| |
Clinical Syndrome
Phantom limb pain is the complaint by a patient of a
painful sensation in an extremity that has been amputated. Phantom
limb pain does not include the pain seen after spinal cord injury,
nor does it include the pain seen after injury to peripheral nerves,
lumbosacral or brachial plexus injury, or dorsal root avulsion when
the body part has not been lost. All these conditions have
similarities to phantom limb pain, on the basis of both clinical and
physiologic evidence, but the differences are significant, and they
should be analyzed and managed independently. Stump pains also must
be differentiated from phantom limb pain; although they may be
related, their management is different.
Virtually all amputees have phantom sensations; a much smaller
percentage have phantom pain. Although phantom sensations and pain
have been described after loss of body parts other than an extremity
(breast, nose, bladder, etc.). It is probable that the neural
substrates for all painful phantoms are similar and management
strategies and results will not differ significantly, but there are
not enough clinical reports about other types of phantom pain to
permit meaningful statements.
Incidence
The incidence of phantom limb pain has been reported to range from
less than 1 to 50 percent in those patients who have lost a limb or
a major portion thereof. Part of this variance can be explained by
sampling techniques and part by variation in patient labeling of
abnormal sensations. Studies based on patients referred to a
treatment center or a particular physician are obviously biased by
the inability to describe the population at risk. Surveys based on a
complete group of amputees at risk suggest that phantom pain
significant enough to cause the patient to seek medical care for the
pain occurs in approximately 5 to 10 percent of the amputee
population. Phantom limb pain appears to be more common in patients
who lose their limb at an older age and in patients with affective
or environmental factors which are likely to lead to chronic pain
behavior independent of the type of injury. The etiology, severity,
and location of extremity pain prior to amputation do not appear to
influence the location or severity of phantom limb pain.
Neurophysiological Basis
The cause of phantom limb pain remains unknown; a variety
of theories have been elaborated, but all lack verification. Any
explanation of phantom limb pain must consider the relevant clinical
phenomena. Phantom limb pain may be present immediately after injury or
may develop weeks, months, or even years later. The phantom is usually
described as foreshortened and in a distorted anatomic position. The
pain may vary from continuous cramping, aching, and burning to electric
shock-like and intermittent, even in the same patient; or it may be
unchanging. Stress, anxiety, fear, or fatigue will usually increase the
patient's complaints. Careful neurological assessment usually reveals
that stimulation of the skin in normal areas remote from the lost part
will often trigger phantom pain, as will percussion of stump neuromata.
A painful phantom can exist with or without stump pain. A final problem
is why only a portion of amputees develop phantom limb pain, even though
almost all develop phantom limb sensation at some time after losing a
limb.
Amputation of a body part clearly results in
deafferentation of dorsal horn neurons and more rostral structures as
well. Central reorganization of synaptic inputs has been shown to occur.
Some neurons may remain deafferentated and become spontaneously
hyperactive and supersensitive to transmitter substances. The behavior
of wide-dynamic range neurons is known to change after peripheral nerve
injury. Stump neuromata are electrically active
and may generate impulses that reach the spinal cord and
cause the patient to describe sensations in the absent part. Altered
sensory input may lead to errors in information processing in higher
centers; phantom phenomena may be a cognitive error. Finally, a
psychiatric theory holds that phantom limb pain is an ego phenomenon;
there is no physiologic substrate but only an affective basis.
There can be no question that environmental and affective
factors play a role in the patient's pain behavior and influence
disability, drug utilization, and demand for health care. These factors
certainly influence the patient's symptom manifestation, but there is
not a shred of evidence that they suffice to explain why phantom limb
pain occurs.
Phantom limb sensations and, rarely, pain have been
described during regional or spinal anesthesia. They always disappear
when the patient recovers from the block. As axonal silence, but not
anatomic disruption, characterizes nerve blocks, this suggests that loss
of afferent information has a central effect which can by itself
generate phantom limb phenomena, including pain. Plasticity, both
through the utilization of anatomically present but functionally silent
synapses and through the growth of new connections after loss of
afferentation, is the probable substrate of phantom limb pain. Neurons
which have suffered major loss of input become spontaneously
hyperactive; their output may be interpreted as pain by higher centers.
The growth of new synaptic connections results in unusual sensory fields
in the deafferented neurons. Downstream modulation can influence the
firing of these neurons; hence the role of stress and relaxation in
augmenting or diminishing phantom pains in some patients. As
hyperactivity, supersensitivity, and new synaptic efficacy involve a
wide array of sensory neurons, it should not be surprising that measures
to alter the pain of acute tissue damage do not succeed in controlling
phantom limb pain. The neural circuits involved in these two types of
pain appear to be dissimilar based on the response to ablative surgical
procedures and systemic opioids.
Natural History
Phantom limb pain may begin at the
time of injury or may start after an interval of weeks, months, or
even, rarely, years. Unlike phantom limb sensations, which are
uniformly present immediately after loss of a limb and then
frequently disappear over weeks or months, once phantom limb pain is
present it is likely to persist for a prolonged period. The
patient's complaints and utilization of health care frequently
diminish over time. but this is often due to environmental and
affective factors rather than spontaneous decrease in the phantom
limb pain itself. In a few patients. however. phantom limb pain is a
transient phenomenon. for reasons which are not understood. Phantom
limb pain is usually a long-term management problem which persists
in spite of attempts at medical management.
Therapeutic Choices and Long-Term
Results
The choice of therapy has been based on the
physician's preferences and resources. not the patients unique
characteristics. How to select a therapy for a particular patient
remains a mystery.
The first step is to collect all the medical,
psychological and environmental data available, Multidisciplinary
pain clinics can often accomplish this goal more effectively than an
individual physician or serial referrals to experts who may try to
synthesize a meaningful comprehensive management plan. The second
step is to address the psychological and environmental factors
which, although not the cause of phantom limb pain, certainly can
play a major role in the genesis of pain behavior. Strenuous efforts
should be made to decrease and abolish the patient's use of
narcotic. sedative-hypnotic and benzodiazepine drugs, which
usually add to the patient's depression and fail to relieve this
type of pain. Getting the patient back into normal daily activities
by any means available is usually helpful. Although there is little
hard evidence to support their efficacy. anticonvulsant medications
have been widely used and they are reasonably safe. They should
however, be discontinued if not clearly helpful. Antidepressants may
be helpful in a nonspecific way: some have claimed success with the
combination of a tricyclic antidepressant and a phenothiazine.
Regional and local nerve blocks offer short-term relief, although
occasional reports have described long-term benefits. Neurosurgical
therapies have included ablative procedures from the stump neuromata
to the forebrain and stimulation of peripheral nerves, spinal cord
and brain. Some have been clearly shown to be rarely beneficial.
Peripheral neurectomy. dorsal rhizotomy and sympathectomy rarely
produce satisfactory results for more than a few weeks. Although
cordotomy has often been performed, most patients show a recurrence
of their pain within a year or two. Recent information about the
presence of afferent fibers in the ventral roots has led to the
modification of dorsal rhizotomy by the addition of ganglionectomy.
Whether this will improve on the results of dorsal rhizotomy alone
is not known at the present time. Thalamotomy rarely produces more
than a year of pain relief. Although initial reports were glowing,
excision of the parietal sensory cortex has also not produced
long-term beneficial results in most patients. Reports on the
efficacy of dorsal root entry zone lesions that destroy substantia
gelatinosa and dorsal horn neurons have suggested a 50 percent
long-term success rate: this operation has not yet stood the test of
time.
Stimulation of peripheral nerves, spinal cord and brain all
have their proponents: data on long-term efficacy are limited and
not encouraging. The target area for brain stimulation should
probably be the lateral thalamus-internal capsule and not the
periaqueductal gray region.
As all forms of surgical therapy
carry some risk to the patient. They should not be offered to the
patient until other therapeutic modalities have been tried. When
surgery is elected. Results seem to be superior if all the
psychological and behavioral issues are addressed in the same time
frame as part of a comprehensive management process. Until the
genesis of phantom limb pain is understood, it is unlikely that
rational highly successful therapies will be available.
| |
