Epidural hematomas occur infrequently although they are not
rare. They can arise spontaneously or after trauma. They are notably
more common in patients with vascular anomalies and in those with coagulation
abnormalities1-6. Treatment usually
involves emergent operative decompression.
We report the case of an otherwise healthy thirty-five-year-old
man in whom a symptomatic cervical epidural hematoma developed after
minimal trauma. The hematoma resolved spontaneously, as documented
with magnetic resonance imaging. The patient remained asymptomatic
over a five-year duration of follow-up.
In April 1994, a thirty-five-year-old man fell while getting
out of bed to answer the telephone. He believed that the fall resulted
in very minor trauma, but he noticed progressive weakness and paresthesias
in the left arm and leg over the next one to two hours. His medical
history was notable for an episode of neck pain associated with
mild spondylosis at the fifth and sixth cervical levels three and
one-half years earlier.
The patient was seen in the emergency room of a local hospital
approximately four hours after the fall. The findings on radiographs
of the cervical spine were unremarkable. A cervical collar was applied,
and the patient was given 20 mg of methylprednisolone orally and
30 mg of Toradol (ketorolac tromethamine) intramuscularly for a
presumed disc problem. The patient reported that, by the time that
he first arrived in the emergency room, his strength had improved
markedly. The initial examination showed 5- strength (of a possible
5) in the left wrist extensors, normal sensation, and symmetrical,
normal reflexes.
Because of his history of upper and lower-extremity weakness,
the patient was transferred to Tampa General Hospital so that magnetic
resonance imaging of the cervical spine could be performed. He arrived
approximately nineteen hours after the onset of the initial symptoms.
Magnetic resonance imaging revealed a large left posterior epidural hematoma,
which was deforming the spinal cord; the hematoma extended from
the fourth to the seventh cervical segment (Figs. 1-AFigs. 1-A and 1-B1-B). By this time,
the weakness had decreased, except that he still had 5- weakness
in the left wrist extensors. Sensation was normal. The reflexes
remained normal and symmetrical; there were no pathological reflexes.
The results of coagulation studies, including prothrombin time,
partial thromboplastin time, and bleeding time, were normal.
Because of the improvement in the neurological findings, no operation
was performed in spite of the dramatic findings on magnetic resonance
imaging. The patient continued to wear the cervical collar and remained
hospitalized for observation. By the following morning (thirty hours
after the injury), muscle strength had returned to normal. The findings
on repeat magnetic resonance imaging at forty-eight hours were unchanged.
The patient was discharged from the hospital while wearing the collar.
Magnetic resonance imaging was performed again fifteen days after
the injury. This showed complete resolution of the hematoma, with
a normal cervical spinal canal and spinal cord (Figs. 2-AFigs. 2-A and 2-B2-B).
At the five-year follow-up, which consisted of a telephone interview,
the patient reported that he had remained asymptomatic and had returned
to normal activities with no restrictions. He was participating in
recreational sports and was working as a manager. There had been
no signs of weakness or any symptoms to suggest a recurrent epidural hematoma.
He described intermittent neck pain similar to that prior to the
hematoma.
A cervical epidural hematoma is believed to be a relatively rare
finding. It occurs more commonly with injuries associated with ankylosing
spondylitis7-9, medication-induced
coagulopathies1,2,7,10,11, hemophilia6, thrombocytopenia12,
postoperative bleeding13, and
Paget disease3. An extradural
hematoma is commonly seen on magnetic resonance imaging after a
cervical fracture or dislocation. In the absence of ankylosing spondylitis
or a coagulation abnormality, it usually is not symptomatic. An
epidural hematoma is considered to be spontaneous when it occurs
after minimal or no trauma.
Operative treatment to decompress the spinal canal is still recommended
for most patients who have a symptomatic cervical epidural hematoma4,14-21. If the patient shows clinical
signs of improvement, then close observation is appropriate. Other
isolated cases of spontaneous resolution of a documented symptomatic
cervical epidural hematoma have been reported by several authors5,8,9,22-28, in ten patients ranging
in age from thirteen to seventy-nine years. In some of the patients,
the neurological deficits developed immediately, and in one patient
the weakness and numbness became apparent two and one-half months
after the onset of neck pain8.
The time until resolution of the symptoms also varied, from hours
to weeks. Connolly et al. reported that their patient's medical
condition (a recent myocardial infarction and treatment with anticoagulants)
precluded acute operative intervention22.
In other patients, the delay in diagnosis allowed enough time for
neurological improvement. Of the previously reported cases, only
two involved a traumatic etiology. The patient reported on by Lefranc et
al. had two inconsequential fractures that were sustained in a motorcycle
accident24. The anterior epidural
hematoma was not noted on the initial diagnostic study but developed
over the next four hours. The hematoma did not deform the spinal
cord. In three patients, a bleeding diathesis (due to anticoagulants22, an arteriovenous malformation26, and hemophilia5)
was implicated as a cause of the hematoma. In six patients, there
was no trauma, coagulopathy, or other apparent predisposing condition.
An epidural hematoma can arise from either arterial or venous
sources. Since our patient and the other ten patients5,8,9,22-28 had resolution without
operative treatment, we can only assume that they all had a hematoma
of venous origin. In several patients, the neurological symptoms
resolved within a few hours after the onset. This finding indicates
that epidural hematomas may be much more common than has been previously
thought.
A patient who presents with neck pain and transient weakness
that began spontaneously or after minor trauma may have an epidural
hematoma. In our patient, magnetic resonance imaging documented complete
resolution of the epidural hematoma after two weeks, which is earlier
than might otherwise be expected. The rich vascularity of the cervical
spinal canal, the presumed venous origin of the hematoma, the motion
of the cervical spine (even in a collar), and the cerebrospinal
fluid pulsations probably contributed to the rapid resorption of
the hematoma.
On the basis of our experience with this patient and our review
of ten other cases of cervical epidural hematoma that resolved without
operative intervention, we believe that, if the neurological deficit
is not severe and if there is clinical improvement, nonoperative
treatment remains an option. Even if the neurological deficit abates,
the patient should be followed with magnetic resonance imaging studies to
confirm resolution of the epidural hematoma. Indeed, Morio et al.
reported a case of a chronic cervical hematoma that did not resorb
completely29.