Phantom sensation involves conscious perception and sensation of body parts that are no longer present. The most common phantom sensations are localized to upper and lower extremities2. Phantom limb patients often experience perceptions of tactile sensation in the absence of external stimulation on the phantom limb1. This subset of tactile hallucinations commonly involves sensory experiences of touch, pressure, temperature, vibration, and most prominently, pain 1,2,3.Phantom limb pain is consciously experienced by 98% of individuals with amputations, spinal cord injury, deafferentation, and congenital limb deficiency1. Phantom limb sensations that lack subsequent pain perception are thought to be an adaptive compensatory mechanism, whereas those invoking pain are considered maladaptive1. Phantom limb pain is often described as a burning, tingling, throbbing or cramping sensation in the nonexistent limb1,2. The source of phantom limb pain has been linked to maladaptive peripheral and central nervous system alterations in brain areas pertaining to perception and body awareness1,2. Additional theories regarding the origin of pain in these hallucinatory limb sensations include body schema and the neuromatrix theory1. Phantom limb pain is speculated to be associated with the reorganization of regions in the primary somatosensory cortex, the thalamus, the insular cortex, the anterior cingulate cortex, as well as in the cerebellum1,2. Moreover, phantom limb sensations can induce a hallucinatory perception of movement in the nonexistent limb3. Treatments for phantom limb sensations, as well as coinciding phantom limb pain,include pharmacological, non-pharmacological and surgical treatment options2. The most promising results have been witnessed in mental imagery and mirror therapy treatment methods2,3,4.

Phantom Limb Syndrome Epidemiology

Phantom limb describes the perception of tactile sensation or pain that is experienced in the limb that is no
longer present. Phantom_Limb_1.jpgPhantom limb patients experience both sensory and motor activation in their phantom limb3. This clinical phenomenon was first described about 450 years ago by Ambroise Pare, a military surgeon5. However, it was not until two centuries later that the name “phantom limb” was created by Silas Weir Mitchell5. Mitchell’s original conception was that phantom limb syndrome represented a sensory hallucination5. The classification of hallucinations as a mental disorder lead to early beliefs that the experience of a phantom limb may be a psychological problem in which the patient either experiences denial regarding the presence of their limb or experiences activation of memories about the phantom limb5. But it was later determined that phantom limb patients do not suffer from a mental illness and they are fully aware that their limb is no longer present, even though they experience vivid phantom sensations5.

It is estimated that 1.6 million individuals in the United States suffer from limb loss2,6. The most common cause of limb loss is from amputation of the extremities2. Amputation can occur as a result of trauma, spinal cord injury, tumor malignancy and inadequate blood flow1,6. Phantom limb experiences can also occur in the absence of amputation and this is often associated with congenital limb deficiency2.

Phantom sensations often appear instantaneously after amputation in 75% of patients; however the onset of these sensations can vary among patients and some may experience phantom limb sensations days, weeks or years after amputation1,7. In some, the phantom sensation is only experienced for a short period but in others it may persist and have a longer duration7. Phantom perceptions are experienced by 80% of patients with limb loss. These sensations include tingling, itching, twitching, gesturing, vibration, pressure, burning, cramping, piercing, hot and cold sensations and phantom limb pain1,2,3,8. Many studies have found that phantom limb patients often perceive their phantom limb as resembling a functional intact limb, whereas some experience the perception of a telescoped or shrunken limb9,10. Telescoping involves the perception that the proximal end of the limb is absent or shortened and that the digits or distal portion of the limb is situated close to or on the stump itself 8,9. Infrequently, patients can also experience the existence of a swollen or enlarged limb3,8.

Phantom Limb Pain

Phantom sensations, following amputation, often result in the experience of residual stump pain and phantom limb pain6. The incidence of phantom limb pain is estimated to be 50-80% in limb amputees1,4. The most common types of phantom limb pain involve the sensation of cramping, tingling, piercing as well as sharp shooting pain1,10.

The two common types of pain conditions that occur after amputation are residual stump pain and phantom limb pain. Residual stump pain is pain associated with the remaining parts of the upper or lower limbs after amputation, whereas phantom limb pain is associated with painful phantom limb perceptions1.

Interestingly, it has been observed that phantom limb pain has an increased prevalence in patients with upper limb amputations compared to amputations of the lower extremities1. The experience of phantom limb pain has also been reported to show gender differences and is more commonly observed in females than males1. Phantom pain, following amputation, is often associated with abnormal posture including twists and bends into fixed unnatural positions9,10.

Neurological Basis

The neurological basis of phantom limb pain involves a combination of many theories that have examined the pathophysiology of phantom limb sensations. Many mechanisms have been proposed in order to explain the phenomenon of phantom limb pain. The neurological basis of phantom limb pain is often related to changes in the central and peripheral nervous systems1,5. Peripheral, spinal and central mechanisms are believed to play an important role in the perception of phantom limb pain; however no theory single-handedly is able to explain the complex neurological basis of phantom limb sensations and perceptions1. It is now believed that the pathophysiology of phantom limb pain is best explained by a multifactorial mechanism5.

Figure 1. Central and peripheral nervous system connections that are associated with phantom limb pain. Neuroma activity from the peripheral fibres alters spinal cord output and in turn effects the cortical structures of the brain, including the thalmus, somatosensory cortex and limbic system.
Peripheral Mechanisms

Amputation results in many peripheral nervous system changes and these changes are a result of severed peripheral nerves after amputation1. The severing of peripheral nerves results in neuronal injury, tissue damage and impairment in afferent input to the spinal cord1,11. These disruptions in afferent connections are known as deafferentation11. After deafferentation, the severed nerves begin to grow towards each other, in an attempt to reform connections, but instead form knots known as neruomas11. Neuromas are often associated with hyperexcitability and unpredictable discharges1,11. This abnormal excitability in phantom limb pain is thought to be caused by the accumulation of sodium channels in the neuromas11,12. This theory is supported by many studies examining the pain alleviating effects of substances that block sodium channels and remove neuromas. Spontaneous discharges from neuromas are often associated with pain, suggesting that peripheral mechanisms may play an important role in the etiology of phantom limb pain1,11,12. A study by Borghi et al indicated that ropivacaine, a sodium channel blocker, abates pain that is associated with the phantom limb. Neuroma removal has also been found to reduce phantom pain in some patients11. The theory of peripheral nervous system changes associated with phantom pain is one of the prominent theories in explaining the pathogenesis of phantom limb pain. However, peripheral nerve sprouting and neuroma formation is not instantaneous; therefore immediate onset of phantom limb pain cannot be explained solely by peripheral nervous system changes1,11,13. In addition to peripheral mechanisms, other mechanisms may play an important role in patients who experience immediate onset of phantom limb pain.

Spinal Cord Mechanisms

After nerve damage, axonal sprouts from damaged peripheral nerves form connections with the spinal cord2. Spinal cord neurons can also extend into the dorsal horn and dorsal root ganglion of the spinal cord, which contain sensory neurons that produce ectopic impulses11. Neuronal activity in regions important for nociceptive input transmission results in the summation of signals and an increase in nociceptive input, which is called central sensitization2. Inhibitory neurotransmitters, such as GABA or glycine, could result in spinal disinhibition11. It has also been determined that interneurons can be damaged by nerve axotomy and hyperexcitability11,9. Nerve tissue damage can also result in an increase in NMDA receptor reactivity, at the level of the dorsal horn, in response to glutamate2,11. This mechanism disrupts the firing pattern of neurons that project from the spinal cord to various cortical areas1,2,11. Studies examining the effect of spinal cord alterations on phantom limb pain have observed that tissue damage and
injury result in the expression of substance P on fibers that transmit proprioceptive information11. Substance P is often associated with
nociceptive stimuli; therefore fibers that, originally, did not transmit nociceptive info now may have the ability to transmit noxious information
in phantom limb patients11.

Central Cerebral Mechanisms

Central nervous system changes in cortical regions are often considered to be the main underlying mechanism of phantom limb pain1,2. Phantom limb pain has been connected to changes in many regions of the brain, including the anterior cingulate cortex, brainstem, thalamus and, most importantly, the primary somatosensory cortex and the primary motor cortex2,11,15. Often these cortical changes are associated with cortical reorganization in the brain15. The cortical reorganization may be a result of axonal sprouting11. Axonal sprouting is thought to be caused by increased excitability, which is a result of increased presence of excitatory neurotransmitter2,11. Increased excitability in various cortical regions can lead to the experience of phantom limb pain11. It has also been suggested that inhibitory fibers are lost due to amputation and this loss may cause cortical reorganization in areas where c-fibers usually have an inhibitory effect11. Cortical reorganization is often observed in areas that are associated with the sensory and motor homunculus1,2,11,15. These cortical changes are often due to maladaptive plasticity, which results in phantom limb pain in many amputees15. The somatosensory cortex displays the most amount of reorganization, and regions adjacent to the lost limb begin to occupy the somatosensory area that was once linked to the phantom limb11,15. These maladaptive neural changes are thought to alter the body schema, which results in the experience of phantom sensations11. Studies have also indicated that the cuneate nucleus of the brainstem and the ventral posterior nuclei of the thalamus show a great deal of restructuring in areas associated with the face and lost limb16. Therefore, the brainstem and thalamus may contribute to changes in body representation in the cortical homunculus.

Figure 2. Summary of central and peripheral nervous system changes in phantom limb patients.

Theories of Phantom Limb Pain

Neuromatrix Theory

The neuromatrix or neurosigniture theory of phantom limb pain was first established by Ronald Melzack and this theory is closely related to the body schema concept2. The neuromatrix is a neural network that produces different impulse patterns depending upon the input received from various cortical areas and the connections that are formed between different neural circuits5. These cortical inputs are received from thalamocortical circuits, reticular formation/limbic system circuits, visual areas and the somatosensory cortex2,5. Inputs from these areas are processed and synthesized to form a characteristic, integrated output5,17. The neuromatrix theory proposes that outputs, or nerve impulses, generated from cortical inputs from the self body schema form the basis of self awareness5. The integrated output is classified as the neurosigniture and is thought to be hardwired5,17. A hardwired neurosigniture may create problems after amputation because of conflict between perceptions regarding the post operational state of the body2,5. The static neurosigniture cannot be altered after amputation even thought there is input indicating that the phantom limb is no longer present5. The differences in input from the neurosigniture and the body gives rise to an overriding response that results in the perception that the body is still unchanged5. The perception of an intact limb, regardless of contrary sensory input from the body, may be the underlying cause of phantom sensation and phantom limb pain1,2,5.

Cortical Reorganization Theory

Cortical Reorganization theory is one of the most cited theories explaining the etiology of phantom limb pain1,2. Merzerich et al examined the validity of the cortical reorganization theory and found that following amputation and deafferentation, adult monkeys displayed cortical changes in the somatosensory homunculus14. Theses finding lead to the belief that cortical reorganization may be a key component in producing phantom limb sensations7,15. Cortical reorganization theory suggests that maladaptive cortical reorganization occurs in the somatosensory cortex (S1), as well as the motor cortex (M1)8. Maladaptive reorganization occurs when surrounding areas take over control of the somatosensory and motor representations of the missing limb2,15. During cortical reorganization it is observed that regions of the face, often, take over homunculi regions associated with the arm in upper limb amputees2,11,15. It is proposed that maladaptive plasticity and cortical reorganization in phantom limb patients explains the cause of phantom limb pain15. Functional magnetic resonance imaging data illustrates that patients who experience phantom limb pain have greater activation in somatosensory and motor cortices, where as amputees without pain and healthy controls show no increase in activation (Figure 4)15. The strong support for the cortical reorganization theory has lead to the development of possible treatment methods that attempt to decrease the maladaptive cortical reorganization in amputees2.

Figure 3. Motor and sensory cortical homunculus. Regions that undergo reorganization in phantom limb patients often involve changes in cortical representation in these sensory and motor areas

Figure 4. Cortical changes associated with phantom limb pain. Functional MRI data indicates that patients with phantom limb pain show greater cortical activity in primary somatosensory and motor cortices. In individuals experiencing phantom pain also show cortical reorganization.


Treatments for phantom limb pain involve different techniques that focus on the various mechanisms associated with phantom limb pain2. Primary intervention strategies include pharmacological treatments and non-pharmacological, behavioural treatments2,9. Surgical treatment interventions have also been utilized; however they are rarely effective in alleviating phantom limb pain and are mainly administered when other treatment methods have failed to yield effective results2,9.

Pharmacological Treatments

Various pharmacological treatment methods have been employed to treat phantom limb pain but effective techniques that completely alleviate symptoms are yet to be determined18. Phantom limb pain is often difficult to treat due to its pathophysiology, which involves central and peripheral nervous system changes2. Therefore, conventional treatment methods have little effect on pain reduction19. Possible pharmacological treatment approaches include the use of pre-emptive analgesia and anaesthesia, opiods, NMDA receptor antagonists and Botulinum toxin type A 2,19,20,21.

Pre-Emptive Analgesia and Anaesthesia

Pre-emptive analgesia and anaesthesia refers to the use of analgesics and anaesthetics before the amputation of the limb takes place2. Experimental studies have observed that the administration of pre-emptive analgesia prevents hyperplasia and neural sensitization that may result from stimuli associated with the amputated limb2. These preventative measures in turn regulate and decrease future activation of painful stimuli associated with the phantom limb. Double blind studies, examining the effects of ketamine and bupivacaine administration, found that preoperative, analgesic injection into the epidural space resulted in a decrease in phantom limb pain2,22. Although pre-emptive analgesia administration helps control phantom limb pain directly after the amputation, the results are not conclusive and pre-emptive pharmacological treatment does not result in long term reduction of phantom limb pain and residual stump pain2,5,23.


Opioid medication is often used to treat phantom limb pain. Opioids are chemicals that bind to central and peripheral nervous system opioid receptors. Huse et al observed the analgesic properties of morphine administration in patients experiencing severe phantom limb pain. It was observed that morphine sulphate treatment resulted in a 50% decrease in phantom limb pain experience in about half of the patients. Morphine was also shown to play a role in the reduction of maladaptive cortical reorganization that is observed in phantom limb patients who experience severe pain20. The etiology of phantom limb pain is often linked to cortical reorganization; therefore opioids could effectively treat phantom limb pain through mechanisms of action that disrupt cortical reorganization in the somatosensory and motor cortex2,20. Although opioids may successfully treat phantom limb pain and decrease maladaptive cortical plasticity, several studies have shown that opioid administration is often associated with greater side effects2,11. Common side effects include drowsiness, dizziness, nausea sweating and vertigo5,20. Opioid medication remains a common treatment method for neuropathic pain conditions, like phantom limb pain, however studies are not completely conclusive and fail to show an effect in some patients with severe phantom limb pain20.

NMDA Receptor Antagonist

N-Methyl-D-aspartic acid (NMDA) receptor antagonists, like Memantine, have been shown to decrease phantom limb pain in many patients and reverse cortical reorganization21. NMDA receptors are thought to be associated with central and peripheral sensitization, which is predicted to be a result of increased glutamate release during and after the injury2,18. Administration of Memantine has been shown to decrease pain in some phantom limb patients, however the results are not definitive and various studies have obtained contradictory results2,21. Hacksworth et al determined that Memantine is more effective than first generation NMDA receptor antagonists (Ketamine)21. First generation NMDA receptor antagonists give rise to many side effects, including headaches, nausea, dizziness and more severe side effects such as loss of consciousness, decreased hearing, and perception of visual hallucinations21. Lastly, studies have demonstrated that NMDA receptor antagonists may be effective as an initial treatment method, as opposed to a long term treatment strategy for alleviating phantom limb pain2,11.

Botulinum Toxin Type A

Botulinum toxin is a powerful neurotoxin produced by the bacterium Clostridium Botulinum24. Once internalized, Botulinum toxin type A inhibits acetylcholinerelease from nerve endings by disrupting mechanisms involved in vesicle docking24. These characteristics allow Botulinum toxin type A to be used in treatment of disorders associated with muscle contraction and various pain syndromes24. Jin et al showed that phantom limb pain and stump pain were successfully treated and pain was reduced after administration of Botulinum toxin type A. This demonstrates that botulinum toxin may provide an addition technique that can be utilized to effectively treat phantom limb pain.

Non-pharmacological Treatments

Mirror Therapy

Figure 5. Mirror Box Therapy. Exmaple of a mirror box that would be used to provide visual feedback in order to treat amputees that experiencing phantom limb pain.

Mirror therapy techniques that are used to treat phantom limb pain make use of visual feedback mechanisms and mirror neuron systems, which play an important role in goal directed behaviour2. Phantom sensations, in amputees, often involve mirror neuron systems. It is thought that phantom limb patients experience mirror neuron activity when phantom limb perceptions are experienced as the patient observes the movement of others around them2. It is thought that mirror neuron activity may reinforce the body schema concept and reinforce representation of the phantom limb within the cortex2,11,25. Therefore, phantom limb perception may be due to the maintenance of limb representation within the cortex. Mirror neuron systems have also been implicated to play a role in touch and pain2. Mirrored sensations are also observed in about 13% of phantom patients2. These sensations involve mirrored perceptions from the intact limb to the phantom limb2. These sensations include thermoception, tactition, proprioception and nociception2,25. Therefore it is proposed that mirror neuron systems involvement in tactile sensation and pain may reinforce and maintain the representation of the phantom in the cortex2. This mirror neuron system and the cortical representation of the phantom limb are manipulated in treatment methods involving mirror therapy25.

Ramachandran was the first to report the effectiveness of mirror therapy and visual feedback in the treatment of phantom limb pain2. Patients observe movements of the intact limb in mirrors placed between the phantom limb and the intact limb. The visual feedback and reflection of the intact limb replaces the phantom limb2,25. Mirror neurons systems in the brain show activation as sensations in the phantom limb are perceived from tactile stimuli associated with the mirrored reflection2. Activation of mirror neurons in patients with amputated limbs results in tactile sensation without the presence of a real tactile stimulus on the phantom limb. Mirror therapy is thought to modulate somatosensory input and as a result this modulation may decrease the perception of phantom limb pain and alleviate painful phantom symptoms2,25.
Rehabilitation methods using the mirror box are mind-body therapies that are effective in treating phantom pain for many patients but these therapies do not alleviate pain in all phantom limb patients2,26. A study by Diers et al looked at the effectiveness of mirror therapy treatment in patients suffering from severe phantom limb pain. They determined that mirror therapy does not show a significant effect in reducing phantom limb pain and it may not be the best form of treatment for patients with severe phantom limb pain2,26. Instead, mirror imagery was found to be effective in alleviating some symptoms in patients with severe pain26. During imagined movements there was no increase in cortical activity in areas that often showed cortical reorganization26. Although the mechanisms are unclear, these experimental findings suggest that mirror imagery may also be an effective treatment for alleviating the painful symptoms experienced by phantom limb patients.

See Also

Phantoms In the Brain Part 1 Episode 1
Phantoms In the Brain Part 2 Episode 1
Phantoms In the Brain Part 3 Episode 1
Phantoms In the Brain Part 4 Episode 1
Phantoms In the Brain Part 5 Episode 1


  1. Giummarra, M. J., Gibson, S. J., Georgiou-Karistianis, N, & Bradshaw, J. L. Central mechanisms in phantom limb perception: The past,present and future. Brain Research Reviews 54, 219-232 (2007)
  2. Subedi, B. & Grossberg, G. T. Phantom limb pain: mechanisms and treatment approaches. Pain Research and Treatment 1, 1-8 (2011)
  3. Raffin, E., Mattout, J., Reilly, K. T, & Giraux, P. Disentangling motor execution from motor imagery with the phantom limb. Brain 135, 582–595 (2012)
  4. Diers, M., Christman, C., Koeppe, C., Ruf, M., & Flor, H. Mirrored, imagined and executed movements differentially activate sensorimotor cortex in amputees with and without phantom limb pain. Pain149, 296–304 (2010)
  5. Reilly, K. T, & Sirigu, A. The motor cortex and its role in phantom limb phenomena. Neuroscientist 14, 195-202 (2008)
  6. Moura, V. L., Faurot, K. R., Gaylord, S. A., Mann, J. D., Sill, M., Lynch, C., & Lee, M.Y. Mind-body interventions for treatment of phantom limb pain in persons with amputation. American J. Phys. Med. Rehabil. 91, 2-14 (2012)
  7. Ramachandran, V. S., & Hirstein, W. The perception of phantom limbs. Brain 121, 1603-1630 (1998)
  8. Sumitani, M., Miyauchi, S., Yozu, A.,Otake, Y., Saitoh, Y., & Yamada, Y. Phantom limb pain in the primary motor cortex. J Anesth 24, 337–341 (2010)
  9. Giummarra, M. J., & Moseley, G. L. Phantom limb pain and bodily awareness: current concepts and future directions. Current Opinion in Anesthesiology 24, 524–531 (2011)
  10. Giummarra, M. J., Georgiou-Karistianis, N., Nicholls, M. E. R., Gibson, S. J., Chou, M., & Bradshaw, J. L. Corporeal awareness and proprioceptive sense of the phantom. British Journal of Psychology 101, 791–808 (2010)
  11. Foell, J., Bekrater-Bodmann, R., Flor, H.,, & Cole, J. Phantom limb pain after lower limb trauma: origins and treatments. The International Journal of Lower Extremity Wounds 10, 224–235 (2011)
  12. Borghi, B., D’Addabbo, M., White, P. F., Gallerani, P., Toccaceli, L., Raffaeli, W., Tognu, A., Fabbri, N., & Mercuri, M. The use of prolonged peripheral neural blockade after lower extremity amputation: the effect on symptoms associated with phantom limb syndrome. Anesthesia 111, 1308-1315 (2010)
  13. Vase, L., ] Egsgaard, L. L., Nikolajsen, L., Svensson, P., Jensen, T. S., & Arendt-Nielsen, L. Pain catastrophizing and cortical responses in amputees with varying levels of phantom limb pain: a high-density EEG brain-mapping study. Exp Brain Res 1, 1-11 (2012)
  14. Merzenich, M. M. et al. Somatosensory cortical map changes following digit amputation in adult monkeys. The Journal of Comparative Neurology 224, 591-605 (1984)
  15. Flor,H., Nikolajsen, L., & Jensen, S. T. Phantom limb pain: a case of maladaptive CNS plasticity?. Nature 7, 873-881 (2006)
  16. Jones, E. G. & Pons, T. P. Thalamic and brainstem contributions to large-scale plasticity of primate somatosensory cortex. Science 282, 1121–1125 (1998)
  17. Melzack, R.Pain and the neuromatrix in the brain. Journal of Dental Education 35, 1378-1382 (2001)
  18. Alviar, M., Hale, T., & Dungca, M.Pharmacologic interventions for treating phantom limb pain. Intervention 1, 2-52 (2011)
  19. Hackworth, R. J., Tokarz, K. A., Fowler, I. M., Wallace, S. C., & Stedje-Larsen, E. T. Profound pain reduction after induction of Memantine treatment in two patients with severe phantom limb pain. Pain Medicine 107, 1377-1379 (2008)
  20. Wu, C. L. et al. Morphine versus Mexiletine for treatment of postamputation pain.Anesthesiology 109(2), 289–296 (2008)
  21. Hackworth, R. J., Tokarz, K. A., Fowler, I. M., Wallace, S. C., & Stedje-Larsen, E. T. Profound pain reduction after induction of Memantine treatment in two patients with severe phantom limb pain. Pain Medicine 107, 1377-1379 (2008)
  22. Wilson, J. A. Nimmo, A. F. Fleetwood-Walker, S. M., & Colvin, L. A. A randomised double blind trial of the effect of preemptive epidural ketamine on persistent pain after lower limb amputation. Pain, 131, 108–118 (2008)
  23. Ypsilantis E, Tang TY. Preemptive analgesia for chronic limb pain after amputation for peripheral vascular disease: a systematic review. Ann Vasc Surg; 24, 1139–1146 (2010)
  24. Jin, L. et al. Treatment of phantom limb pain with botulinum toxin type A. Pain Medicine 10, 300-304 (2009)
  25. Lamont, K., Chin, M., & Kogan, M. Mirror box therapy: seeing is believing. Explore 7, 369-372 (2011)
  26. Diers, M., Christmann, C., Koeppe, C., Ruf, M., Flor, H.Mirrored, imagined, and executed movements differentially activate sensorimotor cortex in amputees with and without phantom limb pain. Pain 149(2), 296–304 (2010)