Extra-axial Hemorrhages

Epidemiology

The majority of spinal epidural hemorrhages are “spontaneous.” Post-traumatic SEH is reported in as few as 1.7% of all spine injuries and is less common than spontaneous SEH. However, recent reports suggest that traumatic SEH is more frequent than previously thought, especially in conjunction with spinal fractures or spinal soft tissue injuries.

Clinical Presentation

SEH is an emergent condition that can result in progressive spinal cord or cauda equina compression and potentially devastating neurologic impairment. There are many causes. Most of these lesions appear to be “spontaneous.” Nonspontaneous, secondary causes of SEH include trauma, coagulopathy, pregnancy, disc herniation, tumor, and arteriovenous fistula or malformation. Clinical signs and symptoms of SEH include back pain, extremity weakness, sensory loss, and bladder or bowel dysfunction. The location and the degree of spinal cord/nerve root compression determine the severity of neurologic symptoms. SEHs resulting in acute or severe neurologic compromise are typically treated by prompt surgical evacuation. If symptoms are not severe or progressive, conservative management may be employed. Close follow-up with frequent clinical examinations and repeat MRIs of the spine ensure that the hematoma is not increasing in size. Follow-up imaging may show reduced size or resolution of the hematoma within 2 weeks. Patients may recover completely without neurologic deficit after conservative management. Patients older than 65 years have an increased incidence of residual symptoms, but there is no correlation between initial size of hematoma and the presence or severity of residual symptoms.

Pathophysiology

SEHs are thought to result from venous hemorrhage. The epidural venous plexus gives rise to the basivertebral veins and communicates freely with both the intracranial venous sinuses and the azygous venous system. The epidural plexus is valveless, rendering it susceptible to rupture from sudden increase in intra-abdominal or intrathoracic pressure. The epidural venous plexus is most prominent in the thoracic spine. Perhaps for that reason spontaneous SEHs are more common in the thoracic and cervicothoracic regions than the lumbar spine. The ventral aspect of the dural sac is strongly adherent to the posterior longitudinal ligament. Therefore, spontaneous SEHs most commonly collect in the dorsal epidural space.

Imaging

CT

In the setting of trauma, CT is often performed before MRI as either the initial screening examination of the spinal column or for further evaluation of an abnormality detected on plain radiographs. On CT, SEHs typically appear as extradural soft tissue masses with attenuation similar to the intervertebral discs (60-70 HU). When the abnormality is ventral, it may be impossible to differentiate a hematoma from an extruded disc fragment ( Fig. 10-1 ). MRI is the most appropriate next step for further evaluation.

MRI

MRI is the primary diagnostic study for evaluating the soft tissues of the spine and spinal cord. It provides rapid, noninvasive evaluation of the entire spinal column and demonstrates any compression of the spinal cord or nerve roots that may necessitate emergent surgical decompression. The craniocaudal extent of an epidural hematoma is variable. Most SEHs extend over two to four vertebral segments. Some are limited to a single vertebral level, whereas others extend the length of the vertebral column. On axial sections, SEHs are most often biconvex with tapered cranial and caudal ends ( Fig. 10-2 ). Replacement of the normal epidural fat signal confirms that the hematoma lies in the epidural space ( Fig. 10-3 ).

The MRI appearance of SEH varies with the age of the lesion. In the hyperacute state (<6 hours), intracellular oxyhemoglobin predominates, so the signal is isointense on T1-weighted (T1W) images and mildly hyperintense on T2-weighted (T2W) images, relative to the spinal cord. From 6 to 72 hours, intracellular oxyhemoglobin within the hematoma deoxygenates to intracellular deoxyhemoglobin. This change causes the characteristic “acute” appearance: isointense to spinal cord on T1W images and markedly hypointense to spinal cord on T2W images. The isointensity on T1W images has been reported to last up to 5 days after the initial trauma. From 3 to 5 days onward, progressive formation of methemoglobin causes high signal intensity on T1W imaging relative to the spinal cord, with variable signal intensity on T2W imaging (depending on the proportions of intracellular and extracellular methemoglobin). In patients followed conservatively, increase in the signal intensity with time is virtually pathognomonic of an epidural hematoma on T1W imaging. Chronic blood products are typically low in signal intensity on both T1W and T2W imaging.