Introduction
Spinal metastases can lead to metastatic spinal cord compression (MSCC) [1,2]. These metastases not only impair quality of life (QOL) due to pain from metastatic lesions and pathological fractures but are also associated with neurological symptoms [3,4]. MSCC occurs when a spinal metastasis compresses the spinal cord and is considered an oncology emergency [5,6]. It affects 5%–10% of patients with spinal metastases and significantly impairs both the patient’s prognosis and QOL [7–9].
Identifying high-risk cases of MSCC is crucial for timely and effective intervention. Several risk factors for MSCC have been reported, including the spinal instability neoplastic score (SINS) and the number of spinal metastases [10,11]. Surgical intervention after the onset of MSCC is reportedly associated with poorer patient prognosis and QOL compared to interventions made prior to the MSCC development [12,13]. However, without proper patient selection, surgical intervention can sometimes compromise QOL [14,15]. Therefore, recognizing the risk factors for MSCC in patients with spinal metastases is essential for making appropriate therapeutic decisions.
The primary mechanism of MSCC is believed to be the extension of the tumor into the spinal canal. Thus, spinal metastasis with epidural involvement may be a significant risk factor for the development of MSCC. However, the relationship between epidural metastasis and the onset of MSCC is poorly understood. We hypothesized that in addition to the presence of epidural metastasis, specific imaging characteristics, such as the degree and circumferential extent of spinal cord compression, may be associated with the development of MSCC.
This study aimed to identify imaging-based risk factors for MSCC in patients with spinal metastases, focusing on the presence and severity of epidural metastases.
Materials and Methods
We retrospectively reviewed patients with spinal metastasis who visited the bone metastasis clinic from 2017 to 2023 at a single institution. Considering the patient bias in the bone metastasis clinic, we selected patients with local symptoms, such as neck pain, back pain, and low back pain. The exclusion criteria were as follows: (1) history of surgical intervention for spinal metastasis, (2) intradural or intramedullary lesions, (3) spinal metastases at the cauda equina level, and (4) missing data. The study adhered to the guidelines of the Declaration of Helsinki, and the study protocol was approved by the Institutional Review Board of Chiba University Hospital (approval number: M10251). Informed consent was obtained from all patients.
Patients’ demographic data, primary tumor type, level of spinal metastasis, and follow-up period were evaluated. The patients were monitored for the development of MSCC during follow-up. We defined MSCC as a decrease of one or more levels on the American Spinal Injury Association (ASIA) Impairment Scale due to spinal metastasis [16]. All patients included in this study were neurologically healthy at baseline (ASIA grade E). The status of systemic therapy administration for primary tumor on initial visit, as well as the current status of systemic therapy for primary tumor, local radiotherapy for spinal metastasis, and surgery for spinal metastasis during follow-up, were also reviewed. The metastatic vertebra was identified as the vertebra with the most tumor invasion. The level of spinal metastasis was classified as cervical (C1–7), upper thoracic (T1–5), lower thoracic (T6–10), or thoracolumbar junction (T11–L2). Primary tumor growth was categorized as fast, moderate, or slow by following the new Katagiri classification [17]. The presence of pathological fracture and SINS was assessed using a plain computed tomography (CT) scan of the spine at the initial visit. The CT images used for evaluation were obtained prior to MSCC, even in cases with MSCC.
Evaluation of epidural metastasis
To assess the potential risk factors contributing to the development of MSCC, epidural metastases were evaluated using images taken before the development of MSCC in follow-ups, even in cases in which MSCC developed during follow-up. The presence of epidural metastasis at the level of spinal metastasis was diagnosed by (1) magnetic resonance imaging (MRI), (2) contrast-enhanced CT, or (3) plain CT interpreted by three radiologists. In cases evaluated using CT alone, epidural metastases were examined based on the consensus of at least two of the three radiologists. Despite evaluating the features of epidural metastases using MRI, the epidural spinal cord compression (ESCC) scale and circumferential angle of spinal cord compression (CASCC) were also assessed [18,19]. The ESCC grade and CASCC were determined only in those cases who underwent MRI; however, CT was not performed for these evaluations. The ESCC was classified into four grades: grade 0, bone-only disease; grade 1, epidural extension without spinal cord compression; grade 2, spinal cord compression with cerebrospinal fluid (CSF) visible around the cord; and grade 3, spinal cord compression with no CSF visible around the cord [19]. Using axial T2-weighted MRI, CASCC was evaluated as the angle between the center of the spinal cord and the points where epidural metastases compressed the spinal cord (Fig. 1). If the spinal cord was deformed by compression, we chose the nearest axial level with a circular spinal cord appearance as a reference (Fig. 1B) [18].
In addition, ESCC and CASCC were evaluated to assess the features of epidural metastases at MSCC; provided that MRI had been performed at that point. In patients, who underwent MRI before MSCC occurred, the interval between the two MRIs was recorded.
Statistical analysis
Values are expressed as mean±standard deviation. Age, SINS, ESCC grade, and CASCC were compared using an unpaired t-test. Categorical variables were compared using the chi-square test. However, if the expected frequency in any cell was less than five, Fisher’s exact test was applied instead. Primary tumor type, primary tumor growth, and metastasis level were compared using the chi-square test for independence. Upon identifying a significant difference in the chi-square test for independence, each item was compared using the chi-square test or Fisher’s exact test. Variables with p<0.1 in the two-group comparisons were included in a multivariable logistic regression analysis to identify the risk factors. In multivariate logistic regression analysis, we categorized primary tumor growth (low, moderate, and high) based on the new Katagiri classification. Although the primary tumor type showed significant differences in the two-group comparison, including each tumor type as an independent variable in the regression model would have made it statistically unstable due to small subgroup sizes. Receiver operating characteristic (ROC) curve analysis was performed for ESCC and CASCC assessment. Additionally, Cox proportional hazards models were constructed to evaluate the association between imaging-based risk factors and time-to-MSCC onset. Separate models were created using sex and either ESCC (dichotomized as ≥3 vs. <3) or CASCC (dichotomized as >180° vs. ≤180°) as covariates. These analyses were limited to cases with evaluable MRI at the initial assessment. Statistical significance was set at p<0.05.
Results
The study enrolled 191 patients (age: 66.4±12.9 years; sex: 120 male, 71 female) (Fig. 2). Patient data are shown in Table 1: 41 patients (21.5%) had MSCC, 35 (19.3%) underwent surgery during follow-up, 138 (71.9%) had epidural metastases (Fig. 2), and 61.5% of epidural metastases were detected on MRI.
A comparison of evaluation items between the enrolled patients with spinal metastases who developed MSCC during follow-up (MSCC group 1) and those without MSCC (Non-MSCC group 1) is shown in Table 2. There were significant differences in primary tumor type, primary tumor treatment status at the initial visit, metastasis level, and presence of epidural metastasis between the two groups (Table 2). Among the patients with spinal metastases who developed MSCC during follow-up, 97.6% had epidural metastases before the onset of MSCC. Multivariate logistic regression analysis was conducted using sex, primary tumor type, untreated primary tumor, metastasis level, and the presence of epidural metastasis as independent variables, with the MSCC development as the dependent variable (Table 3). Breast cancer, which exhibited a complete separation, was excluded from the multivariate logistic regression analysis. Consequently, epidural metastasis was identified as an independent risk factor for MSCC, considering breast cancer a protective factor (Table 3).
The features of epidural metastases were evaluated in 85 patients for whom epidural metastases were assessable on MRI at the time of non-MSCC (Fig. 2). Of these 85 patients, 19 developed MSCC during follow-up (MSCC group 2) and 66 did not develop MSCC (Non-MSCC group 2). As shown in Table 4, there were significant differences in the ESCC grade and CASCC between the two groups. Multivariate logistic regression analysis identified ESCC grade and CASCC as independent risk factors for the development of MSCC (Table 5). Cox proportional hazards models were constructed using sex and ESCC grade (≥3 vs. <3) or CASCC (>180° vs. ≤180°) as covariates to evaluate the association with time-to-MSCC onset. Neither ESCC grade ≥3 nor CASCC >180° showed a statistically significant association with earlier MSCC onset (hazard ratio [HR], 1.25; 95% confidence interval [CI], 0.44–3.51; p=0.67 for ESCC; HR, 1.19; 95% CI, 0.32–4.39; p=0.79 for CASCC). Using ROC curve analysis, cutoffs and area under the ROC curve (AUC) were calculated to be 3 for the ESCC scale (AUC: 0.88, sensitivity: 0.84, specificity: 0.88) and 180° for CASCC (AUC: 0.88, sensitivity: 0.84, specificity: 0.88) (Fig. 3). These findings suggest that a grade-3 ESCC and CASCC >180° represent clinically relevant thresholds for identifying patients at high risk of developing MSCC. These cutoff values showed both high sensitivity and specificity, indicating their potential in guiding clinical decision-making.
Of the 41 patients who developed MSCC, the features of epidural metastases at the time of MSCC were evaluated in 38 patients. At the time of MSCC, the ESCC grade was 2.9±0.4, and CASCC was 249.4°±67.8°. Thirty-six of the 38 cases showed an ESCC grade of 3, and 33 of the 38 cases exhibited CASCC ≥180°. In 19 patients whose pre-MSCC and MSCC MRIs were compared, the ESCC grade increased by 0.2±0.5, and CASCC increased by 42.9°±55.0°; in no case did the two parameters decrease. The mean interval between MRI and MSCC onset was 35.8±23.6 days (range, 7–140 days).
Discussion
Consistent with our hypothesis, the presence of epidural metastasis was a risk factor for the development of MSCC. Furthermore, epidural metastases with severe spinal cord compression and those surrounding the spinal cord by more than 180° were identified as high-risk metastases.
Epidural metastasis was a risk factor for MSCC in patients with spinal metastasis. Previous studies have reported that risk factors for MSCC include SINS, number of vertebral bodies involved, and primary tumor type, but no studies have focused specifically on epidural metastases [10,11]. The leading cause of MSCC is compression of the spinal cord due to the extension of vertebral metastasis [20]. Therefore, epidural metastasis is a plausible risk factor for MSCC. Screening for epidural metastasis in patients with spinal metastasis is essential, as this study found epidural metastases in 71.9% of patients with spinal metastasis and 97.6% of patients with MSCC. We recommend that MRI be performed in patients with spinal metastasis to assess MSCC risk.
Among epidural metastases, those with severe spinal cord compression and those surrounding the spinal cord by more than 180° are considered high-risk metastases. Lesions with an ESCC grade of 3 indicate severe compression of the spinal cord by tumors [19]. Patients with an ESCC grade of 1–2 who do not exhibit neurologic deficits have been reported to have less significance regarding prophylactic decompression surgery. In contrast, while there is no clear evidence, several studies have recommended considering surgery for patients with ESCC grade 3 [21]. This study indicates that spinal metastasis with ESCC grade 3 is a risk factor for imminent paralysis, even if a patient shows no neurological deficits. Therefore, patients with ESCC grade 3 or CASCC >180° may require closer clinical observation and consideration of prophylactic interventions, although further evidence is required to establish treatment guidelines. A CASCC of ≥180° has been identified as a risk factor for poor postoperative outcomes in MSCC cases [18], and our findings suggest that it is also useful for identifying impending paralysis. These thresholds, ESCC grade 3 and CASCC >180°, —provide clear and objective imaging criteria for assessing MSCC risk before the onset of neurological symptoms.
In this study, several patients received palliative treatments, such as radiotherapy or bone-modifying agents, during the follow-up period. Although these therapies were not specifically intended to prevent MSCC, they may have influenced the onset timing to some extent. ESCC and CASCC were significant predictors of the MSCC development in logistic regression analysis, but they did not show a statistically significant association with time-to-MSCC onset in Cox regression models. This discrepancy may reflect variations in follow-up intervals or treatment interventions between imaging and symptom onset. Further prospective studies are warranted to assess the temporal dynamics of MSCC progression. While our findings confirm that epidural metastases are associated with MSCC, a clinically expected relationship, given the mechanism of spinal cord compression. The novel contribution of this study lies in the identification of specific imaging features—ESCC of grade 3 and CASCC of >180°—that characterize high-risk epidural metastases.
We also found that breast cancer is associated with the development of MSCC. Breast cancer is reportedly less likely to lead to MSCC due to the high incidence of blastic lesions, which aligns with this study findings [22]. Our findings are consistent with previous research on the association between breast cancer and the development of MSCC.
Nonetheless, this study had several limitations. First, the patients included in this study were those who visited a bone metastasis clinic, which may have resulted in a heterogeneous patient population. The study focused on symptomatic patients to minimize selection bias; however, this bias could not be fully eliminated, and patients with advanced spinal metastases may have been overrepresented. Second, the timing of MRI was inconsistent, and imaging was not performed in many cases. Therefore, the frequency of epidural metastases could have been underestimated, and could have led to an overestimation of the impact of epidural metastases on the development of MSCC, given that epidural metastases were present in most MSCC cases. Nevertheless, we believe that the characteristics of epidural metastases remain strongly associated with the development of MSCC, as MRI was used to evaluate epidural metastases in all cases. Third, ESCC and CASCC were assessed only in cases using MRI. In approximately 35% of cases, the evaluation of epidural metastases relied on CT, which may have led to an underestimation or overestimation of the extent of spinal cord compression. Future studies should ensure standardized imaging to enhance measurement accuracy.
Conclusions
In conclusion, epidural metastasis was identified as a risk factor for MSCC in patients with spinal metastasis. Additionally, epidural metastasis in those with an ESCC grade of 3 and CASCC >180° were identified as high-risk tumor factors. MRI may be advisable for assessing the risk of MSCC.
Key Points
This study aimed to identify whether the presence and characteristics of epidural metastases are risk factors for the development of metastatic spinal cord compression (MSCC) in patients with spinal metastases showing local symptoms.
Epidural metastasis was a risk factor for MSCC in patients with spinal metastasis.
Among epidural metastases, those with severe spinal cord compression (epidural spinal cord compression grade of 3) and those surrounding the spinal cord by more than 180° were classified as high-risk metastases.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
