Optimizing the Timing of Peripheral Nerve Transfers for Functional Re-Animation in Cervical Spinal Cord Injury: A Conceptual Framework

Abstract

Loss of upper extremity function following spinal cord injury (SCI) can have devastating consequences on quality of life. Peripheral nerve transfer surgery aims to restore motor control of upper extremities following cervical SCI and is poised to revolutionize surgical management in this population. The surgery involves dividing an expendable donor nerve above the level of the spinal lesion and coapting it to a recipient nerve arising from the lesional or infralesional segment of the injured cord. In order to maximize outcomes in this complex patient population, refinements in surgical technique need to be integrated with principles of spinal cord medicine and basic science. Deciding on the ideal timing of nerve transfer surgery is one aspect of care that is critical to maximizing recovery and has received very little attention to date in the literature. This complex topic is reviewed, with a focus on expectations for spontaneous recovery within upper motor neuron components of the injury, balanced against the need for expeditious re-innervation for lower motor neuron elements of the injury. The discussion also considers the case of a patient with C6 motor complete SCI in whom myotomes without electrodiagnostic evidence of denervation spontaneously improved by 6 months post-injury, thereby adjusting the surgical plan. The relevant concepts are integrated into a clinical algorithm with recommendations that consider maximal opportunity for spontaneous clinical improvement post-injury while avoiding excessive delays that may adversely affect patient outcomes.

Introduction

Upper extremity paralysis from spinal cord injury (SCI) can have devastating consequences on function and quality of life.^1

–4 SCI often produces simultaneous, but physiologically distinct lesions within both the upper motor neuron (UMN) and lower motor neuron (LMN) circuits.^5,6 The LMN circuit extends from the anterior horn cell to the distal neuromuscular junction. If the LMN is disrupted by the injury to the anterior horn cells as part of the SCI, or with concomitant peripheral nerve injury, the peripheral targets are said to be denervated. For the purpose of distinction within this discussion, if the LMN fibers remain intact but are no longer receiving descending input because of a more cephalad UMN lesion, the peripheral target is referred to as paralyzed.

The emergence of peripheral nerve transfers as a reconstructive option has the potential to revolutionize treatment and functional recovery for debilitating cervical SCI. The surgical technique to restore upper limb function with a peripheral nerve transfer involves mobilizing an expendable/functionally redundant, intact motor nerve and coapting it to a recipient nerve close to its denervated or paralyzed peripheral target muscle.^7,8 Individuals who have undergone nerve transfers in tetraplegia show significant improvement in muscle strength and functional outcome measures.^9,10 There are many established nerve transfer options for managing complex peripheral nerve injury, but the mainstay nerve transfers to restore function in SCI include: (1) brachialis branch of the musculocutaneous nerve or extensor carpi radialis brevis branch of the radial nerve to the anterior interosseous nerve (AIN), to restore digit flexion; (2) supinator branch of the radial nerve to the posterior interosseous nerve (PIN), for digit extension; (3) posterior deltoid or teres minor branches of the axillary nerve to triceps nerve, to restore elbow extension.^11
–13

Whether for UMN or LMN injury, the primary advantage of performing a nerve transfer close to the denervated or paralyzed recipient target is that it dramatically shortens the distance and time required for re-innervation of motor endplates. If the period of denervation for LMN injury is too long (i.e., >18 months), functional re-innervation will not occur, because of the irreversible resorption of motor endplates, loss of myocyte progenitor (satellite) cells, and fibrofatty replacement of muscle tissue.^14

–18 This principle applies to injury at any position along the length of the LMN and has led to a prevailing philosophy that earlier surgical intervention avoids the harmful effects of chronic denervation.⁸ However, this conflicts with our current understanding of the natural history of spontaneous recovery in SCI within the lesional and infralesional spinal segments, in which clinical improvements can be observed for ≥1 year following injury.^19

–23 This apparent conflict and its impact on the timing of surgical intervention requires careful consideration to optimize outcomes following peripheral nerve transfers in SCI.

The timing of peripheral nerve surgery in SCI patients, while balancing the considerations for UMN and LMN lesions alone or in combination, has not been thoroughly examined. UMN considerations for surgical timing include the fact that although LMNs are not disconnected from their target muscle in the context of isolated UMN injury, they do undergo significant alteration at the neuromuscular junction and muscle composition over time.²⁴ In addition, there is a reduction in the spinal motoneuron pool available to support denervated muscle via “transsynaptic degeneration.”²⁵ Third is the consistent potential for limited, albeit meaningful spontaneous recovery of motor function through a variety of mechanisms over the first 6–12 months post-injury.^19

–23 The purpose of this review is to outline a conceptual framework around the ideal timing of nerve transfers in SCI that considers maximal opportunity for spontaneous clinical improvement post-injury, but also avoidance of excessive delays that adversely affect patient outcomes. These considerations will be discussed around a case example and a timing algorithm presented to facilitate surgical decision making in patients with cervical SCI.

Case Presentation

A 20-year-old male sustained a C5 American Spinal Injury Association Impairment Scale (AIS) A SCI (complete injury) following a motor vehicle accident. He underwent acute spinal stabilization with instrumented fusion from C5 to T2. Imaging at the time of injury demonstrated significant compression and displacement, vasogenic edema extending from C5 to T2 and the hydromyelia cavity into the lower thoracic cord and upper cervical cord (Fig. 1A). He was first evaluated in an interdisciplinary peripheral nerve injury clinic 2.5 months following injury. Motor examination, as per Medical Research Council (MRC) scale, demonstrated the following levels of strength (right and left): shoulder abduction 5 and 5, elbow flexion 5 and 5, elbow extension 4 and 4, wrist extension 5 and 5, wrist flexion 4 and 2, and finger extension (extensor digitorum communis [EDC]) 1 and 0. Digit flexion and intrinsic function were absent bilaterally (grade 0). Pinprick sensation was intact to C6 on the right and C5 on the left, with no sharp or dull discrimination below this level.

FIG. 1.

Recovery of right hand digit extension. Case of a 23-year-old male with C5 American Spinal Injury Association Impairment Scale (AIS) A SCI (complete spinal cord injury). (A) Representative sagittal magnetic resonance imaging (MRI) section from day of injury. On initial evaluation at 2.5 months post-injury, this patient had Medical Research Council (MRC) grade 1 power to right hand digit extension despite needle electromyography (EMG) demonstrating no significant active denervation and intact motor units. After 5 months, the patient had regained antigravity digit extension (grade 3): (B) rest and (C) active digit extension. At 15 months, the patient had regained extension with moderate resistance (grade 4): (D) rest and (E) active digit extension. His left hand had no spontaneous recovery by 7 months (remained at MRC grade 0) and he then underwent supinator-to- posterior interosseous nerve (PIN) nerve transfer with full active digit extension 8 months post-operatively. Color image is available online.

Electrodiagnostic (EDX) studies were performed 2.5 months post-injury including nerve conduction studies (NCS) to evaluate parameters such as conduction latency/velocity and signal amplitude, and electromyography (EMG) to assess muscle resting membrane activity and voluntary activation (Table 1). Pre-operative EDX is an essential element of the planning stages of nerve transfers in cervical SCI, with both NCS and EMG each providing useful information on donor and recipient nerve integrity.^5,26,27 There was evidence of an infralesional LMN injury involving C8-T1 based on the muscles examined. EDC did not demonstrate evidence of significant denervation, despite only a flicker of palpable voluntary activation on clinical examination (grade 1), raising the possibility of a UMN lesion to this myotome. Potential supralesional donors did not demonstrate any evidence of denervation. Note that EDX of several neuromuscular units was limited in this patient because of significant intolerance, for both NCS and needle EMG, and prominent autonomic dysreflexia.

Table 1.

Electrodiagnostic Data of 20-Year-Old Patient with Complete C5 ASIA A Spinal Cord Injury, Demonstrating Electrodiagnostically Normal EDC at 2.5 Months Despite MRC Grade 1 Power

Motor nerve conduction studies
Nerve /site	Muscle	Segments	Distance (mm)	Amplitude (mV)	Latency (ms)	Temp (°C)
Right median – APB
Wrist	APB	Wrist – APB	80	NR	NR	33.1
Elbow	APB	Elbow –Wrist		NR	NR	33.3
Left median – APB
Wrist	APB	Wrist – APB	80	NR	NR	33.9
Elbow	APB	Elbow –Wrist
Right ulnar – ADM
Wrist	ADM	Wrist – ADM	80	NR	NR	33.4
Left ulnar – ADM
Wrist	ADM	Wrist – ADM	80	NR	NR	34

Sensory nerve conduction studies
Nerve /site	Rec. site	Segments	Distance (mm)	Amplitude (mV)	Latency (ms)	Temp (°C)
Right median – Digit III
Wrist	Middle	Wrist – middle	140	29.4	3.3	33.4
Left median – Digit III
Wrist	Middle	Wrist – Middle	140	32.4	3.1	34
Right ulnar – Digit V (antidromic)
Wrist	ADM	Wrist – Dig V	140	20.9	3.1	33.4
Left ulnar – Digit V (antidromic)
Wrist	Dig V	Wrist – Dig V	140	18.5	3.4	34.3

Electromyography (EMG)
	Spontaneous activity				MUAP	Recruitment
Muscle	IA	Fib	PSW	Fasc	Amp	Dur	Config	Pattern
R. Brachialis	Normal	None	None	No	Normal	Normal	Normal	Normal
R. ECRB	Normal	None	None	No	Normal	Normal	Normal	Normal
R. EPL	Incr	1+	2+	No	-	-	-	No MUs
R. FDS	Normal	None	1+	No	-	-	-	1 MU - Pain
R. FDP (Ulnar)	Incr	1+	2+	No	-	-	-	Few distant MUs - pain
R. FDP (Median)	Incr	1+	3+	No	Normal	Normal	Normal	1 MU - pain
L. Brachialis	Normal	None	None	No	Normal	Normal	Normal	Normal
L. ECRB	Normal	None	None	No	Normal	Normal	Normal	Normal
L. FDP (Ulnar)	Normal	2+	1+	No	-	-	-	No MUs
L. FDP (Median)	Normal	2+	3+	No	-	-	-	No MUs
R. Supinator	Normal	None	None	No	Normal	Normal	Normal	Normal
R. EDC	Incr	None	None	No	Normal	Normal	Normal	Decr. activation with pain
L. Supinator	Normal	None	None	No	Normal	Normal	Normal	Normal
L. EDC	Incr	None	None	No	Normal	Normal	Normal	Decr activation with pain

ADM, abductor digiti minimi; amp, amplitude; APB, abductor pollicis brevis; ASIA, American Spinal Injury Association; config, configuration; decr, decreased; dig, digit; dur, duration; ECRB, extensor carpi radialis brevis; EDC, extensor digitorum communis; EPL, extensor pollicis longus; fasc, fasciculations; fib, fibrillations; FDP, flexor digitorum profundus; FDS, flexor digitorum superficialis; incr, increased; IA, insertional activity; MRC, Medical Research Council; MU, motor unit; MUAP, motor unit action potential; NR, no response; PSW, positive sharp wave.

The patient was booked for staged bilateral nerve transfers, including brachialis branch of the musculocutaneous nerve-to-AIN and the nerve to supinator-to-PIN. When the patient presented for his initial right-sided procedure at 5 months post-injury, the supinator-to-PIN nerve transfer was aborted given that he had managed to recover satisfactory anti-gravity metacarpophalangeal (MCP) joint extension (grade 3) at this early time point, an improvement from MRC grade 1 at 2.5 months, and no evidence of denervation on EMG (Fig. 1B, C). He would later go on to recover grade 4 MCP extension of second to fifth digits spontaneously on examination at the 15-month mark (Fig. 1D, E). Ultimately, the patient underwent right sided brachialis-to-AIN nerve transfer, followed by dual transfers on the left side 8 weeks later (brachialis-to-AIN and supinator-to-PIN).

This case illustrates a number of important points in the surgical decision making for nerve transfers in SCI. First, serial examination is essential given the changes in the clinical examination findings within myotomes arising from the lesional segment of the injured cord over the first 6 months following SCI. Second, it demonstrates the important role of early EDX evaluation to distinguish between UMN and LMN patterns of injury. For those myotomes demonstrating UMN involvement, surgery should be delayed for 6 months, as much of the spontaneous recovery typically occurs within this time frame.¹⁹ As seen in this case example, we observed that EDC, with reported C7 (greater than C8) predominant innervation,^28,29 improved from grade 1 (2.5 months post injury), to grade 3 (5 months), and later to grade 4 (15 months post injury). This compares with extensor pollicis longus (EPL) with reported C8 (greater than C7) predominant innervation,^28,30 which was grade 0 at 15 months.

Literature Review

UMN lesion (“paralysis”) influence on timing for peripheral nerve transfers

Consideration of the anatomy of a spinal cord lesion and its relationship to peripheral nerve transfer illustrates several important concepts. The injured spinal cord can be divided into three distinct segments: supralesional, lesional or injured metamere, and infralesional (Fig. 2).^6,31
–33 The supralesional segment of the spinal cord is located cephalad to the level of injury and contains intact, uninjured UMN and LMNs (Fig. 2A). It is from this supralesional segment that expendable donor motor nerves can be mobilized, divided distally, and coapted to recipient nerves within the lesional or infralesional segments to restore function.^11,12 The lesional segment describes the multi-level zone of injured spinal tissue that can involve the descending corticospinal tracts (UMN) and/or the anterior horn cells (LMN).⁶ A mixed UMN and LMN pattern of injury is often observed within the lesional segment (Fig. 2B).³² This zone may extend over several levels cephalad and/or caudad from the location of primary injury as a consequence of the secondary injury cascade caused by edema, venous stasis, compromised blood supply, inflammation, excitotoxins, and other factors.^6,34,35 The infralesional segment features a structurally intact LMN circuit that lacks the descending input/volitional control as a result of an UMN lesion (Fig. 2C).³¹ Within the infralesional segment, corticospinal tract axons are functionally disconnected from the LMN and will not undergo spontaneous reconnection.³⁶ Transferring a donor nerve from a supralesional to an infralesional segment restores volitional control via elongation of donor axons that will re-innervate recipient muscle motor endplates.¹¹

FIG. 2.

Levels of spinal cord injury and timing considerations. (A) Supralesional level. Muscle remains under voluntary control, as the lower motor neuron (LMN) and upper motor neuron (UMN) remain intact. LMN from this level can be used as a donor nerve. (B) Lesional level. At this level, both the UMN and LMN are injured. Consequently, motor endplates become denervated and the affected muscles undergo rapid irreversible atrophy within 12–18 months. Nerve transfers for lesional level insults are considered time sensitive to avoid terminal denervation atrophy. (C) Infralesional level. Although the UMN is injured, and muscles lose descending motor control, the integrity of the LMN and motor endplates are intact. Muscles undergo gradual disuse atrophy over time, which is mostly reversible. Nerve transfers for infralesional level insults are not considered time sensitive. Illustrated by BioHues Digital. Color image is available online.

Delayed peripheral nerve transfers are theoretically possible for UMN lesions given that the integrity of LMNs and motor endplates is maintained for an extended period of time. Muscles remain innervated and do not undergo the rapid irreversible atrophy and fibrosis, as seen in LMN lesions.^31,37,38 In a study of long-term outcomes of people with complete paraplegia, it was found that after an initial phase of rapid muscle atrophy, subsequent atrophy occurred slowly and muscle fibers maintained normal striated appearance after 15–20 years.³⁸ Consequently, restorative nerve transfers into the infralesional segment may be performed many years post-injury and have historically been considered “not time-sensitive.”^11,32,39,40

Despite the conventional notion that peripheral nerve transfers into the infralesional segment are not time-sensitive, there is an abundance of evidence demonstrating time-related changes in candidate recipient muscles within the infralesional segment. First, motor endplate numbers diminish over time with a subsequent gradual denervation of the corresponding muscle fibers.^38,41
–43 Also, predominant type II fiber atrophy is seen within the first several months along with a gradual decrease in oxidative enzymatic activity.²⁴ This is later followed by type I fiber atrophy. Muscle fiber type transitions toward the predominantly fast glycolytic type IIX fibers until the steady state is reached ∼70 months post-SCI.²⁴ The cumulative effect of these time-dependent changes diminishes muscle performance through progressively declining force-output and endurance of muscles.^24,38,42 Although there are currently no data relating to the impact of these changes on muscle performance post-nerve transfer, earlier intervention may be justified for the reasons cited. In summary, preservation of LMN integrity caudal to the level of the spinal injury allows for delayed nerve transfers to restore functional control into the chronic time frame. However, earlier intervention may prove worthwhile by avoiding the changes in recipient muscle that have been noted. The previously described case features a mixture of UMN and LMN lesions within the affected myotomes. Here, the duration in which to observe the UMN injury to EDC is constrained by the adjacent LMN pattern of injury to C8 and T1 levels.

Expectations for recovery following SCI

The potential for meaningful partial recovery during the 1st year post-injury is a concept not often discussed, yet it is highly relevant to the timing of peripheral nerve transfers in tetraplegia, particularly within the lesional segment of the injured cord. This recovery profile is summarized in the Paralyzed Veterans of America (PVA) clinical practice guidelines,⁴⁴ which outline the observed incidence of spontaneous recovery of MRC grade 3 upper extremity power in the SCI population, defined as full range of motion against gravity.⁴⁵ The extent of motor recovery in complete SCI is greatest in the first 3 months post-injury, declining until 6 months post-injury, and plateauing at ∼9–12 months.^19,45,46 Spontaneous recovery is stratified by spinal level in reference to the lowest spinal level demonstrating at least grade 3 power on initial examination, within 3 days post-injury. This level corresponds to the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) level of motor injury, hereafter referred to as the “motor level.” One spinal level below this motor level, 82% of grade 1–2 muscles will reach grade 3 power or better at 6 months post-injury, with 90% reaching antigravity power at 1 year post-injury.^20,44 In comparison, only 36% of grade 0 muscles at this spinal level will reach grade 3 power or better by 6 months and only 45% will reach this level by 12 months.^20,23,44 The likelihood of MRC grade 0 muscles gaining antigravity power at 1 year that are located ≥2 levels below the motor level is close to 1%.^19,44 Although neurological recovery beyond 1 year has been described,⁴⁷ it is likely unreliable for the purposes of timing nerve transfer surgery and should be considered on a case-by-case basis. The mechanisms responsible for this spontaneous recovery after SCI include both passive processes such as resolution of spinal shock and re-myelination, as well as active processes such as spinal circuit reorganization and sensorimotor cortex adaptation.^48
–50 Several of these pathophysiological mechanisms are reviewed in more detail in Figure 3, but are otherwise beyond the scope of this discussion.

FIG. 3.

Mechanisms of spontaneous recovery in spinal cord injury (SCI). Initial short-term improvement in neurological function is attributed to processes that do not rely on neuronal elongation or re-connectivity. These include the resolution of spinal shock and/or focal conduction block, as edema decreases and re-myelination of viable axons occurs.^50,53,72 Neuronal regeneration and mechanisms of spinal cord plasticity also contribute to observed recovery. Local astrocyte proliferation and cellular activity create a dense capsule around the lesion, effectively isolating it from the surrounding tissue. (A) For all practical purposes, spontaneous regeneration of corticospinal neurons through the astroglial scar and lesion core does not contribute appreciably to recovery. However, several mechanisms involving spinal circuit reorganization and sensorimotor cortex adaptation contribute to motor recovery following complete SCI.^73

–76 These mechanisms include (B) neuronal regeneration through spared reactive tissue, (C) collateral sprouting from spared or damaged tracts and spinal interneurons, and (D) reorganization of the spinal neuronal networks via synaptic remodelling and formation of bypass circuits around lesioned tissue.^73,75
–77 Finally, cortical remodelling results in significant synaptic re-connectivity and expanded motor representation of uninjured regions to assume control of the paralyzed territory and permit functional recovery.^49,75 Illustrated by BioHues Digital. Color image is available online.

In contrast to complete SCI, incomplete injuries typically have a more favorable prognosis, and recover more quickly and to a greater extent.^46,51,52 The time frame for observed changes in motor recovery for incomplete injuries is similar, and the rate of improvement is still greatest within the first 3 months and mostly complete by 6 months.⁵³ Mange and coworkers²¹ reported that the median time to reach grade 3 power is 2 weeks for individuals with motor incomplete injuries, versus 2 months for those with complete injuries.⁴⁴ Despite having greater potential, individuals with incomplete SCI may still experience insufficient spontaneous recovery and would be candidates for nerve transfer surgery. Candidacy, timing, and execution would be guided by the same principles as for nerve transfers in complete SCI, as demonstrated in previous case reports.⁵⁴

Given the foregoing discussion outlining spontaneous recovery reaching a maximum at 12 months, nerve transfers for a pure UMN lesion could be delayed until after 1 year post-injury (Fig. 4A). However, in practice, cervical SCI presents with a heterogeneous and complex pattern of UMN and LMN injury;^5,55 therefore, purely UMN injury patterns are rarely encountered. This section highlights the inherent capacity for limited spontaneous recovery within the 1st year after SCI. This fact needs to be balanced against the priority for early surgical intervention to address the LMN components of the injury.

FIG. 4.

Recommended timing for nerve transfer surgery. (A) Isolated upper motor neuron (UMN) injury: after 12 months. (B) Isolated lower motor neuron (LMN) injury: 3–6 months. (C) Mixed LMN and UMN injury: 6–9 months. Green, favorable time frame for a nerve transfer; red, unfavorable time frame for a nerve transfer. Gradient reflects the degree of how favorable/unfavorable a particular time is. Color image is available online.

LMN lesion (“denervation”) influence on timing for peripheral nerve transfers

Injury to the LMN as part of SCI has its own considerations for the timing nerve transfer surgery, including time from injury and distance for donor axons to reach their target muscle after reconstruction.^33,39 Whether at the level of the anterior horn cell or more distally within the peripheral nerve, injury to the LMN leads to Wallerian degeneration and subsequent denervation of motor endplates.^17,56 As a consequence, within 1 month of denervation, muscle loses >60% of its mass and 90% of its maximal force as a result of atrophy.⁸ Irreversible muscle atrophy and fibrosis occurs within 12–18 months from denervation, defining the maximal time frame in which donor axons must reach their targets following nerve transfers.^39,57

–60 In the context of suspected concomitant peripheral nerve injury, an appropriate period of observation for spontaneous recovery is warranted given that spontaneous recovery is often functionally superior than that achieved with nerve transfer surgery.⁶¹ However, once a decision to perform nerve transfers is made, there is evidence that outcomes are superior if surgery is performed within 6 months from the time of injury.^61,62 It follows that a reasonable time frame for surgical intervention for an isolated LMN injury is between 3 and 6 months, with earlier transfers for more proximal injuries (Fig. 4B).

Two additional points bear consideration related to the timing of nerve transfers in the context of LMN injury. First, partial spontaneous recovery may plateau after an appropriate period of observation. In the event that the degree of recovery is not sufficient for function, nerve transfers may be desired, but without compromising the recovery already achieved. In these circumstances, an end-to-side nerve transfer may be employed, whereby the donor nerve is coapted to the side of the recipient nerve via an epineurial window. This technique spares disruption of partially retained or recovered motoneurons and is a well-justified technique to augment native motor recovery.^{39,63

–67} Second, given that the surgical technique involves complete axotomy of the recipient nerve, the design of the nerve transfer must avoid a prohibitively long distance from the location of the nerve transfer to the target musculature. The most common nerve transfers performed for cervical SCI described in the introduction typically allow motor endplate re-innervation within 1 year post injury. Relating back to the case presented, the LMN injuries to C8 and T1 levels dictate a preferred time frame of 6 months within which to perform the brachialis to anterior interosseous nerve transfers.

Mixed UMN and LMN lesions and considerations for peripheral nerve transfer timing

As discussed, cervical spinal cord lesions present with a heterogenous pattern of UMN and LMN involvement throughout the lesional and infralesional segments. Accordingly, selecting the ideal timing for nerve transfers in SCI requires balanced consideration of potential spontaneous recovery of UMN components and expeditious re-innervation for LMN components of the injury. For example, if a given muscle manifesting paralysis exhibits grade 0 power at 6 months, the likelihood of gaining meaningful functional recovery is very low, and therefore surgery is indicated.^31,68,69 The upper time limit in a mixed pattern of injury is bounded by the time frame for irreversible denervation, 12–18 months, while taking into account the anticipated time required by donor axons to reach their target. The lower limit of timing is defined by the window of expected spontaneous recovery in SCI; namely, 6–9 months. Blending these considerations defines an ideal window of nerve transfers of 6–9 months, and ideally no later than 12 months if an associated LMN pattern exists (Fig. 4C).^42,43,60,70 The case presented here highlights the importance of a reasonable period of observation for myotomes arising from the lesional segment of the injured spinal cord. Had the supinator to posterior interosseous nerve transfer been performed at an earlier time point on the right side, the eventual native spontaneous digit extensor function would have been sacrificed along with the resulting natural, intuitive control that the patient has now recovered.

Timing for nerve transfers in the setting of chronic SCI, in which LMN considerations no longer factor into surgical timing, can be performed at the discretion of the reconstructive team. That said, the impact of chronic structural and physiological changes in the recipient muscle and surrounding soft tissues (such as joint contractures, for example) ought to be considered.⁵ Further, general patient-related factors such as age, medical comorbidities, associated injuries, and soft tissue characteristics may also impact surgical candidacy and should factor into timing and decision making. The impact of chronic changes in muscle and the LMN on outcomes following nerve transfers has yet to be defined.

Recommendations

The goal of peripheral nerve transfer surgery in SCI is to restore function to the upper extremities when meaningful spontaneous recovery is unlikely to occur. Optimizing outcomes requires thoughtful decision making around the timing of surgery. Operating too early on recipient nerves destined for meaningful spontaneous recovery exposes patients to unnecessary surgical risk and in the event of a failed nerve transfer, compromises their functional outcome. Delaying a nerve transfer too long for LMN components of the SCI, or concomitant peripheral nerve injury, limits the potential for re-innervation, because of irreversible muscle atrophy and lost motor endplates. The following recommendations are based on the reasonable assumptions that meaningful recovery in this population includes at least antigravity power and that for a given level of power, spontaneous recovery is preferred over that gained by a nerve transfer. Accordingly, the following proposed algorithm considers: (1) the central role of the ISNCSCI intake examination within 3 days post injury, as the most uniformly accepted prognostic parameter for predicting neurological recovery post-SCI;⁷¹ (2) the importance of electrodiagnostic studies to distinguish between paralyzed and denervated muscles, as defined previously; and (3) the need for expeditious timing of nerve transfers for denervated muscles.

All patients being screened for candidacy for peripheral nerve transfer should have an ISNCSCI examination performed, ideally within the first 3 days following injury, consistent with current clinical guidelines. If the ISNCSCI examination is outside of this time frame, then the time from injury must be taken into account to judge expected spontaneous recovery, but ultimately, decision making in this case is reliant on serial examination.

EDX studies should be conducted within an 8–12 week time frame post-injury, which accounts for practical issues such as recovery from other major injuries, improvement in general medical status, intensive care unit (ICU) discharge, and resolution of spinal shock. EDX studies should not be performed earlier than 6–8 weeks post-injury so that LMN injury associated denervation changes can be detected.

The following flow chart incorporates these varying concepts to facilitate decision making around the timing of peripheral nerve transfer surgery in SCI (Fig. 5). In this algorithm, certain principles are emphasized:

○ Observation for 6 months post-injury is indicated for all muscle groups at and above (cephalad) the motor level, as identified at the intake ISNCSCI examination.

○ For muscle groups one level below the motor level, the following should be considered in the context of the EDX findings:

▪ For muscle groups demonstrating grade 1–2 power with no features of denervation/LMN pattern on EDX, clinical reassessment is indicated at 3 and 5 months post-injury given that the UMN component of these injuries is likely to recover antigravity power spontaneously.

▪ Muscle groups demonstrating grade 1-2 power with features of denervation/LMN pattern on EDX should be reassessed clinically at 3 and 5 months post-injury. If no nascent motor units are seen, clinical improvements are less likely and end-to-side nerve transfer is indicated.

▪ Muscle groups demonstrating grade 0 power with features of denervation/LMN pattern on EDX warrant clinical reassessment at 3 and 5 months. Nerve transfers are indicated within 6 months if there is no clinical or electrodiagnostic improvement.

▪ Muscle groups demonstrating grade 0 power with no EDX features of denervation/LMN pattern should be considered for nerve transfers within 1 year.

○ For muscle groups greater than two spinal levels below the motor level:

▪ In these circumstances, without the expectation for meaningful functional recovery in UMN lesions at this level (< 1%), proceed with nerve transfers within 6 months in accordance with the ideal timing for any associated LMN injuries.

FIG. 5.

Algorithm for timing of peripheral nerve transfers in tetraplegia. Color image is available online.

Conclusion

Nerve transfers in tetraplegia have remarkable potential to restore function, yet there is a lack of guidance regarding the timing of peripheral nerve surgery given the balance between allowing spontaneous recovery from paralysis, while avoiding the irreversible effects of denervation. Thoughtful decision making around the timing of peripheral nerve transfers in SCI is critical to achieving good outcomes. The proposed algorithm begins this important discussion and serves as a framework based on existing evidence and conventional practices in SCI medicine and peripheral nerve surgery. In the case presented, serial examination revealed satisfactory spontaneous recovery of digit extensor function one level caudal to the INSCSCI motor level of impairment within 6 months from injury; therefore, the nerve transfer was aborted. This case example illustrates the importance of serial examination as well as carefully balancing spontaneous recovery after SCI with the time sensitivity of nerve transfer surgeries.

Future Directions

This study proposes a conceptual framework to inform the decision-making process when planning peripheral nerve transfers for functional reanimation in cervical SCI. Advancing the surgical care of these patients demands high-quality, prospective data from which these recommendations can be validated, and functional outcomes objectively compared. Comparisons within prospectively acquired natural history databases of cases with and without nerve transfer intervention, or well-executed meta-analyses are potential research designs that future work will explore pre- and post-implementation of this clinical algorithm.

Footnotes

Acknowledgments

The authors thank the Department of Surgery at Queen's University for supporting this publication.

Institutional Review Board Approval

Ethics approval was received from Queen's University's Health Sciences and Affiliated Teaching Hospitals Research Ethics Board (#6030062).

Authors' Contributions

V.C. and J. M.H. were responsible for the study conception, design, manuscript drafting, and substantive revisions. M.B., B.R., and J.T. were responsible for critical analysis and substantive revisions of manuscript. All authors critically reviewed and approved the final manuscript.

Funding Information

The authors acknowledge funding from the Department of Surgery, Queen's University.

Author Disclosure Statement

No competing financial interests exist.

References

Karsy

, Watkins

, Jensen

M.R.

, Guan

, Brock

A.A.

, and Mahan

M.A.

(2019). Trends and cost analysis of upper extremity nerve injury using the national (nationwide) inpatient sample. World Neurosurg. 123, e488–e500.

Taylor

C.A.

, Braza

, Rice

J.B.

, and Dillingham

(2008). The incidence of peripheral nerve injury in extremity trauma. Am. J. Phys. Med. Rehabil. 87, 381–385.

Noble

, Munro

C.A.

, Prasad

V.S.

, and Midha

(1998). Analysis of upper and lower extremity peripheral nerve injuries in a population of patients with multiple injuries. J. Trauma, 45, 116–122.

Wojtkiewicz

D.M.

, Saunders

, Domeshek

, Novak

C.B.

, Kaskutas

, and Mackinnon

S.E.

(2015). Social impact of peripheral nerve injuries. Hand (N Y) 10, 161–167.

Berger

M.J.

, Robinson

, and Krauss

E.M.

(2021). Lower motor neuron abnormality in chronic cervical spinal cord injury: implications for nerve transfer surgery. J. Neurotrauma 38 [Epub ahead of print].

Coulet

, Allieu

, and Chammas

(2002). Injured metamere and functional surgery of the tetraplegic upper limb. Hand Clin. 18, 399–412.

Tung

T.H.

, and Mackinnon

S.E.

(2010). Nerve transfers: indications, techniques, and outcomes. J. Hand Surg. Am. 35, 332–341.

Brown

J.M.

, Shah

M.N.

, and Mackinnon

S.E.

(2009). Distal nerve transfers: a biology-based rationale. Neurosurg. Focus, 26, 1–7.

van Zyl

, Hill

, Cooper

, Hahn

, and Galea

M.P.

(2019). Expanding traditional tendon-based techniques with nerve transfers for the restoration of upper limb function in tetraplegia: a prospective case series. Lancet, 394, 565–575.

10.

Khalifeh

J.M.

, Dibble

C.F.

, Van Voorhis

, Doering

, Boyer

M.I.

, Mahan

M.A.

, Wilson

T.J.

, Midha

, Yang

L.J.S.

, and Ray

W.Z.

(2019). Nerve transfers in the upper extremity following cervical spinal cord injury. Part 1: systematic review of the literature. J. Neurosurg. Spine, 31, 629–640.

11.

Fox

I.K.

(2016). Nerve transfers in tetraplegia. Hand Clin. 32, 227–242.

12.

Fox

I.K.

, Novak

C.B.

, Krauss

E.M.

, Hoben

G.M.

, Zaidman

C.M.

, Ruvinskaya

, Juknis

, Winter

A.C.

, and Mackinnon

S.E.

(2018). The use of nerve transfers to restore upper extremity function in cervical spinal cord injury. PM R, 10, 1173–1184.

13.

Bertelli

J.A.

, and Ghizoni

M.F.

(2017). Nerve transfers for restoration of finger flexion in patients with tetraplegia. J. Neurosurg. Spine, 26, 55–61.

14.

Furey

M.J.

, Midha

, Xu

Q.G.

, Belkas

, and Gordon

(2007). Prolonged target deprivation reduces the capacity of injured motoneurons to regenerate. Neurosurgery, 60, 723–732.

15.

S.Y.

, and Gordon

(1995). Contributing factors to poor functional recovery after delayed nerve repair: prolonged axotomy. J. Neurosci. 15, 3876–3885.

16.

Gutmann

(1948). Effect of delay of innervation on recovery of muscle after nerve lesions. J. Neurophysiol. 11, 279–294.

17.

Gutmann

, and Young

J.Z.

(1944). The re-innervation of muscle after various periods of atrophy. J. Anat. 78, 15–43.

18.

Irintchev

, Draguhn

, and Wernig

(1990). Reinnervation and recovery of mouse soleus muscle after long-term denervation. Neuroscience, 39, 231–243.

19.

Waters

R.L.

, Adkins

R.H.

, Yakura

J.S.

, and Sie

(1993). Motor and sensory recovery following complete tetraplegia. Arch. Phys. Med. Rehabil. 74, 242–247.

20.

Ditunno

J.F.

, Stover

S.L.

, Freed

M.M.

, and Ahn

J.H.

(1992). Motor recovery of the upper extremities in traumatic quadriplegia: a multicenter study. Arch. Phys. Med. Rehabil. 73, 431–436.

21.

Mange

K.C.

, Ditunno

J.F.

, Herbison

G.J.

, and Jaweed

M.M.

(1990). Recovery of strength at the zone of injury in motor complete and motor incomplete cervical spinal cord injured patients. Arch. Phys. Med. Rehabil. 71, 562–565.

22.

Mange

K.C.

, Marino

R.J.

, Gregory

P.C.

, Herbison

G.J.

, and Ditunno

J.F.

(1992). Course of motor recovery in the zone of partial preservation in spinal cord injury. Arch. Phys. Med. Rehabil. 73, 437–441.

23.

, Marino

R.J.

, Herbison

G.J.

, and Ditunno

J.F.

(1992). Recovery of zero-grade muscles in the zone of partial preservation in motor complete quadriplegia. Arch. Phys. Med. Rehabil. 73, 40–43.

24.

Biering-Sorensen

, Kristensen

I.B.

, Kjaer

, and Biering-Sorensen

(2009). Muscle after spinal cord injury. Muscle Nerve, 40, 499–519.

25.

Benecke

, Berthold

, and Conrad

(1983). Denervation activity in the EMG of patients with upper motor neuron lesions: time course, local distribution and pathogenetic aspects. J. Neurol. 230, 143–151.

26.

Jain

N.S.

, Hill

E.J.R.

, Zaidman

C.M.

, Novak

C.B.

, Hunter

D.A.

, Juknis

, Ruvinskaya

, Kennedy

C.R.

, Vetter

, Mackinnon

S.E.

, and Fox

I.K.

(2020). Evaluation for late nerve transfer surgery in spinal cord injury: predicting the degree of lower motor neuron injury. J. Hand Surg. Am. 45, 95–103.

27.

Hill

E.J.R.

, El-Haj

, Giles

J.A.

, and Fox

I.K.

(2020). Using electrodiagnostics to define injury patterns amenable to nerve transfer surgery in tetraplegia: an illustrative case report. Spinal Cord Ser. Cases, 6, 1–6.

28.

Wilbourn

A.J.

, and Aminoff

M.J.

(1998). AAEM minimonograph 32: The electrodiagnostic examination in patients with radiculopathies. American association of electrodiagnostic medicine. Muscle Nerve, 21, 1612–1631.

29.

Preston

, and Shapiro

(2013). Electromyography and Neuromuscular Disorders, 3rd ed. Elsevier Inc.: Amsterdam.

30.

Brendler

S.J.

(1968). The human cervical myotomes: functional anatomy studied at operation. J. Neurosurg. 28, 105–111.

31.

Senjaya

, and Midha

(2013). Nerve transfer strategies for spinal cord injury. World Neurosurg. 80, e319–326.

32.

Brown

J.M.

(2011). Nerve transfers in tetraplegia I: Background and technique. Surg. Neurol. Int. 2, 1–9.

33.

Midha

, and Grochmal

(2019). Surgery for nerve injury: current and future perspectives. J. Neurosurg. 130, 675–685.

34.

Berman

S.A.

, Young

R.R.

, Sarkarati

, and Shefner

J.M.

(1996). Injury zone denervation in traumatic quadriplegia in humans. Muscle Nerve, 19, 701–706.

35.

Ahuja

C.S.

, Nori

, Tetreault

, Wilson

, Kwon

, Harrop

, Choi

, and Fehlings

M.G.

(2017). Traumatic spinal cord injury-repair and regeneration. Neurosurgery, 80, S9–S22.

36.

Serradj

, Agger

S.F.

, and Hollis

E.R.

(2017). Corticospinal circuit plasticity in motor rehabilitation from spinal cord injury. Neurosci. Lett. 652, 94–104.

37.

Wilson

T.J.

(2019). Novel uses of nerve transfers. Neurotherapeutics, 16, 26–35.

38.

Kern

, Hofer

, Modlin

, Mayr

, Vindigni

, Zampieri

, Boncompagni

, Protasi

, and Carraro

(2008). Stable muscle atrophy in long-term paraplegics with complete upper motor neuron lesion from 3- to 20-year SCI. Spinal Cord, 46, 293–304.

39.

Hill

E.J.R.

, and Fox

I.K.

(2019). Current best peripheral nerve transfers for spinal cord injury. Plast. Reconstr. Surg. 143, 184e–198e.

40.

Fox

I.K.

, Novak

C.B.

, Kahn

L.C.

, Mackinnon

S.E.

, Ruvinskaya

, and Juknis

(2018). Using nerve transfer to restore prehension and grasp 12 years following spinal cord injury: a case report. Spinal Cord Ser. Cases, 4, 1–5.

41.

Cisterna

B.A.

, Cardozo

, and Saez

J.C.

(2014). Neuronal involvement in muscular atrophy. Front. Cell. Neurosci. 8, 1–11.

42.

Khalifeh

J.M.

, Dibble

C.F.

, Van Voorhis

, Doering

, Boyer

M.I.

, Mahan

M.A.

, Wilson

T.J.

, Midha

, Yang

L.J.S.

, and Ray

W.Z.

(2019). Nerve transfers in the upper extremity following cervical spinal cord injury. Part 2: Preliminary results of a prospective clinical trial. J. Neurosurg. Spine, 31, 641–653.

43.

van Zyl

, Hahn

J.B.

, Cooper

C.A.

, Weymouth

M.D.

, Flood

S.J.

, and Galea

M.P.

(2014). Upper limb reinnervation in C6 tetraplegia using a triple nerve transfer: case report. J. Hand Surg. Am. 39, 1779–1783.

44.

Consortium for Spinal Cord Medicine (2000). Outcomes following traumatic spinal cord injury: clinical practice guidelines for health-care professionals. J. Spinal Cord Med. 23, 289–316.

45.

Burns

A.S.

, and Ditunno

J.F.

(2001). Establishing prognosis and maximizing functional outcomes after spinal cord injury: a review of current and future directions in rehabilitation management. Spine, 26, S137–145.

46.

Fawcett

J.W.

, Curt

, Steeves

J.D.

, Coleman

W.P.

, Tuszynski

M.H.

, Lammertse

, Bartlett

P.F.

, Blight

A.R.

, Dietz

, Ditunno

, Dobkin

B.H.

, Havton

L.A.

, Ellaway

P.H.

, Fehlings

M.G.

, Privat

, Grossman

, Guest

J.D.

, Kleitman

, Nakamura

, Gaviria

, and Short

(2007). Guidelines for the conduct of clinical trials for spinal cord injury as developed by the ICCP panel: spontaneous recovery after spinal cord injury and statistical power needed for therapeutic clinical trials. Spinal Cord, 45, 190–205.

47.

Choe

A.S.

, Belegu

, Yoshida

, Joel

, Sadowsky

C.L.

, Smith

S.A.

, van Zijl

P.C.

, Pekar

J.J.

, and McDonald

J.W.

(2013). Extensive neurological recovery from a complete spinal cord injury: a case report and hypothesis on the role of cortical plasticity. Front. Hum. Neurosci. 7, 1–15.

48.

O'Shea

T.M.

, Burda

J.E.

, and Sofroniew

M.V.

(2017). Cell biology of spinal cord injury and repair. J. Clin. Invest. 127, 3259–3270.

49.

Anastakis

D.J.

, Malessy

M.J.

, Chen

, Davis

K.D.

, and Mikulis

(2008). Cortical plasticity following nerve transfer in the upper extremity. Hand Clin. 24, 425–444.

50.

Kakulas

B.A.

, and Kaelan

(2015). The neuropathological foundations for the restorative neurology of spinal cord injury. Clin. Neurol. Neurosurg. 129, 1–7.

51.

Kunam

V.K.

, Velayudhan

, Chaudhry

Z.A.

, Bobinski

, Smoker

W.R.K.

, and Reede

D.L.

(2018). Incomplete cord syndromes: clinical and imaging review. Radiographics, 38, 1201–1222.

52.

Ditunno

J.F.

, Cohen

M.E.

, Hauck

W.W.

, Jackson

A.B.

, and Sipski

M.L.

(2000). Recovery of upper-extremity strength in complete and incomplete tetraplegia: a multicenter study. Arch. Phys. Med. Rehabil. 81, 389–393.

53.

Waters

R.L.

, Adkins

R.H.

, Yakura

J.S.

, and Sie

(1994). Motor and sensory recovery following incomplete tetraplegia. Arch. Phys. Med. Rehabil. 75, 306–311.

54.

Bertelli

J.A.

, and Ghizoni

M.F.

(2012). Transfer of nerve branch to the brachialis to reconstruct elbow extension in incomplete tetraplegia: case report. J. Hand Surg Am. 37, 1990–1993.

55.

Steeves

J.D.

, Kramer

J.K.

, Fawcett

J.W.

, Cragg

, Lammertse

D.P.

, Blight

A.R.

, Marino

R.J.

, Ditunno

J.F.

, Coleman

W.P.

, Geisler

F.H.

, Guest

, Jones

, Burns

, Schubert

, Van Hedel

H.J.A.

, and Curt

(2011). Extent of spontaneous motor recovery after traumatic cervical sensorimotor complete spinal cord injury. Spinal Cord, 49, 257–265.

56.

Gaudet

A.D.

, Popovich

P.G.

, and Ramer

M.S.

(2011). Wallerian degeneration: Gaining perspective on inflammatory events after peripheral nerve injury. J. Neuroinflammation, 8, 1–13.

57.

Forli

, Bouyer

, Aribert

, Curvale

, Delord

, Corcella

, and Moutet

(2017). Upper limb nerve transfers: a review. Hand Surg. Rehabil. 36, 151–172.

58.

Fox

I.K.

, Miller

A.K.

, and Curtin

C.M.

(2018). Nerve and tendon transfer surgery in cervical spinal cord injury: individualized choices to optimize function. Top. Spinal Cord Inj. Rehabil. 24, 275–287.

59.

Nath

R.K.

, and Mackinnon

S.E.

(2000). Nerve transfers in the upper extremity. Hand Clin. 16, 131–139.

60.

Bulstra

L.F.

, and Shin

A.Y.

(2016). Nerve transfers to restore elbow function. Hand Clin. 32, 165–174.

61.

Robinson

L.R.

(2018). Predicting recovery from peripheral nerve trauma. Phys. Med. Rehabil. Clin. N. Am. 29, 721–733.

62.

Martin

, Senders

J.T.

, DiRisio

A.C.

, Smith

T.R.

, and Broekman

M.L.D.

(2018). Timing of surgery in traumatic brachial plexus injury: a systematic review. J. Neurosurg. 130, 1333–1345.

63.

Viterbo

, Trindade

J.C.

, Hoshino

, and Mazzoni Neto

(1992). Latero-terminal neurorrhaphy without removal of the epineural sheath. Experimental study in rats. Rev. Paul. Med. 110, 267–275.

64.

Hayashi

, Pannucci

, Moradzadeh

, Kawamura

, Magill

, Hunter

D.A.

, Tong

A.Y.

, Parsadanian

, Mackinnon

S.E.

, and Myckatyn

T.M.

(2008). Axotomy or compression is required for axonal sprouting following end-to-side neurorrhaphy. Exp. Neurol. 211, 539–550.

65.

Dunn

J.C.

, Gonzalez

G.A.

, Fernandez

, Orr

J.D.

, Polfer

E.M.

, and Nesti

L.J.

(2021). Supercharge end-to-side nerve transfer: systematic review. Hand (N Y) 16, 151–156.

66.

Davidge

K.M.

, Yee

, Moore

A.M.

, and Mackinnon

S.E.

(2015). The supercharge end-to-side anterior interosseous-to-ulnar motor nerve transfer for restoring intrinsic function: clinical experience. Plast. Reconstr. Surg. 136, 344e–352e.

67.

Power

H.A.

, Kahn

L.C.

, Patterson

M.M.

, Yee

, Moore

A.M.

, and Mackinnon

S.E.

(2020). Refining indications for the supercharge end-to-side anterior interosseous to ulnar motor nerve transfer in cubital tunnel syndrome. Plast. Reconstr. Surg. 145, 106e–116e.

68.

Cain

S.A.

, Gohritz

, Friden

, and van Zyl

(2015). Review of upper extremity nerve transfer in cervical spinal cord injury. J. Brachial Plex. Peripher. Nerve Inj. 10, e34–e42.

69.

Bertelli

J.A.

, Tacca

C.P.

, Ghizoni

M.F.

, Kechele

P.R.

, and Santos

M.A.

(2010). Transfer of supinator motor branches to the posterior interosseous nerve to reconstruct thumb and finger extension in tetraplegia: case report. J. Hand Surg. Am. 35, 1647–1651.

70.

Bertelli

J.A.

, and Ghizoni

M.F.

(2015). Nerve transfers for elbow and finger extension reconstruction in midcervical spinal cord injuries. J. Neurosurg. 122, 121–127.

71.

Scivoletto

, Tamburella

, Laurenza

, Torre

, and Molinari

(2014). Who is going to walk?. A review of the factors influencing walking recovery after spinal cord injury. Front. Hum. Neurosci. 8, 1–11.

72.

Little

J.W.

, Ditunno

J.F.

, Stiens

S.A.

, and Harris

R.M.

(1999). Incomplete spinal cord injury: neuronal mechanisms of motor recovery and hyperreflexia. Arch. Phys. Med. Rehabil. 80, 587–599.

73.

Onifer

S.M.

, Smith

G.M.

, and Fouad

(2011). Plasticity after spinal cord injury: Relevance to recovery and approaches to facilitate it. Neurotherapeutics, 8, 283–293.

74.

Brown

A.R.

, and Martinez

(2018). Ipsilesional motor cortex plasticity participates in spontaneous hindlimb recovery after lateral hemisection of the thoracic spinal cord in the rat. J. Neurosci. 38, 9977–9988.

75.

Brown

A.R.

, and Martinez

(2019). From cortex to cord: motor circuit plasticity after spinal cord injury. Neural Regen. Res. 14, 2054–2062.

76.

Serradj

, Agger

S.F.

, and Hollis

E.R.

(2017). Corticospinal circuit plasticity in motor rehabilitation from spinal cord injury. Neurosci. Lett. 652, 94–104.

77.

Zavvarian

M.M.

, Hong

, and Fehlings

M.G.

(2020). The functional role of spinal interneurons following traumatic spinal cord injury. Front. Cell. Neurosci. 14, 127.