Athirty-two-year-old man sustained a crush injury of the left forearm and
hand and the medial four digits of the right hand in a motor-vehicle accident.
A guillotine-type below-the-elbow amputation was performed on the left side
and was followed by secondary closure at a later date. The patient had no
stump-related problems and received adequate rehabilitation. He then returned
to his job as a lottery-ticket salesman. Two and one-half years later,
phantom-limb sensations developed in the left upper limb. This uncomfortable
sensation progressed to pain within a few days. The pain was distributed along
the lateral border of the forearm up to the tip of the phantom thumb. The
patient returned to the hospital after failure of symptomatic therapy
prescribed by his general practitioner.
The physical examination revealed that the patient was in good general
condition. Examination of the stump revealed no evidence of infection, undue
osseous prominences, scar tenderness, or neuroma formation. The range of
motion of the elbow and shoulder was normal. Examination of the cervical spine
revealed slight tenderness overlying the posterior aspects of the C5 and C6
vertebrae, without any notable paraspinal muscle spasm. The range of movement
of the cervical spine was normal in all directions, and the patient reported
slight pain at the posterior aspect of the neck on terminal extension. There
was no objective evidence of a motor or sensory deficit. The general
examination revealed enlarged, matted cervical lymph nodes on the left
side.
Hematological investigations showed a moderate lymphocytosis and an
elevated erythrocyte sedimentation rate of 48 mm/hr (normal range, 0 to 30
mm/hr). The Mantoux test was positive, with induration extending up to 16 mm.
The results of the sputum examination were negative for the presence of
acid-fast bacilli. A radiograph of the chest showed that the lungs were clear.
Radiographs of the cervical spine showed loss of the disc space between the C5
and C6 vertebrae. There was mild retrospondylolisthesis of the C5 vertebra in
relation to the C6 vertebra (Fig.
1). Magnetic resonance imaging revealed an epidural abscess in the
intervertebral foramen between C6 and C7
(Fig. 2) but no notable
compression of the spinal cord. A 24-gauge hypodermic needle was used to
perform a fine-needle aspiration of fluid from the matted cervical lymph
nodes, and the diagnosis of tuberculosis was confirmed on cytologic
examination of the aspirate.
Because the patient had only slight local symptoms and because no major
cord compression was evident, we treated the patient nonoperatively with
antituberculous therapy, which included four drugs (isoniazid, rifampicin,
pyrazinamide, and ethambutol) for two months and two drugs (isoniazid and
rifampicin) for another four months. The phantom pain decreased within three
weeks after initiation of the antituberculous drugs, and the patient reported
good relief of symptoms by the end of six weeks. He continued to have
intermittent phantom sensations for about three months. At the one-year
follow-up evaluation, he was completely free of all symptoms, including
phantom-limb sensations.
The management of phantom-limb pain depends on whether the problem is
primary or secondary. Primary phantom-limb pain is often refractory to
treatment. Medications such as antidepressants, transcutaneous electrical
nerve
stimulation15,
thermocoagulation of the dorsal-root entry
zone16, or epidural
spinal cord
stimulation17 may
be used, but no one therapy is uniformly
successful1. Unlike
primary phantom-limb pain, secondary phantom-limb pain can be treated more
successfully by the appropriate management of the inciting abnormality. A
herniated intervertebral disc has been reported to trigger phantom-limb
pain4. Phantom pain
secondary to spinal cord injury and spinal anesthesia has also been
reported5-8.
Phantom pain has also developed in patients who had lesions of the
contralateral cerebral hemisphere and sensory
loss9. Angina
referred to a phantom limb has also been
described13,14.
In addition, diabetic neuropathy has been reported as a cause for this
entity12. Chang et
al. reported the worsening of phantom-limb pain as a result of neoplasia
affecting the L4
vertebra10. The
onset of phantom-limb pain years after resection of a limb also has been shown
to be one of the first symptoms of recurrent disease in the
pelvis11.
To our knowledge, this is the first report of a patient with phantom-limb
problems caused by tuberculosis of the cervical spine and an epidural abscess.
Protracted presentations of spinal tuberculosis are known to occur in patients
who have good immunity. This would explain the fact that the presenting
symptom in our patient was phantom-limb pain rather than any notable local
symptoms. Because of the lack of severe local symptoms and the absence of
spinal cord compression, there was no absolute indication for surgery. Hence,
we adopted a nonoperative approach, which was successful. Because of the
direct association between the cervical tuberculosis and the onset of
phantom-limb problems, treatment of the tuberculosis was successful in
addressing the secondary phantom-limb problems as well.
In conclusion, the development of progressive phantom-limb symptoms in an
amputation stump after an asymptomatic period may indicate an abnormality
proximal to the limb. Appropriate management of the underlying abnormality may
be a successful way of treating the secondary phantom-limb symptoms as
well.