Postamputation pain is a complex condition that includes phantom limb pain (PLP), stump pain and residual limb pain (RLP), the latter of which may be referred from joints, the spine and inflamed bursa and tendons. PLP may have peripheral, spinal and central etiologies. The evidence of peripheral mechanisms includes the relief of both PLP and RLP during local anesthetic (LA) infusions, the relief of PLP and RLP with sympathetic blocks and neuroma injections, and the development of phantom radicular pain in amputees with a herniated disc.

Neurolysis and defunctionalization are long-lasting treatments for pain when LA blocks provide temporary benefit, being most commonly used for cancer pain (e.g., celiac plexus neurolysis). Neurolysis has also been used to treat PAP, with uncontrolled studies showing benefit for both RLP and PLP. However, there are no controlled studies demonstrating efficacy. In this small study, we will evaluate the effectiveness of alcohol neurolysis of lower extremity neuromas (femoral or saphenous; sciatic or common peroneal and/or tibial; obturator and/ or lateral femoral cutaneous when pain is in those distributions) in individuals with RLP and PLP.

For individuals with upper extremity amputation in whom non-selective neurolysis may affect the ability to use certain prosthetics that depend on functioning nerve and muscle signals, high-concentration capsaicin will be injected in an observational arm. The investigators will also examine factors associated with treatment outcome in a subset of patients (e.g., functional MRI, quantitative sensory testing).

Up to 130 patients with lower extremity amputations and PAP will be randomized by a computer-generated randomization table in a 1:1 ratio in blocks of 20 to receive peri-neuroma injections of either lidocaine 2% + 100% ethyl alcohol, or lidocaine 2% + saline, around the following neuromas: 1) Femoral, or saphenous nerve below the adductor canal; 2) sciatic, or common peroneal and/or posterior tibial beneath the popliteal fossa; 3) obturator (in above the knee amputees); and 4) lateral femoral cutaneous (in above the knee amputees). The painful neuromas to be treated will be determined by physical exam (e.g., Tinel's sign, pain reproduction during palpation or use of a prosthesis), and PLP patterns correlating with nerve distributions (e.g., a person with only foot PLP will not have the obturator or lateral femoral cutaneous neuromas injected; a patient who perceives phantom pain only in the top of their foot, or the lateral side of their ankle, may require only neurolysis of the common peroneal nerve or saphenous nerve, respectively). Those with bilateral lower extremity amputations who meet inclusion criteria for both limbs will be suballocated to have an alcohol injection on one side and a lidocaine injection on the other, in random order (estimated 10-20 patients). The side that receives local anesthetic alone and the side that receives local anesthetic and alcohol will be determined by a computer-generated random number table.

The location of the painful neuromas will be identified by physical exam and confirmed via either ultrasound or electrical stimulation (e.g., using a radiofrequency machine or nerve stimulator, with concordant stimulation in the painful area(s) ideally noted at < 0.5 volts). Patients with unilateral lower extremity amputations who meet selection criteria will be allocated via a computer-generated randomization table in blocks of 20 to receive either: 1) an injection of 2 mL lidocaine 2% at each painful neuroma over 5 minutes, followed by 1.5 mL saline within 5 minutes; or 2) 2 mL lidocaine 2% at each painful neuroma over 5 minutes followed by 1.25-3.5 mL 98-100% dehydrated ethyl alcohol (the volume depends on the voltage threshold, i.e., thresholds > 0.5 mL may warrant the 23.5 mL higher volume ). For those suballocated with bilateral lower extremity amputations, both painful sides will receive an injection of 2 mL lidocaine 2% per neuroma site over 5 minutes. Then after approximately 5 minutes, the side allocated to receive alcohol with have that side injected with 1.25-3.5 mL of 98-100% alcohol while the other side will receive 1.5 mL of normal saline; the lidocaine is given first because the alcohol can burn when injected, and normal saline has been shown to provide some therapeutic effect by washing out inflammatory cytokines and breaking up adhesive scar tissue, both of which may mediate neuroma-related pain. The injections will be performed with 20-22-gauge needles or stimulating needles (when a nerve stimulator is used), depending on the means for neuroma location.

After 6 weeks (primary endpoint) in the randomized double-blind portion, those with a negative categorical outcome (< 30% pain relief or < 4/7 on the Patient Global Impression of Change (PGIC) scale) will be unblinded to receive alternative treatments. The next follow-up for those with a successful 6-week outcome will be 12 weeks. For those with a successful 12-week outcome, the final follow-up will occur at 6 months.

Inclusion Criteria:

• 1. Age >/= 18 years 2. At least 1 lower extremity amputation 3. Pain duration >/= 1 month 4. Either average RLP or PLP in one or both (for those who have 2 lower limbs enrolled) amputated extremities >/=4/10 5. Stable analgesic regimen over the past 10 days 6. Failure of physical therapy and at least 2 pharmacological treatments 7. At least 1 suspected painful neuroma, identified by Tinel's sign or pain with pressure or prosthetic use, referred pain in the distribution of the severed nerve, and neuropathic-type symptoms (tingling, shooting or lancinating pain)

Exclusion Criteria:

• 1. Very poorly controlled psychiatric condition (e.g., PCL-5 score > 60, > 15 on the anxiety and/or depression section of HADS) 2. Poorly controlled medical condition that would preclude participation (e.g., heart failure, uncontrolled diabetes) 3. Patients in whom targeted muscle reinnervation or a similar procedure is being considered 4. Systemic infection or infection overlying the stump 5. Clinically-relevant injury to nerve fibers proximal to the amputation 6. Pregnancy

Contingent on case-by-case approval by Ministry of Health

Lidocaine before alcohol neurolysis (which may be painful) will be given to facilitate participant masking