Managing Major Peripheral Nerves in Forearm-Level Amputations With TMR and RPNI: What’s the Best Recipe?

Abstract

Background:

Targeted muscle reinnervation (TMR) and regenerative peripheral nerve interface (RPNI) prevent symptomatic neuroma formation in amputees. Forearm-level amputations present multiple muscular targets, making it challenging to determine the ideal treatment. The purpose of this study was to evaluate the best TMR targets, role of RPNI, and appropriate patient-selection criteria in forearm-level amputations. We hypothesized that deep and distal TMR targets would best prevent symptomatic neuromas, RPNI would prove a success adjunct, and patients with poorly controlled diabetes would not develop symptomatic neuromas regardless of nerve management.

Methods:

We retrospectively identified forearm-level amputations performed between 2017 and 2022. Patients with TMR by outside providers, follow-up <6 months, or insufficient documentation were excluded. Demographics, surgical nerve management, and postoperative complications were collected. The primary outcome was development of a painful neuroma determined by the Eberlin criteria. Patients undergoing TMR were divided a priori into two groups, superficial and proximal versus deep and distal TMR targets, and were compared.

Results:

Thirty-nine patients met inclusion criteria, and 16 developed a symptomatic neuroma. No patients with a deep or distal TMR target developed a symptomatic neuroma. One nerve out of 12 treated with RPNI developed a symptomatic neuroma. No patient with poorly controlled diabetes developed a symptomatic neuroma, despite no advanced nerve management.

Conclusions:

In a case series of forearm amputations, deep and distal TMR targets prevented symptomatic neuroma formation more than superficial and proximal targets. Regenerative peripheral nerve interface is a useful adjunct for neuroma control, especially for the radial sensory nerve. Patients with poorly controlled diabetes may not require advanced nerve management.

Level of Evidence:

Level IV retrospective case series

Introduction

There has been a recent paradigm shift in the management of peripheral nerves during major limb amputation with a shift toward active, rather than passive, nerve management. Active nerve management includes targeted muscle reinnervation (TMR) and regenerative peripheral nerve interface (RPNI). Examples of passive management include traction neurectomy, nerve capping, and burying of nerves into innervated muscle or bone. TMR involves transferring severed nerves to denervated muscular targets.^1,2 In addition to providing additional electromyogram (EMG) signals for intuitive myoelectric prosthesis control, TMR has been shown to reduce neuroma-related pain.^{3 -8} RPNI also reduces the rate of symptomatic neuroma formation^9,10 but does not produce a sufficient myoelectric signal to allow detection by surface electrodes. Despite the success of these two active nerve management strategies, not all TMR and RPNI procedures are successful.^11,12 The rate at which TMR and RPNI failure occurs varies widely in the literature.^7,13,14 Some of this heterogeneity is likely due to the rapid expansion of these procedures into different indications, including acute versus chronic amputation, and amputations for causes including trauma, dysvascularity, infection, and more.^15,16

Targeted muscle reinnervation was first described for proximal-level amputations, including transhumeral amputations and shoulder disarticulations, to improve myoelectric prosthesis control by adding intuitive EMG signals.^3,4,17 TMR was subsequently found to also limit neuroma formation. In contrast, forearm (transradial)-level amputations differ from above-elbow amputations in that there are numerous options for myoelectric signals. Therefore, in forearm-level amputations, TMR is utilized primarily to prevent symptomatic neuroma formation rather than to create additional EMG signals for intuitive myoelectric prosthesis use.

Forearm-level amputations present a unique challenge given the variability in amputation level and therefore variability in TMR targets and subsequent success.^10,11 A plethora of potential targets are described for TMR at the level of the forearm.^{18 -20} However, there is essentially no literature that reports outcomes for the different TMR targets in the forearm. Variability in the level of forearm amputation may place TMR targets at important pressure points when a prosthesis is worn. A typical prosthesis worn after a forearm-level amputation has an anterior cutout to allow for elbow flexion and therefore places pressure over the flexor-pronator and extensor-supinator soft-tissue masses to create a secure fit (Supplemental Figure 1). Distal to this prosthesis compression, there is significantly less pressure on the soft tissues including nerve-management sites.

The purpose of this study was to evaluate which TMR targets in the forearm were most successful at preventing painful neuromas and the role of RPNI in forearm amputations. We also explored which patients benefit most from these advanced nerve-management strategies. Based on our clinical experience, we hypothesized that forearm TMR to superficial and proximal (SP) targets would lead to more symptomatic neuromas than targets that were deep and distal (DD). We further hypothesized that RPNI would prove a useful adjunct to TMR. Finally, we hypothesized that patients undergoing amputation for complications related to underlying diabetes/neuropathy would not develop symptomatic neuromas at a high rate, despite no advanced treatment of their nerves at the time of amputation.

Materials and Methods

This study was conducted as a retrospective case series of patients that presented to a single, large, quaternary, upper extremity surgery group between 2017 and 2022. Approval for this study was granted by the affiliated institutional review board.

Patient Inclusion

Patients were identified via our institution’s amputee registry, which includes all major limb amputees.

All upper extremity amputees were identified for initial screening. Retrospective chart review was used to identify patients who had a forearm-level amputation performed during the study period (Figure 1). Patients with inadequate follow-up (<6 months), age <18 years, or surgery performed by an outside surgeon were excluded. Patients undergoing TMR were divided a priori into two groups: (1) SP TMR targets; and (2) DD targets (Figure 2).

Figure 1.

Consolidated Standards of Reporting Trials (CONSORT Diagram).

Figure 2.

Predetermined groupings for nerve treatment at the time of amputation.

Note. RPNI = regenerative peripheral nerve interface; TMR = targeted muscle reinnervation.

Primary Outcome Measure

The primary outcome measure was development of a symptomatic neuroma. The determination of a postoperative neuroma was established via the Eberlin criteria. ²¹

Data Collection

A retrospective chart review was completed to collect surgical and clinical variables, including preexisting comorbidities, donor nerves, and recipient muscles. Patients who did not have TMR of all nerves had the treatment of the non-TMR nerves recorded as well (eg, RPNI and traction neurectomy). Short-term postoperative complications (eg, infection), long-term nerve-related symptoms, and physical exam findings, as well as prosthesis use and work status at the time of most recent clinic appointment, were also recorded. In patients who had a neuroma by the Eberlin criteria, the specific nerve involved was recorded.

Analysis

For the analysis, TMR targets were determined to be SP or DD (Figure 2). Muscles that were designated as DD were either: (1) muscles that lie in the deep volar compartment (flexor digitorum profundus, pronator quadratus, flexor pollicis longus); or (2) muscles with deep insertions within the forearm itself (pronator teres). Remaining muscles in the superficial volar compartment, the mobile wad, or dorsal compartments were designated as SP. The median, radial, and ulnar nerves were each analyzed individually with the SP group compared to the DD group. A combined analysis was done of all nerves in the SP group compared to all nerves in the DD group with symptomatic neuroma rates for this combined group reported as well. Neuroma formation rates for nerves treated with RPNI, traction neurectomy, or other treatments were recorded. Nerves of preexisting amputations that were not surgically addressed by the included group of surgeons were excluded from analysis. A Fisher’s exact test was run to compare the rate of symptomatic neuroma formation in these groups. Descriptive statistics were generated for demographics, comorbidities, and the remaining nerve treatment options (RPNI, traction neurectomy, etc).

Results

Thirty-nine forearm-level amputations were included. Demographic data are listed in Table 1. Preexisting comorbidities of diabetes and peripheral neuropathy as well as treatment are reported in Table 2. Among all patients, 59 nerves were treated with TMR, 12 were treated with RPNI, 22 were treated with traction neurectomy, and the remaining nerves had a variety of other treatments (Table 3). The treatment of each major peripheral nerve in all patients is detailed in Supplemental Tables 1, 2, and 3.

Table 1.

Demographics.

Sex
Female	22
Male	17
Ethnicity
Hispanic or Latino	0
Not Hispanic or Latino	29
Declined	5
Unknown	5
Race
Black, African American	12
Asian	1
White	19
Declined	2
Unknown	5
Age at surgery (years)
Mean [SD]	45.6 [11.3]
Median	44
Range	20-69
BMI
Mean [SD]	28.6 [7.6]
Median	27.3
Range	16.8-50.6

Note. BMI = body mass index.

Table 2.

Comorbidities and Outcomes by Treatment Algorithm.

Comorbidities		TMR of any nerve (N = 31)	No TMR (N = 11)
Diabetes	Yes	1	5
	No	29	4
Peripheral neuropathy	Yes	0	1
	No	30	8

Note. TMR = targeted muscle reinnervation.

Table 3.

Symptomatic Neuromas by Nerve Treatment at the Time of Amputation.

Nerve	Target	No neuroma	Neuroma
Median nerve	Superficial/Proximal TMR	14	5
	Deep/Distal TMR	8	0
	RPNI	0	0
	Traction neurectomy	6	0
	Other targets	2	2
Ulnar nerve	Superficial/Proximal TMR	17	3
	Deep/Distal TMR	6	0
	RPNI	1	0
	Traction neurectomy	5	0
	Other targets	3	2
Radial sensory nerve	Superficial/Proximal TMR	4	0
	Deep/Distal TMR	2	0
	RPNI	10	1
	Traction neurectomy	9	2
	Other targets	15	0
All nerves combined	Superficial/Proximal TMR	35	8
	Deep/Distal TMR	16	0
	RPNI	11	1
	Traction neurectomy	20	2
	Other targets	20	4

Note. RPNI = regenerative peripheral nerve interface; TMR = targeted muscle reinnervation.

Symptomatic Neuromas

A total of 16 symptomatic neuromas developed as determined by the Eberlin Criteria, 15 of which involved one of the three major peripheral nerves included in this study (Table 4). Of these 16 neuromas, 8 (50%) were associated with nerves that had been treated with TMR. The median nerve was the most common nerve to develop a symptomatic neuroma (Table 4). Analysis of transfers by target location revealed that no nerves in the DD TMR developed a symptomatic postoperative neuroma. Eight of 43 nerves (18.6%) treated with TMR to SP targets developed neuromas (Table 3). Given the development of 0 neuromas in the DD group, the Fisher’s exact test could not be generated to evaluate for a statistical significance between these two groups. Only one nerve (8.3%) out of 12 treated with RPNI developed a symptomatic neuroma (Table 3).

Table 4.

Location of All Neuromas.

Involved nerve	Patients
Median nerve	7
Ulnar nerve	5
Radial nerve	3
Medial antebrachial cutaneous nerve	1
Total	16

Diabetes, TMR/RPNI, and Symptomatic Neuroma Formation

Patients with a preexisting diagnosis of diabetes were less likely to be treated with TMR or RPNI by our group during the study period (Table 2, Supplemental Table 4). Of the 6 patients with diabetes, 4 required amputations due to complications of poorly controlled diabetes (all infections). None of these patients were treated with TMR, and none developed symptomatic neuromas. Two patients had well-controlled diabetes and sustained traumatic injuries that led to their amputations. One of these patients was treated with TMR and developed a symptomatic neuroma. The other was not treated with TMR (coaptation of the median and ulnar nerves) and also developed a symptomatic neuroma. Patients with a diagnosis of diabetes developed symptomatic neuromas at a rate of 20% (1 of 5 patients) when not treated with TMR or RPNI. Nerves of preexisting amputations that were not surgically addressed by the included group of surgeons were excluded from analysis.

Evolution of Technique Over Time

Treatment of major peripheral nerves evolved over time during the study. The midpoint of the study period was the end of 2019. Prior to 2020, the median and ulnar nerves were transferred to a DD target in 15.8% of cases. Since 2020, this rose to 70% of cases for the median nerve and 60% of cases for the ulnar nerve. The radial sensory nerve (RSN) was treated with RPNI in 10.5% of cases before 2020 and in 90% of cases since.

Discussion

These results of this study can help guide upper extremity surgeons with the complex decision-making involved with peripheral nerve management at the time of amputation. Previous literature suggesting higher rates of symptomatic neuroma formation ²² combined with our clinical suspicion over the years that forearm amputees with superficial TMR targets developed symptomatic neuromas led to the development of this study. The delineation of SP versus DD targets represented an evolution in our group’s algorithm and approach to these procedures. This study is unique in both its size and granularity in demonstrating that no nerves treated with TMR to DD targets went on to develop a symptomatic neuroma. The selection of TMR sites has been considered nonprescriptive as myoelectric sensors can be placed wherever necessary to detect signals, yet the predilection of specific targets’ impact on the development of symptomatic neuromas has not been considered. The results of this study suggest that a surgeon deciding between two otherwise equal TMR targets in the forearm might best prevent painful neuroma formation by selecting a DD target.

One potential explanation for the propensity to develop painful neuromas after SP forearm TMR may lie in the design of many forearm prostheses that place pressure on the SP muscles in the flexor-pronator and extensor-supinator soft-tissue masses. Of note, we favor performing forearm-level TMR, including TMR to a pedicled pronator quadratus and flexor pollicis longus, through a single distal incision in order to minimize prosthesis irritation on a proximal incision (Figure 3).

Figure 3.

TMR of pedicled pronator quadratus (PQ) performed through a single distal incision.

Note. TMR = targeted muscle reinnervation.

While this study demonstrates that DD TMR targets in the forearm prevent symptomatic neuroma formation better than SP targets, TMR to DD targets may limit EMG signal detection when using a myoelectric prosthesis. Therefore, surgeons must still consider utilizing SP TMR targets if additional signals are needed to power a myoelectric prosthesis.

In this study, RPNI proved to be a useful adjunct to TMR in forearm-level amputees. In a total of 12 nerves treated with RPNI, only one nerve developed a symptomatic neuroma. In this series, RPNI was used almost exclusively for the RSN and was typically placed into a deep location within the residual limb. RPNI seems to be sufficient at preventing painful neuroma formation of the RSN as this nerve has a relatively low number of axons and is not needed to generate a motor signal to power a myoelectric prosthesis. Of note, we typically use a single muscle graft for RPNI of the RSN as one RPNI appears to adequately harnesses the volume of the RSN. For RPNI of a larger mixed motor-sensory nerve with more axons, such as the median and ulnar nerves, the surgeon should consider dividing the nerve into several groups and using multiple RPNI grafts in order to maintain an appropriate nerve to RPNI ratio and adequately prevent painful neuroma formation. However, this series did not employ RPNI for larger mixed motor-sensory nerves, and we therefore cannot comment on its efficacy in that specific scenario. In low-demand or highly comorbid patients who are unlikely to ever meaningfully use a myoelectric prosthesis, this study leads credence to the growing evidence that RPNI remains a good option for neuroma control.^23,24

Lastly, the results of this study provide some insight into patient selection for TMR and RPNI. We have not routinely performed TMR for patients with poorly controlled diabetes or a history of peripheral neuropathy, ⁸ as our suspicion has been that these patients do not develop symptomatic neuromas at the same rate that non-diabetic patients do. Although limited by a very small sample size, the results of this study do weakly support that notion. Further work here is warranted—especially considering reports of successful management of neuroma pain in diabetic patients who underwent TMR and/or RPNI.^{25 -27}

As a single-center retrospective case series, there are several limitations to our study. The first of which is its size, due to the relative rarity of patients who need advanced nerve handling of forearm-level amputations. Furthermore, the study is inherently biased by the evolution and improvements in technique by the surgeons in our group. The use of DD TMR targets evolved over time and undoubtedly shares correlation with other improvements in technique, approach, and patient selection that cannot be fully accounted for by this study model. The absence of patient-reported outcomes is another limitation of this study. However, the strict use of the Eberlin criteria at each subsequent office visit for diagnosis of symptomatic neuroma formation was used in an effort to offset this limitation as it relies in part on patient-reported symptoms. Despite these limitations, the results of this review provide promising areas for further, more robust investigation into the ideal recipe for the handling of the major peripheral nerves in forearm-level amputees.