Pain Management for the Lower-Limb Amputee

Nicolas E Walsh MD
University of Texas Health Science Center
San Antonio, Texas

Limb pain is a common problem following amputation. The etiology of post amputation pain is often multifactorial and its satisfactory resolution presents a significant challenge to the healthcare provider. The treatment of pain in each individual must begin with understanding of pain components, accurate differentiation of the multi-factorial etiology and appropriate diagnosis.


The International Association for the Study of Pain defines pain as an unpleasant sensory and emotional experience associated with actual or potential tissue damage or described in terms of such damage.

  • Acute pain is a biologic symptom of an apparent nocioceptive stimulus such as tissue damage due to disease or trauma. The pain may be highly localized and/or may radiate. It is generally sharp, crushing, or burning and persists only as long as the tissue pathology itself persists. Acute pain is generally self limiting and as the nocioceptive stimulus lessen, the pain decreases. Acute pain usually last a few days to a few weeks. If it is not effectively treated it may progress to a chronic form.

  • Chronic pain is a disease process in which the pain is a persistent symptom of an autonomous disorder with neurological, psychological, and physiological components. Chronic pain is defined as pain lasting longer than anticipated within the context of the usual course of an acute disease or injury. Chronic pain may be associated with continued pathology or may persist after recovery from a disease or injury. As with acute pain, treatable chronic pain is due to organic disease and is managed by effectively treating the underlying disorder. Chronic pain is often poorly localized and tends to be dull, aching, and constant. The associated signs of autonomic nervous system response may be absent, and the patient may appear, exhausted, listless, depressed, and withdrawn.

Pain is a central perception of multi-primary sensory modalities. This interpretive function is complex involving psychological, neuroanatomical, neurochemical, and neurophysiological factors of both the pain stimulus and the memory of past pain experiences as well as cultural conditioning. Conceptually pain can be thought of as being composed of three hierarchical levels: a sensory-discriminatory component (e.g., location intensity quality), a motivational-affective component (e.g., anxiety, depression), and a cognitive-evaluative component (e.g., thoughts concerning the cause and significance of the pain). Clinically, this concept of pain serves to focus attention on the broad range of factors which contribute to the pain experienced by the patient.

Post-amputation Limb pain is often the result of surgical trauma, wound healing complications, tissue loading effects, local scarring, and central neuropathic phenomenon. Limb pain may be categorized as immediate post-operative pain, extrinsic residual limb pain, intrinsic residual limb pain, and phantom limb pain. Unfortunately, most authors have not discriminated carefully between residual limb pain and phantom limb pain which makes the literature confusing as to the incidence and prevalence of each of these postamputation pain problems.


Immediate postoperative pain is almost always present following amputation and is the direct result of the surgical trauma to bone, nerve, and soft tissue. Postoperative pain can be expected to resolve within three weeks or less, as with pain following any major surgical procedure. This pain is often described as sharp, is localized to the surgical site, is usually self limiting, and resolves as the edema decreases and the surgical wound heals. In the immediate postoperative period the primary method of pain control is medication utilizing intravenous or epidural delivery of pain medication via patient controlled analgesia (PCA pump). The patient is commonly transitioned to oral analgesic medication by post-operative day 3 or 4. The adjunctive measures for edema include elastic wraps, elastic stump socks, semi-rigid dressings, and rigid plaster casts. Immediate postoperative pain is expected to improve rapidly as the incision site heals.


Extrinsic residual limb pain is usually mechanical in origin related to the prosthetic socket or other prosthetic components. It is often the result of a mismatch between residual limb tissue tolerance and the prosthetic loads on the soft tissues which may be the result of inappropriate alignment, design, or fit of the socket and less often due to prosthetic dysfunction of the other component parts. The underlying disease process as well as surgical trauma reduces the margin of safety between tissue tolerance and prosthetic socket loads. A patient is initially evaluated for the source of pain by examination of the residual limb, making check sockets as necessary, and reconsidering the appropriate mechanical components.

Clinical manifestations of misalignment of the lower limb prosthesis are often manifest by pain. Misalignment of the prosthesis may result in increased or prolonged loading forces in the residual limb even with appropriate socket fit. In the individual with transtibular amputation pain in the distal tibia region is often due to sagittal plane misalignment and fibular pain is often due to frontal plane misalignment. In the individual with transfemoral amputation pain in the distal femur is often the result of misalignment. Clinical manifestations of excessive prosthetic load to tissue tolerance are often manifest by skin changes to include erythema, blisters, calluses, or ulceration.

Treatment normally involves changing the number of sock plies, padding the socket, modifying the socket to relieve high pressure areas, or utilizing specialized prosthetic components to relieve pain or correct skin injury through the use of silicone or gel to sensitive areas. If these fail, then fabrication of a new socket to relieve pressure over sensitive areas may be required. The most common causes are of poor socket fit are body weight changes or residual limb atrophy. Appropriate alignment and socket modifications usually resolve these problems.


Intrinsic residual limb pain is often due to underlying disease process, surgical trauma, bone abnormality, local scar, neuroma, or central neuropathic phenomenon. Some amputees have residual limb pain that outlasts normal healing time. The reported incidence varies between 13% and 71%, however, detailed studies suggest 57% of patients have residual limb pain eight days post-operatively; 22% six months post-operatively; and only 10% have residual limb pain two years post operatively.1,2 Approximately of these individuals report pain as moderate to severe. Pain before amputation is not correlated with residual limb pain after amputation by the majority of studies. Intrinsic residual limb pain is pain perceived as originating in and affecting the residual portion of the limb. It is often described as sharp, aching, throbbing, or burning in nature.

Underlying Disease Process

Residual limb pain may result from infection, ischemia, tumor recurrence, joint dysfunction, or stress fractures. Infection may be superficial resulting from local skin ulceration or deep, which is often the result of osteomyelitis. Pain secondary to limb ischemia is the result of inadequate blood supply often due to peripheral vascular disease. Residual limb pain may also result from tumor recurrence or joint dysfunction. The pain is often described as generalized and usually requires medical or surgical intervention.

Surgical Trauma

Intrinsic residual limb pain resulting from surgical trauma may be due to poor surgical technique such that the bone is improperly trimmed, wound dehiscence, as well as ischemia resulting in inadequate closure due to poor vascularization of the muscles and skin. A residual limb length that is less than optimal for the prosthesis of choice may also result in mechanical residual limb pain. Treatment may require surgical revision or a prolonged course of wound care prior to refitting of the socket.

Bone Abnormality

Bony overgrowth at the distal end of the residual limb most often occurs in children and only occasionally in adults. This bony overgrowth often results in a bone spicule. Normal bone growth may result in irregular edges or changes in length which also produces residual limb pain. Heterotopic ossification in the soft tissues of the residual limb results in excess bone formation usually around the distal end of the surgical site. Any bony contour irregularity may result in areas of high pressure manifest by the complaint of pain. Diagnosis of bony abnormality is routinely made utilizing routine x-rays of the residual limb. The pain is often managed by socket modifications to offload pressure over painful areas. When prosthetic modification is not satisfactory surgical intervention is usually required.

Local Scar

Entrapment of nerves in scar tissue occurs within the surgical incision at all levels. This pain is usually exacerbated with shear force or pressure directly to the healed scar tissue. Mechanical forces in soft tissue deformation cause pain with almost any movement. Extensive scarring is often found in individuals with land mind injuries or delayed wound closure. Initial treatment usually includes prosthetic modifications to reduce loading of pressure sensitive areas through a better distributed load or reduced shear force on adjacent tissues. Treatment may also include injections as well as medication intervention. Surgical intervention rarely provides adequate relief for intrinsic residual limb pain secondary to scar formation.


Neuromas at the surgical site are the most common etiology of intrinsic residual limb pain. It is critical that residual limb pain be recognized as distinct from phantom limb pain as the pain management strategies for these two conditions are very different. Neuromas result of the normal nerve regrowth during the healing process. The lack of an axonal sheath results in a balllike structure at the end of the transected nerve. All amputations result in neuroma formation, however, not all amputees have residual limb pain. The incidence of painful neuromas is thought to be between 10 to 25%. Since only a fraction of amputees develop residual limb pain the presence of a neuroma per say cannot be the cause of pain. The etiology of residual limb pain from neuromas appears to be due to cross talk between nerve which results in extreme sensitivity of the axonal sprouts of the neuroma to mechanical and chemical stimuli. This sensitivity is often accentuated by entrapment of the neuroma in scar tissue or the location of a neuroma where it is exposed to excessive external mechanical loading. Limb pain from neuroma is often variable in intensity, of intermittent duration, with changeable qualities of aching, cramping or shooting. The diagnosis is often confirmed by tenderness to palpation at a single point and by injection local anesthetic directly into the site of the neuroma with resulting pain relief. Treatments of the neuroma have included surgical, physical modalities, medicine, and neurolysis. Surgical Treatment -- It is unknown if encapsulation of the cut nerve, fascicle ligation for self implantation of the nerve during surgery decreases neuromal formation. Research to date suggests that the initial resection of a neuroma provides only occasional relief, i.e., usually the neuroma returns and is painful. Sympathectomy (surgical transection of the sympathetic nerves) provides relief only when pain resembles complex regional pain syndrome (CRPS) type 1 or causalgia. Cordotomy (selective surgical transection of the spinal cord) provides pain relief secondary to neuroma in the residual limb for less than 2 years. Repeated resections of the neuroma, neurectomies at higher levels, reamputation for pain relief, periarterial sympathectomy, intrathecal alcohol, or dorsal rhizotomy do not provide adequate pain relief for intrinsic residual limb pain due to a neuroma. Physical Modalities --Physical modalities have been utilized to treat limb pain due to neuroma which includes ultrasound, massage, vibration, percussion, acupuncture as well as modification of socket for pressure relief. Most physical modalities have had very limited success in reducing pain due to neuroma. Medical Interventions -- Medical intervention for limb pain resulting from neuromas have utilized nonsteroidal anti-inflammatory drugs, tricyclic anti-depressants, and anti-convulsants with limited success. Most other medicine regimes have been the subject of sporadic case reports or case series and have not been shown to be effective. Neurolysis -- Minimally invasive neurolysis with the use of phenol and cryo (freezing of the nerve) nerve blocks have been shown to be effective in the pain control for neuroma. Radio frequency neurolysis is effective but difficult to perform due to technical impedence problems in many individuals. Neurolysis is usually effective for three to five months with longer results reported on repeated procedures.

Central Pain

Residual limb pain may also be the manifestation of autonomic nervous system abnormalities involving the sympathetic post-ganglion neurons after peripheral nerve injury. The manifestation is often burning pain as well as vasomotor, pilomotor, and pseudomotor abnormalities. The residual limb is often exquisitely sensitive to light touch and less sensitive to deep palpation. This manifestation is classified as Complex Regional Pain Syndrome (CRPS) or causalgia. The treatment of limbs with allodynia or dysesthesia often responds to sympathetic blockade and sympathectomy. This form of limb pain is rare but most debilitating.


The phantom limb is the perceived presence of the amputated body part. Phantom limb experience was first referenced by d'Ambroise Pare, a 16th century military surgeon for Napoleon, in 1551. Silas Weir Mitchell, a Civil War surgeon, was the first to use the term phantom limb pain in a scientific report in 1871. Phantom limb sensation appears to occur in all amputees. The phantom sensation is usually more intense in the distal portion of the extremity with many patients reporting the ability to move the phantom by conscious thought. Phantom sensation usually decreases during the first year after amputation. Phantom limb sensation is often a well defined image of the lost part thought to be secondary to the spinal segmental apparatus and higher brain centers. Lesions in the spinal cord and brain will abolish phantom sensations; however, they do not often eliminate phantom limb pain. Congenitally absent limbs and early childhood amputations are unlikely to lead to phantom limb pain which suggests that the phantom depends upon an intact body map in the cerebral hemispheres. The phantom sensations are not painful and all amputees should be made aware that this is a probable outcome of limb loss.

Phantom limb pain is a chronic pain perceived by the individual as originating from the amputated body part. Reported incidence of phantom limb pain is highly variable often due to the setting and circumstances in which the data was collected. Prospective research suggests approximately 82% of patients experience phantom pain following amputation. It is estimated that 65% experience phantom pain at six months and 59% two years after amputation.1,2 Severe disabling phantom limb pain incidence is reported from 10% to 25% after several years. Of note, are reports that severe pain occurs more often with proximal amputations. The incidence of phantom limb pain is independent of age, gender, level of amputation, employer, or age after 8 years old. 3,4 Most studies have found no significant relationships between pre-operative limb pain and persistent phantom pain. Factors contributing to phantom limb pain include peripheral and spinal innervation, central neural components, as well as psychological and cognitive factors.

Peripheral Factors

Peripheral components of phantom limb pain appear to be the result of sensation due to loss of previously present peripheral nerve activity triggering changes in the central nervous system. Development of abnormal activity in a neuroma or dorsal root ganglion results in pain fiber transmission and an alteration in ion channel activity at site of injury resulting in pain fiber activation. Micro-neurographic studies clearly showed spontaneous c-fiber activity in the transected nerves in amputations with phantom limb pain. Local anesthesia to the neuroma stops stump neuroma pain when mechanically stimulated, however, phantom pain and proximal discharge of c-fibers continued (i.e. activity that was proximal to the neuroma).

Spinal Factors

Spinal level factors involved in phantom limb pain include deafferentation resulting from the loss of large A-afferent nerve fiber input in the dorsal horn cells with an unopposed c-fiber input. Experiments have found that the residual limb is much more sensitive to stimuli than that of the homologous region of the opposite limb. N-methyl-D-aspartate (NMDA antagonist) blocks the glutamine receptors in the central nervous system. Glutamine is an excitatory neurotransmitter which is thought to activate the central nervous system after noxious stimuli. The use of N-methyl-D-aspartate (NMDA) receptor blockade for relief of phantom limb pain supports the premise that central sensitization by peripheral nerve injury. Tricyclic antidepressants inhibit the uptake of mono amines and are believed to augment the effects of amines in inhibitory descending spinal pathways Central Factors

Cortical reorganization has been observed in the primary motor and somatosensory cortex in humans after amputation. These reorganizational changes appear to be a significant factor in the occurrence and intensity of phantom limb pain. The remapping of regions of the brain receiving touch, pain and temperature input from the amputated limb to other parts of the brain appears to be an etiology of phantom limb pain. Data indicates that phantom limb pain is related to and possibly a consequence of the plastic changes in the primary somatosensory cortex.

Both animal and human studies have demonstrated that the cortex, when lacking normal sensory input from the extremity, reroutes the pathways to new areas of the brain. The brains prior image or "primordial genetically determined image" of the body is imprinted on the cortex and limbic regions. This image is partially created through sensory experiences in lifeand represents a prior image in the brain for each body part. Some hypothesize that phantom limb pain is the result of abnormal reorganization in the neuromatrix. The lack of proprioceptive feedback for movement or position of the limb appears to be a component of phantom limb pain as well as sensory memories of pain in the original limb that are often carried over to the phantom limb. Cortical reorganization and phantom pain may be maintained by peripheral inputs in some individuals and by central input in others. The primary somatosensory and motor cortex in patients who had unilateral arm amputations have shown extensive reorganization in those with phantom limb pain and minimal reorganization in those without phantom limb pain. Those without phantom limb pain maintain a cortical organization similar to the normal controls. These findings suggest somatosensory pain memory and altered homunculus structure in the somatosensory cortex to be a significant etiology of phantom limb pain. This also suggests that peripheral component may sustain this pain memory.


In patients with phantom limb pain, the sensory-discriminative dimension (localization) of the pain decreases while the cognitive-affective dimension (suffering) of the pain increases. There are no consistent personality defects reported in patients with phantom limb pain; nor does there appear to be an increased incidence of neurosis or other psychological disorders in amputees with phantom limb pain. In the non-amputee population with chronic pain, pain is magnified by stress, anxiety, and depression. No correlation has been found between acceptance of the impairment due to amputation and the frequency of phantom limb pain. It is possible that phantom limb pain is the misinterpretation of phantom limb sensation, however, to date no studies have significantly supported this etiology. Psychodynamic explanations have not been consistent with the cumulated physiological and psychological data in individuals with phantom limb pain.

Phantom limb pain appears to be a central neuropathic pain disorder maintained by peripheral input. The etiology of phantom limb pain is multi-faceted involving deafferentation, pain memory, and cortical reorganization.. It is more than likely a variable combination of the above etiologies in any single individual.


The intensity of phantom limb pain varies from an annoyance to totally devastating with the quality of pain variable and described as burning, cramping, gnawing, crushing, pulling, stabbing, or a wire like band. One quarter of the patients describe phantom limb pain in causalgia terms of burning, aching, and throbbing while another third describe this as an extremely abnormal position of the limb.


Though a large number of medical and surgical interventions for phantom limb pain have been reported, only a few studies represent randomized controlled trials. In the randomized controlled trials few treatments have proven to be effective. Recommended treatments for phantom limb pain suggest regimes similar to the management of neuropathic pain states. Unfortunately the trials for treatment of neuropathic pain have rarely included patients with phantom limb pain. Tricyclic antidepressants -- Tricyclic anti-depressants (Elavil [amitriptylene], Norpramin [desipramine]) have shown to be effective treatment of neuropathic pain. Unfortunately, newer antidepressants ( Prozac [fluoxetine], Paxil [paroxetine], and Zoloft [sertraline]) have shown little or no efficacy in the treatment of neuropathic pain. With tricyclic antidepressants the primary problem is their adverse side effect profile and must be used cautiously in patients with a history of cardiovascular disease, glaucoma, urinary retention, or autonomic neuropathy. Tricyclic antidepressants may also block the effects of certain antihypertensive medications and they interact with drugs metabolized by cytochrome P450. Anticonvulsants -- Anticonvulsants (Tegretol [carbamazipine], Neurontin [gabapentin], Keppra [levetiracetam]) have been shown to be effective in neuropathic pain by blocking nerve impulses in pain fibers. Gabapentin has been shown to be effective for chronic neuropathic pain in eight published double-blind placebo controlled randomized clinical trials including patients with phantom limb pain. Local anesthetics -- 5% lidocaine topical patches have been shown to be effective in two double-blind randomized studies for the treatment of neuropathic pain. Local anesthetics such as lidocaine and its oral congener mexiletine have also been reported to be effective in neuropathic pain. In an open label study mexiletine was reported to provide pain relief in 18 of 31 patients with phantom limb pain. NDMA receptor agonists -- Ketamine was shown to be effective in reducing wind-up, allodynia, and spontaneous pain in different neuropathic pain conditions including phantom limb. In a double-blind placebo controlled study, intravenous Ketamine was shown to reduce pain and hyperalgesia in 11 amputees.5 Opioids -- Opiate analgesics interact with opioid receptors in the spinal cord and inhibit central nocioceptive neurons. Opioid analgesics have been shown to be effective in peripheral neuropathy in five double-blind randomized trials in which they significantly relieved pain, disability and allodynia compared to placebo. Risks of substance abuse must always be considered in patients with neuropathic pain. There is a substantial risk in prescribing opioids to patients with a history of substance abuse and doing so requires very close monitoring. Longacting oral opioids (methadone or MS-Contin), rather than short-acting, are recommended as the drugs of choice for this type of pain control. Ultram [tramadol] is an analgesic with both monoaminergic and opioid activity which may prove to be an alternative to strong opioids as tolerance and dependence during long-term treatment with tramadol appears to be uncommon. Other medications -- A large number of other treatments to include beta-blockers, topical application of Zostrix [capsaicin] and various anesthetic blocks, have been claimed to be effective in phantom limb but none of them have proven to be effective in well controlled trials. Surgical intervention -- No surgical interventions for phantom limb pain, including cordotomy, thalamotomy, and sympathectomy, are an effective intervention in phantom limb pain. During a century of attempts, surgery provided short term relief at best. Physical modalities -- Physical modalities including transcutaneous electrical stimulation (TENS), vibration, acupuncture, hypnosis, biofeedback and electro convulsive therapy are rarely effective in the treatment of phantom limb pain. CNS electrical stimulation -- Dorsal column stimulation is a procedure in which electrodes are implanted surgically in the dura region of the spinal cord and has shown limited efficacy in patients with phantom limb pain. Stimulation of the deep brain structures and the motor cortex have been shown to reduce primary somatosensory cortex reorganization secondary to limb amputation for many years after initial limb amputation. The effect of the treatment appears to decrease over time and no controlled studies have been conducted to date. Electrical stimulation of the brain reveals that phantom limb pain will most likely respond to stimulation of the lateral and medial thalamic structures though initial success rates have been reported very high with deep brain stimulation (approximately 25%) of the patients have good long term pain relief. In spinal cord stimulation approximately 50% of the patients initially have good pain relief and 25% have long-term benefit. Psychological Treatment -- The use of psychotherapy, biofeedback, and relaxation training, as well as hypnosis, generally short lived results at time of follow-up. Electrical stimulation of the motor cortex has been shown to reduce phantom limb pain. Experiments that involve a "virtual reality box" then makes use of mirrors are used to trick the brain into thinking that the phantom limb is moving. The amputee looks into a mirror at the contralateral intact hand while it is positioned to coincide spatially with the felt position of the phantom limb. The individuals see their remaining limb mirror reversed to look like the amputated limb. This appears to recreate a coherent body image thus decrease pain as a result of reduced and disordered brain reorganization and input. This has been a relatively simple treatment that appears to have significant pain relieving benefits. This technique was developed by Dr. Ramachandran.6 Its repeated use diminishes the perception of the phantom limb itself which not only diminishes the phantom limb pain but with repeated use diminishes the perception of the phantom limb itself enabling the brain to accurately remap the change in the body. This method still requires randomized controlled treatment studies. On a clinical basis, this technique has been found to be effective in many patients. Prevention In studies using preemptive epidural analgesia to prevent phantom limb pain in patients with limb pain before surgery, 52% of the patients had phantom limb pain after surgery and 56% of the controls had phantom limb pain after surgery. No statistically significant differences were noted between the two groups for relief pain or improved function.


A detailed history and physical examination are needed to differentiate residual limb pain from phantom limb pain. The correct diagnosis is essential to treat the patient appropriately. The etiology of phantom limb pain remains allusive; hence strategies for the treatment of limb pain should include the safest, least invasive, and least expensive therapies. Pain relief and functional improvement are the primary goals


  1. Jensen TS, Krebs B, Nielson J, Phantom Limb, Phantom Pain, and Stump Pain in Amputees During the First Months Following Limb Amputation. Pain 1983, 17:243- 256.

  2. Jensen TS, Krebs B, Nielson J, Rasmussen P: Immediate and Long Term Phantom Pain in Amputees: Incidence - Clinical Characteristics in Relationship to Pre-amputation Pain. Pain 1985, 21:256-78.

  3. Nikolajsen L, Jensen TS. Phantom Limb Pain. British Journal of Anaesthesia 2001, 87:107-116, 2001.

  4. Kalauokalani DAK, Loeser JD: Phantom Limb Pain. In: Crombie IK: Epidemiology of Pain. IASP Press: Seattle. pp.143-53, 1999.

  5. Nikolajsen L, Hansen CL, Nielsen J, et. al: The Effect of Ketamine on Phantom Limb Pain: a Central Neuropathic Disorder Maintained by Peripheral Input. Pain 1996, 67:69- 77.

  6. Ramachandran VS, Rogers-Ramachandran D: Synaesthesia in Phantom Limbs included with Mirrors. Biological Sciences 1996, 263:377-86.


Dworkin RH, Backonja M, Rowbotham MC, Advances in Neuropathic Pain: Diagnosis, Mechanisms, and Treatment Recommendations. Arch Neurol 2003, 60: 1524-34.

Flor H: Phantom Limb Pain: Characteristics, Causes, and Treatment. Lancet Neurology 2002, 1:182-189.

Sherman RA: Postamputation Pain. In: Jensen TS, Wilson PR, Rice ASC (eds.): Clinical Pain Management: Chronic Pain. Arnold Press: London. pp. 239-49, 2003.