Traumatic loss of a hand or arm represents a decisive event resulting in tremendous functional and psychological impairment influencing the quality of life in affected patients.1 Above-elbow or even shoulder level amputations pose a major disability leading to a challenging prosthetic replacement.2 Although homologous limb transplantation at the above-elbow level of amputation is possible, the functional outcome is questionable and the risks of immunosuppression still remain high; thus, prosthetic limb replacement represents the standard of care to date.3,4
Until the advent of targeted muscle reinnervation technique, myoelectric upper arm prostheses were operated by two surface electrodes that are controlled by two separately innervated muscle groups.2,5 The various prosthetic joints were chosen by co-contractions and controlled linearly by these two muscles. To overcome these limitations, in targeted muscle reinnervation surgery, amputated nerves of the brachial plexus, median, ulnar, musculocutaneous, and radial nerve, respectively, are selectively transferred to the muscle branches of the remaining stump muscles. It is hereby possible to create up to six myosignals for intuitive and simultaneous control of the different prosthetic joints.5 In this way, an efficient and more harmonious control of the prosthetic device should be achieved without the need to change between the different levels and prosthetic joints.6
However, almost 15 years after the advent of targeted muscle reinnervation, abandonment or nonuse of prosthetic devices still represents a commonly faced problem and a financial burden for the health care systems.7 In this article, we present our outcomes and range of indications in addition to experiences and pitfalls after 30 targeted muscle reinnervation cases at above-elbow and shoulder disarticulation levels of amputation.
PATIENTS AND METHODS
In the time between January of 2012 and December of 2017, a total of 73 patients with above-elbow and shoulder disarticulation amputations consulted the Center for Extremity Reconstruction at the Medical University of Vienna. In 10 patients (14 percent), targeted muscle reinnervation was not indicated, as these patients needed robust devices mostly for agricultural work and were fitted with body-powered systems. Of the 63 patients for whom targeted muscle reinnervation was indicated, 38 (60 percent) mentioned the lack of function and 25 (40 percent) mentioned pain as the main cause for admittance. However, most patients are referred for improvement both in function and in pain. In total, 30 upper limb amputees (48 percent of patients in whom targeted muscle reinnervation is indicated) underwent targeted muscle reinnervation surgery. Mostly financial issues of the service providers prohibited surgery in the other 33 patients; notably, 49 of the 73 patients (67 percent) were from foreign countries. This study was reviewed and approved by the local institutional review board, and all patients gave informed consent to be enrolled. Surgery was performed by the senior author (O.C.A.) in all 30 cases.
The operations were performed under general anesthesia without muscle relaxation (so motor nerves could be stimulated) in the supine position. Depending on the circumstances of the accident, the surgery began with a supraclavicular exploration of the brachial plexus. Especially after high-velocity trauma and potential traction injuries of the brachial plexus, it is important to evaluate and also treat possible proximal nerve damage. The individual nerves of the brachial plexus were prepared and dissected. Dissection and separation of the different branches is important, and stimulation of the different branches should only provoke twitches in the targeted muscles selectively. The donor nerves have to be neurotomized at least to a level of palpable healthy fascicles. With accurate and individualized planning, a primary tensionless coaptation could be achieved even in patients previously treated with neuroma resections. However, a standard nerve transfer scheme has been established both for above-elbow and for shoulder disarticulation amputees (Tables 1 and 2).
Table 1. -
Standard Nerve Transfer Scheme for Above-Elbow Amputees
|Biceps long head
|Biceps short head
|Triceps long and medial head
|Triceps lateral head
||Split deep radial branch
||Split deep radial branch
Table 2. -
Standard Nerve Transfer Scheme Shoulder for Disarticulation Amputees
|Pectoral major clavicular part
|Pectoral major sternocostal part
||Hand close/wrist rotation
|Pectoral major abdominal part
||Deep radial branch
In patients with long above-elbow stumps and the presence of the brachioradialis muscle, the distal radial nerve could be split along the fascicles in two parts, one to reinnervate the lateral head of the triceps and one to reinnervate the brachioradialis muscle to achieve separate signals for hand opening and supination of the prosthetic device.
For all nerve transfers, the motor branches of the targeted muscles were transected close to the muscle to achieve a short regeneration time. The proximal part of the motor branches was transected a few centimeters back and buried deep to prevent it from reinnervating the targeted muscles. All nerve transfers were performed under loupe magnification in an end-to-end fashion using 8-0 or 9-0 nylon sutures and fibrin glue. The distal neuromas were not excised, as this would necessitate additional and not beneficial further dissection, whereas transferring the nerve to a different target prohibits reinnervation of the distal neuroma.
Indications and Adapted Concepts
The ideal patient for targeted muscle reinnervation surgery has a history of a sharp amputation without traction or soft-tissue injuries.8 In these cases, the brachial plexus should be intact, the residual muscles healthy, and the nerves of sufficient length to facilitate transfers. In above-elbow amputees, 50 to 70 percent of the normal length of the remaining humeral bone are necessary to achieve best prosthetic fitting and functional outcome. Preferably, the remaining joints should reveal full range of motion.9,10 In addition, the patient should be motivated and cognitively capable for the time-consuming rehabilitation process.
There are, however, patients who do not fit these standard inclusion criteria who can still be candidates for adapted targeted muscle reinnervation surgery and prosthetic fitting. For patients with a short above-elbow stump, lack of muscles at the stump region, soft tissue deficits, or brachial plexus injuries, treatment concepts such as adapted nerve transfer schemes or additional soft-tissue surgery are able to improve prosthetic function.
Case 1: Brachial Plexus Injury
We present a case of a 24-year-old man who suffered a car accident resulting in an above-elbow amputation and infraclavicular brachial plexus injury of his left arm. He was referred to our clinic 4 months after the injury with a complete palsy of his left residual limb, and suspected damage of at least the axillary, radial, and musculocutaneous nerves. However, there were no signs of root avulsions; therefore, adapted nerve transfers were possible. Brachial plexus reconstruction had not been attempted.
Four weeks after initial presentation, brachial plexus reconstruction and selective nerve transfers were performed. The supraclavicular exploration showed all roots in continuity, although only the pectoral major and latissimus dorsi muscles showed response to electrical stimulation. The infraclavicular exploration revealed the axillary, radial and musculocutaneous nerves embedded in scar tissue. Through an additional incision on the medial aspect of the stump, the different nerves could be identified and dissected proximally into the scar tissue. Median and ulnar nerve, now without distal targets, seemed to be unaffected; nevertheless, they had lost their distal targets. Therefore, the ulnar nerve was used as a graft to reconstruct the musculocutaneous and the axillary nerves. After complete neurolysis of the radial nerve, the long head of the triceps showed slight response to stimulation and therefore was not grafted. The median nerve was transected proximal from its neuroma and coapted to the muscle branch of the short head of the biceps. Nine months after surgery, the patient was able to move his residual limb in three-dimensional space, and the triceps muscle showed two individual myosignals. In addition, the reconstructed musculocutaneous nerve reinnervated the long head of the biceps, providing a fourth individual myoelectric signal.
Case 2: Very Short Transhumeral Amputation
A 29-year-old man presented after undergoing a very short above-elbow amputation after an agricultural accident (Fig. 1). Only the coracobrachial muscle and the long head of the triceps remained. Because these two muscles did not offer enough myosignal options, nerve transfers were performed, as for shoulder disarticulation amputees.6 The residual activity of the coracobrachial muscle and the long head of the triceps, respectively, provided myosignals for prosthetic elbow flexion and extension. The ulnar nerve was transferred to the clavicular part of the pectoral major muscle, the median nerve was transferred to the sternocostal part of the pectoral major muscle, and some small fascicles of the median nerve were transferred to the pectoral minor muscle. Distal radial nerve fascicles (past the division to the triceps) were coapted to the thoracodorsal nerve to create a hand-opening signal within the latissimus dorsi muscle. Therefore, the patient was provided with five individual myoelectric sites.
Case 3: Insufficient Soft-Tissue/Bony Overgrowth
A 14-year-old boy suffering progressive lymphaticovenous malformation of his left upper extremity underwent an above-elbow amputation after failing multiple sclerotherapy treatments and debulking operations. Because of appositional bone growth, the humerus perforated the skin 7 months after the amputation (Fig. 2). In consideration of the planned prosthetic reconstruction, we did not want to shorten the stump. However, as the skin was already perforated, Marquardt procedures for this patient were not applicable.11 Therefore, we performed a tissue enlargement with a pedicled myocutaneous latissimus dorsi flap to cover the bone and create a proper stump. In this case, the new muscle provided an additional target, and selective nerve transfer of the distal radial nerve to the thoracodorsal nerve was performed in conjunction with the standard nerve transfers for above-elbow level amputations. Almost 5 years after this operation, having reached maturity, the bone is adequately covered, without any soft-tissue problems (Fig. 3).
Because the long head of the biceps and the long and medial heads of the triceps maintain their original innervation, above-elbow amputees can use their conventional myoelectric prosthesis throughout the rehabilitation process. However, shoulder disarticulation amputees have to wait for reinnervation of target muscles before they are able to use a myoelectric device. As soon as the new myosignals are active, between 3 and 9 months postoperatively, a complex neurorehabilitation program is needed for the patient to learn how to activate and separate the different signals, especially in patients who have lost their arm many years previously. For this purpose, biofeedback systems are used to visualize the individual myosignals. This process needs professional guidance of a physiatrist and an occupational therapist and is often underestimated.12 Especially in patients with up to six myosignals, a tight interaction between the therapist and an experienced orthopedic technician is necessary to fabricate a socket with accurate placement of the surface electrodes.
Global upper extremity function was evaluated using the Action Research Arm Test,13 the Southampton Hand Assessment Procedure,14 and the Clothespin-Relocation Test,5 which monitor hand and extremity function closely related to activities of daily living. The Action Research Arm Test was performed according to the standardized approach from Yozbatiran et al.15 except from standing position of the patients because of limited range of motion of the shoulder joint. The Southampton Hand Assessment Procedure has been validated for assessment of pathologic and prosthetic hand function, where normal hand function is regarded as equal to or above 100 points.14 An outcome study of 17 below-elbow prosthetic users has shown Action Research Arm Test scores of approximately 35 of 57 points, a mean time for fulfilling the Clothespin-Relocation Test of approximately 23 seconds, and Southampton Hand Assessment Procedure scores of 65 points.16 In addition, patients who finished prosthetic fitting were asked about the average length of time per day they wore the prosthetic device.
Patient Demographics and Indications
The mean patient age at the time of targeted muscle reinnervation surgery was 38.6 ± 10.4 years, and 83.3 percent were male patients (n = 25). Targeted muscle reinnervation surgery was performed after an average of 5.7 ± 14.5 years after the amputation, which was at the above-elbow level in 18 and at the shoulder level in 12 patients. The mean follow-up from targeted muscle reinnervation surgery was 5.3 ± 2.4 years. Causes of amputation were car accidents (n = 10), work-related accidents (n = 11), battlefield injuries (n = 3), tumor resections (n = 2), motorcycle accidents (n = 2), gangrene after a 15-m fall (n = 1), and polytrauma after a boat accident (n = 1).
In 11 of these 30 patients (37 percent), neuroma pain was the main indication for targeted muscle reinnervation surgery. For these patients, prosthetic replacement was not an issue. Of the remaining 19 patients, five suspended prosthetic fitting either because of financial or time issues or because of the heavy weight of the prosthesis. In one glenohumeral patient, none of the targeted muscles showed response during intraoperative stimulation 4 years after amputation; thus, no nerve transfers were performed and the patient returned to his body-powered device. From the remaining 13 fitted targeted muscle reinnervation patients, 10 were available for final evaluation of prosthetic function. The patients of this study were fitted with a Dynamic Arm Plus, Electric Wrist Rotator and Sensor Hand Speed (Ottobock Healthcare Products GmbH, Duderstadt, Germany).
In all patients, follow-up of at least 12 months was available. All 38 nerve transfers were successful and showed sufficient electromyographic amplitude at the 9-month follow-up visit. The 10 patients available for final evaluation of prosthetic fitting showed mean Action Research Arm Test results of 20.4 ± 1.9 of 57 and mean Southampton Hand Assessment Procedure results of 40.5 ± 8.1 (normal healthy score, approximately 100). The Clothespin-Relocation Test showed a mean time of 34.3 ± 14.4 seconds. The patients wore their prosthesis between 3 and 10 hours/day, depending on their activities. All these outcome measures were performed at least 6 months after final fitting within a routine follow-up visit. The patients received no financial reimbursement. Of the 10 patients available for final functional evaluation, two were tested preoperatively with their standard myoelectric device and the results compared. The Southampton Hand Assessment Procedure showed an improvement of 131.8 percent for the first patient and 146.2 percent for the second patient. The Action Research Arm Test showed only minimal improvement of 0 and 4.6 percent, respectively. This minimal difference was attributable to the nature of this test. As the Action Research Arm Test was originally not designed to evaluate prosthetic hand function, time limits for maximum scores as proposed by Yozbatiran et al.15 are reached by only the most skilled prosthetic users and therefore cause a floor effect.
Function and comfort are the most important factors for successful prosthesis use, from both the amputees’ and prosthetic experts’ perspectives.17 The conventional two-signal control is limited in function, is unnatural, and is unintuitive.6,18 Amputees often do not experience sufficient improvement in their daily lives with these prosthetic devices, resulting in up to 50 percent abandonment rates.7,18,19 Although, the neural signals for hand and arm function can be manifested within the targeted muscles after targeted muscle reinnervation, function, compliance, and reliability of use of prostheses after high-level amputations remain significant concerns.18
In above-elbow amputees with a long stump and existing brachioradialis muscle, six individual and intuitive myosignals can be achieved with the above-described nerve transfers. Thus, one separate signal for each degree of freedom of the prosthetic device can be established.2 In shoulder disarticulation amputees, a maximum of five myoelectric signals can be achieved.3 However, in patients not fitting the standard inclusion criteria described above, the focus should lie on the creation of four strong individual myoelectric signals to control hand and elbow function independently. Two additional signals for wrist control are advantageous, but not a priority. Thus, in challenging cases, the main goal is to perform one transfer of the median or ulnar nerve and one of the distal radial nerve. Especially in patients suffering additional nerve damage requiring repair, the nerve transfer scheme has to be adapted to minimize nonsynergistic matching between cortical organization and target muscle function. Therefore, only cognitively “simple” nerve transfers should be performed. Transfers of both median and ulnar nerves may be difficult to incorporate into prosthetic function because even amputees without nerve damage have difficulties in cognitive separation of these two signals.
Although the results of this study showed overall improved functional outcome scores after cognitive nerve transfers, 32 percent of the patients in our already highly selected cohort did not finish prosthetic rehabilitation with insufficient or nonuse of the final fitting. However, because of the myriad reasons for prosthetic abandonment, this rate of nonuse cannot be interpreted as failure of targeted muscle reinnervation. The weight of the prosthetic device (and thus uncomfortable wearing), no reliable feedback, and insufficient help in activities of daily life are reported as the main reasons in our patients.7
Especially after targeted muscle reinnervation surgery with up to six myoelectric sites, the functional outcome of prosthetic limb replacement strongly depends on a stable stump-socket interface.10 As signals are currently recorded by means of transcutaneous surface electrodes, the position of these sensors must be very stable, as relative movements of the socket could result in not only vague position control of the device but also severe malfunction because of signal loss.20 To overcome this limitation, different systems using implanted sensors are under investigation.20,21 The use of such sensors will stabilize signal quality independent of load or environmental conditions. However, to ensure constant electrode positions and therefore limit relative movements between stump and socket, straps and harnesses to the contralateral shoulder are needed within conventional socket fittings. These harnesses limit the range of motion of the shoulder joint, and can evoke skin irritation and lead to discomfort when wearing the prosthesis.22–24 Therefore, direct prosthetic bone anchorage may achieve the best possible stability and range of motion of remaining joints.25 Moreover, patients report the ability to identify various sensations through their prosthesis-bone interface, called osseoperception.26 As long as the prosthesis itself is not able to provide useful feedback, this is of great advantage. In addition, with this attachment, wires can pass through the percutaneous port of the osseointegrated system.20,27 Although this solution is promising, percutaneous interfaces disturb the skin barrier, resulting in the risk of superficial and deep infections in addition to wire breakage, unstable connectors, and subsequent need for surgical intervention.28–30 Still, a combination of future technologies may be able to enhance overall prosthetic function. However, the incorporation of these new technologies will need to be carefully weighed against the already cost-intensive process of prosthetic fitting.
In addition, the lack of sensory feedback represents one of the major ongoing obstacles to long-term prosthetic adoption.31 Patients are currently forced to rely on visual feedback only. Despite this, targeted sensory reinnervation does not solve this problem, as current prosthetic systems are not able to incorporate this interface, and patients have reported long-lasting dysesthesia and pain in the reinnervated areas.
Compared with myoelectric devices, body-powered prostheses have shown advantages in durability; frequency of adjustments; maintenance; grip force regulation; sweat/temperature independence; and, most importantly, feedback.32 Body-powered control transmits adequate proprioception, particularly regarding grip strength, even under hot or wet conditions. This feedback is essential, especially in bilateral amputees. However, the present literature is insufficient to conclude that either system provides a significant general advantage.33 Thus, prosthetic prescription should be based on the patient’s individual functional needs and expectations. For some patients, even after targeted muscle reinnervation surgery, a hybrid device combining the advantages of both systems may be a valuable option.
However, even if prosthetic rehabilitation is not successful in all cases, a prospective randomized clinical trial has recently shown that targeted muscle reinnervation surgery improves phantom limb pain significantly.34 Thus, targeted muscle reinnervation surgery is worth the effort, at least for pain management and patient comfort, even if this newly established neuromuscular interface is not used for prosthetic control.
Targeted muscle reinnervation technique has improved prosthetic function and revolutionized neuroma treatment in above-elbow and in shoulder disarticulation amputees. Therefore, it represents a worthwhile procedure, but displays other limitations of prosthetic control. The rate of abandonment—even after targeted muscle reinnervation surgery—has been shown to be high, and several advances within the biotechnological interface to improve prosthetic function and acceptance will be needed in this special patient cohort. Meanwhile, the capabilities of body-powered or hybrid systems should not be underestimated.
1. Kovacs L, Grob M, Zimmermann A, et al. Quality of life after severe hand injury. J Plast Reconstr Aesthet Surg. 2011;64:1495–1502.
2. Dumanian GA, Ko JH, O’Shaughnessy KD, Kim PS, Wilson CJ, Kuiken TA. Targeted reinnervation for transhumeral amputees: Current surgical technique and update on results. Plast Reconstr Surg. 2009;124:863–869.
3. Salminger S, Sturma A, Herceg M, Riedl O, Bergmeister K, Aszmann OC. Prosthetic reconstruction in high amputations of the upper extremity (in German). Orthopade 2015;44:413–418.
4. Salminger S, Sturma A, Roche AD, et al. Functional and psychosocial outcomes of hand transplantation compared with prosthetic fitting in below-elbow amputees: A multicenter cohort study. PLoS One 2016;11:e0162507.
5. Kuiken TA, Dumanian GA, Lipschutz RD, Miller LA, Stubblefield KA. The use of targeted muscle reinnervation for improved myoelectric prosthesis control in a bilateral shoulder disarticulation amputee. Prosthet Orthot Int. 2004;28:245–253.
6. Aszmann OC, Dietl H, Frey M. Selective nerve transfers to improve the control of myoelectrical arm prostheses (in German). Handchir Mikrochir Plast Chir. 2008;40:60–65.
7. Biddiss EA, Chau TT. Upper limb prosthesis use and abandonment: A survey of the last 25 years. Prosthet Orthot Int. 2007;31:236–257.
8. O’Shaughnessy KD, Dumanian GA, Lipschutz RD, Miller LA, Stubblefield K, Kuiken TA. Targeted reinnervation to improve prosthesis control in transhumeral amputees: A report of three cases. J Bone Joint Surg Am. 2008;90:393–400.
9. Mixter RC, Rao VK, De Angelis A, Donald CE. Salvage of proximal humeral amputations with a remnant forearm flap. Plast Reconstr Surg. 1991;87:965–968.
10. Salminger S, Gradischar A, Skiera R, et al. Attachment of upper arm prostheses with a subcutaneous osseointegrated implant in transhumeral amputees. Prosthet Orthot Int. 2018;42:93–100.
11. Marquardt E. Plastic surgery in imminent bone perforation at the infantile humerus stump: Preliminary report. Z Orthop Ihre Grenzgeb. 1976;114:711–714.
12. Sturma A, Herceg M, Bischof B, Fialka-Moser V, Aszmann OC. Jenson W, Anderson KO, Akay M. In: Replace, Repair, Restore, Relieve: Bridging Clinical and Engineering Solutions in Neurorehabilitation. Proceedings of the 2nd International Conference on NeuroRehabiliation (ICNR2014); Aalborg, Denmark; June 24–26, 2014. 2014:New York: Springer; 169–177.
13. Lyle RC. A performance test for assessment of upper limb function in physical rehabilitation treatment and research. Int J Rehabil Res. 1981;4:483–492.
14. Kyberd PJ, Murgia A, Gasson M, et al. Case studies to demonstrate the range of applications of the Southampton Hand Assessment Procedure. Br J Occup Ther. 2009;72:212–218.
15. Yozbatiran N, Der-Yeghiaian L, Cramer SC. A standardized approach to performing the action research arm test. Neurorehabil Neural Repair. 2008;22:78–90.
16. Salminger S, Vujaklija I, Sturma A, et al. Functional outcome scores with standard myoelectric prostheses in below-elbow amputees. Am J Phys Med Rehabil. 2019;98:125–129.
17. Schultz AE, Baade SP, Kuiken TA. Expert opinions on success factors for upper-limb prostheses. J Rehabil Res Dev. 2007;44:483–489.
18. Farina D, Aszmann O. Bionic limbs: Clinical reality and academic promises. Sci Transl Med. 2014;6:257ps12.
19. Wright TW, Hagen AD, Wood MB. Prosthetic usage in major upper extremity amputations. J Hand Surg Am. 1995;20:619–622.
20. Ortiz-Catalan M, Håkansson B, Brånemark R. An osseointegrated human-machine gateway for long-term sensory feedback and motor control of artificial limbs. Sci Transl Med. 2014;6:257re6.
21. Pasquina PF, Evangelista M, Carvalho AJ, et al. First-in-man demonstration of a fully implanted myoelectric sensors system to control an advanced electromechanical prosthetic hand. J Neurosci Methods 2015;244:85–93.
22. Gradischar A, Skiera R. Subcutaneous implant supported attachment of exoprostheses for transhumeral amputees. Orthopädie Technik
23. Neusel E, Traub M, Bläsius K, Marquardt E. Results of humeral stump angulation osteotomy. Arch Orthop Trauma Surg. 1997;116:263–265.
24. Datta D, Kingston J, Ronald J. Myoelectric prostheses for below-elbow amputees: The Trent experience. Int Disabil Stud. 1989;11:167–170.
25. Brånemark R, Brånemark PI, Rydevik B, Myers RR. Osseointegration in skeletal reconstruction and rehabilitation: A review. J Rehabil Res Dev. 38:175–181.
26. Häggström E, Hagberg K, Rydevik B, Brånemark R. Vibrotactile evaluation: Osseointegrated versus socket-suspended transfemoral prostheses. J Rehabil Res Dev. 2013;50:1423–1434.
27. Mastinu E, Branemark R, Aszmann O, Ortiz-Catalan M. Myoelectric signals and pattern recognition from implanted electrodes in two TMR subjects with an osseointegrated communication interface. Conf Proc IEEE Eng Med Biol Soc. 2018;2018:5174–5177.
28. Bergmeister KD, Hader M, Lewis S, et al. Prosthesis control with an implantable multichannel wireless electromyography system for high-level amputees: A large-animal study. Plast Reconstr Surg. 2016;137:153–162.
29. Tsikandylakis G, Berlin Ö, Brånemark R. Implant survival, adverse events, and bone remodeling of osseointegrated percutaneous implants for transhumeral amputees. Clin Orthop Relat Res. 2014;472:2947–2956.
30. Weir RF, Troyk PR, DeMichele GA, Kerns DA, Schorsch JF, Maas H. Implantable myoelectric sensors (IMESs) for intramuscular electromyogram recording. IEEE Trans Biomed Eng. 2009;56:159–171.
31. Biddiss E, Chau T. Upper-limb prosthetics: Critical factors in device abandonment. Am J Phys Med Rehabil. 2007;86:977–987.
32. Schweitzer W, Thali MJ, Egger D. Case-study of a user-driven prosthetic arm design: Bionic hand versus customized body-powered technology in a highly demanding work environment. J Neuroeng Rehabil. 2018;15:1.
33. Carey SL, Lura DJ, Highsmith MJ; CP; FAAOP. Differences in myoelectric and body-powered upper-limb prostheses: Systematic literature review. J Rehabil Res Dev. 2015;52:247–262.
34. Dumanian GA, Potter BK, Mioton LM, et al. Targeted muscle reinnervation treats neuroma and phantom pain in major limb amputees: A randomized clinical trial. Ann Surg. 2019;270:238–246.