Phantom Limb Pain: A Literature Review of Theories and Treatments

Phantom Limb Pain: A Literature Review of Theories and Treatments
Submitted by: Tammie Liszka McElhinny
Required by NCOPE
Residency Program in Prosthetics

Abstract

Phantom limb pain is reported in as many as 80% of amputees, adversely disrupting the progression of rehabilitation and ultimately activities of daily living. Since phantom limb pain is not caused by an external stimulus to the residual limb, several theories exist in an attempt to explain its etiology. Various treatments including physical, pharmacological, psychotherapeutic and surgical techniques are utilized in an effort to provide symptomatic relief. Success of these treatments differs between amputees with only a small percentage of patients reporting lasting relief from the pain. Both theories and treatments to phantom limb pain will be presented in this literature review.

Introduction

Phantom limb pain is a common resulting condition to amputation effecting 60% to 80% of all amputees (Sherman et al. 1984). This phenomenon is defined as pain in the amputated portion of the extremity, first described by Andre Pare in the 16^th century and named by Weir Mitchell in 1871. The physiological mechanism causing phantom pain is unclear and somewhat controversial. Several theories have been presented with both psychological and physiological implications. Due to the elusiveness of this causal mechanism, medical teams have implemented approximately 50 different types of treatments in order to provide relief to the pain. The treatments include physical, pharmacological, psychotherapeutic and surgical techniques. Unfortunately, these treatments have yielded limited success rates let alone long-term relief. This paper presents current theories to the causal factor of phantom pain along with common treatments implemented for pain management.

Phantom Pain

Amputees often describe phantom pain as a continuous or episodic, burning or cramping sensation experienced distal to proximal in the phantom limb. Jensen et al. (1983) found that proximal pain was vague and disappeared quickly unlike phantom phenomena localized distally, which evoked vivid and longer lasting pain. Observations in the amputee population support their findings. Upper extremity amputees complain of pain in missing fingers, to the hand and then wrist whereas lower extremity amputees complain of pain in their toes, then ankle and calf. A possible reason for this is that these regions have greater cortical representation and field a larger amount of sensory information from afferent peripheral neurons.

The intensity of phantom pain varies from severe in only 0.5%-5% of all amputees to 51 % (Sherman, 1980). Severe pain may be debilitating, hindering the daily activities of the amputee and forcing lifestyle changes. Hill (1999) reported a study of 2694 amputees where 27% of the sample population had severe pain 15 hours each day and an additional 21 % had pain over periods of 10-14 hours.

Phantom limb pain generally begins within one week after amputation and seems to diminish over time after surgery. Jensen et al. (1983) reported a decline in pain from 72% 8 days post amputation to 67% after a 6-month follow-up. However, some amputees have reported a delayed onset of several months and others have dealt with phantom pain for 30 years or more (Davis, 1993, Montoya et al. 1997). Once phantom pain persists longer than six months, treatment becomes increasingly difficult (Patterson 1988). These amputees risk having phantom phenomena throughout their lifetimes.

Melzack proposes three major characteristics to phantom limb pain (Bowker, 1992). First, the pain endures after surgical healing is complete. Second, trigger points in healthy tissue areas or certain actions such as mictrition, yawning or sneezing may stimulate the pain. Stress, fear, fatigue or insomnia may also influence phantom limb pain. Finally, patients who had pain in the limb prior to surgery are more likely to experience a phantom phenomenon. Jensen et al. (1983) studied 58 non-traumatic amputees at three specific intervals: prior to amputation, shortly after surgery and at a 6-month follow-up. They reported that phantom pain was significantly more frequent in patients with pain in the preamputated limb than in patients without preoperative pain.

The etiology of phantom phenomena is unclear and controversial. The formation of post-surgical neuromas within the residual limb is a specific stump pathology that seems to correlate with phantom pain. However, phantom limbs have not been associated with the reason for amputation or the location of the amputation. An amputee's age, sex, marital or socioeconomic status does not predict phantom phenomena. Patterson (1988) reports the incidence of phantom pain is similar in both military and civilian populations suggesting that persons with traumatic amputations do not suffer from phantom limbs more often than those with a disease necessitated amputation. Interestingly, persons born with a congenital anomaly have even experienced phantom pain (Hill, 1999). It is often the misconception that because there is no damage to existing nerves, congenital deficiencies do not precipitate phantom phenomena.

Theories

Interaction between the peripheral and central nervous systems is implicated in current theories responsible for phantom pain. This interaction is assumed in the observation of complex perceptual, emotional and cognitive components to phantom phenomena. The peripheral component implies that surgical techniques during amputation damage afferent sensory neurons. Normal neurons generate activity at terminal sites whereas damaged develop abnormal properties and begin firing spontaneously. Damaged neurons have the ability to generate activity at other levels such as dorsal root ganglia (Nikolajsen et al. 1996). These changes may lead to phantom pain. The input from peripheral neurons initiates sensitization of afferent neurons in the dorsal horn of the spinal cord. Normally, these neurons respond to noxious stimulation however,due to the abnormal peripheral neurons, these neurons begin responding to lowlevel stimulations. Peripheral involvement is assumed because as damaged peripheral nerves heal the barrage of information to the central nervous system decreases and phantom pain subsides. Those patients in whom phantom pain eventually diminishes supports this theory. However, phantom pain can not be completely explained by the peripheral nervous system. Baron et al. (1998) reported a case study of one patient with a spinothalamic injury previous to amputation who experienced phantom pain. No treatment to peripheral neurons relieved his pain. This study is strong evidence that phantom pain consists of some supraspinal involvement as well as peripheral.

In 1992, Melzack presented his Neuromatrix Theory of Pain which consists of three major neural circuits. The first pathway carries information from the periphery through the thalamus to the somatosensory cortex. The thalamus directs all afferent sensory information to appropriate regions of the cortex for interpretation. The second pathway is critical for emotion and may account for the affective descriptions of phantom limbs. This pathway goes through the reticular formation to the limbic system. As sensory impulses reach the reticular formation, it activates the cerebral cortex. When the reticular formation is damaged, the cortex remains unaware of stimulation and does not carry out any thought processes. Finally the last pathway includes the parietal lobe which is responsible for sensations to temperature, pressure, touch and pain. Melzack's theory states that abnormal information sent by damaged peripheral neurons alters the normal operating pattern of the neuromatrix. Abnormal input may be either a lack of normal sensory input due to the amputation or an over-excitement of the neuromatrix from excessive firing of the damaged nerves.

Another facet of the Neuromatrix Theory is a concept called Somatosensory Pain Memories. Pain memories are experienced as sensory and affective and both these aspects may be encoded within the neuromatrix. Continuous afferent sensory information may produce extensive changes in central nervous structures causing a pain memory (Baron et al.1998). These pain memories reinforce the occurrence of phantom phenomena. This concept was further strengthened by a study conducted by Hill et al. (1993).

Treatments

Management of phantom pain is especially difficult because the mechanism causing the pain is unknown. Due to peripheral nerve irritation, abnormal sensory input and psychological factors that contribute to the phenomena, each needs to be considered when suggesting a treatment. Approximately 50 different treatments are currently being used to manage the pain (Sherman et al. 1980). However, in a study of 2500 veterans, only 1 reported lasting relief to the pain attributed to the treatment (Urban et al. 1986). This review highlights several areas but does not include all 50 treatments.

Physical Treatments

The most simple and noninvasive treatment suggested by physical therapists and prosthetists is manual stimulation of the residual limb by massaging, tapping or light slapping (Sherman, 1980). This not only helps to desensitize the residual limb, but also provides stimulation to the nerve endings. However, there have been reports of ineffectiveness as well as cases where manipulation actually evokes phantom pain. Nikolajsen et al. (1996) used manual stimulation to produce phantom pain in order to study the effects of ketamine, a pharmacological treatment.

Another noninvasive treatment is transcutaneous electrical nerve stimulation (TENS). An electrical current is passed through the residual limb at a comfortable level in order to stimulate nerve endings. TENS is a moderately successful treatment due partly to the lack of adverse effects and complications (Baron et al. 1998). This treatment is also easily repeated at a reasonable cost. A second electrical treatment is electroconvulsive therapy which has been used since 1938 on persons suffering from severe depression. Rasmussen and Rummans (1999) studied the effects of electroconvulsive therapy on severe phantom pain in two case studies. Neither patient had responded to previous treatments including TENS, multiple pharmacological treatments, hypnosis, or biofeedback. The patients reported a significant lessening of pain over the course of treatment and both cases were determined to be successful. One patient reported lasting relief to phantom pain for 18 months following the conclusion of treatment. The second patient was not followed.

Biofeedback and relaxation techniques teach amputees to voluntarily relax muscles throughout the body. As previously mentioned, stress influences phantom pain, therefore stressmanagement via biofeedback has shown reduction in phantom pain. Unfortunately, there have been no substantial long-term studies.

Ultrasound treatment was studied in the 1950's with reasonable success. The ultrasound device is applied to the residual limb for specific intervals of time at varying intensities. Again, patients were evaluated within short follow-ups.

Finally, acupuncture has been tried on a small number of amputees with some success. Further studies should be conducted for more conclusive results.

Pharmacological Treatments

Nonsteroidal Anti-Inflammatory Drugs (NSAIDs) such as ibuprofen (Excedrin, Motrin) and naproxen (Aleve) relieve inflammatory pain but are generally ineffective on phantom limb pain (Baron, 1998).

Carbamazepine is an analgesic that acts on the central nervous system. It is also considered an anti-convulsant, often used in the treatment of epileptic seizures (McVan, 1995). The action of carbamazapine is to stabilize neuronal membranes during nerve impulses. Few scientific studies evaluate its effects on the treatment of phantom pain. Patterson (1988) reported one case study of severe pain. Carbamazepine was successfully prescribed and the pain diminished. However, once the patient quit the medication, the pain returned. The patient was prescribed the same medication and the pain subsided. It appears that carbamazepine suppresses or inhibits the mechanism causing phantom pain.

Gabapentin (neurotin) is another drug used to treat epileptic seizures. It binds to a protein located on neurons only in the brain (Chadwick, 1994). Epilepsy and phantom pain are both characterized by hyperexcitability and abnormal discharge of signals by neurons (Morris, 1999). However, the mechanism influencing phantom pain is unclear.

Ketamine is an anaesthetic shown to reduce neuropathic pain including phantom pain. The mechanism acts antagonistically at receptor sites in dorsal horn neurons. Ketamine binds their sites to block the hyperexcitability of the neurons and decrease clinical manifestations of phantom pain. Nikolajsen et al. (1996) reported that ketamine produced significant relief of phantom pain in eleven patients. As with carbamazepine, cessation of treatment caused a reoccurrence of phantom pain.

Methadone is a narcotic that acts on the central nervous system. This drug binds to opiate receptors in the CNS to alter both perception and emotional response to pain. Methadone has a synergistic relationship with antidepressants. Antidepressants increase the effectiveness of the narcotic at lower dosages. Risk of addiction is less at lower dosages of methadone. Urban et al. (1986) reported the successful management of five patients over a 22-month period concluding that narcotics may be safely utilized for management of phantom limb pain.

All analgesics reviewed have seen success in the treatment of phantom pain. However, pharmacological intervention has disadvantages including side effects, drug interactions, habituation, addiction, and long term effects on other organs such as the liver and kidneys.

Psychotherapeutic Treatment

Hypnosis has been used as a psychosomatic treatment of phantom pain with variable success. Muraoka et al. (1996) reported that phantom pain was significantly reduced in one patient by using hypnosis. The therapy lasted for five years. Hypnosis is generally used with pharmacotherapy including anticonvulsants (clonazepam) or antidepressants (amitryptilin).

Surgical Treatment

Surgical revisions due to neuromas are shown to be the most effective procedure, although not always a permanent solution to phantom pair. The pain often recurs, sometimes at a more intense level. In an effort to eliminate incoming nerve signals to the brain, various surgical procedures have yielded little success in the literature. Chordotomies are the surgical division of the anterolateral tracts of the spinal cord. Most studies show negative outcomes (Sherman, 1980). Cutting the dorsal roots of the spinal cord such as in rhizotomies yields similarly poor results. Prefrontal lobotomies incise the white matter of the frontal lobes of the brain and also have limited success. The reviews state that patients report the cessation of the pain, however phantom sensations are still present. Due to the limited success reported in the literature, surgical intervention should be reserved to the most severe phantom pain cases.

Conclusion

Phantom pain disrupts many lives. Not only does it interfere with activities of daily living but complicates and impedes the progression of rehabilitation. As evident in the literature review, there is no one successful treatment to phantom pain. This is largely due to its unknown origin. Symptomatic treatment should be utilized to its full benefit before considering invasive techniques such as surgery, especially in the elderly. A high percentage of elderly comprise the amputee population. This age group generally has multiple chronic diseases or illnesses and takes several medications making drug therapy difficult to manage and invasive techniques risky.

Prosthetic fitting should be started as soon as the residual limb is suitable following amputation and should continue as long as the phantom pain can be tolerated. Some patients find that the incoming peripheral nerve signals generated by the pressures of the prosthesis reduce the pain. Others are not able to proceed with prosthetic treatment. For these patients, rehabilitation is most challenging and frustrating.

References

1. Baron, R., G. Wasner and B. Lindner. Optimal Treatment of Phantom Limb Pain in the Elderly. Drugs and Aging. 1998;12(5):361-376.

2. Bowker, John and John Michael (editors). Atlas of Limb Prosthetics. 1992. Mosby Company. St. Louis, Missouri. 930 pgs.

3. Chadwick, David. Gabapentin. The Lancet. 1994;343:89-91

4. Hill, Anne. Phantom Limb Pain: A Review of the Literature on Attributes and Potential Mechanisms. Journal of Pain and Symptom Management. 1999;17(2):125-142.

5. Hole, John and Karen Koos. Human Anatomy. 1994. Wm. C. Brown Publishers. Dubuque, Iowa. 662 pgs.

6. Jensen, Troels, Borge Krebs, Jorn Nielsen and Peter Rasmussen. Phantom Limb, Phantom Pain and Stump Pain in Amputees during the First 6 Months Following Limb Amputation. Pain. 1983;17:243-256.

7. McVay, Barbara (publisher). Nursing 95 Drug Handbook. 1995. Springhouse Corporation. Springhouse, Pennsylvania. 1335 pgs.

8. Montoya, Pedro, Wolfgang Larbig, Norbert Grulke, Herta Flor, Edward Taub, Niels Birbaumer. The Relationship of Phantom Limb Pain to Other Phantom Limb Phenomena in Upper Extremity Amputees. Pain. 1997;72:87-93.

9. Morris, George. Gabapentin. Epilepsia. 1999;48:563-570.

10. Muraoka, Mamoru, Hiroaki Komiyama, Masako Hosoi, Kazunori Mine, and Chiharu Kubo. Psychosomatic Treatment of Phantom Limb Pain with Post-Traumatic Stress Disorder: a Case Study 1996;66:385-388.

11. Nikolajsen, Lone, Carsten Hansen, Jesper Nielsen, Johnny Keller, Lars Arendt-Nielsen, and Troels Jensen. The Effect of Ketamine on Phantom Pain: A Central Neuropathic Disorder Maintained by Peripheral Input Pain. 1996;67:69-77.

12. Patterson, John. Carbamazepine in the Treatment of Phantom Limb Pain. Southern Medical Journal 1988;81(9):1100-1102.

13. Rajbhandari, S.M., J.A. Jarratt, P.D. Griffiths, J.D. Ward. Diabetic Neuropathic Pain in a Leg Amputated 44 Years Previously. Pain. 1999;83:627-629.

14. Rasmussen, Keith and Teresa Rummans. Electroconvulsive Therapy for Phantom Limb Pain. Pain. 2000;85:297-299.

15. Sherman, Richard. Published Treatments of Phantom Limb Pain. American Journal of Physical Medicine 1980;59:5:232-244.

16. Sherman, Richard, Crystal Sherman and Norman Gall. A Survey of Current Phantom Limb Pain Treatment in the United States. Pain. 1980;8:85-99.

17. Sidebotham, David, Stephan Schug, Eugen Petcu. Pain Control. The New Zealand Medical Journal 1997;110:1039.

18. Urban, Bruno, Randal France, Eileen Steinberger, Dianne Scott and Allan Maltbie. Long-Term Use of Narcotic/Antidepressant Medication in the Management of Phantom Limb Pain. Pain. 1986;24:191-196