences (Table 4).

Table 4.

Univariate Analysis for Possible Factors Affecting Residual Gait Disturbance After Tumor Resection in the Mild Group.

	Final Follow-Up MMCS = I or II n = 91	Final Follow-Up MMCS = III–V n = 18	P Value
Age at surgery (y/o)	44.5 ± 15.6	52.4 ± 17.4	0.044^*
Sex, no. (%)			0.630
Men	50 cases (54.9)	11 cases (61.1)
Women	41 cases (45.1)	7 cases (38.9)
Location, no. (%)
Cervical	62 cases (68.1)	5 cases (27.8)	0.001^**
Thoracic	29 cases (31.9)	13 cases (72.2)	0.001^**
Tumor size (mm)
Sagittal max	21.7 ± 21.5	34.2 ± 23.1	0.044^*
Axial max	8.7 ± 4.5	10.4 ± 3.7	0.113
Surgical time (min)	336.6 ± 134.3	431.7 ± 127.4	0.010^*
EBL (mL)	74.1 ± 81.0	130.5 ± 115.9	0.070
Approach for tumor resection, no. (%)
DMS	75 cases (82.4)	15 cases (83.3)	0.925
DREZ	16 cases (17.6)	3 cases (16.7)	0.925
Intraoperative MEP disappearance, no. (%)
(+)	24 cases (26.4)	7 cases (38.9)	0.282
(−)	67 cases (73.6)	11 cases (61.1)	0.282
Resection, no. (%)
GTR	82 cases (90.1)	15 cases (83.3)	0.401
PR	9 cases (9.9)	3 cases (16.7)	0.401
Histopathology of tumor, no. (%)
Ependymoma	44 cases (48.3)	8 cases (44.3)	0.036^*
Cavernous hemangioma	23 cases (25.3)	2 cases (11.1)
Low grade astrocytoma	7 cases (7.7)	5 cases (27.8)
Hemangioblastoma	13 cases (14.3)	1 case (5.6)
Lipoma	1 case (1.1)	1 case (5.6)
Myxopapillary ependymoma	1 case (1.1)	0 case (0.0)
Ganglioglioma	2 cases (2.2)	0 case (0.0)
Subependymoma	0 cases (0.0)	1 case (5.6)

Values indicate mean ± standard deviation, ^*P < 0.05, ^** < 0.01, statistically significant difference. IMSCTs, intramedullary spinal cord tumors; MMCS, modified McCormick scale; EBL, estimated blood loss; DMS, dorsal median sulcus; DREZ, dorsal root entry zone; MEP, motor-evoked potential; GTR, Gross total resection; PR, partial resection.

Multivariate logistic regression analysis among the variables with a P-value of < 0.25 in the univariate analysis revealed that location (β = 2.369, P = 0.009, odds ratio [OR]: 10.69, 95% confidence interval [CI]: 2.88 – 40.36) and age at surgery (β = 0.044, P = 0.026, OR: 1.045, 95% CI: 1.005 – 1.086) were factors influencing residual gait disturbance after tumor resection in the Mild group (Table 5).

Table 5.

Multivariate Logistic Regression Analysis of MMCS in the Mild Group at the Final Follow-Up.

Characteristic	β	95% Confidence Interval	Odds Ratio	P Value
Location (Thoracic vs Cervical)	2.369	2.651 – 44.617	10.690	0.009^**
Age at surgery (y/o)	0.044	1.005 – 1.086	1.045	0.026^*
Surgical time (min)	0.005	0.999 – 1.011	1.005	0.094
Histopathology of tumor	−1.977	0.006 – 3.403	0.138	0.287
Tumor size (sagittal max, mm)	−0.015	0.939 – 1.034	0.985	0.145
Tumor size (axial max, mm)	0.224	0.899 – 1.740	1.251	0.562
Volume of EBL (ml)	0.006	0.998 – 1.013	1.006	0.905

^*P < 0.05.** < 0.01, statistically significant difference. MMCS, modified McCormick scale; EBL, estimated blood loss.

Tumors localized at the thoracic spine led to worsening of gait status immediately after surgery in 17 out of 42 cases in the Mild group. Additionally, the symptoms of gait disturbance remained in 13 out of 42 cases at the final follow-up. In contrast, in the Severe group, only 1 out of 4 patients with cervical lesions and 2 out of 15 patients with thoracic lesions showed improvement in neurological status immediately after surgery. Furthermore, at the final follow-up, 2 out of 4 patients with cervical lesions presented with MMCS grade I or II. However, gait disturbance remained without improvement in cases of thoracic lesions. In the Mild group, there were statistically significant differences in each level of tumor localization according to MMCS grade immediately after surgery or at the final follow-up, but not in the Severe group (Mild group, P < 0.001, 0.003; Severe group, P = 0.530, 0.178) (Table 6).

Table 6.

Comparison of Postoperative Outcomes According to the Severity of Preoperative Gait Disturbance in Patients With Cervical or Thoracic IMSCTs.

	Mild Group (Pre-operative MMCS = I or II)
MMCS	Cervical n = 67	Thoracic n = 42	P Value
Immediate post-operation, no. (%)
I or II	62 cases (92.5)	25 cases (59.5)	<0.001^***
III-V	5 cases (7.5)	17 cases (40.5)	<0.001^***
Final follow-up, no. (%)
I or II	62 cases (92.5)	29 cases (69.0)	0.003^**
III-V	5 cases (7.5)	13 cases (31.0)	0.003^**

	Severe group (Pre-operative MMCS = III-V)
MMCS	Cervical n = 4	Thoracic n = 15	P Value
Immediate post-operation, no. (%)
I or II	1 case (25.0)	2 cases (13.3)	0.530
III-V	3 cases (75.0)	13 cases (86.7)	0.530
Final follow-up, no. (%)
I or II	2 cases (50.0)	2 cases (13.3)	0.178
III-V	2 cases (50.0)	13 cases (86.7)	0.178

Values indicate mean ± standard deviation, ^**P < 0.01, ^*** < 0.001, statistically significant difference. MMCS, modified McCormick scale.

Discussion

The present study investigated whether the severity of preoperative gait disturbance remains after tumor resection in patients with IMSCTs using the MMCS grade. At the final follow-up, 74.2% of patients with IMSCTs were capable of walking without support. However, concerning the improvement of gait disturbance, only 10.2% of all cases demonstrated improvement, while the remaining 89.8% either remained unchanged or worsened. Specifically, 39.1% of patients with IMSCTs experienced gait disturbance immediately after surgery, and 37.5% of the patients continued to experience gait disturbance at the final follow-up. Additionally, tumor location and age at surgery were identified as factors influencing poor improvement in postoperative gait disturbance in patients with IMSCTs who had mild gait disturbance preoperatively. Our findings bear significant implications for studies investigating the efficacy of tumor resection, as few studies have comprehensively examined changes in postoperative neurological status based on the degree of preoperative gait disturbance in patients with IMSCTs.

To date, numerous studies have described the pathogenesis, treatment, and surgical outcomes of IMSCTs.^3-9 In a retrospective review focused on cervical IMSCTs, Xu et al²⁰ observed improved neurological outcomes in 77.6% of their study patients. Karikari et al³ noted that there was no statistically significant difference in functional outcomes based on tumor location. Baig et al demonstrated that IMSCTs spanning a greater number of vertebral levels were significantly associated with poor MMCS outcomes. They also found that the presence of pre-operative gait disturbance was related to unfavorable outcomes in both postoperative MMCS and Nurick scores.⁴ Suzuki et al⁷ showed that the higher the lower-limb function scores in the early phase, the better the improvement in walking ability was predicted 1 year after IMSCTs resection. Endo et al⁵ reported that 251 patients with IMSCTs showed improvement, 500 remained unchanged, and 160 experienced worsened MMCS grades after 6 months postoperatively. Conversely, Alizada et al⁶ found that despite immediate postoperative worsening of neurological status, a great number of patients with IMSCTs showed improvement at the last follow up. However, only a limited number of reports have comprehensively summarized postoperative changes in neurological status based on preoperative gait disturbance or clearly indicated the presence or absence of changes in gait disturbance after surgical resection. Furthermore, the factors influencing postoperative MMCS grade have not been fully elucidated. Therefore, this study assessed in detail whether the severity of preoperative gait disturbance affects changes in neurological status after surgery by comparing patients with IMSCTs in the Mild and Severe groups.

Most IMSCTs belong to histological types such as ependymomas and astrocytomas, but there are also various other types, including cavernous hemangiomas and hemangioblastomas.^10-18 Alhalabi et al¹⁰ found that the presence of gait disturbance or ataxia increased the risk of incomplete long-term recovery after spinal ependymoma resection. Tsuji et al¹¹ identified worse preoperative MMCS as independent risk factors for motor deterioration after ependymoma resection. Eroes et al¹³ highlighted that an MMCS of less than 3 and a tumor extent of less than 5 levels were crucial factors for a favorable postoperative functional outcome in low-grade astrocytoma and ependymoma. Nagoshi et al¹⁵ demonstrated that among patients with cavernous hemangiomas, those with unstable gait prior to surgery had a higher frequency of hemorrhagic episodes, as assessed by the MMCS. Schwake et al¹⁶ reported that surgery on patients with hemangioblastomas displaying no or minor neurological deficits (MMCS grade I or II) was associated with a favorable postoperative outcome and overall higher quality of life compared to those with severe neurological deficits (MMCS grade III or IV). While the above results focus on specific histological types of IMSCTs, in a comprehensive review of IMSCTs as a whole, Karikari et al³ showed that tumor histology played a predictive role in functional neurological outcomes, suggesting that most patients with non-astrocytic IMSCTs undergoing surgical resection would maintain their neurological status. Endo et al⁵ reported that the histopathological type of IMSCTs was a significant factor associated with mortality and functional outcomes. However, in this study, the histopathology of the tumor was not a factor influencing residual gait disturbance after tumor resection in the Mild group. This difference may be attributed to the study’s focus on the Mild group, as its purpose was to determine whether the severity of preoperative gait disturbance remains after tumor resection. Another possibility is the exclusion of cases where only a biopsy was performed (eg, high-grade astrocytoma), despite no significant difference in tumor histology between the Mild and Severe groups.

Recently, several reports have assessed the clinical characteristics of pediatric IMSCTs and identified differences between pediatric and adult cases.^21-23 McGirt et al noted that approximately one-third of pediatric patients might experience improvement in motor, sensory, and urinary dysfunction years after IMSCTs resection, whereas the majority would continue to experience long-term dysesthetic symptoms.²¹ Ahmed et al²² described that pediatric IMSCTs patients with MMCS grade I at presentation were more likely to show functional improvement by final follow-up compared to those with MMCS Grades II-V. Umebayashi et al²³ found that preoperative MMCS were compatible between pediatric and adult patients with IMSCTs. The proportions of patients with MMCS grade V neurological function increased immediately after surgery, while the proportions of those with MMCS grade I and II neurological function increased as well. Postoperative deterioration of MMCS at discharge was more common in pediatric patients than in adult. In contrast, Pennington et al²⁴ suggested that advanced age (≥ 63 years) and worse baseline neurological function were the strongest predictors of long-term neurological worsening. Tsuji et al¹¹ reported that age at surgery was independent risk factors for motor deterioration after ependymoma resection. In our current study, multivariate logistic regression analysis revealed that age at surgery influenced residual gait disturbance after IMSCTs resection in the Mild group. The reasons for these results include the fact that there were only six cases of pediatric IMSCTs patients in this study, which is extremely low. Additionally, in the Severe group, there was a relatively high proportion of elderly individuals aged 65 and above, with 8 out of 19 cases (42.1%). Therefore, based on the results of this study, further research should be validated in the future, including more pediatric patients, regarding whether postoperative MMCS for each age group will be changed over long-term periods.

In patients with spinal cord tumors, the MMCS serves as the standard clinical outcome tool—a physician-rated assessment encompassing global functional impairment, including neurological status and walking ability.^17-19 While there is a growing emphasis on employing patient-rated outcome measures to evaluate clinical outcomes in various spinal surgeries, a limited number of reports have explored postoperative neurological symptoms using patient-rated outcome scales specifically in IMSCTs resection surgery.^25-28 Nakamura et al demonstrated that all 36-item Short-Form (SF-36) Health Survey subscores of the patients with IMSCTs were significantly lower than the national average, postoperatively. They also identified a significant negative correlation between SF-36 and Neuropathic Pain Symptom Inventory (NPSI) subscores.²⁵ Myrseth et al²⁶ similarly discovered significantly lower SF-36 results across all five subdomains related to physical function, while scores for mental health and role emotional showed no significant differences compared to Norwegian norms. Xiao et al²⁷ reported no significant changes between preoperative and postoperative EuroQol 5-Dimensions (EQ-5D), Pain Disability Questionnaire (PDQ), or Patient Health Questionnaire-9 (PHQ-9). Tufo et al²⁸ highlighted a significant interaction between MMCS grade and scores on the Pain Catastrophizing Scale (PCS) and Pittsburgh Sleep Quality Index (PSQI). Higher MMCS grades were associated with increased pain perception and poor sleep quality perception long after surgery. Simultaneously, SF-36 subscales exhibited a significant interaction with the degree of the role-physical subscale, correlating lower scores with MMCS grades II and III. Unfortunately, in this study, postoperative outcomes for patients with IMSCTs were assessed solely using the physician-rated MMCS. Consequently, direct comparisons with patient-rated outcomes like SF-36, NPSI, EQ-5D, and others could not be made. To comprehensively gauge the impact of the disorder on QOL and daily functioning, it is imperative to evaluate outcomes from the patient’s perspective. Additionally, for future research, we believe that more comprehensive clinical studies should be conducted with larger sample sizes and evaluations from the patient’s perspective, using patient-rated questionnaires. This approach would offer a more thorough assessment of clinical outcomes associated with the surgical resection of IMSCTs.

Our study had several important limitations that should be considered. First, the retrospective design of the study inherently reduced the level of evidence. Second, the sample size was relatively small, and the statistical power was not robust enough to draw definitive conclusions regarding accurate clinical outcomes in patients with IMSCTs. Third, the patients with IMSCTs included in this study underwent tumor removal within approximately one year of symptom onset. Unfortunately, due to the absence of data on patients experiencing symptoms for more than one year, we were unable to explore the relationship between symptom duration and postoperative neurological outcomes. Finally, the surgical outcomes may have been influenced by the diverse techniques, experiences, and skills of the four orthopedic spine surgeons who performed IMSCTs resection. Despite these limitations, the current study can provide important insights into the surgical and clinical outcomes of tumor resection in patients with IMSCTs, particularly based on the severity of preoperative gait disturbance. This study is primarily focused on novel analyses that center on the postoperative course of patients with IMSCTs who exhibit mild or severe preoperative gait disturbance.

Conclusion

Here, in the present study, with a focus on the severity of preoperative gait disturbance, we performed a follow-up study on 128 patients with IMSCTs to evaluate postoperative clinical outcomes. At the final follow-up, 74.2% of patients with IMSCTs could walk without support. However, only 10.2% showed improvement in gait disturbance, while 89.8% either remained unchanged or worsened. Specifically, 39.1% of patients with IMSCTs experienced immediate postoperative gait disturbance, and 37.5% continued to experience it at the final follow-up. Factors influencing poor improvement in postoperative gait disturbance in patients with mild preoperative gait disturbance included tumor location and age at surgery. The results of this study will provide valuable information for patients with IMSCTs considering tumor resection, particularly regarding postoperative outcomes and potential complications. They will also assist spine surgeons in explaining surgical plan and clinical outcomes, such as during preoperative informed consent with IMSCTs patients. While the current study involved a relatively small number of patients, future research with a larger sample size is warranted to yield more precise and comprehensive conclusions.

Footnotes

Acknowledgments

The authors would like to express our gratitude to Dr Hirokazu Fujiwara, the radiologist from our institution, for evaluating the extent of tumor resection.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Toshiki Okubo

Narihito Nagoshi

Kazuki Takeda

Kota Watanabe

References

McCormick

Stein

. Spinal cord tumors in adults. In: Youmans

, ed. Neurological Surgery. Philadelphia, PA: WB Saunders; 1996:3102-3122.

Fischer

Brotchi

. Intramedullary Spinal Cord Tumors. Stuttgart, Germany: Thieme; 1996.

Karikari

Nimjee

Hodges

, et al. Impact of tumor histology on resectability and neurological outcome in primary intramedullary spinal cord tumors: a single-center experience with 102 patients. Neurosurgery. 2015;76 Suppl 1(Suppl 1):S4-S13. doi:10.1227/01.neu.0000462073.71915.12

Baig Mirza

Gebreyohanes

Knight

, et al. Prognostic factors for surgically managed intramedullary spinal cord tumours: a single-centre case series. Acta Neurochir (Wien). 2022;164(10):2605-2622. doi:10.1007/s00701-022-05304-9

Endo

Inoue

Mizuno

, et al. Current trends in the surgical management of intramedullary tumors: a multicenter study of 1,033 patients by the neurospinal society of Japan. Neurospine. 2022;19(2):441-452. doi:10.14245/ns.2244156.078

Alizada

Kemerdere

Ulu

, et al. Surgical management of spinal intramedullary tumors: ten-year experience in a single institution. J Clin Neurosci. 2020;73:201-208. doi:10.1016/j.jocn.2019.12.054

Suzuki

Tsuji

Ichikawa

, et al. Early phase functional recovery after spinal intramedullary tumor resection could predict ambulatory capacity at 1 Year after surgery. Asian Spine J. 2023;17(2):355-364. doi:10.31616/asj.2022.0068

Tobin

Geraghty

Engelhard

Linninger

Mehta

. Intramedullary spinal cord tumors: a review of current and future treatment strategies. Neurosurg Focus. 2015;39(2):E14. doi:10.3171/2015.5.FOCUS15158

Juthani

Bilsky

Vogelbaum

. Current management and treatment modalities for intramedullary spinal cord tumors. Curr Treat Options Oncol. 2015;16(8):39. doi:10.1007/s11864-015-0358-0

10.

Alhalabi

Heene

Landré

, et al. Association of early surgery and absence of ataxia with full recovery after spinal intramedullary ependymoma resection. J Neurosurg Spine. 2023;40:185-195. doi:10.3171/2023.8.SPINE23606

11.

Tsuji

Nagoshi

Ishii

, et al. Poor prognostic factors for surgical treatment of spinal intramedullary ependymoma (world health organization grade II). Asian Spine J. 2020;14(6):821-828. doi:10.31616/asj.2020.0064

12.

Yang

Sun

Xie

, et al. Multisegmental versus monosegmental intramedullary spinal cord ependymomas: perioperative neurological functions and surgical outcomes. Neurosurg Rev. 2022;45(1):553-560. doi:10.1007/s10143-021-01567-5

13.

Eroes

Zausinger

Kreth

Goldbrunner

Tonn

. Intramedullary low grade astrocytoma and ependymoma. Surgical results and predicting factors for clinical outcome. Acta Neurochir (Wien). 2010;152(4):611-618. doi:10.1007/s00701-009-0577-x

14.

Nagoshi

Tsuji

Suzuki

, et al. Clinical outcomes and a therapeutic indication of intramedullary spinal cord astrocytoma. Spinal Cord. 2022;60(3):216-222. doi:10.1038/s41393-021-00676-8

15.

Nagoshi

Tsuji

Nakashima

, et al. Clinical outcomes and prognostic factors for cavernous hemangiomas of the spinal cord: a retrospective cohort study. J Neurosurg Spine. 2019;31(2):271-278. doi:10.3171/2019.1.SPINE18854

16.

Schwake

Ricchizzi

Krahwinkel

, et al. Resection of intramedullary hemangioblastoma: timing of surgery and its impact on neurological outcome and quality of life. Medicina (Kaunas). 2023;59(9):1611. doi:10.3390/medicina59091611

17.

Okubo

Fujiyoshi

Kobayashi

, et al.

Does the degree of preoperative gait disturbance remain after tumor resection in patients with intradural extramedullary spinal cord tumors?

Spinal Cord. 2023;61(12):637-643. doi:10.1038/s41393-023-00931-0

18.

McCormick

Werner

Shimer

. Patient-reported outcome measures in spine surgery. J Am Acad Orthop Surg. 2013;21:99-107.

19.

Imagama

Ito

Ando

, et al. Optimal timing of surgery for intramedullary cavernous hemangioma of the spinal cord in relation to preoperative motor paresis, disease duration, and tumor volume and location. Global Spine J. 2017;7(3):246-253. doi:10.1177/2192568217707938

20.

Bao

Mao

Yang

. Aggressive surgery for intramedullary tumor of cervical spinal cord. Surg Neurol. 1996;46(4):322-328. doi:10.1016/s0090-3019(96)00040-7

21.

McGirt

Chaichana

Atiba

Attenello

Yao

Jallo

. Resection of intramedullary spinal cord tumors in children: assessment of long-term motor and sensory deficits. J Neurosurg Pediatr. 2008;1(1):63-67. doi:10.3171/PED-08/01/063

22.

Ahmed

Menezes

Awe

Torner

. Long-term disease and neurological outcomes in patients with pediatric intramedullary spinal cord tumors. J Neurosurg Pediatr. 2014;13:600-612. doi:10.3171/2014.1.PEDS13316

23.

Umebayashi

Naito

Kurokawa

, et al. Epidemiology and comparative analysis of outcomes of intramedullary spinal cord tumor between pediatric and adult patients. Spine. 2024;49(2):107-115. doi:10.1097/BRS.0000000000004775

24.

Pennington

Mikula

Lakomkin

, et al. Impact of tumor-associated syrinx on outcomes following resection of primary ependymomas of the spinal cord. J Neuro Oncol. 2022;160(3):725-733. doi:10.1007/s11060-022-04194-2

25.

Nakamura

Tsuji

Iwanami

, et al. Central neuropathic pain after surgical resection in patients with spinal intramedullary tumor. J Orthop Sci. 2012;17(4):352-357. doi:10.1007/s00776-012-0236-6

26.

Myrseth

Habiba

Rekand

Sætran

Mørk

Grønning

. Intramedullary spinal cord and filum tumours-long-term outcome: single institution case series. Acta Neurochir (Wien). 2022;164(11):3047-3056. doi:10.1007/s00701-022-05350-3

27.

Xiao

Miller

Abdullah

Lubelski

Mroz

Benzel

. Quality of life outcomes following resection of adult intramedullary spinal cord tumors. Neurosurgery. 2016;78(6):821-828. doi:10.1227/NEU.0000000000001147

28.

Tufo

Grande

Bevacqua

, et al. Long-term quality of life and functional outcomes in adults surgically treated for intramedullary spinal cord tumor. Front Psychol. 2023;14:1136223. doi:10.3389/fpsyg.2023.1136223