Abstract
Background/Aim: Spontaneous spinal epidural hematoma (SSEH) is a rare but serious condition, accounting for less than 1% of spinal lesions, with an incidence of 0.1 per 100,000 annually. Discovered by Jackson in 1869, around 40-50% of SSEH cases often lack a definitive cause, though risk factors, such as anticoagulant usage, vascular malformations, and hypertension are recognized. Symptoms vary from mild pain to severe neurological impairments like paraparesis, depending on the spinal cord compression level. Prompt treatment, usually involving spinal decompression and hematoma removal, is crucial, especially in cases of neurological decline. The study aims to provide comprehensive analysis of SSEH through examination of by patient cases, critical prognostic factors, and therapeutic strategies, based on demographics, clinical data, and outcomes observed at the Tri-Service General Hospital. Patients and Methods: This retrospective study, spanning 2003-2023 at the Tri-Service General Hospital, analyzed 14 patients with SSEH. It examined demographics, risk factors, clinical and radiological profiles, treatments, outcomes, and prognoses, using SPSS software (version 22.0) and adhering to the Modified Rankin Scale (mRS) and the American Spinal Injury Association (ASIA) impairment scale guidelines for data analysis. Results: In this study of 14 patients with SSEH, 93% underwent urgent surgery, including total laminectomy or open-door laminoplasty, while 7% received conservative treatment. Post-surgery, 69.2% showed favorable outcomes (mRS ≤2) in the one-year follow-up, while 30.8% had poorer results (mRS 3-4). A significant negative correlation was noted between initial ASIA scores and one-year mRS outcomes, suggesting less initial impairment predicts better recovery. These findings indicated that a moderate positive correlation between treatment delay and one-year mRS scores. Nevertheless, factors, such as age, antiplatelet use, spinal levels with hematoma localization, and myelopathy signs observed before treatment did not demonstrate any significant effects on neurological outcomes during the one-year follow-up. Conclusion: Patients with minor initial deficits or those receiving early surgery, preferably within 12-36 h of symptom onset, exhibit better neurological recovery. Poor prognosis correlates with high International Normalized Ratio (INR) on anticoagulants, hematoma size, lumbar involvement, or severe motor issues. Rapid surgical hematoma evacuation is advised. Our study supports recovery of neurological function following surgical intervention in all cases, highlighting the potential efficacy of surgical decompression even in severe and prolonged instances of SSEH.
Spontaneous spinal epidural hematoma (SSEH) is an infrequent yet critical medical condition, representing less than 1% of spinal space-occupying lesions with an annual incidence of about 0.1 per 100,000 individuals in Sweden (1). It involves the accumulation of blood in the epidural space without an evident traumatic or iatrogenic origin, potentially leading to significant spinal cord compression. First identified by Jackson in 1869, SSEH’s precise etiology remains unclear in approximately 40-50% of cases (2). Identified risk factors include hemophilia, vascular malformations, coagulopathies, anticoagulant use, neoplasms, hypertension, and conditions that elevate intra-thoracic or intra-abdominal pressure, like coughing, physical strain or the Valsalva maneuver (3, 4).
Clinically, SSEH presents a spectrum of symptoms ranging from mild radicular pain to severe neurological deficits, such as paraparesis or quadriparesis, contingent upon the level and severity of spinal cord compression (5). The condition’s manifestation is often acute, and approximately 40-60% of cases are deemed idiopathic, lacking an identifiable cause (6). Magnetic resonance imaging (MRI) is the diagnostic modality of choice, offering crucial insights for accurate diagnosis (1). Immediate and appropriate medical intervention is vital in the management of SSEH, particularly for patients exhibiting progressive neurological deterioration. Treatment typically involves urgent spinal cord decompression and hematoma evacuation (7). For patients with negligible or no neurological impairment, conservative monitoring may be considered. The extent of neurological deficits prior to intervention significantly influences the prognosis (8).
However, the rarity of SSEH poses challenges in gathering comprehensive data, which is crucial for delineating clear treatment guidelines. Therefore, meticulous documentation of each case, including patient history, symptomatology, hematoma characteristics, and post-treatment outcomes, is imperative to enhance our understanding and management of this rare, but potentially debilitating condition (7). This study aimed to enhance SSEH knowledge by reviewing patient cases at our Institution, identifying key variables impacting prognosis, and suggesting optimal treatment strategies based on demographic, clinical, and outcome data analysis.
Patients and Methods
Patients. This study was a retrospective descriptive analysis conducted at the Tri-Service General Hospital, Taipei, Taiwan, spanning from 2003 to 2023, to investigate cases of SSEH. The patient cohort was identified from the hospital’s database using specific diagnostic terms including ‘non-traumatic spinal hemorrhage’, ‘non-traumatic spinal hematoma’, and ‘non-traumatic spinal epidural hemorrhage’. The data utilized in this research were extracted from various sources within the patient’s medical records, encompassing discharge summaries, clinical narratives, and documentation from the intensive care and surgical departments. This analysis focused on a range of variables including demographic information, risk factors, clinical presentations, radiological findings, therapeutic strategies, clinical outcomes, and prognostic evaluations. The study’s inclusion criteria were rigorously restricted to patients diagnosed with SSEH, explicitly excluding any cases associated with spinal trauma, tumor, vascular lesion, or related to surgical complications.
Outcomes. MRI studies were conducted utilizing a 1.5 Tesla MRI scanner. The diagnosis of SSEH was confirmed through identification of a space-occupying lesion within the epidural space of the spinal canal, characterized by isointense to hyperintense signals on T1-weighted images and hyperintense signals on T2-weighted images. These lesions often exhibited heterogeneous signal intensities on T2-weighted images, indicative of hemoglobin degradation. Myelopathy was diagnosed based on the presence of spinal cord hyperintensities observed in T2-weighted images and Short Tau Inversion Recovery (STIR) sequences.
One year post SSEH onset, patients’ functional status was evaluated using two scales: the Modified Rankin Scale (mRS) and the Modified Rankin Scale (ASIA) impairment scale. The mRS, which ranges from 0 (no symptoms) to 6 (severe disability), and the ASIA scale, ranging from A (complete impairment) to E (normal function), were employed to assess the extent of recovery. Patients achieving an mRS score between 0 and 2 and an ASIA impairment scale rating of D or E were classified as having a favorable functional status. Conversely, those with an mRS score ranging from 3 to 6 or an ASIA impairment scale rating between A and C were considered to have an unfavorable functional status.
Statistical analysis. The data encompassing the mRS and ASIA impairment scores, alongside variables, such as the utilization of antiplatelet or anticoagulant medications, patient age, and time elapsed prior to treatment, were subjected to rigorous statistical analysis to determine their significance. The Chi-Squared Test was employed for this purpose, and a p-value threshold of less than 0.05 was established as the criterion for statistical significance. All statistical analyses were executed using the Statistical Package for the Social Sciences (SPSS) software, version 22.0, provided by SPSS Inc.
Results
Baseline patient characteristics. This study encompassed a total of 14 patients, with a mean age of 59.8±19.21 years, a median age of 67.5 years, and an interquartile range (IQR)=25.2 years. Female participants constituted 42.9% of the study cohort. The predominant comorbid conditions identified in this group were hypertension, diabetes mellitus, and the utilization of antiplatelet medications. Specifically, hypertension was observed in 43% (6 patients), antiplatelet medication usage in 29% (4 patients), diabetes mellitus in 14% (2 patients), ischemic heart disease in 7% (1 patient), and carotid atherosclerosis in 7% (1 patient). Furthermore, 14% of patients exhibited two or more vascular risk factors. In cases where patients were receiving anticoagulant therapy, the INR was measured upon their admission to the emergency department. It was noted that all patients included in this study had an INR value of less than 2. The principal demographic data of the study population are concisely presented in Table I.
Main clinical, diagnostic, therapeutic, and prognostic characteristics for the 14 patients.
Clinical manifestations. The predominant clinical manifestation observed in 79% of the cases was the abrupt onset of backache, correlating with the location of the hematoma. A total of 57% of the participants (eight individuals) reported experiencing radiating pain with a non-metameric distribution. The majority of the patients (93%) exhibited neurological deficits that either coincided with or followed the onset of pain. These deficits were categorized as sensorimotor syndromes in 79% of cases, pure motor syndromes in 14%, and pure sensory syndromes in 7%. There was a single case in which the patient presented with isolated neck pain and hypesthesia following neck extension, without any additional symptoms. The motor deficits varied among the patients, including paraplegia in five cases, hemiparesis in four cases, paraparesis in three cases, and quadriplegia in one patient. Furthermore, 43% of the patients (six individuals) experienced a loss of sphincter control. The principal clinical findings from this study are systematically compiled in Table I.
Diagnostic assessment. The median duration from the onset of symptoms to receipt of medical care at the emergency department was recorded at 3.9±1.93 h. The interval between symptom emergence and evaluation by the on-call neurologist was 6.8±2.71 h. A notable 93% of the patients (thirteen individuals) underwent MRI of the spinal cord. The median time from the onset of symptoms to the performance of the MRI scan was 8.7±3.21 h. Prior to the spinal cord MRI, eleven patients were subjected to additional imaging studies, which included simple radiography of the vertebral column in nine cases and computed tomography (CT) of the vertebral column in seven cases. Among the seven patients who underwent vertebral column CT, five exhibited radiological findings suggestive of a spinal epidural hematoma, characterized by a hyperdense extramedullary mass. The average time taken from arrival at the emergency department to the completion of the MRI scan was 4.8±2.71 h. The key radiological findings from this study are concisely summarized in Figure 1, Figure 2 and Figure 3 which provide illustrative examples of the spinal cord MRI images obtained.
Thoracic spine MRI of an 80-year-old female. (A) T1-weighted imaging, sagittal view, revealing an isointense to hyperintense lesion in the dorsal epidural space from T6-T10; (B) T2-weighted imaging, sagittal view, revealing the lesion to be hyperintense; (C) T2-weighted imaging, axial view, arrow: epidural hematoma occupies the spinal canal and compresses the spinal cord. MRI: Magnetic resonance imaging.
A 69-year-old male with hematoma from C4 to T1. (A) T1-weighted imaging, sagittal view, revealing an isointense to hyperintense lesion; (B) T2-weighted imaging, sagittal view, revealing the lesion to be hypointense; (C) T2-weighted imaging, axial view, revealing an isointense lesion occupies the dorsal spinal canal and compresses the spinal cord.
A 44-year-old male with SSEH. (A) T1-weighted imaging, sagittal view, revealing an isointense lesion in the ventral epidural space from T6-T2; (B) T2-weighted imaging, sagittal view, revealing the lesion to be hyperintense; (C) T2-weighted imaging, axial view, revealing spinal cord compression on the ventral side. SSEH: Spontaneous spinal epidural hematoma.
Therapeutic strategies and neurological function follow-up. Of the 14 patients included in this study, a significant majority (93%) received urgent surgical intervention. This encompassed total laminectomy with hematoma evacuation in 11 patients and open-door laminoplasty in 2 patients. One patient (7%) underwent conservative treatment, which included corticosteroids and symptomatic management. Among the 13 patients who received surgical treatment, 9 (69.2%) achieved a mRS score of 2 or less at the one-year follow-up, indicating a favorable functional outcome. Conversely, 4 patients (30.8%) had mRS scores ranging from 3 to 4, which is indicative of a poor functional outcome. Rehabilitation was provided to all patients, except for one individual who did not experience any motor deficits.
These findings at the first time demonstrated that a significant negative correlation between the ASIA impairment scale at initial contact and the mRS score at one-year follow-up, with a p-value of less than 0.05 (Table II). This negative correlation suggests that higher ASIA scores, indicating less severe impairment, are linked to lower mRS scores, representing better functional outcomes one year after SSEH insult. Additionally, a moderate positive correlation was observed between the time elapsed before treatment and the one-year mRS score. However, this correlation did not reach statistical significance, as evidenced by a p-value of 0.11 (Table III). However, there is no statistically significant correlation between the one-year mRS scores and the potentially poor prognostic factors, such as patient age, the use of antiplatelet agents, the level of spinal epidural hematoma, and the presence of myelopathy signs (Table III). Although there is an observed trend suggesting that a prolonged time to treatment might be associated with poorer outcomes, this observation is not statistically significant according to the data available.
Outcome-related patient characteristics compared with mRS score at 1 year.
Outcome-related patient characteristics.
Discussion
SSEH are uncommon neurological disorders, more frequently observed in adults over 40, with a modest male predominance (9, 10). Although rare in children, 30 pediatric cases have been documented (11, 12). The average patient age is 71, with predisposing factors including hypertension and usage of oral anticoagulants, particularly with elevated INR levels (13). SSEH are associated with various conditions, including arteriovenous malformations, coagulopathies, therapeutic thrombolysis post-acute myocardial infarction, hemophilia B, factor XI deficiency, chronic aspirin use, and vertebral hemangiomas (3, 14). However, nearly half of the cases present no clear risk factors. Despite their rarity, accounting for only 0.1% of spinal epidural lesions, the severe potential outcomes of SSEH necessitate their consideration in differential diagnoses for spinal cord-related symptoms (1).
The cervicothoracic and thoracolumbar regions are the most common sites for SSEH (9). The debate over the origin of these hematomas continues, with most researchers suggesting they arise from the epidural venous plexus, which lacks venous valves and is susceptible to undulating pressure changes (1, 15). This venous plexus, positioned between the dura and ligamentum flavum, is more prone to rupture compared to the anterior venous plexus shielded by the posterior longitudinal ligament. Communication between the intrathoracic veins and the venous plexus facilitates pressure transfer, making areas like the cervicothoracic and thoracolumbar junctions vulnerable, especially under conditions of increased pressure, such as heavy lifting, straining, or hypertension (10, 16). Some theories suggest that the hematomas are venous in nature, caused by increased intrathoracic or intraabdominal pressure leading to venous rupture. This idea, however, is contested due to the low pressure in the epidural venous plexus and the absence of valves, which would presumably limit the extent of a venous hemorrhage (17). Alternatively, an arterial origin is proposed, attributing the hemorrhages to ruptures in the radicular arteries, possibly triggered by trauma or abrupt movements (18). However, most surgical observations report epidural venous bleeding, supporting the venous origin theory. While significant, rapid neurological decline might indicate an arterial source, the slow progression of symptoms in many cases, along with the anatomical characteristics of the internal epidural plexus, support a venous source. The dorsal epidural space is more frequently affected, possibly due to the larger size of the dorsal plexus compared to the ventral, and its location beneath the posterior longitudinal ligament (19). The susceptibility to rupture in certain spinal regions may be influenced by factors like increased mobility and the cumulative effect of gravity on vessel dilation (20).
SSEH typically manifest as sudden, severe neck or back pain, often radiating to the extremities, and may progress from nerve root irritation to complete neurological impairment (21, 22). Patients may initially experience sharp pain, progressing to varying degrees of paralysis. Symptoms typically reflect lower motor neuron pathology, characterized by hyporeflexia and flaccid paralysis. The onset of neurological decline post back pain can vary, spanning hours to months (23). Early detection and imaging are vital, as SSEH can lead to severe, permanent neurological deficits, including persistent paresis or death. Neurological deficits in SSEH are primarily attributed to direct neural compression, which can block conduction and damage spinal cord myelin, leading to rapid symptom progression. Secondary factors like venous congestion and white matter edema, resulting from impaired spinal cord venous drainage, contribute to slower symptom progression (7). High cervical SSEH may lead to spinal shock, a potentially fatal condition (5). Prognosis is better with less severe initial neurological dysfunction and quicker surgical intervention. Complete sensorimotor deficits at onset typically lead to poorer outcomes, whereas partial motor deficits show better recovery (13). Patients with no pain often delay seeking treatment, resulting in worse long-term outcomes. The ASIA scale, ranging from A (most severe) to E (normal), is a crucial tool for assessing spinal cord injury severity and prognosis in SSEH patients. Surgical intervention is recommended even in severe cases (ASIA A), with studies indicating potential recovery to ASIA E status (24).
MRI is the preferred diagnostic method for SSEH, often revealing biconvex hematomas in the epidural space with distinct borders and characteristic signal intensities (25). Within 24 h of symptom onset, the hematoma appears isointense on T1-weighted and hyperintense on T2-weighted MRI images. After 24 h, it often becomes hyperintense on both T1- and T2-weighted images. Chronic hematomas eventually show hypointensity on both types of images. Fat suppression images can help differentiate hematoma from epidural fat (22, 26). Occasionally, active bleeding may be identified through central enhancement in contrast-enhanced images, while epidural hematomas may exhibit enhancement due to hyperemic or hypertrophic meninges (24, 27). However, no specific MRI characteristics have been identified that can predict the outcome of SSEH.
If MRI is not available, especially in health centers without such facilities, spinal CT scan serves as an alternative. Arteriography can also be employed to exclude dural fistulas and other malformations not evident in MRI scans (9, 28). Differential diagnoses for SSEH include acute herniated intervertebral discs, spinal cord ischemia, epidural tumors or abscesses, spondylitis, transverse myelitis, dissecting aortic aneurysms, and acute myocardial infarction (9). The increasing incidence of SSEH diagnoses coincides with the widespread use of MRI, emphasizing its importance in early detection and treatment, which can lead to complete recovery and excellent neurological outcomes (3).
Early surgical intervention, typically involving decompressive laminectomy and hematoma removal with irrigation and debridement, is the primary treatment for SSEH (29). Surgical treatment is generally advised within 12 to 24 h of motor symptom onset, or 48 h for milder cases (29). Complete recovery has been reported in surgeries performed as late as 96 h post-symptom onset (30). The duration of pre-operative paralysis is inversely related to recovery prospects, emphasizing the importance of prompt treatment. This relationship underscores the need for rapid decision-making in SSEH management, based on the patient’s initial presentation and progression of symptoms (7). There is no consensus on SSEH treatment due to its varied etiology and symptoms. Over 50% of patients retain some level of sensorimotor deficiency post-treatment. The degree of pre-operative neural deficit and the time between symptom onset and surgery are significant prognostic factors (31, 32). The ASIA score, indicating the severity of neural deficit, is critical in prognosis and treatment planning. Studies suggest that patients with higher pre-operative ASIA scores show more improvement post-surgery. Therefore, immediate assessment with the ASIA score upon SSEH symptom presentation is crucial (24, 29). Conservative treatment is recommended for patients with minimal symptoms, significant early improvement, bleeding disorders, or health issues contraindicating surgery. Conservative management is advisable only if there’s clinical improvement (12, 33, 34). The decision between surgical and conservative treatment depends on the patient’s neurological status prior to intervention, which is a crucial prognostic indicator. Non-operative management, involving bed rest and monitoring, is reserved for non-surgical candidates or asymptomatic patients, although outcomes are generally worse without surgery (27, 35).
Post-operative recovery is influenced by both the patient’s pre-operative neurological function and the time elapsed between symptom onset and surgery (28, 36). Our review corroborates that patients with higher pre-operative Frankel grades (ASIA C or D) tend to have better functional recovery. Furthermore, a shorter interval to surgery correlates with more favorable outcomes, particularly in cases with complete neurological dysfunction, where time to surgery is crucial (19, 37). Initial neurological dysfunction emerges as the strongest outcome predictor, with patients presenting severe or worsening symptoms generally recovering less effectively than those with milder symptoms. Larger hematomas, especially involving four or more spinal segments, and a lack of sensory sparing indicate a worse prognosis (7, 23). Thoracic spine areas, with limited space for cord expansion, particularly in the middle and lower thoracic regions, are associated with worse outcomes, likely due to their vulnerability to ischemic insult. However, lower thoracic spine location’s association with worse outcomes in surgery patients is not statistically significant (7, 38). Patients on anticoagulants typically fare worse, though this is not identified as an independent risk factor in prognosis. The data analysis from our study indicates the absence of a statistically significant correlation between the use of antiplatelet agents and patient outcomes. Nevertheless, the correction of coagulopathy prior to surgical intervention, especially among patients using anticoagulants or antiplatelet agents, remains a critical consideration. This approach is essential for halting the progression of hematomas and enhancing the efficacy of surgical outcomes (27).
The primary limitation of the study is its small size, due to the rarity of the condition. This small sample size limits the strength of our statistical analysis and conclusions. Due to the limited number of participants, the impact of other potential risk factors may have been underrepresented. Additionally, being a retrospective study, it is subject to inherent biases and limitations in assigning specific treatments or rehabilitation programs. Another challenge is the variety of surgical procedures used, which affects our ability to establish standardized treatment guidelines. Also, the short follow-up period for some patients makes it difficult to assess the long-term outcomes of SSEH. Future research should involve larger, more robust studies that include both surgical and conservative treatments. This will help improve our understanding and refine treatment approaches.
Conclusion
In this retrospective study, we demonstrated that minor initial neurological deficits represent a better prognosis for SSEH patients based on one-year follow-up data. Importantly, prompt surgical evacuation of hematoma leads to a significant improvement in neurological functions among SSEH patients, even in cases characterized by complete or severe neurological deficits and extended duration of symptoms.
Acknowledgements
The Authors are grateful to the members and fellowship of Department of Neurological Surgery, Tri-Service General Hospital.
Footnotes
Authors’ Contributions
CJ Hsu: Data acquisition, data analysis and interpretation, drafting the article, statistical analysis, reviewing submitted version of manuscript; PZ Lin: Data acquisition, data analysis and interpretation; DT Ju: Critical revision, administrative/technical/material support; DY Hueng: Critical revision, administrative/technical/material support, study supervision; KY Tseng: Conceptualization and design, critical revision.
Conflicts of Interest
The Authors have no conflicts of interest regarding this study.
- Received May 1, 2024.
- Revision received May 28, 2024.
- Accepted June 13, 2024.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
