Closed reduction of the cervical spine is a commonly performed method of treatment for acute subluxations or dislocations. Although the recommendation has been debated by many authors3,10, a magnetic resonance image or a myelogram of the cervical spine has been advised for evaluation of lesions occupying the canal before closed reduction in patients who are neurologically intact or who have an incomplete injury of the spinal cord. A sudden or gradual worsening of the neurological status of a patient during reduction should alert the physician to the presence of high-grade compression of the spinal cord.
Causes of neurological compression include a spinal epidural hematoma, a herniated intervertebral disc, direct compression of the cord due to osseous fracture fragments, edema of the cord, hemorrhage of the cord, and subdural or subarachnoid hemorrhage13.
The prevalence of spinal epidural hematoma associated with fracture of the cervical spine was 4 per cent (two of fifty-two) in the study reported by Wortis and Sharp and 1.7 per cent (six of 357) in that reported by Foo and Rossier. Traumatic causes of spinal epidural hematoma include vertebral fracture-dislocation, birth trauma, lumbar puncture, postoperative bleeding, and missile injury1,4,15-17. Spinal epidural hematoma due to trauma has been reported relatively infrequently in the orthopaedic and neurosurgical literature, and it usually results in progressive neurological deficits that necessitate operative decompression8.
We report the case of a patient who had bilateral subluxation of the facet of the fifth cervical vertebra on that of the sixth cervical vertebra. The patient was initially neurologically intact and had a gentle closed reduction with slow extension of the cervical spine. Subsequent acute neurological deterioration indicated the need for urgent acquisition of a magnetic resonance image of the cervical spine, the findings of which were consistent with a cervical epidural hematoma.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Department of Orthopaedic Surgery, Thomas Jefferson University Hospital, 850 Walnut Street, Walnut Towers, Philadelphia, Pennsylvania 19107.
An eighty-five-year-old woman was in a motor-vehicle accident while wearing a seat belt. At the time of the injury, there was no loss of consciousness. She reported severe neck pain with transient numbness and paresthesias in both upper extremities. She subsequently was transferred to our facility for evaluation.
Physical examination revealed tenderness to palpation in the posterior aspect of the caudad region of the neck. Neurologically, there was no evidence of motor weakness or sensory changes. The history of numbness and paresthesias, which had resolved by the time that she was seen at our institution, was suggestive of neurological compression.
An initial plain lateral roentgenogram demonstrated a flexion-distraction injury at the fifth and sixth cervical levels, consistent with bilateral facet subluxation (Fig. 1). Closed reduction with gentle physical manipulation of the cervical spine into extension was successful in reducing the subluxation. Within three to five minutes after completion of the reduction maneuver, progressive weakness developed in both the upper and the lower extremities, resulting in complete motor quadriplegia. A second plain lateral roentgenogram was made immediately, and it revealed adequate alignment of the cervical spine without evidence of excessive cervical extension.
In response to the acute neurological deterioration, the mechanism of reduction (extension) was reversed. The neck was subsequently manipulated into a slightly flexed position, with rapid improvement (within five minutes) to the baseline neurological status. The cervical spine was then stabilized in a halo device, and a magnetic resonance image was made.
The magnetic resonance image of the cervical spine revealed an epidural hematoma posteriorly, at the fifth and sixth cervical levels, which extended posteriorly into the substance of the ligamentum flavum (Fig. 2).
The patient was taken immediately to the operating room, where posterior cervical decompressive laminectomy of the fifth and sixth cervical vertebrae was performed with evacuation of the epidural hematoma. A concomitant posterior arthrodesis of the fourth, fifth, and sixth cervical vertebrae was performed with use of Axis plates and screws (Sofamor-Danek, Memphis, Tennessee) and autologous bone graft from the iliac crest (Fig. 3). Postoperatively, normal motor function of the upper and lower extremities was maintained, and there was no recurrence of numbness and paresthesias in the upper extremities.
Roentgenograms made at six months revealed a solid fusion. There had been no change in the neurological status. Clinically, the patient was doing well at that time.
The primary goal of treatment of a traumatic subluxation or dislocation of facets of the cervical spine is to achieve reduction and stability while preserving neurological function. The literature is filled with a myriad of studies concerning the treatment of a traumatically dislocated spine. Tator and Fehlings suggested that reduction may facilitate neurological recovery by decompression of the blood supply to the spinal cord, thereby minimizing ischemic damage. It is still debated whether dislocations should be reduced at the time of presentation, thereby hastening neurological recovery3, or whether a magnetic resonance image should be made before an attempted reduction5. Moreover, the mechanism of reduction, consisting of manipulation of the cervical spine, closed reduction with skeletal traction over a short or long time-period, or operative reduction, has been questioned by several authors2,3,5.
After careful scrutiny of the roentgenograms, it was determined that our patient had bilateral facet subluxation. The injury was isolated to a single level, less than three millimeters of anterolisthesis was observed, the inferior facets of the fifth cervical vertebra were posterior to the superior facets of the sixth cervical vertebra, and no rotational malalignment of the vertebral bodies or the spinous processes had occurred. Thus, in our alert, cooperative patient, because the paresthesias had resolved at the time of presentation, we thought that only gentle extension of the cervical spine, with continuous neurological evaluation, would be necessary to improve cervical alignment with minimum risk neurologically. Therefore, given the risk of dislocation with transport to a magnetic resonance imaging scanner, we decided to attempt immediate reduction and stabilization in the emergency room.
Loss of neurological function after manipulation of the cervical spine should raise the suspicion of compression of the spinal cord from a lesion occupying the canal, such as a herniated disc, buckling of the ligamentum flavum, an epidural hematoma, or bone fragments due to malalignment. Immediate measures must be taken to determine the etiology of the compression; these include the making of a plain lateral roentgenogram followed by magnetic resonance imaging or myelography. The method of treatment is determined according to the etiology of the compression. In our patient, gentle flexion of the cervical spine, a reversal of the mechanism of reduction, allowed immediate return of motor function. A magnetic resonance image made after this maneuver revealed an extensive posterior epidural hematoma.
Post-traumatic spinal epidural hematoma is uncommon in adults but not rare in neonates1. In neonates who have had birth trauma, epidural hematoma is a major factor in spinal cord injury17. In children, a traumatic spinal epidural hematoma may occur without fracture because of the greater elasticity of the spinal column in this age-group. In elderly patients, a post-traumatic spinal epidural hematoma often occurs in the presence of certain inflammatory or metabolic spinal diseases, such as ankylosing spondylitis, rheumatoid arthritis, or diffuse idiopathic skeletal hyperostosis (Forestier disease)8.
According to Foo and Rossier, post-traumatic spinal epidural hematoma associated with fracture is most common at the cervical level. The bleeding usually arises from disrupted epidural veins, but in patients who have rheumatoid arthritis it frequently originates from oozing of exposed cancellous bone of the fractured vertebrae7. A symptomatic epidural hematoma of the cervical or thoracic spine is considered an operative emergency with the potential for progressive neurological deterioration. However, it must be recognized that neurological deficits may develop as long as one week after the initial injury8.
Thus, we believe that intensive operative decompression in patients who have a traumatic spinal epidural hematoma and neurological compromise associated with subluxation or dislocation is essential to prevent additional neurological deterioration or to improve motor function. However, there have been sporadic case reports of traumatic epidural hematomas of the cervical spine that have resolved spontaneously without deleterious neurological sequelae9,13.
Varying rates of success have been reported after manipulation of the cervical spine. Burke and Berryman reported success in thirty-seven (90 per cent) of forty-one patients who had a flexion-rotation dislocation of the cervical spine; Kleyn, in eighty-two (81 per cent) of 101 patients who had a traumatic dislocation of the cervical spine; and Osti et al., in 151 (90 per cent) of 167 traumatically dislocated cervical spines.
Few authors have described the morbidity associated with closed reduction of the cervical spine. Mahale et al. reported on one patient who had complete paralysis and three patients who had complete motor paralysis but some sensory sparing after manipulation for reduction of a traumatically dislocated cervical spine. The etiology of the neurological compromise in these patients was not discussed. Lee et al. reported success in sixty-six (73 per cent) of ninety-one patients who had had manipulation under anesthesia and improvement of the neurological status in twenty-nine (32 per cent).
Neurological deterioration is the most disastrous complication of a manipulative reduction of a traumatically injured cervical spine. The etiology may include extrusion of a disc, partial or complete transection of the spinal cord, and epidural hematoma. Other complications include vertebral fracture, ligamentous damage, and additional dislocation. Mahale et al. reported on a unilateral facet dislocation that became bilateral after manipulation. Burke and Berryman reported a 5 per cent rate of mortality (two of forty-one patients) after manipulation under anesthesia for the treatment of a flexion-rotation dislocation of the cervical spine. It should be noted that both of these reports described manipulation under anesthesia, during which a continuous neurological assessment could not be performed. Uncontrolled manipulative procedures without close monitoring of neurological function are thought to be the major cause of secondary neurological compromise6. Our patient was awake, alert, cooperative, and neurologically intact at the time of presentation. A plain roentgenogram revealed no evidence of fracture or bone fragments in the spinal canal. Because close evaluation of the neurological status was possible, the decision was made to gently manipulate the cervical spine.
We present this case to remind orthopaedic surgeons of the potential for disastrous complications after a manipulative reduction of the cervical spine. There are few guidelines and much controversy concerning the mechanism for reduction of a subluxated or dislocated cervical spine. If the decision has been made to manipulate the cervical spine, the procedure should be performed gently in an awake, alert, cooperative patient who is constantly being evaluated neurologically. The maneuver should be performed by an experienced spinal surgeon who has a heightened awareness of the potential complications.