Subscription is FREE for qualified healthcare professionals in the US.
14 Articles in Volume 19, Issue #5
Agonism and Antagonism of the Muscles of the Shoulder Joint: An SEMG Approach
Analgesics of the Future: The Potential of IV Formulations for Post-Op Treatment of Pain
Blood Biomarkers Show Promise for Precision Pain Management
Can I Call Myself a “Pain Specialist?”
Cases in Urine Drug Monitoring Interpretation: How to Stay in Control (Part 2)
Fear-Avoidance and Chronic Pain: Helping Patients Stuck in the Mouse Trap
How to Avoid Patient Alienation When Discussing Stress
Managing Phantom Limb Pain with Medication
Nerve Blocks Lead to Improved Quality of Life
Sacroiliac Joint Dysfunction: New Methods in Evaluation and Management
SCS Therapy in a Patient with Advanced Bilateral Kienbocks
Thoracic Epidural Abscess with Cord Compression Following a High-Frequency SCS Trial
What is the evidence to support clonidine as an adjuvant analgesic?
What’s In A Name? In This Case, That Which We Call Addiction Is Not Dependence

Managing Phantom Limb Pain with Medication

Pharmacologic options for the treatment of this multifactorial, neuropathic condition, including gabapentin, TCAs, opioids, ketamine, memantine, lidocaine, and bupivacaine.
Pages 54-57

Phantom limb pain (PLP) is described as pain present in the part of a limb that has been removed. Affecting 60 to 80% of patients following amputation, PLP is characterized by burning, tingling, warmth, aching, cramping, shooting, sharp, stabbing, and constriction in the missing portion of the limb. Most amputees experience the onset of PLP within a week following amputation, but it may present intermittently throughout life. Risk factors for developing PLP include female sex, pre-amputation pain, and depression. The condition may negatively affect a patient’s well-being, functioning, activity, employment, and overall quality of life.1-4

The etiology and pathophysiology of PLP are not well understood, but the condition is thought to be multifactorial and to involve both the peripheral and central nervous systems.2,5 This proposed pathophysiology has led to treatment of PLP with those drugs commonly used for neuropathic pain, such as antidepressants, opioids, anticonvulsants, N-methyl-D-aspartate (NMDA) receptor antagonists, and anesthetics. However, there is insufficient evidence for pharmacologic interventions that effectively reduce PLP after amputation.1-5 In addition, there are no specific guidelines for the management of PLP; however, the US Department of Veterans Affairs and Department of Defense have issued guidelines on the rehabilitation of lower limb amputation which include a segment on pain management. Rather than supporting one pain control method over another, the guidelines recommend a multimodal approach involving both pharmacologic and nonpharmacologic therapies.3

Given that most amputations in the United States are performed due to vascular disease,1,3 it is important for healthcare providers to become aware of various treatment options for PLP and their efficacy. It is also worth pointing out that nonpharmacologic measures have been studied in the management of phantom limb pain, including transcutaneous electrical nerve stimulation (TENS), mirror therapy, deep brain stimulation, acupuncture, relaxation, and biofeedback.3-5 In this review, however, we will focus on the numerous pharmacologic therapies used in the management of phantom limb pain.

The lack of clinical guidelines for phantom limb pain makes it difficult to prioritize treatments. (Source: Shutterstock)

Anticonvulsants: Gabapentin

The efficacy of the anticonvulsant gabapentin in the treatment of PLP was studied in a double-blind, placebo-controlled crossover study consisting of 19 patients. Subjects received up to 2400 mg of gabapentin daily during a gabapentin phase.6 The primary endpoint was pain intensity difference (PID) compared with baseline at the end of each treatment. Although the mean ± SD PID was significantly greater in the gabapentin group (3.2 ± 2.1 versus 1.6 ± 0.7) during Week 6, there was no impact on activities of daily living. Participants reported few adverse events, such as dizziness, headache, nausea, and somnolence. Somnolence occurred more frequently in the gabapentin arm compared to the placebo arm, but the results were not statistically significant.7

Another placebo-controlled, double-blind crossover trial involving 24 patients and lasting 5 weeks, found no difference in pain intensity (primary outcome) between placebo and treatment with up to 3600 mg of gabapentin, mean numerical rating scale ± SD (3.60 ± 2.67 versus 3.43 ± 2.54). More than half of the participants reported a meaningful pain decrease during the gabapentin phase compared with about one-fifth who reported a meaningful decrease in pain during the placebo phase (P < 0.05).8 A separate study (n = 41) found no reduction in incidence or intensity of post-amputation pain between placebo and patients treated with 2400 mg of gabapentin (titrated) in the first 30 days post-surgery. Primary outcome measures included the rates and intensities of phantom pain at the end of the 30-day treatment period and after 6 months. The incidence of phantom pain was 55% versus 52.6% (risk difference, 2.4%; 95% CI-28.9 to 33.7%; P = 0.88; 30 days) and 58.8% versus 50% (risk difference 8.8%; 95% CI-23.3 to 40.9%; P = 0.59; 6 months), for gabapentin versus placebo, respectively. The median intensities of PLP were similar between groups at 30 days and 6 months, ranging from 0.5 to 1.5 on a numeric rating scale. Additionally, there were no significant differences in opioid consumption between the groups. Side effects included nausea, fatigue, and ataxia.9

Most of the studies evaluating gabapentin are limited by small sample size and short study duration; these points, along with mixed results, make it difficult to draw definitive conclusions regarding the efficacy of this medication for phantom limb treatment.1 Other anticonvulsants, such as carbamazepine, oxcarbazepine, and pregabalin may have a role in PLP management; however, there is limited available evidence to support their use.5

NMDA Receptor Antagonists: Ketamine and Memantine

In a double-blind, crossover study assessing 11 patients with PLP, ketamine was infused at a bolus dose of 0.1 mg/kg/5 min followed by an infusion of 7 ug/kg/min. The medication was shown to significantly increase pressure pain thresholds in all patients.10 More studies are needed to establish efficacy of ketamine in PLP; therefore, it should be used with caution considering its potentially significant adverse effects, such as delirium, hallucinations, cystitis, cardiovascular effects, and diplopia. The risk of ketamine-induced adverse effects for PLP may be amplified by the higher dosing required compared to lower dosing for other conditions.11

Memantine has been investigated in PLP as well, producing mixed results. In a case report involving two patients with severe PLP that was refractory to opioids, antidepressants and gabapentin, initiation of memantine at a daily dose of 20 mg resulted in marked reduction in PLP. Both patients were weaned off of opioids and had no significant complaints of pain in the follow-up period.12 Limitations of the study included that causality could not be established and that memantine was administered concurrently with other adjunctive medications (ie, NSAIDs, gabapentin, and antidepressants). In contrast, memantine 30 mg daily had no effect on chronic PLP in a placebo-controlled crossover trial including eight patients. There were no significant differences in intensity rating of PLP. Side effects, in this case, included headache, nausea, and vertigo, with no significant difference between treatments.13 Again, conclusions were limited by the small sample and study designs, conflicting evidence, and potentially serious adverse effects for the use of NMDA receptor antagonists for phantom limb pain.

Antidepressants: TCAs and SNRIs

Tricyclic antidepressants (TCAs) are commonly prescribed off-label to treat neuropathic pain.14 A randomized controlled study of amitriptyline titrated up to 125 mg/day given for six weeks in 39 participants found no significant difference in average PLP intensity when compared to placebo, benztropine mesylate. At Week 6, the average PLP intensity was 3.1 for both arms on a 0-10 numeric rating scale. Common side effects included dry mouth, drowsiness, constipation, and dizziness, with dry mouth being significant and severe with amitriptyline.15

In contrast, a controlled trial of amputees treated with tramadol (n = 33), placebo (n = 31), and amitriptyline (n = 30, titrated up to 75 mg daily) revealed significant efficacy of amitriptyline in reducing phantom pain intensity. After initial titration phase, the number of patients who had a decrease of at least 10 mm in phantom pain visual analog scale (VAS) compared with baseline was 25, 22, and 2 for the amitriptyline, tramadol, and placebo groups, respectively. Following the titration phase, nonresponders were switched to receive an alternative analgesic. When nonresponders from the tramadol group (n = 11) were switched to amitriptyline group, 8 out of 11 patients had phantom scores decrease to 0 at the end of one month. Common side effects in the amitriptyline group included tiredness, headache, dizziness, constipation, and nausea.16 The conflicting results of these trials, small sample size, and the negative safety profile of TCAs may limit their use in phantom limb pain.

Serotonin-norepinephrine reuptake inhibitors (SNRIs) have demonstrated efficacy for the treatment of several chronic pain conditions. In a case series of four patients with PLP, duloxetine 30 mg daily reduced pain intensity as measured by VAS. In three cases, pain intensity decreased by ≥5 points on VAS over ≥ 1 month, and by 3 points in one patient 15 days after initiation of duloxetine, who was later lost to follow-up.17 Similarly, in a case series of three patients with PLP, treatment with milnacipran (dosed at 25 to 75 mg daily for varying durations) resulted in reduced pain severity, with pain scores eventually decreased to ≤ 1 on VAS.18 Duloxetine and milnacipran may be preferred over TCAs for PLP due to less side effects. However, the low methodological quality of case series implies that no clinical decisions can be made based on these study results alone.


Multiple studies have shown effectiveness of opioids in the treatment of phantom limb pain. In a double-blind, placebo-controlled study of crossover design involving 12 patients, treatment with extended oral morphine at doses of 70 to 300 mg daily resulted in reduced PLP intensity. During the morphine phase, 42% of patients were classified as responders and showed PLP reduction of more than 50% compared to the placebo phase where only one patient showed PLP reduction of more than 50% (8%; t(11) = 2.60, P < 0.05).19

In another, double-blind, placebo-controlled crossover trial (n = 60) comparing morphine (mean dose 112 mg) to mexiletine (mean dose 933 mg) in patients with post-amputation pain of six months or longer, morphine, but not mexiletine, led to decreased pain intensity. The mean percent pain relief during treatment with placebo, mexiletine, and morphine was 19%, 30%, and 53%, respectively (P < 0.0001, morphine vs. placebo and mexiletine), and morphine treatment provided lower pain scores compared with placebo and mexiletine (P = 0.0003) during the post hoc analysis. However, morphine was associated with higher rates of side effects (ie, constipation, drowsiness, and nausea), and no improvement in functional activity compared to mexiletine and placebo groups. Stump and phantom pains were not analyzed separately; hence, a differential analgesic response of stump and phantom pains to morphine or mexiletine might not have been detected by the study.20 Although opioids provide analgesia in PLP patients, their long-term use may be limited by the high risk of side effects, tolerance, and psychological dependence.

Anesthetics: Lidocaine and Bupivacaine

The efficacy of anesthetics such as lidocaine and bupivacaine for PLP have been investigated as well.21,22 In one study, 32 participants were randomized to one of six possible combinations of intravenous placebo, lidocaine, and morphine. For three days participants received one of three solutions (bolus dose of 1 mg/kg of lidocaine, 0.05 mg/kg of morphine, or 10 mg of diphenhydramine) over 2 min, followed by an infusion (4 mg/kg of lidocaine, 0.2 mg/kg of morphine, or 40 mg of diphenhydramine) over 40 minutes. The maximum infusion dose was 400 mg of lidocaine and 25 mg of morphine, and the interval between the infusions was at least 24 hours to minimize carry-over effects. Morphine reduced stump and phantom pains significantly compared with placebo. However, lidocaine decreased stump (P < 0.01), but not phantom pain. Self-reported stump pain relief on 0-100 VAS, was significantly greater for lidocaine compared with placebo, 32.8 ± 33.6 versus 8.2 ± 15.9, respectively. Self-reported stump and phantom pain relief were significantly greater for morphine compared to placebo, 44.8 ± 35.4 versus 8.2 ± 15.9, and 47.9 ± 38.2 versus 3.2 ± 10.1, respectively.21

A randomized, double-blind crossover study evaluating the efficacy of contralateral injections of bupivacaine showed promising analgesic effect on PLP. Either 1-mL injections of bupivacaine 0.25% or 0.9% saline were administered alternatively in each point with 72 hours between injections, in eight lower limb amputees with PLP in the past six months. Sixty minutes following injection, phantom limb pain was significantly reduced in patients treated with bupivacaine compared to placebo, with the mean (SD) intensity of PLP of 2.6 (1.2) for bupivacaine treatment versus 6.1 (1.6) for placebo, on 0 – 10 VAS.22 The small sample size and short follow-up period of these studies warrant further investigations to validate the use of anesthetics in phantom limb pain.


Management of PLP is complex and the pathophysiology of the disease is not well understood. Pharmacologic therapies that have been studied so far have limitations, including inconsistent pain relief, intolerable and concerning adverse effects, and lack of improvement in daily function to justify the expense in many patients. Additionally, most studies are limited by small sample size, heterogeneous populations, low-quality methodological design, short follow-up, and conflicting results.

For drug classes such as opioids, which have concerning safety profile, balance of benefits and risks should be considered. The chosen drug regimen should be individualized, and a multimodal approach combining pharmacologic and nonpharmacologic options remains a standard of care for PLP. The lack of clinical guidelines in the management of PLP makes it difficult to prioritize first-line treatments, such as nonpharmacologic treatments versus pharmacologic options. However, nonpharmacologic options, especially TENS, virtual reality, and mirror therapy, which have shown promising results, may have an advantage over their pharmacologic counterparts.23 Controlled studies with larger sample sizes and longer follow-up, are needed to determine the most effective pharmacologic therapies for phantom limb pain.

Last updated on: August 2, 2019
Continue Reading:
Current Treatments for Phantom Limb Pain
close X