Advances in the repair of segmental nerve injuries and trends in reconstruction

Abstract

Despite advances in surgery, the reconstruction of segmental nerve injuries continues to pose challenges. In this review, current neurobiology regarding regeneration across a nerve defect is discussed in detail. Recent findings include the complex roles of non-neuronal cells in nerve defect regeneration, such as the role of the innate immune system in angiogenesis and how Schwann cells migrate within the defect. Clinically, the repair of nerve defects is still best served by using nerve autografts with the exception of small, non-critical sensory nerve defects, which can be repaired using autograft alternatives, such as processed or acellular nerve allografts. Given current clinical limits for when alternatives can be utilized, advanced solutions to repair nerve defects demonstrated in animals are highlighted. These highlights include alternatives designed with novel topology and materials, delivery of drugs specifically known to accelerate axon growth, and greater attention to the role of the immune system.

Introduction

Traumatic peripheral nerve injuries are common and caused by factors ranging from acts of violence, motor vehicle accidents, and recreational activities, to iatrogenic injuries during surgery. The majority of nerve injuries occur in the upper extremity^1,2. It is estimated that 1–3% of all upper extremity trauma patients are diagnosed with nerve injuries during the first few months after trauma^3–5. These injuries are often severely debilitating, resulting in lifestyle disruptions from loss of function, both at work and in leisure^6–12. Moreover, traumatic nerve injuries frequently affect relatively young individuals, resulting in lifelong reductions in quality of life and income^13–15,4.

Despite advances in surgery and neuroscience, the improvement of patient outcomes by the surgical reconstruction of nerve injuries continues to pose challenges. The most severe nerve injuries, in which trauma to the nerve generates a defect between the nerve ends, remain an area in significant need of improvement and further research. This review will discuss the current known neurobiology of regeneration across a nerve defect, and present current clinically available options for nerve defect repair. Additionally, we will provide a brief overview highlighting select areas of the large body of experimental animal work towards advanced solutions to repair nerve defects.

Biology of nerve regeneration across a defect

Peripheral nerve is capable of robust regeneration following injury. The molecular and cellular mechanisms have primarily been studied in rodent models. Following injury, neurons and their axons and the non-neuronal cellular environment distal to the injury undergo immediate morphological and molecular changes. Within the axon, there is a rapid influx of ions, principally calcium, as well as a disruption of transport proteins signaling a disruption to homeostasis with its end-organ. This multifactorial injury response from axon damage is rapidly transported to the neuron cell body leading to an upregulation of a regeneration associated gene (RAG) program¹⁶. This RAG program includes the upregulation of inflammatory genes, neurotrophic factors, and cytoskeletal protein-related genes serving to promote axon outgrowth from the damaged end of the proximal axon^16,17. While this RAG program is upregulated in surviving neurons, some neurons will not survive the injury. Specifically, motor neurons are generally spared from cell death if the axonal injury is distal enough from the cell bodies¹⁸, while sensory neuron death can be as high as 40% regardless of injury location^19–22.

Distal to the injury site, the process of Wallerian degeneration is initiated, which involves the fragmentation of axons disconnected from the neuron, and non-neuronal cell activation that primes the distal nerve environment for new axon outgrowth. Schwann cells (SCs) sensing the axon damage dedifferentiate and adopt a unique “repair” phenotype^23–26. SC signaling involved in this dedifferentiation relies on Notch signaling and ERK-mediated signaling pathways^27–29, and is orchestrated by the phosphorylation of c-Jun^30,23. Following c-Jun phosphorylation, myelination genes are down-regulated while growth promoting genes are up-regulated³⁰. These activated repair SCs serve multiple functions. First, SCs begin to phagocytose axon and myelin debris around the site of injury to permit future axon growth. This debris is also mitogenic and thereby plays a role in SC proliferation post-injury^31,32. Secondly, SCs begin to express neurotrophic factors and inflammatory factors, such as chemokines and cytokines^33–36, through activation of signaling pathways, such as Sox-2³⁷, which serve to recruit cells of the innate immune response, such as neutrophils³⁸ and macrophages³⁹. Both hematogenous and resident macrophages are recruited to the site of injury. These innate immune cells provide critical phagocytic functions⁴⁰. But additionally, these innate immune cells provide other functions essential to nerve regeneration, including re-myelination and functional recovery, which are not yet entirely understood^41,42,40. After this Wallerian degeneration process is complete, SCs progressively assume long processes and align on the basal lamina of the intact distal nerve environment (Bands of Bungner), providing a permissive growth environment for the regenerating axons that emerge from the proximal nerve stump^24,43. As axon growth proceeds from proximal to distal nerve, remyelination of the axons is initiated primarily by axon-derived neuregulin-1 signaling through SCs’ ErbB receptors^44,45. At the target end organs, specialized SCs have already begun priming this environment for reestablishment of axonal connection⁴⁶. For example, in muscle a unique type of SC, terminal SCs, extend cytoplasmic processes within days after injury, which serve as platforms for incoming axonal growth^47,48,46.

Our understanding of the regenerative processes enabling nerve regeneration across a small nerve defect or cell-free nerve graft, where the entire nerve tissue must regenerate, is incomplete (Figure 1). To first allow for cell migration and infiltration into this defect, an endogenous matrix must form. The repair of nerve defects using a pseudosynovial sheath demonstrated that plasma exudates from the proximal and distal nerve stumps fill the empty tube volume, and provide a deposition of extracellular matrix (ECM), including a fibrin matrix, that allows for innate immune system cell migration^49–56. While the earliest (<4 days) macrophages are derived from tissue resident macrophages, subsequent macrophages are primarily hematogenous-derived⁴⁰. Due to the hypoxic nature of this environment, macrophages support a substantial amount of angiogenesis enabling endothelial cell recruitment and vessel formation⁵⁶ (Figure 1A). These blood vessels become polarized allowing for the migration of SCs⁵⁶, which have a distinct phenotype similar to a stem cell in this ^57,26, and in turn form cellular cords through their interactions with fibroblasts⁵⁸ (Figure 1B). After axons cross this bridge (Figure 1C), the regenerative processes described in the distal nerve promote the growth of axons to their down-stream targets.

Figure 1.

Schematic of nerve regeneration across a short nerve defect. After formation of an endogenous matrix, the innate immune system infiltrates the defect. (A) Due to the hypoxic nature of this environment, macrophages produce factors to support angiogenesis, such as vascular endothelial growth factor (VEGF). The secretion of VEGF from macrophages recruits endothelial cells into the injured nerve bridge⁵⁶. (B) As blood vessels form, Schwann cells migrate upon this network, and in turn fibroblasts interact with Schwann cells forming cellular cords through EphrinA-EphrinB signaling⁵⁸. (C) These processes enable axon growth across the nerve defect.

Identifying and managing traumatic nerve injuries

In the clinic, nerve injuries present on a spectrum, which includes damage to the many axons contained within nerve that may be recoverable or non-recoverable depending on the type and degree of injury. Therefore, the extent of nerve injury must first be determined before any surgical decisions are made. For severe nerve injuries requiring repair, this identification may be immediately obvious due to substantial soft tissue damage resulting in loss of nerve, or less obvious and result from damage to nerve that generates a “zone of injury.” This zone of injury, which refers to the scarring of nerve following damage, is not fully present until approximately 3 weeks after the initial injury^59,60. It is critical to identify this zone, as this scar inhibits the endogenous processes for regeneration. Nerve repair performed before this is present has the risk that the proximal or distal end of the nerve still contains scar that would block endogenous regeneration facilitated by the repair from proceeding. To identify this zone of injury, the use of electrodiagnostics, clinical examination, and in the operating room, a technique termed “bread-loafing,” as it dissects the nerve back to the start of scarred tissue, are commonly used^59,60. Imaging modalities can be used as well, but these imaging modalities are generally of minimal clinical utility. Current clinical imaging modalities are unable to provide direct correlation to axonal injury and cannot adequately evaluate peripheral nerve injury in which the damaged tissue recovers over time⁶¹.

After identifying the zone of injury together with any original gap, the effective defect size is revealed. A defect of even a few millimeters will often prevent the previously described endogenous mechanisms for regeneration, and therefore, nerve repair acts primarily to facilitate this regenerative process. However, even with a nerve defect, nerve grafting to bridge the defect is not the only option to repair the nerve and restore function. After identifying the injury, the location of the injury is also considered. Nerve transfers, whereby the proximal portion of an expendable nerve is sutured to the damaged distal nerve end, are frequently used for most proximal nerve injuries. Transfers avoid not only the regenerative processes involved to bridge the nerve defect, but also reduce the long period of axon growth needed to reestablish axon reinnervation of end-organs^62–64. It also obviates the need for grafts. Therefore, distal nerve injuries are typically reconstructed by direct coaptation of the proximal and distal nerve ends, either directly or with a grafting material. Direct epineurial or grouped fascicular repair is possible when a defect is small, only a few millimeters, so as to avoid inducing longitudinal tension^65,66. For other injuries, nerve grafting is then employed^67–69.

Bridging materials to repair nerve defects

While the repair of nerve defects using a bridging material dates to the 1800s, it was not until the 1970s that the current standard of nerve defect management was initiated with the advent of the nerve autograft. The efforts of Huber and Bunnell established nerve autografting for nerve defect repair, while further understanding of anatomy and microsurgical techniques from Sunderland, Millesi, and Buncke advanced this technique⁶⁰. Since then, several alternatives have become clinically-approved for repair of nerve defects. But, while this section will detail the clinically available options to repair a segmental nerve injury, the autologous nerve graft, nerve tissue harvested from the patient, remains the gold standard of nerve gap repair.

Nerve autografts

From all the biological principles just described in the previous sections, the nerve autograft serves in theory as an ideal bridging material. Nerve autograft is nerve tissue harvested from the patient taken from an area in the body where the loss of function from its harvest is thought to be minimal. Donor nerves that are commonly used as autografts are expendable sensory nerves, such as the sural nerve or the medial antebrachial cutaneous nerve⁶⁰. The nerve autograft provides not only a suitably matched scaffolding structure for regenerating nerve, principally ECM arranged to include endoneurial tubes, but also a diverse cell supply, including a vascular network and support cells of nerve, critically SCs. While blood flow is initially disrupted during the grafting procedure, functional blood flow is rapidly restored within days via inosculation, or spontaneous end to end repair of existing vasculature, which occurs from both the tissue bed and reconnected nerve ends^70–72. Furthermore, these graft undergo remodeling in the weeks that follow, including Wallerian degeneration, thus mimicking the scenario just described following injury for the distal portion of an injured nerve. These processes include an invasion of macrophages, proliferation of the donor graft SCs, degeneration of axons and myelin debris within the graft, and even a robust migration of SCs from the graft to the repaired nerve ends⁷³. While similar in many ways to the processes affecting nerve distal to an injury site, there is one critical difference. It is highly likely that a portion of non-neuronal support cells within the graft die because of disrupted blood flow and tissue oxygenation before inosculation is complete. But, this cell death does not appear to be a major issue that disrupts its capabilities to facilitate robust regeneration. Therefore, an autograft currently represents the best available, while imperfect, option for a grafting material.

Despite the inherent advantages of providing a cellular supplemented scaffold that promotes robust nerve regeneration across a defect, autografting comes with significant drawbacks. The donor harvest will entail morbidity at the site, including loss of donor nerve function, and its harvest entails additional operations that can include complications, scarring, and even the potential for neuroma formation at this additional surgical site^74,75. Furthermore, there is a limited supply of expendable donor nerves, and the harvesting of these nerves adds to the overall operative time for the patient. Therefore, there has been a desire for alternatives, and a continued push to use these alternatives even if the outcomes do not yet match the autograft.

Cadaveric nerve allografts

Following the advent of adequate immunosuppressive regimens in the 1980s, cadaveric nerve allografts became a feasible clinical alternative to autografts. Nerve allografts, i.e. nerve tissue from organ donors, represent an analogous form of bridging material to the autograft while sparing the donor from loss of function. However, as their use requires immunosuppression to avoid rejection and regeneration failure⁷⁶, and the use of immunosuppression is associated with significant clinical morbidity⁷⁷, the use of allografts in peripheral nerve repair is quite rare, and limited to the repair of only the most severe cases of nerve injuries, with a considerably long nerve defect length⁷⁸.

Nerve guidance conduits

Conduits, also referred to as nerve guidance conduits, are the most diverse and deeply researched category of nerve autograft alternatives. Conduits consist of tubular structures to encapsulate and facilitate cellular and axon growth across the nerve ends, as these tubes serve to provide a protective environment for nerve regeneration. The idea of repairing a nerve using a conduit was largely driven from the previously-discussed animal studies that determined that encapsulating a nerve can promote a robust endogenous mechanism that drives an entire nerve to regenerate its structure. These experiments provided a framework demonstrating that even in the absence of a scaffold or cells, peripheral nerves are capable of producing their own scaffolding, and in turn, use that scaffolding to support cell migration and ultimately axon regeneration^50,49,79,53.

From this animal work, conduits were the first translated work providing “off the shelf” alternatives to treat nerve defects⁸⁰. However, non-degradable conduits demonstrated significant limitations arising from the side-effects to nerve. Silicone conduits have been shown to cause significant chronic nerve compression and irritation at the implantation sites, requiring removal^81–83. This incompatibility from the long-term presence of a material surrounding the nerve drove research into alternative materials that could act as more natural or temporary biodegradable conduits. Naturally-derived conduits, such as arteries, muscles and tendons, have been studied for bridging nerve defects, but vein tubes have received the most attention among researchers and even are used clinically⁸⁴. Far and away, manufactured conduits synthesized with properties that yield biodegradable structures represent the majority of modern products available in the clinic. At least seven synthetic nerve conduits have been approved by the US Food and Drug Administration (FDA) for clinical use in nerve reconstruction⁸⁵. However, these conduits are generally similar in their principles and features, as they are biodegradable through various mechanisms and promote regeneration via first the formation of endogenous ECM within the empty tube structure. Extensive reviews of the wide range of conduit material properties are available^68,85–87.

Processed or acellular nerve allografts

More recently, processed or acellular tissue scaffolds have gained prominence for clinical reconstructions specifically in the United States (U.S.). For example, a recent survey estimated that ~70% of U.S. hand surgeons use processed nerve allografts in nerve defect repair⁸⁸. Acellular tissue scaffolds are generated using techniques to retain a large portion of native ECM proteins while minimizing cellular debris and undesired immunological response (i.e. rejection). Acellular tissues, despite their lack of cells and processing techniques, typically maintain a highly organized extracellular matrix structure providing an ideal scaffold structure for regeneration. As such, acellular nerve allografts (ANAs) or processed nerve allografts (PNAs), which are generally synonymous, have become a prominent nerve autograft alternative choice due to these features^89,69,88. Acellular nerve grafts are conceptually appealing not only because they provide a scaffold for immediate cell migration and angiogenesis, making them distinct from clinically-available conduits^90–92, but their ECM structure is similar to native nerve.

Numerous experimental methods have been used to generate these nerve grafts, which all entail some technique to remove cells and antigens from the nerve. This topic was reviewed extensively by Szynkaruk et al ⁹², but examples of decellularizing techniques include repeated freeze-thaw cycles, exposure to radiation, lyophilization, extended storage in cryopreservation solution, and decellularization with detergents^93–99. However, a detergent based protocol developed by Hudson et al.¹⁰⁰ has been the only processing technique to translate to a clinical nerve product with FDA approval. A variation of this protocol was used in 2008 to develop the first commercially available acellular peripheral nerve allograft for clinical use. These PNAs are produced from harvested cadaveric human nerves, processed to remove cells using a human variation of Hudson et al.¹⁰⁰, as well as enzymatic removal of chondroitin-6-sulfate proteoglycan, a known inhibitor of axonal regeneration^101–103.

Guidelines for nerve graft repairs

Any form of nerve repair is affected by a range of variables that ultimately limits the quantity of axons regenerating and reaching their target. These include the distance for axon growth, a general decline in the capacity to promote axon growth over time, axonal misdirection from their appropriate end-organ targets, as well as atrophy of the end-organ targets^91,104–108. For nerve grafting repairs, outcomes are also additionally affected by the bridged defect size (i.e. both diameter and length) and whether the nerve defect repaired is supporting primarily sensory or motor functions. Furthermore, the choice of material to repair the defect affects the outcome.

The nerve autograft provides generally superior and more consistent outcomes compared to any currently available clinical alternatives. Animal studies have demonstrated the superiority of nerve autografting over alternatives to support axon regeneration and functional recovery. Two independent research groups have demonstrated in rodent models that autografts support a greater extent of axon regeneration across a defect compared to any clinically available alternative, as well as more rapid axon growth across the graft reaching the distal nerve^109,110. As autografts supported axon regeneration across a defect more rapidly and to a greater extent than alternatives, this outcome has critical implications for functional recovery. Reducing the time of distal nerve and end-organ denervation is the best known strategy to achieve reinnervation of end-organ targets and functional recovery^104,91. From these specific research studies, it was also determined that a hierarchy between grafting procedures exists. Autografts supported the greatest extent of axon regeneration across a defect, followed by ANAs, and then empty conduits^109,110.

Rather critically, there has yet to be a head-to-head prospective clinical study comparing the results of autografts to any existing nerve graft alternatives, which would provide the best evidence to support or oppose a change in practice. A comparison of nerve autografting to current clinical autograft alternatives has only been directly demonstrated in animal studies, as just described. Thus, any evidence demonstrating alternatives are comparable to autografts in outcomes has been derived from individual studies of alternatives used in nerve repair that are then compared to historical autograft data. Not only do these study designs introduce additional confounding factors, such as differences in surgical techniques and patient population, but comparisons to historic data can make the outcome comparisons appear biased.

While autografts remain the best option for managing nerve defect repairs, there are criteria that specify situations in which the use of autograft alternatives is preferred. First, the use of nerve graft alternatives in the reconstruction of small and short segmental sensory nerve injuries can be supported for logical reasons regardless of evidence of their efficacy. There are situations where harvest of a nerve autograft would be counterintuitive: for example, sacrificing a non-critical, small sensory nerve to fix a single non-critical, small sensory nerve defect. Another example is autograft harvest in a patient with an established pain syndrome, where pain at a nerve donor site is more likely to occur. If the reconstructed nerve is non-critical, more harm may be done than good with autograft harvest. In clinical situations such as these, alternatives are better justified so long as the risks of complications are minimized.

Short or small sensory nerve defects

A growing body of evidence has now demonstrated that nerve graft alternatives support adequate recovery (i.e. comparable recovery to an autograft based on historic data on outcomes) to treat short and small diameter segmental sensory nerve injuries. While defining these parameters can be slightly subjective, in general there is a consensus that short length defects are less than 30 mm^111,60,112. The most common metric used for these comparisons has been meaningful recovery based on the Medical Research Council (MRC) scale (Table 1), set for sensory function¹¹³ at ≥S3 and motor function¹¹⁴ at ≥M4¹¹⁵. For digital nerves repaired with an autograft, the reported ranges of meaningful recovery are nearly 100%^116,117. For comparison, commercial conduits yielded meaningful recovery in 44–75% of repairs^81,118,119. Alternatively, the data for PNAs suggests superior clinical results when compared to conduits for the reconstruction of digital nerve defects. Multiple independent studies demonstrated meaningful recovery in 80–89% of digital nerve defect repairs using PNAs meeting similar criteria^120–122. And finally, while there is limited data demonstrating a major difference in the degree of recovery when regeneration is successful, evidence demonstrates that autografts and PNAs provide more consistent recovery compared to conduits for treating short length or small diameter sensory nerve defects^123,111,124.

Table 1.

Medical Research Council (MRC) scale

Sensory Function113	S0	absence of sensibility in the autonomous area of the nerve
	S1	recovery of deep cutaneous pain and tactile sensibility
	S1+	recovery of superficial pain sensibility
	S2	recovery of some degree of superficial cutaneous pain and tactile sensibility
	S2+	as in S2, but with over response
	S3	return of pain and tactile sensibility with disappearance of over response, s2-PD: >15mm
	S3+	return of sensibility as in S3 with some recovery of 2-point discrimination, s2-PD: 7–15mm
	S4	complete recovery, s2-PD: 2–6mm
Motor Function114	M0	no contraction
	M1	flicker or trace of contraction
	M2	full range of active movement, with gravity eliminated
	M3	active movement against gravity
	M4	active movement against gravity and resistance
	M5	normal power

Abbreviations: S2-PD: static two-point discrimination test

Short or small motor / mixed nerve defects

In general, regardless of grafting option (autograft or alternative), anticipated motor functional recovery rates will be appreciably reduced compared to sensory recovery when repairing motor and mixed segmental nerve injuries. Furthermore, the evidence to use any nerve graft alternative for their repair is much more contentious, and there are no evidence-based guidelines to use nerve graft alternatives instead of nerve autograft to reconstruct motor or mixed nerve defects¹¹¹. For the repair of mixed or motor nerves, such as the median and ulnar nerves, studies have demonstrated that autograft repair results in meaningful recovery in 60–80% of radial and median nerves^125,126 and 57–60% of ulnar nerves¹²⁷. For conduits, while few studies exists, the outcomes are dismal. In one study, the repair of median or ulnar nerve (defects less than 30 mm) resulted in only 8% meaningful recovery¹¹⁹. For the other prominent category of alternatives (PNAs), the outcomes from studies have been more promising. For median nerve repairs up to 50 mm, 75% of patients experienced meaningful recovery, and for ulnar nerve repairs of similar criteria, up to 67% of patients experienced recovery¹²⁰. Furthermore, in a more recent study, a motor recovery rate of ~67% was appreciated across a variety of repaired nerve defects up to 30 mm¹²². Therefore, recent data suggest that even mixed or motor nerve defects less than 30 mm could be repaired equally well using PNAs compared to autografts. But, these outcomes represent few studies with select surgeons. As reviewed in greater detail by Rbia and Shin, there is still insufficient evidence at this time to support the widespread use of alternatives to repair mixed or motor nerve defects¹¹¹.

Long or large nerve defects

There is still a critical unmet need for improved nerve grafting options to treat long (>30 mm) segmental nerve injuries. Regardless of considerations for sensory or motor functions, neither is anticipated to recover well. And, these anticipated poor outcomes result even from repair using a nerve autograft, while it remains the best available option. There is no consensus on the maximum gap that can be bridged by a nerve autograft, and in fact, widely varying degrees of success have been reported in autografts up to 200 mm¹²⁸. However, there is a general consensus that any regenerative success declines as autografts go beyond 60 mm⁶⁰. As an example, meaningful sensory recovery from a digital nerve defect repair using autografts was observed in 100% of patients with defect lengths less than 21 mm, while this recovery rate fell to 67% for lengths between 21 and 49 mm and only 9% for lengths greater than 49 mm¹¹⁷. This relationship regarding defect length and recovery rates has also been observed for conduits⁸¹ and PNAs¹²². Similar to data on autograft repairs for long defects, studies using PNAs to repair long nerve defects demonstrate that sensory and motor recovery can vary considerably among studies. In one study, a high level (~86%) of meaningful sensory recovery was observed in patients with nerve defects repaired using PNAs at lengths between 40 and 50 mm¹²⁹. Yet, more recent studies from Leckenby et al significantly temper these promising outcomes. In their studies, PNAs used to repair nerve defects between 30–49 mm yielded ~75% meaningful sensory recovery but only 38% motor recovery. Similarly, defects greater than 50 mm yielded 53% sensory recovery, but only 10% recovery ofmeaningful motor function¹²².

Concerns regarding alternatives for repairing defects

A prevailing concern for caution regarding the use of alternatives is due to the general variability and inconsistency of results with alternatives compared to autografts. Alternatives can fail to facilitate any nerve regeneration across a bridged nerve defect, which is distinct from autograft repair. Specifically, a failure to regenerate across an alternative can result in no appreciable functional recovery, while a poor outcome from autografting will still yield some recovery, even if not ≥S3+ or M4. This phenomenon whereby nerve fails to regenerate any axons across a defect repaired using an alternative is well-documented in the animal literature,^130–132 but now there is an increasing body of evidence of these issues present in the clinical literature. Indeed, Moore et al, found that use of conduits can result in inferior quality nerve regeneration, which is most prevalent when conduits are used to repair nerve defects beyond their length indications. This failure can not only lead to lack of recovery, but can even result in formation of a painful neuroma in the repaired nerve defect¹³³. Specific examples of this issue have been published on PNAs as well. Nietosvaara et al. presented three cases of PNAs ranging from 20 mm to 50 mm failing to facilitate any measurable axon regeneration across these PNAs. In fact, they document that the PNAs resorbed leaving behind scar tissue, and in one case, had thickened into a neuroma-like stump¹³⁴. Overall, this evidence is perhaps the strongest reason to take a conservative approach to utilizing alternatives in the clinic until these circumstances are better understood.

Experimental nerve graft alternatives

Since the development of biodegradable conduits and identification of their limitations, experimental work to develop improved nerve graft alternatives has been ongoing. As the body of work on experimental alternatives is substantial, this section will highlight specific recent areas that show promise for translation. For a more comprehensive review of bioengineered nerve graft alternatives, see Pfister et al.⁶⁸, Boecker et al⁸⁶, and Kornfeld et al⁸⁷.

Scaffolding and topology

The use of synthetic conduits containing internal scaffolding that has included ECM, such as collagen, laminin, and fibrin^131,135,136, has demonstrated advantages in improving regeneration compared to empty conduits. However, there are limits to these internal scaffolds compared to acellular nerve. These scaffolds develop with a random arrangement of the molecular fibers, which differs significantly from the organized and longitudinal arranged structure of nerve. Instead, recent advances that arrange the topology of the scaffold fibers holds promise, and the most interesting developments are the use of scaffolds with a longitudinal organization. This scaffolding can be achieved through a variety of means, including the use of electromagnetic fields and photolithography, but electrospinning techniques are becoming increasingly well-developed for this approach^137,138. These techniques allow for aligned fibers, synthetic or “naturally-derived”, to be deposited, resulting in longitudinally oriented pores or channels as small as 1 μm, which are comparable to endoneurial tubes, ranging from 1–20 μm¹³⁷. These designs allow for scaffolds that not only more closely mimic the nerve’s structure^139,140 but could lead to more rapid cell migration and polarization of cells^141,142. As endothelial cell polarization is important for both SC migration and then axon outgrowth⁵⁶, this could lead to a major translational development. In addition, the use of synthetic fibers and/or conduits also holds an advantage in that drugs for improving regeneration (detailed in the next section) could be readily incorporated into the underlying scaffold.

Local drug delivery

The delivery of biologically active molecules locally during regeneration is another strategy that has been extensively pursued for several decades. Nerve scaffolds or the conduit itself can be incorporated with bioactive molecules via chemical interactions, via crosslinking or affinity-based interactions, or physically encapsulated. While chemical interactions can provide drugs immediately to cells or as cells proceed to grow within a scaffold, physical encapsulation of drugs provides a greater range of delivery options, such as long-term or sustained release. Numerous molecules or drugs have been delivered to nerve for regeneration across a nerve defect^143,144, but the most researched category for delivery has been neurotrophic factors. Of these, glial cell line-derived neurotrophic factor (GDNF) is a promising example that promotes regeneration because it targets both axons and SCs. Both sensory and motor neurons express receptors for GDNF (Ret/GFRα1)¹⁴⁵, where GDNF signaling promotes axon outgrowth and neuronal survival^146–151. In addition, SCs also express receptors for GDNF (NCAM/GFRα1)¹⁵², where GDNF signaling activates pathways in SCs implicated in cell migration, differentiation, and trophic factor production^152–163. Exogenous GDNF delivered to nerve has been shown to improve not only axon regeneration but functional recovery^131,164.

More recently, the local delivery of drugs that accelerate axon outgrowth have been investigated. FK-506, an immunosuppressive drug, enhances nerve regeneration, as it increases the axonal growth rate in animal models^165–172. Now, approaches have been developed to provide FK506 locally to enhance axon regeneration given its abilities to stimulate more rapid axon outgrowth^173–177. This strategy is quite intriguing given that it reduces the potential for any systemic toxicity from its immunosuppression. Furthermore, while this delivery strategy has demonstrated its potential to improve nerve regeneration in animal models¹⁷⁴, the sustained local delivery of FK506 from a nerve graft alternative bridging a nerve defect could translate to benefits even after axons cross the bridged region, as it could continue to stimulate accelerated axon growth through the distal nerve.

Alternatively, based on our increasing knowledge of the immune response during regeneration across nerve defects, recruiting immune cells, such as macrophages, is a promising strategy. While angiogenic factors, such as vascular endothelial growth factor (VEGF), have been locally delivered from conduits to promote improved angiogenesis within bridged nerve gap^178,179, modulating the local immune system in order to recruit cells to promote endogenous angiogenesis could have advantages, such as a greater degree of endothelial cell polarization. Furthermore, the immune system has a critical role in resolving inflammation following injury, where macrophages also have additional roles in this aspect of tissue regeneration. Macrophages alter their secreted cytokines based on their phenotype. While a simplification, macrophages are broadly classified as classically activated (M1) or alternatively activated (M2) phenotypes. The M1 macrophage response predominates during the onset of injury while the M2 polarization or subtypes generally promotes healing, remodeling, and resolution of regeneration^180,181. Studies using conduits releasing factors promoting a more M1-like phenotype (IFNγ) versus a more M2-like phenotype (IL-4, collagen VI, or fractalkine) have demonstrated improved nerve regeneration when factors promoting a greater accumulation of M2-like phenotype macrophages were used to repair a nerve defect compared to conduits lacking this ability^182–184. Furthermore, the nerve autograft alternative itself could be designed to promote macrophage polarization to the more M2-like phenotype through its inherent biomaterial properties, which has been observed upon conduit materials such as chitosan¹⁸⁵. Indeed, these conduits are now undergoing clinical trials to determine their efficacy in nerve repair¹⁸⁶.

Supplementing cells

Match and surpassing the outcomes of the autograft will likely entail alternatives that include a cellular component. SCs and stem cells have been shown to be the most promising sources for this supplementation¹⁸⁷. However, several challenges remain: improving the survival of transplanted cells and determining the cells to transplant. Supplementing ANAs with SCs has been shown to result in very low survival of the transplanted cells, which can be as low as 2% of transplanted cells surviving up to 7 days^188,189. And while SCs in culture do not transform or reach a proliferative limit¹⁹⁰, the difficulty in SC isolation and culture has impeded clinical translation^191,95,192. Instead, stem cell transplantation seems a more promising avenue. A more extensive review of stem cells used for nerve graft alternatives can be found in Johnson et al.⁶⁹. Recent work from Shin and colleagues is noteworthy. They developed an approach to seed mesenchymal stromal (stem) cells upon ANAs with improved viability and greater reproducibility. A bioreactor was used to seed the mesenchymal stromal cells onto ANA¹⁹³, a technique that has led to cells surviving up to 29 days¹⁹⁴. Overall, supplementation of cells to nerve graft alternatives seems an inevitable translational advance once the processes to reliably seed these cells to the alternatives are robust.

Future research needed to improve alternative designs

While these experimental strategies are promising, unaddressed issues remain, and it is unclear whether these will be adequately addressed by new designs. The goal of these new designs should be not only to meet the autograft “gold standard” but to also understand what drives and inhibits nerve regeneration in order to surpass the results of an autograft. To do this requires looking beyond new designs to also consider what are the factors limiting the current designs.

How is consistent regeneration and motor recovery promoted?

As demonstrated by clinical evidence, sensory recovery can generally be achieved using nerve graft alternatives, but motor recovery is highly inconsistent. Specifically, repairing motor nerve defects with alternatives yields meaningful motor function recovery at rates ranging from 8–75%^{133,69,111,195}. We need to understand how to promote consistent regeneration across alternatives with robust recovery. A critical component to achieving any level of functional recovery is regeneration of an adequate number of axons to their target (i.e. skin or muscle) in a timely manner^196–201,17, but the promotion of functional recovery also requires other processes. Axonal guidance and non-neuronal cell signaling, including that occurring through the immune system, all contribute to regeneration leading to functional recovery^{197,198,200,104,42}. The essential mechanisms involved in promoting nerve regeneration across alternatives yielding robust motor functional recovery are unresolved, despite many animal studies demonstrating return of motor recovery. Uncovering what factors are missing in the translation of animal to human research will be critical to advancing the field.

Why is regeneration functionally limited by graft length?

Animal studies have yet to demonstrate that nerve graft alternatives can reliably facilitate axon regeneration across longer defects (>30mm) despite their robust abilities to promote regeneration across shorter defects^202–206. A review of this specific topic is available: see Kornfeld et al⁸⁷. However, there is limited understanding as to why few axons regenerate across these long graft alternatives despite cues to promote axon growth and regeneration that succeed at shorter lengths. Effectively, alternatives are functionally limited by length.

While the cause of limited axon regeneration across long nerve graft alternatives is still controversial, a series of recent studies has illuminated some novel and insightful findings. Using ANAs as a model nerve graft alterative, these studies compared the regenerative processes within short (non-critical length) vs long (critical-length) ANAs revealing that cell repopulation of ANAs due to length was the primary reason for limited axon regeneration^202,203,206 (Figure 2). In these studies, to establish whether the environment that develops in long nerve graft alternatives – not the ability of neurons to regenerate their axons over long distances – is responsible for axon growth arrest²⁰³, experiments were performed whereby long alternatives were shortened <30mm in length. These shortened alternatives, which should normally facilitate axon regeneration across a short defect, still contained this altered cellular environment and ECM structure. These now “stressed” short grafts halted axon growth and regeneration²⁰³. Furthermore, it was also established that this poor regenerative outcome across long grafts is not caused by a failure of neurons to regenerate their axons. Using a sciatic nerve defect repair with a long acellular nerve, regenerating axon growth was first arrested and then proceeded by replacing the remaining scaffold with an isograft (animal equivalent to an autograft). In this scenario, axon growth resumed²⁰³. Overall, these results demonstrated the environment of long nerve graft alternatives generates a barrier to axon growth.

Figure 2.

Long nerve graft alternatives develop an environment limiting axon regeneration. Using acellular nerve allografts (ANAs) as a model, studies have compared the regenerative processes within short (non-critical length) vs long (critical-length) ANAs to discover factors contributing to successful or failed nerve regeneration across ANAs^202,203,206. (A) Short ANAs are repopulated with cells similar to previously described in Figure 1. Conversely, (B) long ANAs are repopulated with an environmental imbalance consisting of altered populations of immune cells, cells expressing markers of senescence²⁰³, and delayed angiogenesis compared to their shorter counterparts^72,70,71,206. These cumulative changes to long ANAs limit axon regeneration and represent a “barrier” to axon regeneration^203,202.

Recently, immune cells other than macrophages were determined to have additional, undescribed processes that contributed to regeneration across ANAs based on length. Specifically, a novel role for T cells was elucidated. T cells accumulated within short (20mm) ANAs after macrophages, but before appreciable SC accumulation or axon growth. T cell accumulation within short ANAs was associated with robust nerve regeneration. Conversely, few T cells accumulated within long (40mm) ANAs, which were associated with minimal axon regeneration across these grafts. Furthermore, a direct causal relationship between T cells and nerve regeneration across ANAs was demonstrated. Nerve regeneration across short (20mm) ANAs, which normally support robust axon regeneration across the gap, was decreased by ~50% in T cell deficient rats²⁰⁶. Overall, research regarding the role of the immune system represents an important area that needs further examination in this context (Figure 3). Given the prominent role that the immune system plays in tissue regeneration, there is still limited knowledge on its role during regeneration across a nerve defect.

Figure 3.

Temporal relationship between immune system cells within nerve defects and regeneration. Each cell type migrates within a defect, where its accumulation peaks at different times during regeneration across nerve defects and precedes axon growth. Prior to substantial Schwann cell migration within a nerve defect, the innate immune system plays a pivotal role. Neutrophils are the first cells to infiltrate the defect, and have key processes involved in debris clearance, as well as yet unresolved roles in further cell recruitment. Macrophages shortly follow neutrophils and are responsible for the majority of blood vessel formation, as well as providing signaling for Schwann cell functions. Cells of the adaptive immune system, such as T cells, arrive at the same time as Schwann cells, and regulate axon myelination. The nature of how T cells regulate myelination is not yet resolved.

Abbreviations

ANA

Acellular nerve allograft

ECM

Extracellular matrix

FDA

Food and Drug Administration

FK506

Tacrolimus

GDNF

Glial cell line-derived neurotrophic factor

MRC

Medical Research Council

PNA

Processed nerve allograft

RAG

Regeneration associated gene

SC

Schwann cell

VEGF

Vascular endothelial growth factor