Targeted muscle reinnervation in oncologic amputees: Early experience of a novel institutional protocol

Abstract

Background:

We describe a multidisciplinary approach for comprehensive care of amputees with concurrent targeted muscle reinnervation (TMR) at the time of amputation.

Methods:

Our TMR cohort was compared to a cross-sectional sample of unselected oncologic amputees not treated at our institution (N = 58). Patient-Reported Outcomes Measurement Information System (NRS, PROMIS) were used to assess postamputation pain.

Results:

Thirty-one patients underwent amputation with concurrent TMR during the study; 27 patients completed pain surveys; 15 had greater than 1 year follow-up (mean follow-up 14.7 months). Neuroma symptoms occurred significantly less frequently and with less intensity among the TMR cohort. Mean differences for PROMIS pain intensity, behavior, and interference for phantom limb pain (PLP) were 5.855 (95%CI 1.159-10.55; P = .015), 5.896 (95%CI 0.492-11.30; P = .033), and 7.435 (95%CI 1.797-13.07; P = .011) respectively, with lower scores for TMR cohort. For residual limb pain, PROMIS pain intensity, behavior, and interference mean differences were 5.477 (95%CI 0.528-10.42; P = .031), 6.195 (95%CI 0.705-11.69; P = .028), and 6.816 (95%CI 1.438-12.2; P = .014), respectively. Fifty-six percent took opioids before amputation compared to 22% at 1 year postoperatively.

Conclusions:

Multidisciplinary care of amputees including concurrent amputation and TMR, multimodal postoperative pain management, amputee-centered rehabilitation, and peer support demonstrates reduced incidence and severity of neuroma and PLP.

1 |. INTRODUCTION

Despite a paradigm in the United States that has favored limb salvage and advances in limb salvage techniques, amputation remains a cornerstone in the management of extremity malignancies at time of index resection, for the management of locally recurrent disease, and for failed limb salvage. Approximately 18,000 patients undergo amputation for cancer diagnoses annually, representing a small but significant portion of the amputee population.¹ In light of heterogeneous evidence favoring limb salvage over amputation, Robert et al,² summarized it best when they concluded that regardless of the surgical strategy (limb salvage or amputation), patients with more functional lower limbs had a better quality of life. Therefore, we submit that we must refocus our efforts away from the decision of limb salvage versus amputation and instead strive towards developing techniques and strategies to achieve “limb optimization.”

Several strategies may be considered to increase the odds of a favorable outcome after amputation. Herein we present a multidisciplinary approach for the care of the oncologic amputee with a foundation built upon the surgical innovation of targeted muscle reinnervation (TMR) and synergistic adjuvants including multimodal acute pain management, emphasizing the role of oncologic rehabilitation, and psychosocial support.

1.1 |. Management of postamputation pain through surgical innovation

One reason to avoid amputation has historically been the possibility of intractable postamputation pain. People living with limb loss may experience pain in the form of residual limb pain, phantom limb pain (PLP), and axial musculoskeletal pain.^3–5 As few as 9% of all amputees report living pain-free,⁵ leaving a significant number of amputees living with chronic pain. Symptomatic neuromas are responsible for chronic localized pain within the residual limb in approximately 30% of major limb amputees.⁶ These neuromas can make prosthetic wear uncomfortable or even impossible. Traditionally, traction neurectomy is performed at the time of amputation; however this results in a neuroma as the transected axons attempt to regenerate.⁷ Numerous surgical techniques have been proposed⁸ including excision of the neuroma and burying the nerve in muscle,⁹ bone,^10–12 or vein,^13,14 and coaptation of two nerve ends.^15,16 To date, no method provides consistently effective relief of neuroma-related postamputation pain. In addition to neuroma-related residual limb pain, up to 85% of all major limb amputees experience PLP.^17–23 PLP is defined as burning, tingling, discomfort, or electrical shooting pain in the missing limb.²⁴ Oncology patients experience similar, if not higher, rates of PLP after amputation compared to the general amputee population.^25–30 Krane and Heller²⁵ noted PLP in up to 90% of pediatric extremity sarcoma survivors, while others have noted PLP in 60%-87% of proximal oncologic amputees.^29,30 Some evidence suggests that chemotherapy may play a potentiating role in the development of PLP.²⁶ While the exact mechanism of PLP and phantom limb sensations are poorly understood, it is surmised that spontaneous and abnormal peripheral nerve signals along with associated central changes including cortical reorganization and gray matter plasticity may play crucial roles.^31–35 Proposed strategies for the management of PLP include peri-operative catheters, pharmacological therapies, cognitive and mirror therapy, and transcutaneous electrical nerve stimulation.^36–45 Although these modalities have demonstrated some benefit, no intervention has been shown to clearly prevent or treat PLP.

TMR has recently been adopted as a strategy for the management and prevention of postamputation pain. TMR was originally performed by Dumanian as a secondary procedure to improve myoelectric prosthetic control in proximal upper extremity amputees.⁴⁶ It involves the transfer of transected peripheral nerves to otherwise redundant target muscle motor nerve units.^46–53 The central principle underlying nerve transfers in TMR surgery is to reestablish a physiologic end organ for the transected nerve through the organized reinnervation of denervated target muscle units. Details of the surgical technique have been previously published.^54–56 Nerve coaptation provides the nerve with a function as it heals, thus increasing the likelihood that neuromas do not form and that subsequent neuroma related symptoms are minimized. This clinical effect has been observed in animal models that demonstrate normalization of nerve morphology after neuroma excision and subsequent TMR.⁵⁷ In a study focusing on the effect of TMR on postamputation neuroma pain, Souza et al⁵⁸ demonstrated successful treatment of neuroma-related symptoms in 93% of patients who had neuroma-related pain before TMR. A recent randomized control trial for the treatment of established residual limb and neuroma pain in amputees with TMR versus neuroma excision and muscle burying showed improved PLP with TMR.⁵⁹ Furthermore, Serino et al⁶⁰ objectively demonstrated the beneficial effect of targeted muscle and sensory reinnervation on the normalization of motor and sensory cortical organization and the prevention of cortical reorganization that occurs after amputation, confirming similar findings by Chen et al⁶¹ The potential for this strategy in the oncology population was recently highlighted.⁶²

With the recognition that chronic pain is difficult to treat, we have instituted TMR at the time of major limb amputation as routine practice at our institution, in particular for oncologic amputees, for the preemptive management of postamputation pain.

1.2 |. Management of postamputation pain through multimodal pharmacology

With few historical options for neuroma-related residual limb and PLP, opioid analgesics remain a central component of postamputation pain management.^43,63 The short-term effects of various interventions on inpatient opioid consumption are a frequently reported outcome in the literature^42,64–66; however, the reporting of chronic opioid dependence after amputation is inconsistent. Although recent literature supports its use in the treatment of PLP in a subset of patients who appear to be opioid-responders,^37,67 awareness of the growing opioid epidemic and resultant legislative changes discourage and limit the use of opioids for the management of chronic pain, thus necessitating the development of alternative strategies to prevent phantom limb and neuroma-related residual limb pain.

Neuromodulators, such as gabapentin and pregabalin, are commonly utilized in the management of neuropathic pain, including postamputation PLP. Literature supporting the use of neuromodulators in the prevention and treatment of PLP is inconclusive. Given its associated side effects, it is uncertain if the risk-benefit favors its long-term daily use.^36,43,68,69

The effect of TMR on the use of postoperative narcotic dependence is investigated in our patient population.

1.3 |. Oncology rehabilitation: optimizing the benefit of TMR

Oncology rehabilitation teams consisting of physical and occupational therapists have an essential role in the integrative care model of cancer patients. Research has shown that oncology-specific rehabilitation decreases disability and overall health care costs when utilized throughout the cancer treatment continuum.^70,71 A study from the Journal of the National Cancer Institute projects there will be 18.1 million cancer survivors in the United States in 2020 with a total cost of $173 billion, representing a 39% increase from the year 2010.⁷² With this increase, due in large part to more effective treatment measures, the health care system is met with the challenge of survivorship care. Thus it is crucial to minimize the functional impairment and disability experienced by patients during and after treatment and facilitate patients’ reintegration into society.⁷³ Furthermore, oncologic amputees have comorbidities that may affect or slow down their postoperative rehabilitation course, including chemotherapy-related fatigue and neuropathy that can cause increased pain and decreased strength and endurance.⁷⁴ Similarly, radiation-associated fibrosis can slow down the rehabilitation process and limit prosthetic utilization.⁷⁵ Addressing these hurdles requires a coordinated team of physical and occupational therapists dedicated solely to the care of oncology patients who are aware of the unique needs and burdens of this population.

In addition to their oncology-specific needs, patients undergoing amputation with TMR have tailored rehabilitation that evolves as the nerve coaptations heal. Our institutional rehabilitation protocol is available in Appendix 1. Throughout the recovery process our therapist continues to assess functional outcomes at 2-month intervals until patients achieve their functional goals and then ideally annually thereafter.

Our collaborative multidisciplinary team is essential to create continuity of care and effectively manage the patient’s rehabilitation needs throughout their recovery.

1.4 |. Peer support

The emotional and physical strain endured by an oncology patient necessitates a supportive cast of caregivers, family, friends, healthcare providers, and peers. Furthermore, a strong peer support network for major limb amputees is known to be associated with improved outcomes.^76,77 In the cancer population, this sense of community from a peer support network is associated with improved quality of life for many participants.⁷⁸ The plight of the oncologic amputee is truly unique with patients battling a cancer diagnosis during ongoing treatment, as well as the pain, change in body image, and loss of function associated with a major limb amputation. Therefore, our patients are offered the opportunity to speak with TMR amputees preoperatively, and at follow-up appointments are encouraged to seek out support groups.

2 |. METHODS

Patients during the sample period followed our institutional protocol as described in detail in Appendix 2. Following Institutional Review Board approval, patient-reported outcomes were collected using a previously developed amputee survey for patients who underwent major limb amputation with concurrent TMR for an oncologic diagnosis between November 2015 and November 2018. Opioid prescription patterns were examined using the Ohio Automated Rx Reporting System (OARRS). Patients were excluded if they were under 18 years of age at the time of follow-up, had a cognitive impairment, were enrolled in other studies relating to neuropathic pain, had open wounds, or were actively undergoing radiation therapy. Minimum follow-up time was 3 months. The TMR cohort was compared to a cross-sectional sample of unselected amputees not treated at our institution recruited from across the United States via prosthetic clinics, pain clinics, amputee clinics, and activity clubs, as well as amputee conferences and trade shows. The survey was also advertised online at amputee-coalition.org. The result of this effort was 727 completed surveys, of which 60 identified “Cancer” as the reason for amputation. Two were excluded for age <18 years for a final control group N = 58.

The survey collected demographic information (age, gender, education, employment) and amputation information (level of amputation, time since amputation, and reason for amputation). Participants were asked to rate the frequency and intensity of neuroma symptoms, including tingling sensations, burning sensations, sudden episodes of pain, light touch causing pain, hot or cold temperatures causing pain, and light pressure causing pain. Patient-Reported Outcomes Measurement Information System (PROMIS) instruments for pain intensity (short form 3a), pain behavior (short form 7a), and pain interference (short form 8a) were used to assess residual limb and PLP separately.^79–82 Pain outcomes were analyzed only for patients with greater than 1 year follow-up. TMR patients with greater than 1 year follow-up (N = 15) were compared to TMR patients with less than 1 year follow-up (N = 12) in a secondary analysis.

Categorical variables were analyzed by the Chi-squared statistic. Ordinal variables were analyzed by the Mann-Whitney U nonparametric test. PROMIS scores were analyzed by independent samples t test comparison of means. All statistical analyses were performed using SPSS (IBM, Armonk, NY).

3 |. RESULTS

During the study period, 31 patients with malignant or benign aggressive tumor diagnoses were treated at our institution with amputation and concurrent TMR. Four patients were excluded from patient-reported outcomes analysis according to our exclusion criteria–two patients with open wounds and two patients with inadequate patient-reported outcomes follow-up. Of those included in the patient-reported outcomes study, the average patient age was 53.5 years (range 18-85 years). Average follow-up time was 14.7 months (range 3-40.2 months). Six upper extremity (3 transhumeral and 3 shoulders disarticulation) and 21 lower extremity amputations were performed (7 below the knee, 13 above knee, and 1 hip disarticulation).

Our cohort of patients undergoing amputation with concurrent TMR included 31 patients with a wide range of histologic diagnoses, including bone and soft-tissue sarcomas as well as other non-sarcoma malignancies and benign aggressive tumors (Table 1). With respect to oncologic follow-up, average follow-up was 16 months with three patients lost to follow-up with no evidence of disease at 0, 5, and 9 months respectively. Nine upper extremity amputations (1 transradial, 4 trans-humeral, and 4 shoulder disarticulations) and 22 lower extremity amputations (7 below the knee, 14 above the knee and 1 hip disarticulation) were performed. Sixteen amputations were performed at the index resection for a malignant diagnosis. Fifteen patients underwent secondary amputation; 13 for local recurrence and two for infection-related failures of distal femoral replacements for a large distal femoral B-cell lymphoma and remote distal femoral chondrosarcoma. Eleven patients received neoadjuvant (within 8 weeks of amputation) chemotherapy, and 12 patients received adjuvant chemotherapy (started before 8 weeks postoperatively) after amputation. Seven patients received radiation to the operative extremity at any time before amputation and two patients received postoperative radiation (within 8 weeks of amputation).

TABLE 1.

Patient demographics including age at the time of amputation, gender, oncologic diagnosis, level of amputation, the use of adjuvant therapies, and oncologic outcome at final follow-up

	Level of amputation	Age, y	Gender	Oncologic diagnosis	Adjuvant therapy				Oncologic outcome
					Preoperative		Postoperative
					Chemotherapy	XRT	Chemotherapy	XRT
1	Trans-radial	49	M	Synovial sarcoma	–	–	Yes	–	Metastatic/died of disease
2	BKA	42	M	Leiomyosarcoma	Yes	–	–	–	No evidence of disease
3	Shoulder disarticulation	58	M	Recurrent chondroblastic osteosarcoma	–	–	–	–	No evidence of disease
4	BKA	32	M	Clear cell sarcoma	–	–	–	–	Lost to follow-up (NED)
5	BKA	18	M	Ewing sarcoma	Yes	–	Yes	–	No evidence of disease
6	AKA	51	M	Recurrent malignant peripheral nerve sheath tumor	Yes	Yes	Yes	–	No evidence of disease
7	AKA	62	M	B-cell lymphoma	Yes	Yes	Yes	–	No evidence of disease
8	Shoulder disarticulation	62	M	Recurrent metastatic colonic adenocarcinoma	Yes	Yes	Yes	–	No evidence of disease
9	AKA	45	M	Pseudomyogenic hemangioendothelioma	–	–	–	–	No evidence of disease
10	BKA	69	F	Synovial cell sarcoma	–	–	–	–	No evidence of disease
11	AKA	44	F	Recurrent squamous cell carcinoma	Yes	–	Yes	–	No evidence of disease
12	BKA	51	M	Clear cell sarcoma	–	–	–	–	No evidence of disease
13	AKA	83	F	Recurrent high-grade myxofibrosarcoma	–	–	–	–	Local recurrence/metastatic disease
14	Shoulder disarticulation	54	M	Osteosarcoma	Yes	–	Yes	Yes	Local recurrence/metastatic disease/died of disease
15	Shoulder disarticulation	66	F	Recurrent undifferentiated pleomorphic sarcoma	–	Yes	–	–	Local recurrence
16	AKA	17	M	Osteosarcoma	–	–	Yes	–	No evidence of disease
17	AKA	47	M	High-grade myxofibrosarcoma	Yes	Yes	Yes	–	No evidence of disease
18	Trans-humeral	60	F	Recurrent epithelioid sarcoma	Yes	–	–	Yes	No evidence of disease
19	AKA	26	M	Recurrent low-grade parosteal osteosarcoma	–	–	–	–	Lost to follow-up (NED)
20	BKA	22	M	Recurrent high-grade osteosarcoma	–	–	Yes	–	No evidence of disease
21	Trans-humeral	68	F	Recurrent high-grade myxofibrosarcoma	–	Yes	–	–	Local recurrence/metastatic disease/died of disease
22	AKA	59	M	Interme intermediate-grade MPNST	–	–	–	–	Lost to follow-up (NED)
23	AKA	84	M	High-grade myxofibrosarcoma					No evidence of disease
24	AKA	75	F	Chondrosarcoma	–	–	–	–	No evidence of disease
25	AKA	68	M	High-grade spindle cell sarcoma of bone	Yes	–	Yes	–	No evidence of disease
26	Trans-humeral	78	M	Recurrent intermediate-grade myxofibrosarcoma	–	–	–	–	No evidence of disease
27	BKA	70	F	Recurrent high-grade myxofibrosarcoma	–	–	–	–	No evidence of disease
28	AKA	22	F	Recurrent desmoid tumor	–	Yes	–	–	No evidence of disease
29	AKA	50	F	Alveolar soft part sarcoma	–	–	–	–	No evidence of disease
30	Hip disarticulation	46	M	Chondroblastic osteosarcoma	Yes	–	Yes	–	No evidence of disease
31	Trans-humeral	57	F	Squamous cell carcinoma	–	–	–	–	No evidence of disease

At clinical last follow-up, 23 patients (74%) were without evidence of disease, one patient (3.2%) had an isolated local recurrence, one patient developed metastatic disease without local recurrence and unfortunately died of his disease, and three patients (9.7%) developed a local recurrence with concurrent metastatic disease, two of whom died of their disease. Three patients were lost to clinical follow-up, but at the time of the last follow-up were without evidence of disease. These results are similar to those observed in similar studies of patients undergoing amputation for oncologic reasons with local recurrence occurring in 7%-36% of patients.^{28–30,83–86} Wound complications requiring a return to the operating room occurred in 16% (5/31) of patients, including one patient who initially underwent a below-knee amputation with TMR who required conversion to an above knee amputation with TMR because of a nonhealing stump wound. Two patients returned to the operating room for neuroma excisions and TMR of symptomatic neuromas that developed in pure sensory nerves (medial antebrachial cutaneous nerve and lateral femoral cutaneous nerve) that were not included in the initial nerve transfer. Nineteen (61.3%) patients were fitted with and received their prosthesis on average 3.6 months postoperatively (range 2-7 months). Fifteen of 27 patients (56%) were taking opioids before amputation. At 6 weeks, 56% continued to require opioid prescriptions; this percentage decreased to 26% by 3 months and plateaued to 22% by 1 year (Figure 1).

FIGURE 1.

Opioid prescriptions for TMR-treated oncologic amputees. Time zero represents the day of amputation and TMR. TMR, targeted muscle reinnervation

Demographics and details of amputation were similar between TMR and general oncologic amputee cohorts with the exception of time since amputation (Table 2). Outcomes were analyzed for patients with greater than 1-year follow-up. Neuroma symptoms, specifically burning sensations (Figure 2), tingling or crawling sensations (Figure 2), pain with light touch (Figure 2), and sudden pain episodes at the amputation site (Figure 2), occurred less frequently and with less intensity in the TMR cohort (all P < .05; Table 3). No differences were observed in hot or cold sensitivity or light pressure causing pain. Mean differences for PROMIS pain intensity, behavior, and interference for PLP were 5.855 (95%CI 1.159, 10.55; P = .015), 5.896 (95%CI 0.492, 11.30; P = .033), and 7.435 (95%CI 1.797, 13.07; P = .011), respectively, with lower scores for the TMR cohort (Table 4). For residual limb pain, PROMIS pain intensity, pain behavior, and pain interference mean differences were 5.477 (95%CI 0.528, 10.42; P = .031), 6.195 (95%CI 0.705, 11.69; P = .028), and 6.816 (95%CI 1.438, 12.2; P = .014), respectively. Comparison of TMR patients with greater than 1 year follow-up to TMR patients with less than 1 year follow-up demonstrated no significant differences (Supporting Information Table). Minimally important difference (MID) for health-related quality of life measures may be estimated as one-half of a standard deviation or five t-score points.⁸⁷ Among advanced-stage cancer patients, MID for PROMIS pain interference is estimated to be between 4.0 and 6.0.⁸⁸ In summary, pain measures across the board were statistically and meaningfully favorable for the TMR cohort.

TABLE 2.

Demographic and amputation variables

	All			>1 y
Variable	TMR, N = 27 N (%)	General, N = 58 N (%)	P value	TMR, N = 15 N (%)	General, N = 53 N (%)	P value
Age (mean, range)	53.5 (18-85)	53.6 (25-77)	.990	48.5 (18-85)	55.2 (25-77)	.148
Male	17 (63%)	31 (53%)	.410	12 (80%)	27 (51%)	.045
Race			.733			.727
Black/African American	2 (7%)	1 (2%)		0 (0%)	1 (2%)
White	24 (89%)	53 (91%)		14 (93%)	48 (91%)
Other	1 (4%)	4 (7%)		1 (7%)	4 (8%)
Time since amputation			<.05			<.05
<1 y	11 (41%)	5 (9%)	Excluded
1-4 y	16 (59%)	18 (31%)		15 (100%)	18 (34%)
>5 y	0 (0%)	35 (60%)		0 (0%)	35 (66%)
Level of amputation			.274			.339
Below elbow	0 (0%)	1 (2%)		0 (0%)	1 (2%)
Below knee	7 (26%)	15 (26%)		6 (40%)	14 (26%)
Above elbow	3 (11%)	2 (3%)		0 (0%)	2 (4%)
Above/through knee	13 (48%)	29 (50%)		7 (47%)	26 (49%)
Shoulder disarticulation	3 (11%)	2 (3%)		2 (13%)	2 (4%)
Hip disarticulation/ hemipelvectomy	1 (4%)	9 (16%)		0 (0%)	8 (15%)

Abbreviation: TMR, targeted muscle reinnervation.

FIGURE 2.

Frequency (A) and intensity (B) of neuroma symptoms at last follow-up. Dark solid bars represent the general cohort; lighter shaded bars represent the TMR cohort. All general versus TMR comparisons are statistically significant by nonparametric regression with P < .05. TMR, targeted muscle reinnervation [Color figure can be viewed at wileyonlinelibrary.com]

TABLE 3.

Neuroma symptom frequency and intensity for oncologic amputees at least 1 year from amputation

	Median Frequency/intensity TMR, N = 15	Median Frequency/intensity General, n = 53	U Frequency/intensity	P Frequency/intensity
Burning sensations	Never/none	Rarely/uncomfortable	226.5/256.5	.006/.029
Tingling/crawling sensations	Rarely/uncomfortable	Sometimes/mild	183.5/253.5	.001/.027
Pain with light touch	Never/none	Never/none	279/292	.028/.044
Sudden pain episodes	Never/none	Sometimes/moderate	180/159.5	.001/<.001
Hot/cold pain sensitivity	Never/none	Never/none	322.5/315	.072/.057
Pain with pressure	Never/none	Never/none	362/384.5	561/.829

Note: P < .05 denotes significantly different distributions Abbreviation: TMR, targeted muscle reinnervation.

TABLE 4.

PROMIS pain outcomes for oncologic amputees at least 1 year from amputation

	Mean difference	95%CI lower, upper	P
Phantom limb pain (PLP)
Pain intensity	5.855	1.159, 10.55	.015
Pain behavior	5.896	0.491, 11.30	.033
Pain interference	7.435	1.797, 13.07	.011
Residual limb pain (RLP)
Pain intensity	5.477	0.528, 10.42	.031
Pain behavior	6.195	0.705, 11.69	.028
Pain interference	6.816	1.438, 12.19	.014

Note: P < .05 denotes significance.

Abbreviation: CI, confidence interval.

4 |. DISCUSSION

4.1 |. Defining an oncologic amputee center of excellence

When amputation is indicated for oncologic reasons, a multidisciplinary team approach can maximize outcomes. The Sarcoma Division at The James Cancer and Solove Research Institute at The Ohio State University Wexner Medical Center has developed an oncologic amputee program in which limb optimization begins with prehabilitation to set patient expectations, continues with intraoperative collaboration between the oncology and plastic surgery teams to perform the resection and subsequent TMR, and a patient-specific postoperative rehabilitation. All of this occurs in the setting of a comprehensive cancer center where a close relationship exists between medical, radiation, and surgical oncologists who work together to address cancer-specific considerations in the oncologic amputee population.

In our study, TMR reduced patient-reported phantom and residual limb pain behavior and interference compared to unselected general oncologic amputee controls beyond the clinically meaningful threshold for this population. Raw PROMIS pain interference scores in our study population undergoing amputation with concurrent TMR for oncologic diagnoses demonstrated favorable outcomes in comparison to both oncologic amputees and limb salvage patients in the recent study by Wilke et al.⁸⁹ In a population with similar follow-up, they demonstrated PROMIS pain interference scores of 53.8 and 53.3 for limb salvage and amputation respectively. These are both higher (more pain interference) than our findings of phantom pain related pain interference of 42.1 and 46.2 for >1 year and <1 year follow-up, respectively, and residual limb pain related pain interference of 44.0 and 44.1 for >1 year and <1 year follow-up, respectively, after TMR. In addition, TMR patients experience less frequent and intense residual limb sensations and sudden pain and relatively low rates of opioid utilization. Notably, we also found that when pure sensory nerves such as the medial antebrachial cutaneous and lateral femoral cutaneous nerves are not included in the initial nerve transfer, symptomatic neuromas may develop leading to reoperation. This finding triggered a change in our practice to include these nerves in subsequent transfers. Finally, postoperative care can be enhanced through multimodal pain management, patient-centered rehabilitation, and finally, a peer support network that plays a crucial role throughout the entire process.

Our multidisciplinary approach was informed by lessons from the military community where recent conflicts have necessitated medical advance. Programs such as the US Armed Forces Amputee Patient Care Program at Walter Reed National Military Medical Center, the Center for the Intrepid at Brooke Army Medical Center, and Veterans Administration Regional Amputation Centers provide a comprehensive medical home for the rehabilitation of combat-related amputees and have been shown to have a positive impact on their patients’ recovery by providing durable provider-patient relationships, open communication, easy accessibility, and peer support.^76,90 The application of military medical practices in civilian medical institutions, specifically those caring for oncology populations, requires acknowledgment of their similarities and differences. Both oncologic and combat-related amputees may have high expectations for postamputation activity, and both may include young men and women. However, oncology patients may have additional medical comorbidities, metastatic disease, need for adjuvant therapies including chemotherapy and/or radiation, different emotional coping strategies due to the reason for their amputation, and overall life expectancy. With these differences in mind, it is recognized that the oncology population requires a dedicated team at a facility equipped and invested in their care to optimize both their oncologic and functional outcomes after amputation.

We acknowledge several limitations to our study. We have developed our institutional protocol based on the experience of our team members and patients over time. The patient population is heterogeneous and relatively small due to the rare nature of extremity-based malignancies necessitating amputation. Patient-reported data was analyzed at last follow-up and not tracked longitudinally to examine the effects of specific rehabilitative interventions or medication changes on pain. Surgical details, oncologic details, and opioid use were unavailable for the control cohort. We assumed that the cross-sectional survey represented the current standard of care in the United States. Furthermore, there is a significant difference in the follow-up between the control and TMR cohorts. However, previous studies have shown that the prevalence of postamputation pain, in particular, phantom limb and residual limb pain, improves with time in the general population,¹⁷ but plateaus to slightly increases with greater than 1-year follow-up.⁵⁹ Most patients experience mild symptoms in the long-term, but approximately 25%-33% continue to experience debilitating symptoms.⁴ In contrast, our experience with TMR has shown continued improvement after 1 year in NRS pain scores after the treatment of symptomatic neuromas⁵⁹ and our unpublished experience has found continued improvement in postamputation pain 1 year after primary TMR. Due to these findings, we would expect the differences between the control and TMR group to be greater with more pain experienced by the control group with shorter follow-up. Since a statistically and clinically significant difference was identified in our study between the intervention group with shorter follow-up and the control group with longer follow-up) we would expect our results to remain significant with matched follow-up of the control group.

Given the lack of long-term follow-up for patients undergoing TMR, the durability of the TMR effect is unknown. However, given the biologic healing that occurs at the nerve endings, we anticipate the effect to be long-lasting.

The use of the OARRS for assessing opioid use is limited by the nature of the data generated. The OARRS report represents prescriptions filled by the patient but does not necessarily reflect daily or actual opioid consumption. In our previous work, however, the OARRS data were consistent with self-reported use with greater than 90% agreement.⁹¹ Finally, formal assessments including functional outcomes measures and detailed prosthetic use information were beyond the scope of our investigation. Despite these limitations, this is the largest case series to-date of concurrent TMR in patients undergoing amputation for malignant diagnoses and our findings provide early support for the strategy and the basis for further clinical investigations.

5 |. CONCLUSIONS

Multidisciplinary care of the oncologic amputee that includes TMR at the time of resection, multimodal postoperative pain management, amputee-centered rehabilitation, and peer support has demonstrated favorable outcomes for reducing the incidence and severity of painful neuroma and PLP. The multidisciplinary team of medical, radiation, orthopedic and surgical oncologists, plastic surgeons, and oncology rehabilitation therapists working in concert in the care of an oncologic amputee defines, for us, an Oncologic Amputee Center of Excellence. Dedicated programs such as these will ignite innovation and serve as proving grounds for new technologies and treatment strategies for the oncologic amputee.