While the pain benefits of TMR are largely independent of prosthesis use, the prosthetic control benefits of TMR are clearly dependent on the use of a myoelectric prosthesis. By expediting donning and doffing, enabling increased range of motion, and eliminating socket-related discomfort, osseointegration of upper extremity prostheses has proven to have a synergistic benefit when combined with TMR.¹⁵ By decoupling prosthetic suspension from myoelectric control, osseointegration offers greater signal fidelity than can be achieved with a conventional liner and socket construct that must house the surface electrodes necessary for myoelectric control while also providing suspension and spatial control of the prosthetic. It is not uncommon for relative motion between the socket and residual limb to be perceived as electromyographic activity that prompts errant activity in the prosthesis. While the surface control band used in conjunction with osseointegration may be subject to a similar phenomenon, it is often less frequent. Optimal signal fidelity can be achieved through implantation of electrodes directly into the native and reinnervated residual limb muscles. Osseointegration offers a conduit for passage of the wired components via the percutaneous abutment, as shown in Figure 1. The e-OPRA (Osseointegrated Prostheses for the Rehabilitation of Amputees) system has been reported to provide improved prosthetic control and stable long-term outcomes in a small group of patients.¹⁵ As the safety of osseointegration becomes more clearly defined and the cost of the technology becomes less prohibitive, it is likely to be more broadly used. Because osseointegration techniques involve resection of muscle distal to the residual skeleton, it is important that the TMR nerve transfer patterns anticipate potential conversion to osseointegration and avoid use of distal targets or mobilize distal muscle targets proximally. In the setting of prior TMR at the transhumeral level, the reinnervated brachialis muscle has been successfully transferred based on its donor ulnar nerve. Alternatively, the coracobrachialis muscle has served as a useful target for revision TMR in the setting of osseointegration.

Sensory feedback is a critical, and yet inadequately addressed, component of prosthetic control. Sensory feedback is a fundamental component of safe and efficient ambulation and is cited by upper limb amputees as one of the major gaps in the capabilities offered by commercially available prosthetics.^16,17,18 Tremendous effort and investigational funds have been expended in an effort to devise strategies capable of providing meaningful touch perception. Targeted sensory reinnervation offered a promising strategy to restore sensation.¹⁹ Stimulation of the reinnervated skin results in the sensation of the patient’s hand being touched, thus providing cortical feedback to the hand representation on the cortical homunculus. There is restoration of all modalities of cutaneous sensation, including pressure, vibration, and thermal sense. These patients gained an ability to discriminate gradations of force that matched that of their uninjured skin. However, the sensory reinnervation maps generated by targeted sensory reinnervation have proven less predictable than expected and commercially available prostheses are still unable to capitalize on the surface sensory information. Alternative sensory strategies have included implantable nerve interfaces, artificial skin, peripheral nerve stimulation, and the agonist-antagonist myoneural interface technique.²⁰ By recreating the dynamic agonist-antagonist relationship that exists between opposing muscle groups in the extremity, the agonist-antagonist myoneural interface technique offers a means to restore some degree of proprioceptive sensation.²¹ The clinical impact of this technique on prosthetic function requires further exploration, and existing techniques require the distal joint anatomy to be intact at the time of amputation. TMR may prove to be a valuable means of adapting a regenerative form of the agonist-antagonist myoneural interface concept for use in those who have already experienced limb loss.