17 Spinal Epidural Compression
Introduction
Nontraumatic spinal epidural compression can result from several different entities, but acute deterioration almost always occurs as a result of a few conditions, three of which are highlighted in this chapter: spontaneous epidural hematoma, spinal epidural abscess, and metastatic epidural spinal cord compression syndrome.
Incidence
Spontaneous Spinal Epidural Hematoma
Spinal epidural hematomas (SEHs) are a rare cause of spinal cord compression. However, they constitute the majority (up to 75%) of spinal hematomas. The peak incidence occurs in patients in their sixth decade of life, though a second peak is seen in adolescents between 15 and 20 years of age. A male predominance has frequently been documented.
Spinal Epidural Abscess
Spinal epidural abscesses (SEAs) are an infrequent cause of spinal cord compression, representing 0.2 to 2 per 10,000 hospital admissions. The majority of affected patients are between 30 and 60 years old, though they span a wide range from neonates to geriatric. A male predominance—approximately twice as common as in women—exists. Risk factors include diabetes mellitus, end-stage renal disease, HIV or other immune-compromised states, intravenous (IV) drug use, and alcoholism. Local factors additionally include spine surgery or trauma, and catheter placement into the vertebral canal.
While nearly one-third of affected patients died at the beginning of the twentieth century, the mortality is now less than half of that number given improvements in antibiotic therapy and surgical technique. Correspondingly, the percentage of patients with either complete recovery or only minor residual neurologic deficit has more than doubled. 1
Metastatic Epidural Spinal Cord Compression
In the United States, there are ~1.4 million new cases of cancer annually and every year over half a million cancer patients succumb to metastatic disease. The skeletal system serves as the third most common site of metastatic spread (after pulmonary and hepatic), and within the skeletal system the spinal column is most frequently affected.
Current estimates based on postmortem studies imply 30–90% of cancer patients (large variability depending on primary) will have metastatic spinal disease. A total of 5 to 10% of cancer patients have metastatic epidural spinal cord compression (MESCC), with this proportion increasing to 40% in those with other, nonspinal bony metastases. These numbers translate into <25,000 cases of symptomatic MESCC per year, an incidence that is rising as antineoplastic therapies evolve and life expectancies increase. MESCC is an epidural lesion causing true displacement of the spinal cord from its normal position in the spinal canal.
Etiologies
Spinal Epidural Hematoma
Spontaneous SEH can be divided into traumatic and nontraumatic. Causes of traumatic SEH include lumbar puncture or epidural anesthesia, fracture, spinal surgery, physical exertion, birth trauma, and chiropractic manipulation. Causes of spontaneous SEH include hemorrhage from an arteriovenous malformation (AVM), hemangioma, or tumor. In up to 30% of cases, no etiology is discerned. 2 Following these idiopathic cases, anticoagulant therapy and vascular malformations are most often implicated. Anticoagulation or any bleeding diathesis is a risk factor for SEH. 3
Spinal Epidural Abscess
Infection can spread hematogenously or contiguously. Any distant site of infection can spread hematogenously; however, skin and soft tissue infections represent the most common sources. SEAs arising in this fashion generally develop in the posterior epidural space. SEAs that spread by direct extension predominantly originate from a vertebral body focus, or less commonly from adjacent soft tissue. This vector of spread usually involves the anterior aspect of the spinal canal.
Inoculation can also occur iatrogenically. In a large meta-analysis of over 900 cases, epidural anesthesia or analgesia were associated with a 6% rate of infection, and invasive procedures, either spinal or extra-spinal, with 14–22%. 4 Usually a severe pyogenic infection with Staphylococcus aureus is the most common causative agent. Streptococcus species and coagulase-negative Staphylococcus follow in frequency. Gram-negative rods such as Pseudomonas and Escherichia coli, account for a small fraction, being more prevalent with IV drug use. Finally, Mycobacterium tuberculosis, fungal species, and parasitic organisms are rare except for immune-compromised states.
MESCC
Metastatic disease spreads to epidural space in two ways: (1) directly into the spinal canal through intervertebral foramen from a paravertebral mass (15% of metastatic cord compression); and (2) the remaining 85% from hematogenous spread (historically thought via Batson’s plexus, now believed to be more likely arterial) to the vertebral body, from where the lesion grows posteriorly into the epidural space. These metastatic lesions can cause bone erosion, pathologic fractures, and extrusion of bony fragments into canal, which can all further compound canal narrowing or cord compression.
Pathophysiology
Spinal Epidural Hematoma
Bleeding is generally the result of tearing of epidural veins, although tearing of epidural arteries or hemorrhage from a malformation is also possible. Even in circumstances involving anticoagulant therapy, other factors are posited to contribute, including increased pressure in the interior vertebral venous plexus and foci of vascular “decreased resistance.”
Spinal Epidural Abscess
As with any form of compression, vascular compromise with consequent hypoxia has been one favored pathogenetic explanation. However, in animal models of S. aureus epidural abscesses, even when SEAs caused para- or quadriplegia, no compression of spinal arteries was noted, 5 thus supporting a paramount role for direct mechanical compression.
MESCC
Hypothesized mechanisms by which damage occurs include (1) direct compression that leads to demyelination and axonal damage; (2) vascular compromise, where occlusion of venous plexus promotes breakdown of cord–blood barrier and thus vasogenic edema; and (3) terminal arterial occlusion with ischemia/infarction may follow leading to irreversible damage. Certain authors have hypothesized that in patients rapidly deteriorating arterial infarction may underlie decline whereas venous congestion may initially be more relevant in patients with slow decline. 6 This disparity may explain the worse outcome associated with a more rapid evolution of motor weakness. 7
Presentation
Spinal Epidural Hematoma
SEH is usually acute and progressive, leading to permanent neurologic deficit if not managed immediately. Symptoms consistently begin with severe back pain in the location of the hemorrhage, with or without a radicular component. The common segmental levels involved vary by age; in the patients of the 46- to 75-year-old year age group, the lower thoracic and lumbar regions are most common, with a smaller frequency maximum in the cervical levels. 8 A pain-free interval may occur, but then is generally followed by progression of neurologic deficit over hours to days toward flaccid paresis or plegia.
Spinal Epidural Abscess
Seventy-one percent of patients present with back pain as the initial symptom; 66% have fevers. This proceeds to radicular irritation, with subsequent neurologic deficits, including muscle weakness, sensory disturbances, and sphincter incontinence. Progression to frank paralysis occurred only in one-third of patients. 9
MESCC
Pain (83–95%) is a common presentation. Local pain is thought to be related to periosteal stretching or local neoplastic inflammatory process. This pain responds well to steroids and is worse with recumbency. Mechanical pain is pain that is exacerbated by movement/activity and is often caused by pathologic fracture or vertebral body collapse, and indicative of spinal instability. This pain is recalcitrant to steroids/narcotics; radicular pain is that which involves nerve root compression and usually conforms to a dermatomal distribution.
Motor dysfunction is present in 60–85% of patients and is characterized by weakness and long tract signs. There are correlations between neurologic status at time of diagnosis (particularly with respect to motor function) and prognosis from MESCC. Sensory loss is in close proximity to motor findings and autonomic/sphincter dysfunction is a later finding, with bladder dysfunction being the most common. Though the rate varies, patients with these deficits inevitably progress to paralysis without intervention.
Indications
Spinal Epidural Hematoma
Most SEHs are located dorsal to the spinal cord, with a large meta-analysis quoting < 75% in this sagittal location. 8 Emergent or urgent decompression within hours is associated with better outcomes. In the same meta-analysis, for patients who received treatment within 12 hours of onset of symptoms, 66% recovered completely, 13% recovered with mild residual neurologic deficit, and 13% continued to have severe impairment or show no improvement. In contrast, for patients whose treatment was initiated 13–24 hours after symptom onset, 64% had severe deficits or no improvement, versus 36% with substantial recovery. Therefore, the treatment of choice is immediate decompression in those patients that can tolerate surgery. Asymptomatic patients without neurologic deficit can be considered for observation, especially in children and teenagers in which a laminectomy may destabilize the posterior column.
Spinal Epidural Abscess
The first operative intervention—a laminectomy—for SEA was performed in 1892; after increasing reports of successes, surgery became the mainstay of treatment by the 1930s. An early series 10 noted that SEA patients without paralysis or whose paralysis had developed less than 36 hours before the operation had better postoperative outcomes with respect to survival and function. In contrast, in patients whose paralysis developed more than 48 hours before surgery, none recovered neurologic function; all mortalities in the series were reported in this latter group. This correlation of outcome with time to intervention has been repeatedly confirmed. 11 , 12 Conservative treatment is rarely indicated: either for those who cannot tolerate surgery, or who have large abscesses extending a considerable length of the spinal cord.
MESCC
Consensus and expert opinions regarding indications for surgery largely derive from studies investigating the prognostic value of surgical intervention given various patient group attributes. The evidence dictating the appropriate approach to tumor decompression has evolved significantly over the past 50 years. Early treatment underscored indirect decompression of the epidural space via straight laminectomy, followed by radiation therapy (RT). 13 , 14 However, later studies 15 , 16 demonstrated no advantage for laminectomy, rendering radiation alone the preferred therapeutic strategy for a period of years. More recent studies with modern anesthetic and imaging techniques have led to a resurgence of surgical decompression as part of the treatment strategy. 6 , 17 A large randomized control trial 6 assessed decompressive resection in conjunction with RT versus RT alone. Criteria for study inclusion required MESCC restricted to a single area; acceptable surgical candidates with life expectancy > 3 months; one neurologic symptom (including pain); not totally paraplegic for > 48 hours. Radiosensitive tumors and sole root compression or cauda equina syndromes were excluded; 84% of the surgery group versus 52% of the RT group were able to walk after treatment, 62% versus 19% regained ambulation whence lost, and 94% versus 74% remained ambulatory. Additionally, the study revealed significant differences between treatment groups with respect to maintenance of continence; muscle strength; functional ability; and increased survival (126 versus 100 days), with ambulation and continence persisting for the lifetime of the surgery group.
Spinal instability can independently contribute to symptoms, by directly causing mechanical injury to the spinal cord. As RT is unlikely to ameliorate spinal instability, surgery may be more appropriate in these circumstances. An analysis focusing on forms of compression for patients who were, at the onset, either independently ambulatory, assisted ambulatory, paraparetic, and paraplegic: without bony compression, post -RT ambulation rates were 100%, 94%, 60%, 11%, respectively. These rates dropped to 92%, 65%, 43%, and 14%, respectively, when all patients (with bony and nonbony compression) were considered. 18
A comprehensive literature review 19 suggested that with RT alone, 36% subjects improved while 17% worsened; with decompressive laminectomy ± RT, 42% improved while 13% worsened; with posterior decompression with stabilization, 64% improved; and finally with an anterior approach, 75% improved with <10% mortality.
Prevailing conviction holds that if compression is of short duration, neurologic deficits may be reversible, as re-myelination and recovery of function are possible. However, with prolonged compression, secondary vascular injury with infarction of the spinal cord may occur with irreversible consequence.
Based on these and similar studies, generally accepted indications for surgery include: the need for tissue for diagnosis; spinal instability; cord compression with dysfunction from bone or tumor not radiosensitive; and deterioration or recurrence during/despite RT. Surgical decompression to prevent irreversible damage should be immediate. Conversely, RT is a reasonable alternative for patients with radiosensitive tumors, stable neurologic status, no spinal instability, no significant bony compromise of canal, or life expectancy less than 3 months.
The location of the origin of the tumor (isolated epidural disease versus arising from osseous lesion with extension) as well as considerations of spinal stability should dictate choice of operative procedure. A thorough description of all surgical approaches is beyond the scope of this chapter. However, a simple laminectomy should be reserved for dorsally located disease, and a posterolateral or ventral approach should be utilized whenever ventral disease is present, as tumors may continue to grow or swell and thus without a direct removal of the offending pathology, an indirect decompression will result in further deformation of the spinal cord. At the spinal cord level (occiput to bottom of conus medullaris), the cord should never be retracted to gain access to ventral tumor; the approach should be selected that obtains the most advantageous angle to access the tumor instead.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
