The effect of timing on rehabilitation results in tumor-related spinal cord injuries

Abstract

Objectives

This study aims to investigate the impact of early initiated rehabilitation on outcomes in patients with spinal cord injuries related to tumors.

Patients and methods

In this retrospective cohort study conducted between January 2015 and March 2021, a total of 31 patients with tumor-related spinal cord injury were assessed. Two patients were excluded due to incomplete clinical records, and the remaining 29 patients were divided into the early rehabilitation group (n=15; 8 males, 7 females; mean age: 45.53±16.73 years; range: 8 to 68 years) and the late rehabilitation group (n=14; 9 males, 5 females; mean age: 54.86±14.76 years; range: 28 to 78 years). Demographic and clinical characteristics, Functional Ambulation Category (FAC), and Barthel Index scores for outcome measurement were recorded.

Results

The neurological level of the patients was predominantly at the thoracic level. The majority of patients were in the American Spinal Injury Association D category. In the evaluation of treatment outcomes, a significant difference was observed within each group of FAC and Barthel Index scores before and after rehabilitation (p=0.034, p=0.002; p=0.007, p=0.005 in early and late rehabilitation groups). In the comparison of FAC and Barthel Index scores before and after rehabilitation between groups, there was a significant difference in favor of the late rehabilitation group in both scores, both before and after rehabilitation (respectively, before rehabilitation p=0.032, p=0.005; after rehabilitation p=0.026, p=0.003). However, there was no significant difference in the FAC difference and Barthel Index scores percentage changes between groups.

Conclusion

This study demonstrates that functional outcomes improved with rehabilitation in patients with spinal cord injuries related to tumors. Although the timing of rehabilitation initiation did not result in significant differences in functional outcomes, early rehabilitation did not appear to increase the risk of surgery-related adverse events.

Traumatic spinal cord injuries account for approximately 60% of all spinal cord injury cases, whereas non-traumatic injuries comprise the remaining 40%, representing a substantial proportion of the overall incidence.^[1] Hospital-based studies have reported that the proportion of tumor- or metastasis-related cases within non-traumatic spinal cord injuries ranges from approximately 15 to 33%; these studies exhibit heterogeneity due to differences in center characteristics and sample selection.^{[2, 3, 4]} Such data underscore the clinical importance of providing rehabilitation to non-traumatic spinal cord injury patients with tumor- or metastasis-related injuries, as early intervention can improve functional outcomes and quality of life, challenging the perception that rehabilitation unnecessarily prolongs hospital stay or is reserved for patients with particularly good prognoses.^[5]

Epidemiological data and rehabilitation outcomes for spinal cord injuries are predominantly derived from traumatic spinal cord injury cases, whereas evidence regarding non-traumatic spinal cord injuries remains limited and heterogeneous. This is largely attributable to the diverse etiologies of non-traumatic spinal cord injury, including tumors, vascular disorders, infections, degenerative conditions, and inflammatory diseases, as well as variations in study design, underscoring the need for comparative and population-based research specifically focused on non-traumatic spinal cord injury.^{[1, 6]} Current guidelines, primarily based on studies of traumatic spinal cord injury patients, recommend early initiation of rehabilitation, as delays from injury to rehabilitation are associated with reduced performance in activities of daily living and lower quality of life. The main potential drawback of early rehabilitation is an increased risk of retransfer to acute care if the patient is not medically stable enough to tolerate the intensity of therapy.^[7] However, functional outcomes, length of hospital stay, and rehabilitation requirements in patients with tumor- or metastasis-related spinal cord injuries differ not only from traumatic cases but also from other types of non-traumatic spinal cord injuries, suggesting that the optimal timing for initiating rehabilitation may likewise differ in this patient group. Given the limited and heterogeneous data on rehabilitation in patients with spinal cord injuries due to tumors, the aim of this study is to investigate the impact of early rehabilitation on functional outcomes in this underrepresented population, as such research is crucial to inform clinical decision-making and optimize patient outcomes.

PATIENTS AND METHODS

This retrospective cohort study was conducted at the Health Sciences University, Haydarpaşa Training and Research Hospital Faculty of Medicine, Department of Physical Medicine and Rehabilitation between January 2015 and March 2021. A total of 31 patients with tumor-related spinal cord injury were assessed. Two patients were excluded from the final analysis due to incomplete clinical records. Of the remaining 29 patients, 15 (8 males, 7 females; mean age: 45.53±16.73 years; range: 8 to 68 years) were classified as the early rehabilitation group and 14 (9 males, 5 females; mean age: 54.86±14.76 years; range: 28 to 78 years) as the late rehabilitation group.

Patients with tumor-related spinal cord injuries, defined as injuries directly resulting from primary or metastatic spinal tumors, as well as iatrogenic spinal cord injury occurring during surgical intervention for these tumors, who received rehabilitation services at our inpatient clinic were included in the study. Only patients with complete and accessible clinical records available for retrospective analysis were considered. Patients were excluded if they had traumatic spinal cord injuries (e.g., resulting from traffic accidents, falls, or sports injuries), non-tumor, non-traumatic spinal cord injuries (including vascular, inflammatory, infectious, or degenerative causes), or had received radiotherapy and/or chemotherapy within the last three months prior to admission were excluded from the study. Patients with incomplete or insufficient clinical records were also excluded. A written informed consent was obtained from each patient. The study protocol was approved by the Haydarpaşa Numune Training and Research Hospital Clinical Research Ethics Committee (Date: 29.03.2021, No: HNEAH-KAEK 2021/123). The study was conducted in accordance with the principles of the Declaration of Helsinki.

During the inpatient screening period corresponding to the study dates, 31 patients who received rehabilitation for tumor-related spinal cord injury were identified. Of these, two patients were excluded due to insufficient clinical records. Patient demographics, including age, sex, tumor pathological diagnoses, grades, tumor location (intradural/extradural), malignant/benign characteristics, recurrence status, additional treatments received (chemotherapy, radiotherapy, chemotherapy+radiotherapy combination), spinal cord injury levels (preoperative and postoperative, prior to rehabilitation), American Spinal Injury Association (ASIA) classifications, presence of proprioception impairments, pre- and post-rehabilitation functional ambulation category (FAC) and Barthel Index scores were recorded. The joint position sense was assessed using manual clinical testing to evaluate proprioceptive function. The assessments were conducted on both the upper and lower limbs, depending on the level and severity of injury. All assessments were performed individually by an experienced physiotherapist during the initial rehabilitation evaluation, prior to the start of the therapy program. Pre- and post-rehabilitation FAC and Barthel Index scores were recorded to evaluate ambulation and activities of daily living. All assessments were conducted by the same experienced physiotherapist using standardized procedures across all time points to ensure consistency. The FAC was classified on a six-point scale (0-5), where 0 indicates non-ambulatory status and the patient cannot walk or requires total assistance; 1 indicates ambulation with continuous manual contact from another person; 2 indicates ambulation with intermittent manual contact or supervision; 3 indicates ambulation with verbal supervision only on level surfaces; 4 indicates independent ambulation on level surfaces, but assistance may be needed on stairs, slopes, or uneven surfaces; and 5 indicates fully independent ambulation on all surfaces without supervision or assistance.^[8] The original 10-item version of the Barthel Index was used for all patients.^[9] Complications that occurred during the rehabilitation process (such as wound site complications, pulmonary complications, etc.) were also documented. The patients were divided into two groups based on the time of admission to rehabilitation after surgery: the early rehabilitation group (n=15) (within the first 15 days) and the late rehabilitation group (n=14) (after 15 days).^[10] The obtained data were evaluated with the aim of comparing rehabilitation outcomes before and after in both groups and investigating whether there were any differences between the two groups.

The rehabilitation protocols applied in this study were developed by our clinic in accordance with the relevant literature and tailored to the severity of the spinal cord injury. The program comprised a combination of active and passive physical therapy exercises and neuromuscular electrical stimulation (NMES), aiming to maintain joint mobility, strengthen muscles, improve balance and coordination, and enhance neuroplasticity and functional independence. Passive range-of-motion exercises were performed by the physiotherapist to preserve joint flexibility, while active exercises required voluntary muscle contraction to perform functional tasks. Therapy sessions were conducted five times per week, each lasting approximately 45-60 min, with frequency, intensity, and duration adjusted according to each patient’s tolerance and clinical status. Where appropriate, specific neurorehabilitation techniques, including the Bobath approach and proprioceptive neuromuscular facilitation patterns, were employed to facilitate motor re-education and functional movement patterns. Neuromuscular electrical stimulation was applied to targeted limb muscles using surface electrodes (pulse duration: 200-300 μs; frequency: 20-35 Hz; intensity adjusted to elicit visible muscle contraction without discomfort; on/off cycle: 5-10 sec) to enhance muscle activation, sensory feedback, and support trunk or limb stability during early mobilization and gait training (SAMMS Mod Professional, Chattanooga Group Inc., USA). All interventions were delivered individually by an experienced physiotherapist under the supervision of a physiatrist to ensure standardized application across sessions.

Statistical analysis

The data were analyzed using IBM SPSS Statistics for Windows, version 22.0 (IBM Corp., Armonk, NY, USA). Whether the data were normally distributed was evaluated using the Shapiro-Wilk test. Descriptive statistics were presented as mean ± standard deviation (SD) or median (min-max), as appropriate, for continuous variables and as frequency and percentage for categorical variables. Between-group comparisons were performed using Fisher’s exact test for categorical variables and the Mann-Whitney U test for continuous variables. For the Functional Ambulation Category (FAC), which is an ordinal scale, between-group comparisons were based on the distribution of categories rather than median values. Within-group comparisons were conducted using the Wilcoxon signed-rank test. Statistical significance was accepted as p < 0.05.

As no a priori sample size calculation was performed due to the retrospective nature of the study, a post hoc power analysis was conducted using G*Power software version 3.1.9.7 (Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany). The analysis was based on post-rehabilitation Barthel Index scores.

A two-tailed independent samples t-test was selected (test family: t tests; statistical test: means-difference between two independent means). The observed effect size was calculated as Cohen’s d based on the post-rehabilitation Barthel Index means and standard deviations of the two groups (d=1.22). The allocation ratio (N2/N1) was 0.93, reflecting the unequal group sizes (n=15 and n=14). With an alpha level of 0.05, the achieved post hoc power was 89%.

RESULTS

The mean age of the patients was 50.03±16.24  (8-78) years, with a male/female ratio of 12/17. Details of the pathological diagnoses of the patients are provided in Table 1. The majority of patients, 82.6% (n=24), had tumors located intradurally. In terms of tumor pathology, 51.7% of patients (n=15) exhibited malignant characteristics. Among the 29 patients included in the study, one patient was diagnosed with an endodermal cyst on histopathological examination. Of the remaining 28 patients, six (21%) were diagnosed with metastatic tumors classified as Grade 4, two (7%) as Grade 3, seven (25%) as Grade 2, and 13 (46%) as Grade 1. No statistically significant difference in tumor grading was observed between the early and late rehabilitation groups (p=0.728). Seven (24.1%) patients had undergone surgery due to recurrent tumors. While 69% of patients (n=20) did not receive any additional treatment after surgery, one (3.4%) patient received chemotherapy only, three (10.3%) patients received radiotherapy (including proton therapy), and five (17.2%) patients received a combination of chemotherapy and radiotherapy. Among the nine patients who received radiotherapy and/or chemotherapy, four patients underwent these treatments for their primary tumor diagnoses, while the remaining five with tumor recurrence who received the therapy following their initial surgical interventions. No statistically significant differences were observed between the early and late rehabilitation groups with respect to adjuvant therapies received (p=0.688). All radiotherapy and/or chemotherapy treatments were completed at least three months prior to the start of the rehabilitation program. Most patients had thoracic-level neurological injury and were classified as ASIA D (both 69%, n=20). Six (20.7%) patients were classified as ASIA C, two (6.9%) patients as ASIA B, and one (3.4%) patient as ASIA A. Preoperative ASIA scores of the cases revealed that one (3.4%) patient was classified as ASIA A, two (6.9%) patients as ASIA B, three (10.3%) patients as ASIA C, 17 (58.6%) patients as ASIA D, and six (20.7%) patients as ASIA E. Accordingly, seven (24.1%) patients were reported to have developed new neurological deficits or experienced progression of their pre-existing neurological status as a complication. Proprioception impairment was detected in 34.5% (n=10) of patients.

TABLE 1. Etiology of tumors related to spinal cord injury.

Pathological diagnosis	n	%
Meningioma	7	24.1
Ependymoma	5	17.2
Schwannoma	4	13.8
Hemagioblastoma	3	10.3
Dermoid tumor	1	3.4
Endodermal cyst	1	3.4
Chordoma	1	3.4
Astrocytoma	1	3.4
Mesenchymal cell tumor	1	3.4
Prostate cancer metastasis	2	6.9
Lung cancer metastasis	2	6.9
Breast cancer metastasis	1	3.4

The initiation time for rehabilitation after the operation had a mean of 21 ± 16.15 days (range, 5-63), and the mean length of hospital stay was 26.44 ± 13.43 days (range, 3-45), and the number of rehabilitation sessions received was 19.22 ± 9.19 (range, 1-30). There were no statistically significant differences between the early and late rehabilitation groups in terms of age, sex, tumor pathology, tumor location, malignant/benign ratio, proprioception impairment, sensory-motor-neurological levels, ASIA levels, length of hospital stay, and the number of rehabilitation sessions received (p>0.05). Patients operated on due to tumor recurrence were more frequently observed in the late rehabilitation group (n=6, 42.9%) compared with the early rehabilitation group (n=1, 6.7%; p=0.023). The statistical comparisons of demographic and clinical data between the groups are provided in Table 2. In the comparison of FAC and Barthel Index scores before and after rehabilitation between groups, there was a significant difference in favor of the late rehabilitation group in both scores, both before and after rehabilitation (respectively, before rehabilitation p=0.032, p=0.005; after rehabilitation p=0.026, p=0.003). Group comparisons and the data comparing FAC and Barthel Index scores before and after rehabilitation within each group are presented in Table 3. In the evaluation of treatment outcomes, a significant difference was observed within each group of FAC and Barthel Index scores before and after rehabilitation (Table 3). Given the baseline differences in FAC and Barthel Index scores between the early and late rehabilitation groups, functional improvement was compared using the absolute change in FAC score and the percentage change in Barthel Index score. The change in FAC was calculated as the absolute difference between post-rehabilitation and pre-rehabilitation scores, reflecting the number of levels gained in ambulation capacity. The percentage change in Barthel Index score was computed using the following formula: (Post-rehabilitation score-Pre-rehabilitation score)/Pre-rehabilitation score, providing a standardized measure of functional improvement relative to each patient’s initial status. However, there was no significant difference when comparing the pre-post rehabilitation FAC score difference and Barthel Index score percentage changes between groups (respectively, p=0.409 and p=0.455).

TABLE 2. Inter-group demographic and clinical comparison.

-	Early rehabilitation group (n=15)					Late rehabilitation group (n=14)					-
-	n	%	Mean±SD	Median	Min-Max	n	%	Mean±SD	Median	Min-Max	p
Age (year)	-	-	45.53±16.73	52	8-68	-	-	54.86±14.76	55	28-78	0.186*
Sex Female Male	- 7 8	- 46.7 53.3	-	-	-	- 5 9	- 35.7 64.3	-	-	-	0.837**
Location Intradural Extradural	- 12 3	- 80 20	-	-	-	- 12 2	- 85.7 14.3	-	-	-	0.697**
Characteristic Malignant Benign	- 6 9	- 40 60	-	-	-	- 9 5	- 64.3 35.7	-	-	-	0.191**
Recurrence Yes No	- 1 14	- 6.7 93.3	-	-	-	- 6 8	- 42.9 57.1	-	-	-	0.023**
Proprioception impairment Yes No	- 6 9	- 40 60	-	-	-	- 4 10	- 28.6 71.4	-	-	-	0.518**
Sensory level Servical Thoracal Lumbar	- 3 11 1	- 20 73.3 6.7	-	-	-	- 3 9 2	- 21.4 64.3 14.3	-	-	-	0.779**
Motor level Servical Thoracal Lumbar	- 2 11 2	- 13.2 73.3 14.3	-	-	-	- 2 10 2	- 14.3 71.4 14.3	-	-	-	0.993**
Neurologic level Servical Thoracal Lumbar	- 4 10 1	- 26.7 66.7 6.7	-	-	-	- 3 10 1	- 21.4 71.4 7.1	-	-	-	0.947**
ASIA A B C D	- 1 2 3 9	- 6.7 13.3 20 60	-	-	-	- 0 0 3 11	- 0 0 21.4 78.6	-	-	-	0.366**
Length of stay (day)	-	-	27.93±14.25	28	3-43	-	-	26±12.67	25	5-45	0.703*
Rehabilitation session number	-	-	18.93±10.22	18	1-30	-	-	20.29±8.00	20	1-30	0.696*

SD: Standard deviation; ASIA: American Spinal Injury Association.* Fisher exact test; ** Mann-Whitney U test. Bold values indicate statistically significant p-values (p<0.05).

TABLE 3. Between-group and within-group comparisons.

-	Early rehabilitation group (n=15)			Late rehabilitation group (n=14)			-
-	Mean±SD	Median	Min-Max	Mean±SD	Median	Min-Max	p
FAC (pre-rehabilitation)	-	0	0-4	-	1	0-4	0.032†
FAC (post-rehabilitation)	-	0	0-4	-	2	0-4	0.026†
P**	0.034§			0.007§			-
Δ FAC score (post-pre)	-	0	0-2	-	1	0-2	0.409‡
Barthel score (pre-rehabilitation)	29.33±31.10	20	0-90	69.64±33.02	80	5-100	0.005‡
Barthel score (post-rehabilitation)	49.33±27.57	50	5-95	81.07±24.03	82.5	15-100	0.003‡
P**	0.002§			0.005§			-
Barthel Index percentage change (% of total score)	-	36.36	0-1100	-	6.70	0-1000	0.455†

SD: Standart deviation; FAC: Functional ambulation category; * Between-group comparisons; ** Within-group comparisons; † Fisher’s exact test; ‡ Mann-Whitney U test; § Wilcoxon Signed Ranks test; Δ Indicates change from pre- to post-rehabilitation. Bold values indicate statistically significant p-values (p<0.05). Although median (min-max) FAC values appear identical between groups, statistical comparisons were based on category distributions due to the ordinal nature of the FAC scale.

Since the only significant difference between the two groups in terms of demographic data and clinical characteristics was the presence of recurrence, the patients were divided into two subgroups-those with recurrence and those without-to evaluate the potential effect of this variable on the outcomes. The two subgroups were then compared with respect to the change in FAC scores and the percentage change in Barthel Index scores. No significant difference was found between the subgroups for either outcome measure (p=0.661, p=0.342, respectively).

During the rehabilitation period, complications were observed in only three patients-two in the early rehabilitation group and one in the late rehabilitation group. In the early rehabilitation group, one patient experienced respiratory arrest on the 5^th day of treatment. The patient received immediate resuscitative intervention, was transferred to the intensive care unit, and was subsequently referred to the pulmonology service, from which he was later discharged. This patient had a diagnosis of metastatic lung carcinoma. The second complication in the early rehabilitation group occurred in a patient with a Grade I meningioma who developed a postoperative hematoma in the second week following surgery. Rehabilitation was discontinued, and the patient was referred back to the neurosurgical team, where she reportedly underwent a second surgical procedure. In the late rehabilitation group, the only complication was a seroma that developed on the 10^th day of treatment in a patient who had undergone surgery for a Grade 1 hemangioblastoma. Despite pressure dressing applications, the seroma increased in size; therefore, rehabilitation was discontinued, and the patient was referred to the surgical team. The seroma was subsequently managed with percutaneous drainage, followed by clinical monitoring.

DISCUSSION

In this study, the rehabilitation outcomes of 29 patients with tumor-related spinal cord injuries were discussed. The results demonstrated that rehabilitation interventions, implemented at any stage during the postoperative period, could lead to significant improvements in both ambulation levels and daily life activities. In both groups, only a small number of complications that could affect the rehabilitation process were encountered.

Tumors are among the most common etiologies for spinal cord injuries, constituting 15-30% of patients in spinal cord rehabilitation units.^[10] With the aging of the population and advancements in early cancer diagnosis and treatment, the number of patients with spinal cord injuries related to tumors who have the capacity to benefit from rehabilitation is steadily increasing.^[12]

Having both diseases that can lead to devastating outcomes is as challenging for patients as it is demanding for physicians to manage the rehabilitation process in this group of patients. The availability of literature information for consultation during the planning of this process is significantly restricted. In the few studies reporting rehabilitation outcomes in patients with spinal cord damage due to neoplastic compression, patient groups are heterogeneous concerning tumor histopathology.^{[13, 14, 15, 16, 17]} In this study, patients with tumor-related spinal cord injuries, defined as injuries directly resulting from primary or metastatic spinal tumors, as well as iatrogenic spinal cord injury occurring during surgical intervention for these tumors, were evaluated. Although most studies investigating rehabilitation outcomes in spinal cord injury have focused on cases secondary to metastatic tumors,^{[13, 15, 17, 18]} similar to the more limited cohorts reported in the literature,^{[14, 16]} our study also predominantly included patients with spinal cord injuries caused by primary spinal tumors. The fact that most of our patients were classified as ASIA-C and ASIA-D was also consistent with the literature.^{[3, 5, 15, 16, 19, 20]} While it may imply that spinal cord injuries caused by tumors result in less damage compared to other etiologies, this outcome could also be attributed to the fact that patients with more severe injuries may have lower expectations and therefore be less frequently referred to rehabilitation units. Another possible explanation is that the gradual increase in tumor size often leads to a slowly progressive neurological deficit, allowing diagnosis to be made before severe impairment develops, especially when compared with other causes of spinal cord injury that typically present with a more abrupt onset.

The first challenging decision related to the rehabilitation of patients with spinal cord injuries due to tumors is determining who will be included in the rehabilitation program. Considering the varying survival models of patients with different tumor etiologies, rehabilitation teams should exercise caution in selecting only those patients who have the potential to benefit from inpatient rehabilitation programs and should recommend palliative care in cases where the prognosis is poor. Naturally, individuals with tumors tend to have lower functionality during hospital admission and discharge, and the overall change in functionality is likely to be lower.^[11] When admitting these patients to the rehabilitation program, the entire rehabilitation team, particularly the rehabilitation physician, should bear in mind that these patients may deteriorate more quickly than expected, the rehabilitation process may be disrupted due to complications, necessitating transfer to other departments or palliative care related to the patient's illness. Additionally, it should be considered that some patients may have a diminished ability to participate in treatment compared to non-tumor-related spinal cord injury patients, and in some cases, patients may even pass away during inpatient rehabilitation.^[13] When evaluating a patient with spinal cord injury due to a tumor for rehabilitation, consideration should be given to other diseases accompanying their neurological and oncological condition; accordingly, reports involving similar patient populations indicate that rehabilitation admission typically requires patients to be medically stable, able to tolerate 2-3 h of intensive therapy per day, not dependent on mechanical ventilation, and expected to survive at least 3-6 months.^{[14, 16, 17]} Consistent with current recommendations, rehabilitation interventions in this study were delivered with careful attention to patients’ neurological severity and overall clinical status. Therefore the patient group consists of individuals who are younger, predominantly have primary tumors rather than metastatic tumors, and have better neurological levels. On the subject the rehabilitation of these patients, the second challenge encountered is determining how long the rehabilitation process should be. Particularly, considering the limited survival in spinal cord injuries due to tumors, there is a need to strike a balance between a longer rehabilitation process for better functional outcomes and expediting the process to allow patients more time at home with their loved ones. Hence, optimizing rehabilitation management is necessary.^[5] Sometimes, planning short-term rehabilitation programs, including caregiver training, may be necessary to ensure that the patient can be managed at home with appropriate support or assistance.^[11] Data on patients' length of stay in hospital in the literature vary significantly, with average values ranging from 27 to 90 days.^{[12, 16, 17, 21, 22]} In this study, the patients' average hospitalization duration was found to be 26 days, aligning with data reported by McKinley and colleagues^[17] from a tertiary healthcare institution comparable to us. One significant gap in the literature is when to initiate rehabilitation for patients with tumor-related spinal cord injuries. In previous studies, although there has been no direct assessment of this issue, a small number of reports have provided information on the elapsed time before admission to a rehabilitation unit. In one study that included all non-traumatic etiologies, the mean elapsed time was 22.5 days.^[22] In contrast, among studies focusing specifically on neoplastic spinal cord injury, one reported a markedly longer interval of 161.2 days,^[14] while another found an average of 74.4 days.^[16] These findings indicate that the initiation of rehabilitation in our cohort occurred considerably earlier than what has been reported in the existing literature on tumor-related spinal cord injury. Therefore, to define the early postoperative period, we drew upon findings from a study conducted in our country involving patients with stroke, a neurologic condition with comparable rehabilitation dynamics. In that study, the average time for initiating post-stroke rehabilitation in tertiary healthcare institutions was found to be approximately 15 days.^[10] When patients were grouped based on the onset of rehabilitation and the outcomes were evaluated, both groups showed significant improvements in post-treatment ambulation and independence in daily activities compared to pre-treatment values, consistent with findings reported in the literature.^{[14, 16]}

In the literature, the frequency and nature of complications reported during rehabilitation in patients with tumor-related spinal cord injury vary widely across studies. In one study involving 63 patients, 21 complications were documented, most commonly pressure sores, deep vein thrombosis, pulmonary embolism, heterotopic ossification, and various urological complications (excluding urinary tract infections).^[16] Another study of 32 patients reported a notably high rate of medical complications, with pain observed in 66% of cases, followed by urinary tract infections in 14 patients, pulmonary complications in six, and less frequent events such as gastrointestinal bleeding, pressure sores, wound dehiscence, depression, neutropenia, and hyponatremia.^[17] A third study reported 29 complications among 43 patients, predominantly deep vein thrombosis, pressure ulcers, and urinary tract infections.^[14] In another cohort, most patients (65%) experienced no complications; however, those reported were largely related to the underlying malignancy, including anemia, neutropenia with infection, venous thromboembolism, pressure ulcers, and various infectious complications.^[15] In comparison with these studies, the overall complication rate in our cohort was notably low, with only three complications observed among 29 patients. Although a single life-threatening complication occurred in the early rehabilitation group, this was considered to be primarily related to the patient’s underlying metastatic tumor and a complication of the primary malignancy (lung carcinoma). Even in this patient, as the event occurred in a fully equipped hospital setting, timely and appropriate intervention was possible and was likely life-saving. Other complications were of a nature that could be resolved with timely interventions. The lower rate of complications observed in our study may be attributable to the fact that common but less severe events, such as urinary tract infections, which were frequently reported in previous studies, were not systematically recorded. However, differences in demographic and clinical characteristics, as well as variations in parameters such as the time elapsed before rehabilitation and length of hospital stay, may also have contributed to this discrepancy. Nevertheless, despite the findings of the aforementioned studies, the remarkably low number of complications observed during a rehabilitation process initiated at a much earlier stage is noteworthy. This underscores the importance of appropriate patient selection and the optimization of hospital stay and peri-rehabilitation management.

In intergroup comparisons, the post-treatment FAC and Barthel Index scores of the late rehabilitation group were found to be significantly higher than those of the early rehabilitation group. However, this result is likely due to the fact that the functional scores of the late rehabilitation group were significantly better even before treatment. We believe that this result is attributed to the early initiation of treatment sought by patients with more significant functional loss. Furthermore, the time that transpired after surgery in the late rehabilitation group might have contributed to the observed improvement, potentially creating this difference. However, when we compared the FAC differences and percentage changes in Barthel Index scores between the two groups in our study, no significant difference was found. This result suggests that the timing of rehabilitation did not make a difference in the outcomes.

To our knowledge, this is one of the few studies in the literature to examine the timing of rehabilitation in tumor-related spinal cord injury. This study has several limitations, including the relatively small sample size, the retrospective design, and the presence of uncontrolled baseline differences between the groups, all of which may limit the generalizability of the findings.

In conclusion, the role of rehabilitation in improving functional outcomes among patients with tumor-related spinal cord injury is well supported. In the present study, early initiation of rehabilitation was associated with meaningful functional improvement, and while postoperative complications cannot be entirely excluded, early rehabilitation did not appear to heighten the risk of surgery-related adverse events. Furthermore, given that potential complications can be promptly identified and managed within a rehabilitation setting, concerns regarding their occurrence should not constitute a reason to delay referral. Accordingly, following a comprehensive assessment of the patient’s primary medical condition, timely integration into a rehabilitation program is warranted. Nevertheless, in light of the limitations of this study larger, methodologically robust prospective studies are required to more precisely elucidate the effects of rehabilitation timing and other clinical determinants on functional outcomes.