Targeted Muscle Reinnervation (TMR) & Regenerative Peripheral Nerve Interface (RPNI)

When a peripheral nerve is transected (as in amputation or nerve injury) and the regenerating axons have no distal target to reinnervate, they sprout into a disorganised tangle — a neuroma — a hypersensitive free nerve ending that can cause debilitating residual-limb (stump) pain and neuropathic symptoms. Amputees may additionally experience phantom limb pain (pain perceived in the missing part). Together these are a major cause of pain, poor prosthesis use and disability after amputation, and no traditional technique has given consistent relief.

Traction neurectomy, burying the nerve in muscle or bone, capping and simple excision all leave the nerve with no distal target, so the axons sprout into a fresh neuroma — high recurrence, inconsistent relief. Burying a nerve in muscle is not the same as TMR: there is no motor-nerve coaptation and no reinnervation target, just a deeper neuroma. This distinction is a classic viva trap.

In TMR, the cut major (mixed) nerve is coapted (transferred) to a small, expendable motor nerve that supplies a nearby muscle (which is thereby reinnervated). This gives the regenerating axons a target to reinnervate — "somewhere to go and something to do" — so they do not form a neuroma. TMR was originally developed to create additional myoelectric control sites for advanced prostheses (each reinnervated muscle segment becomes a new EMG signal source), and was then found to prevent and treat neuroma and phantom limb pain as well.

In RPNI, the transected nerve end is implanted into an autologous free skeletal muscle graft — a small, devascularised strip of muscle that revascularises and is reinnervated by the implanted axons. The graft provides the axons with a target (preventing neuroma) and acts as a biological signal transducer that amplifies neural signals for prosthetic control. Because it uses a small free muscle graft rather than a nerve transfer, RPNI does not require an expendable motor nerve, is technically simpler in some respects, and can be applied to multiple individual nerves or fascicles — each fascicle wrapped in its own graft.

TMR = nerve transferred to a muscle's motor nerve. RPNI = nerve buried in a free muscle graft. Both give the axons a target (so no neuroma) and both generate myoelectric signals for prostheses. TMR needs a spare motor nerve; RPNI does not, and scales to many nerve ends.

Both are microsurgical procedures done under loupe or microscope magnification, and both end by burying the construct in a deep, padded, non-weight-bearing bed away from the scar:

• TMR: identify the cut major (mixed) nerve and a nearby expendable motor nerve to a recipient muscle; trim both back to healthy fascicles; perform a tension-free epineurial coaptation, accepting the large-donor-to-small-recipient size mismatch because only a target is needed, not a one-to-one repair.
• RPNI: harvest a small free skeletal muscle graft (from a locally expendable muscle), then implant or wrap the nerve end - or each individual fascicle - into the graft; the initially avascular graft revascularises and is reinnervated by the implanted axons. Because each fascicle gets its own graft, RPNI scales to multiple nerve ends without needing a donor motor nerve. The two are not mutually exclusive - large mixed nerves can be managed with TMR and smaller sensory branches or spare fascicles with RPNI in the same limb.

A point examiners increasingly probe: TMR provides a motor signal and RPNI a graft target, but neither restores the natural proprioceptive feedback (the muscle-spindle and Golgi-tendon length-tension sense) that is lost at amputation. The agonist-antagonist myoneural interface (AMI) addresses this by surgically linking an agonist and antagonist muscle pair - often within an RPNI-type construct - so that contraction of one stretches the other, recreating the physiological length-tension feedback loop. This gives the patient a sense of joint position and finer, more intuitive prosthetic control, and is the emerging extension beyond TMR and RPNI - useful to name as "what comes next" once you have covered the two core techniques.

A systematic review of TMR and RPNI for pain (17 studies; 14 TMR / 366 patients, 3 RPNI / 75 patients; one randomised controlled trial) found that, for treatment of established pain, the techniques improved neuroma pain in about 75–100% of patients and phantom limb pain in about 45–80%, averaging a 2.4–6.2-point reduction on the numeric rating scale. Used prophylactically at amputation, many patients reported no neuroma pain (48–100%) or no phantom limb pain (45–87%) at follow-up, and PROMIS patient-reported outcomes consistently improved. Complication rates ranged 13–31%, most commonly delayed wound healing. The evidence base is still maturing (mostly cohort studies with limited randomised data and relatively short follow-up), so longer-term randomised comparisons with traditional management are awaited.

TMR and RPNI are not first-line for every residual ache - they sit at the surgical end of a multimodal pathway that an examiner expects you to outline. Post-amputation neuroma and phantom pain are first managed with neuropathic pharmacotherapy (gabapentinoids - gabapentin/pregabalin; the tricyclic amitriptyline; the SNRI duloxetine), physical and central strategies for phantom pain (mirror therapy and graded motor imagery, desensitisation, TENS), good perioperative analgesia and meticulous surgical technique, and prosthetic/socket optimisation. TMR or RPNI is the step for refractory neuroma/phantom pain - or for prophylaxis at the time of amputation - once these measures are in place or have failed. Framing surgery within this ladder (rather than as a stand-alone fix) is what gains marks.

• Neuroma = a cut nerve with no target → disorganised painful bulb; phantom limb pain often coexists.
• Traditional methods fail (traction neurectomy, bury, cap) — no target, high recurrence.
• TMR = nerve transferred to a muscle's motor nerve; RPNI = nerve into a free muscle graft.
• Both also generate myoelectric signals for intuitive prosthesis control (TMR's original purpose).
• Best prophylactically at amputation, but both also treat established neuroma/phantom pain.
• Main complication: delayed wound healing (overall 13–31%); evidence is encouraging but still maturing.

The head-to-head comparison with traditional burying comes from the Dumanian RCT (Annals of Surgery, DOI), the pooled outcome and complication figures from the Mauch systematic review (DOI), and the RPNI rationale/technique from the Ganesh Kumar & Kung Hand Clinics review (DOI). The neuroma mechanism and the TMR/RPNI principles are well-established. (See also our Foot & Ankle Amputation Levels topic for amputation principles.)