Residency · Residency · Chronic Pain Management

Phantom Limb Pain

Introduction

Phantom limb pain (PLP) is a neuropathic pain condition in which patients experience painful sensations that they perceive as originating from an amputated or deafferented limb. PLP affects 50-80% of amputees, with onset typically within the first week following amputation. It is distinct from residual limb (stump) pain and non-painful phantom sensations. Despite decades of research, PLP remains one of the most challenging chronic pain conditions to treat, and effective management requires a multimodal approach that addresses peripheral, spinal, and supraspinal mechanisms simultaneously.

Terminology

Four terms must be clearly distinguished. Phantom limb pain refers to painful sensations perceived in the missing limb — burning, cramping, shooting, stabbing, or squeezing. Phantom limb sensation is a non-painful awareness of the missing limb, including its shape, position, movement, temperature, or itching. Residual limb (stump) pain is pain localized to the remaining portion of the amputated limb, often caused by neuroma formation, infection, ischemia, or prosthetic issues. Telescoping describes the perception that the phantom limb progressively shortens over time, with the distal portion (hand or foot) gradually approaching the stump.

Peripheral Mechanisms

At the peripheral level, severed axons form disorganized neuromas at the stump. These neuromas contain upregulated sodium channels (Nav1.3, Nav1.7, Nav1.8) and alpha-adrenergic receptors that generate spontaneous ectopic discharges. Ectopic firing — both spontaneous and mechanically evoked action potentials — arises from the neuroma and proximal nerve segments. In the dorsal root ganglion, neurons corresponding to the amputated limb develop increased excitability, spontaneous firing, and cross-excitation between adjacent neurons. Sympatho-afferent coupling, in which sympathetic efferents sprout around DRG neurons and neuroma sites, links sympathetic nervous system activity to nociceptive signaling and explains why phantom pain often worsens with stress and temperature changes.

Spinal Mechanisms

Loss of peripheral afferent input triggers dorsal horn hyperexcitability through NMDA receptor activation, wind-up, and loss of inhibitory interneuron function — the hallmarks of central sensitization. Deafferented dorsal horn neurons develop expanded receptive fields and begin responding to input from adjacent intact dermatomes, a form of spinal cord reorganization. Disinhibition occurs because loss of C-fiber input reduces the tonic inhibition normally mediated by GABAergic and glycinergic interneurons in the superficial dorsal horn. Glial activation compounds the problem: microglia and astrocytes in the dorsal horn are activated following amputation and release pro-inflammatory cytokines (TNF-alpha, IL-1beta) and brain-derived neurotrophic factor (BDNF) that maintain the sensitized state.

Supraspinal Mechanisms

The most extensively studied supraspinal mechanism is cortical reorganization. After amputation, the somatosensory cortical representation of the amputated limb is invaded by adjacent cortical areas — for example, the face or upper arm representation may expand into the hand area after upper extremity amputation. The landmark work by Flor and colleagues demonstrated that the degree of this cortical reorganization correlates with PLP intensity. The Ramachandran hypothesis proposes that incongruence between motor intention and sensory feedback — where the motor cortex generates movement commands to the phantom limb but receives no proprioceptive or visual confirmation — creates a sensorimotor conflict that may itself generate pain. At the thalamic level, altered firing patterns in the ventral posterior lateral (VPL) thalamus, particularly increased burst firing, correlate with phantom pain. Underlying all of these changes is a fundamental disruption of body schema: the brain's internal representation of the body fails to update following amputation, maintaining a neural representation of the missing limb.

<image>Three-level illustration of phantom limb pain mechanisms: Top panel shows the supraspinal level with a somatosensory cortex homunculus demonstrating invasion of the hand representation zone by the face and upper arm areas following upper extremity amputation, with fMRI activation maps; Middle panel shows the spinal level with dorsal horn central sensitization, NMDA receptor activation, microglial activation, and loss of inhibitory interneuron input; Bottom panel shows the peripheral level with neuroma formation at the stump, ectopic sodium channel upregulation, and sympathetic fiber sprouting around the neuroma.</image>

Risk Factors

Pre-amputation pain is the strongest predictor of PLP — patients with severe limb pain before amputation are significantly more likely to develop phantom pain afterward. Proximal amputation levels (above-knee, above-elbow) carry higher risk than distal amputations. Bilateral amputation increases risk compared to unilateral. Traumatic amputations may carry higher risk than surgical ones due to the acute neural injury involved. Psychological factors — anxiety, depression, catastrophizing, and poor coping — are associated with greater PLP severity.

Mirror Therapy

Mirror therapy, introduced by V.S. Ramachandran, is the most studied non-pharmacological intervention for PLP. A mirror is placed vertically in the patient's midsagittal plane, and the patient views the reflection of the intact limb superimposed over the phantom limb space. Performing movements with the intact limb creates the visual illusion of normal movement in the phantom limb. The proposed mechanism is resolution of the sensorimotor conflict by providing visual feedback that matches motor intention, promoting cortical reorganization and normalization of the body schema.

Multiple randomized controlled trials demonstrate significant pain reduction (NRS reduction of 2-3 points), and systematic reviews support its efficacy. The standard protocol involves 15-30 minutes of mirror therapy daily for 4-8 weeks, with movements such as opening and closing the hand, wrist flexion and extension, and finger individuation. Limitations include reduced effectiveness for lower extremity PLP, the requirement for an intact contralateral limb, and the fact that not all patients respond.

<image>Clinical setup illustration of mirror therapy for a patient with left upper extremity amputation: the patient is seated at a table with a vertical mirror placed in the midsagittal plane. The intact right hand is positioned in front of the mirror, and the patient views the mirror reflection, perceiving it as the missing left hand. The illustration shows the patient performing finger extension exercises with the right hand while viewing the reflection, with visual feedback pathways shown from the mirror to the visual cortex and then to the motor and somatosensory cortices, promoting cortical reorganization.</image>

Pharmacotherapy

First-Line Agents

Gabapentin (900-3600 mg/day) is the most studied pharmacological agent for PLP, with RCT evidence showing modest efficacy through alpha-2-delta calcium channel modulation that reduces central sensitization. Pregabalin (150-600 mg/day) has a similar mechanism and strong neuropathic pain evidence, though fewer studies specifically address PLP. Tricyclic antidepressants (amitriptyline, nortriptyline at 25-150 mg/day) have limited RCT evidence specifically for PLP but draw on a strong neuropathic pain evidence base and offer additional benefit for comorbid depression and insomnia.

Second-Line and Adjunctive Agents

Tramadol (200-400 mg/day) has RCT evidence in PLP through its dual mu-opioid and norepinephrine reuptake inhibitor mechanism. NMDA receptor antagonists — ketamine given as intravenous subanesthetic infusions or intranasally, and memantine — address central sensitization directly. Ketamine infusion protocols at 0.1-0.5 mg/kg/hour for 4-24 hours show efficacy in refractory cases. Botulinum toxin type A can be injected into the residual limb stump for neuroma-related PLP. Beta-blockers such as propranolol may reduce the sympathetically mediated component. Calcitonin, administered intravenously or intranasally, has some evidence for acute PLP in the early post-amputation period.

Perioperative Prevention

Pre-emptive analgesia using epidural analgesia or a continuous peripheral nerve catheter initiated before amputation and maintained for 48-72 hours postoperatively may reduce PLP incidence. The evidence is mixed but the approach is favored in practice. Optimized perioperative regional anesthesia with perineural catheters providing continuous local anesthetic infusion is recommended. Perioperative intravenous ketamine during and after surgery may reduce PLP by blocking NMDA receptor-mediated central sensitization.

Neuromodulation

Spinal Cord Stimulation (SCS)

Epidural electrodes are placed targeting the dorsal columns at the level corresponding to the amputated limb. Conventional tonic stimulation produces paresthesias perceived in the phantom limb, and the presence of phantom paresthesia coverage correlates with analgesic efficacy. Programming can be challenging because the cortical representation of the missing limb may shift. Case series and retrospective studies report approximately 50% responder rates, though no large RCTs exist.

Dorsal Root Ganglion (DRG) Stimulation

Targeted stimulation of the DRG at levels corresponding to the amputated limb territory may provide more focused stimulation than traditional SCS, with the advantages of less susceptibility to postural changes and greater dermatomal specificity. A growing case series literature supports its efficacy in PLP.

Peripheral Nerve Stimulation (PNS)

Electrodes are placed along the residual peripheral nerve proximal to the neuroma, using either percutaneous or implanted systems. This approach may directly address peripheral ectopic firing.

Transcranial Magnetic Stimulation (TMS) and Transcranial Direct Current Stimulation (tDCS)

These non-invasive neuromodulation approaches target the motor cortex contralateral to the amputated limb. Repetitive TMS involves multiple sessions (5-10) of high-frequency stimulation over M1, with modest and variable efficacy. Anodal tDCS over M1 is low-cost and safe but has limited evidence specifically for PLP.

Emerging Virtual Reality Approaches

Virtual reality-based phantom motor execution uses motion sensors on the residual limb to decode myoelectric signals from residual muscles and control a virtual limb in real time. Augmented reality systems project a virtual limb onto the stump using camera and screen technology, allowing patients to control virtual limb movements through muscle signals. Immersive VR environments let patients see and interact with a virtual body that includes the missing limb, and embodiment of the virtual limb may correct body schema and reduce maladaptive cortical reorganization. Ortiz-Catalan and colleagues demonstrated significant PLP reduction using myoelectric pattern recognition-based AR/VR in patients refractory to conventional treatments, with sustained benefits at six-month follow-up. These approaches offer several advantages over mirror therapy: they are applicable to bilateral amputees, allow more complex and engaging motor tasks, improve compliance through gamification, and work for lower extremity amputees.

<image>Illustration of a virtual reality phantom limb pain treatment session showing a patient with a below-elbow amputation wearing a VR headset and EMG sensors on the residual limb. The VR display (shown as an inset) depicts a first-person view of two virtual hands performing a reaching and grasping task, with the virtual left hand controlled by the patient's residual limb muscle signals. Arrows trace the pathway from residual limb EMG signals through signal processing and machine learning decoding to real-time virtual hand movement, with feedback loops to visual cortex and motor cortex illustrated.</image>

Clinical Pearls

Pre-amputation pain is the strongest predictor of PLP, so pain control should be optimized before elective amputation with regional anesthesia techniques and multimodal analgesia. Distinguishing PLP from residual limb pain is essential because their treatments differ substantially — neuroma-related stump pain may respond to targeted interventions such as nerve blocks, neuroma excision, or targeted muscle reinnervation. Mirror therapy should be offered to all PLP patients as a first-line non-pharmacological intervention because it is safe, low-cost, and effective, and early initiation may prevent chronic PLP from becoming entrenched. Targeted muscle reinnervation performed at the time of amputation can reduce neuroma formation and PLP incidence, and surgical colleagues should be consulted about this technique for planned amputations. Cortical reorganization is potentially reversible with effective treatment — therapies that restore normal cortical maps, including mirror therapy, VR, and effective prosthesis use, may provide lasting benefit. PLP management should always be multimodal, combining pharmacotherapy, rehabilitation, psychological support, and consideration of neuromodulation for refractory cases.

References

1. Flor H, Nikolajsen L, Jensen TS. Phantom limb pain: a case of maladaptive CNS plasticity? Nat Rev Neurosci. 2006;7(11):873-881.
2. Ortiz-Catalan M, Guethmundsdottir RA, Kristoffersen MB, et al. Phantom motor execution facilitated by machine learning and augmented reality as treatment for phantom limb pain. Lancet. 2016;388(10062):2885-2894.
3. Karanikolas M, Aretha D, Tsolakis I, et al. Optimized perioperative analgesia reduces chronic phantom limb pain intensity, prevalence, and frequency. Anesthesiology. 2011;114(5):1144-1154.
4. Alviar MJM, Hale T, Dungca M. Pharmacologic interventions for treating phantom limb pain. Cochrane Database Syst Rev. 2016;10:CD006380.

Read this lecture as Markdown