Advancement of Electrospun Nerve Conduit for Peripheral Nerve Regeneration: A Systematic Review (2016–2021)

Introduction

Peripheral nerve injury (PNI) is a worldwide problem which hugely affects the quality of patients’ life. As a consequence of the injuries, patients often suffer painful neuropathies with reduction in motor function and sensory perception.¹ In severe situations, the nerve is completely ruptured, resulting in a gap, necessitating surgical procedures to restore nerve functionality.² Currently, there are different surgical procedures available, including end-to-end suture which is preferably used in short nerve gaps (<5 mm).³ In larger nerve gaps, an autologous nerve graft procedure is currently the gold standard⁴ despite its limitations, such as additional surgery, scarring and donor-site morbidity, and limited source of donor nerves.⁵ Thus, there is a strong need to find innovative therapies for treatment of PNI.

Figure 1 Flowchart for the identification and selection of studies.

Figure 2 Schematic illustration of the promise of electrospun nerve conduit.

There are alternative nerve conduits available on the market to mimic autologous nerve graft, which are made of synthetic polymers such as collagen, PLGA, silicone; or decellularized nerves. Decellularized nerves are obtained from the human dead body, decellularized, retaining some extracellular matrix-related protein. They are cleaned of their antigenic component yet retain their 3D structure which serves as scaffolds for axonal growth.⁶ Although the decellularized nerve graft has been rated as the next best procedure for PNI treatment, it has limitations such as its source from a dead body and its expensive price (around RM 15k–20k). This has led to the further research of nerve conduits using synthetic polymers, which are cheaper and easily obtained. Therefore, the search for an ideal nerve conduit is necessary to replace autologous nerve graft in a similar effectiveness, with longer length and additional biological additives such as cell-seeded nerve conduit or neurotrophic factors-/drugs-/bio-functional peptides-added nerve conduit.

There are many techniques that can be used to fabricate nerve guidance conduits. In comparison to other methods, electrospinning is relatively easy and less labor intensive.^7,8 This review will focus on electrospun nerve conduits due to the numerous advantages of the electrospinning technique. Electrospinning produces randomly or aligned fibrous mats with fiber diameters ranging from nanometers to micrometers. This technique allows the use of different synthetic and natural polymers and results in three-dimensional fibrous microstructures capable of mimicking the native extracellular matrix (ECM). Moreover, electrospinning meets the requirement of porous and flexible mats that have a high surface area to sufficiently enhance cell–substrate interactions.⁹ Physically and linearly guiding axon growth increases the probability of functional recovery of the injured nerve.¹⁰ The use of electrospun conduits with an aligned structure has advantages over the commercial hollow tube conduits, such as high porosity, high specific surface area which increases the area available for protein absorption, Schwann cell migration and axon regeneration. Additionally, aligned nanofibers promote cell proliferation and growth, and guided axonal growth,^1,11 preventing the formation of neuroma.

The objective of our study is firstly, to provide an update on the progress of electrospun nerve conduit over the last 5 years (2016–2021). Secondly, is to gather and review all preclinical and clinical data on electrospun nerve conduit to date.

Materials and Methods

Keywords Used

The following keywords were being used for searching literatures through databases:

“nerve tissue engineering” OR “tissue-engineered nerve” OR “nerve conduit” OR “nerve guidance conduit” OR “nerve scaffold” OR “tissue scaffold” AND “in vivo nerve regeneration” OR “peripheral nerve regeneration” OR “nerve regeneration” AND “electrospun nanofiber” OR “electrospun nanofibers” OR “electrospinning nanofiber” OR “electrospinning nanofibers” AND “transplantation” OR “cell transplantation” OR “cells transplantation”.

Selection Process and Data Extraction

The following databases were being used for searching literatures for this study: PubMed database, Scopus database, Science Direct database and Ovid database. After searching by the keywords above, the results were further filtered by years (2016–2021), language (English) and article type (only research articles were included). After applying those filters, each article or literature was included based on their: title, which stated “nerve conduit”, “in vivo” or “nerve regeneration”; abstract, which mentioned “electrospinning nerve conduit”, “in vivo study” or “animal defect model”; and method, which clearly listed the procedures and protocols of fabrication of electrospun nerve guidance conduit or animal surgery, to ensure the articles were related to “electrospun nerve conduit for peripheral nerve regeneration”.

The screening of the research articles was performed by two reviewers (SYL and MHN). Any disagreement during the screening process were resolved through discussion between the two reviewers. Relevant information from the selected articles were extracted and documented in the data extraction sheet (Figure 1).

Results and Discussion

Selection of Articles

According to the keywords and criterion above, there are 39 literatures in total that were included in this study, which could be further categorized by the 4 varieties of nerve conduit. The yields of each category are as follows: 18 for nerve conduit with nanofibers only,^12–29 6 for nerve conduit with cells seeded,^30–35 12 for nerve conduit with additional factors (neurotrophic factors, drugs and bio-functional peptides)^36–49 and 3 for nerve conduit with the combination of cells seeded and additional factors.^48–50

Study Characteristics

In this study, the data and information extracted were tabulated and included the review of material used as nerve conduit, their properties and their in vitro biocompatibility, the additional factors incorporated such as cells and non-cellular factors, and the functional and morphological assessment of in vivo implantation of the nerve conduit will be presented as Table of Comparison, which are shown as Tables 1–5 as follows.

Table 1 Comparison of the Fabrication Method, Material Used and Characterization (Including Biocompatibility) of Nerve Conduit

Table 2 Comparison of the Sources of Cells Seeded on Nerve Conduits and Their Manipulations

Table 3 Comparison of Non-Cellular Factors (Growth Factors, Drugs, Peptides) Added to the Nerve Conduits

Table 4 Comparison of Functional Outcomes of Peripheral Nerve Repair of All Studies

Table 5 Comparison of Gene and Protein Expression and Morphometric Assessment Outcome of All Studies

Nerve Conduit Materials

Fabrication Method

The electrospinning techniques that were used for nerve conduit fabrication included classic electrospinning and coaxial electrospinning. The classic electrospinning only allows one ejection of one polymer solution at one time, while coaxial electrospinning allows two polymer solutions to be ejected at the same time, resulting in nano- or micro-fibers with a “core/shell” structure. Core and shell represent the inner layer and outer layer of a nerve conduit respectively. Thus, the different structures of nerve conduit fabricated by different electrospinning techniques could be classified as: single layer of fibers or core/shell layers of fibers. Moreover, not only the “hollow” nerve conduit that consists of only layers of fibers (with nothing added in the lumen of conduit), but also the nerve conduit that is filled with freeze-drying nano-sponge (NS) or hydrogel (in the lumen) could be used as the scaffold for peripheral nerve regeneration study.

Material Used

According to Table 1, there are various types of materials that can be used for the fabrication of nerve conduit, which can be categorized to either natural, synthetic or hybrid.

Among these polymers, PCL, with its ease of processability, mechanical strength and proper stability under ambient conditions,⁵¹ is the most popular polymer to be used in studies from year 2017 onwards. Its usage continues to increase in the later years either independently, or incorporated with gelatin, PLGA, CNF, CL or chitosan. Recently studies are focusing on PLLA on account of of its good biocompatibility, acceptable biodegradability and thermoplastic processability.⁵² Nevertheless, in electrospinning, PLGA, gelatin and chitosan are the polymers that people are familiar with and continue to use until now. However, PLCL, PPY, CL, PHBV and various types of silk fibroin seems to reach the “bottle neck” as they have not been used since 2019, probably due to the development of more popular polymers, PCL and PLLA.

Interestingly, there are several polymers or materials that are having specific function or characteristics, which attracted much attention by researchers, such as: PFTBA which could provide short term oxygen supply to Schwann cells (SCs) and counteract the detrimental effects of hypoxia on SCs during the early stages of nerve injury;⁵³ PLATMC, a kind of shape memory polymer, which can realize shape recovery through an entanglement of the molecular chains without chemical and physical cross-linking;⁵⁴ and MAP that displays negative charge that can confer cell adhesion, cell proliferation and tissue regeneration due to its unique adhesive property and biocompatibility.⁵⁵

Characteristics of Nerve Conduit

The average diameter of electrospun micro- or nano-fiber is the most important factor that concerns researchers. The diameter of electrospun fiber should be similar with or close to that of native epineurium which is 67 ± 25 nm.²⁹ Among the 39 studies, Niu et al²⁸ using PLLA and gelatin produced nano-fibers closest to the native value with the average diameter of 77 ± 35 nm. Neshat et al²⁴ produced the nano-fibers with average diameter 71 ± 14 nm, by adopting PCL & DSC cross-linked sodium alginate. Cheong et al⁴² produced the lowest diameter nano-fibers that could be electrospun, 20–30 nm, by adopting PLGA (75:25) and 10% MAP.

The mechanical properties of a native nerve or a nerve conduit can be expressed as tensile strength, Young’s modulus, elongation at break and elasticity. Tensile strength and Young’s modulus are the more common indicators. The tensile strength and Young’s modulus value of a native sciatic nerve of rats are approximately 6.5–11.7 MPa and 0.58 MPa respectively.⁵⁶ By comparing the data of mechanical properties of nerve conduit in Table 1, only 6 out of 39 studies have fabricated a nerve conduit having the similar mechanical properties with the native sciatic nerve: Chang et al³⁷ (2017; PEO/gelatin), Jing et al¹⁸ (2018; PLGA/PPY), Wang et al²⁵ (2020; PLATMC), Chen et al⁴⁵ (2020; PCL), Amini et al⁴⁶ (2020; PCL) and Samadian et al⁴⁷ (2020; PCL/gelatin). We can conclude that recent studies have been more successful in developing nerve conduits which have mechanical properties closer to native nerve. The key polymer used was PCL which could account for its popularity.

It was found that conductivity is not a necessary characteristic for nerve conduit, thus, only studies that used conductive materials presented the need to test conductivity. Conductive materials such as PPY and carbon nanotubes (CNTs) are commonly used, owing to their excellent electrical properties and biocompatibility (Table 1).⁵⁷ As electrical signals might be generated between the neural cells themselves during the nerve regeneration process, electrospun conductive conduits may have the potential to propagate these electrical stimuli, and consequently, stimulate nerve regeneration.⁵⁸ However, the necessity and feasibility of conductive material for a nerve conduit graft should be further validated through the results of in vivo implantation in an animal nerve defect model, which will be discussed later.

A nerve conduit could be said as the “bridge” to connect the nerve gap and provide topological support for axonal outgrowth or cell proliferation and attachment, thus the incorruptibility of scaffold before the nerve regeneration has completed is extremely important. It is known that the nerve regeneration rate for human nerve is 1–3 mm per day, and slightly faster in rat.⁵⁹ As a result, the degradation rate of a nerve conduit should be as long as possible in order to provide enough support for axonal outgrowth through the defect nerve gap. For instance, a nerve conduit should ideally maintain its scaffold for 5 days to provide enough support for a 5 mm nerve defect model. Studies often use the weight loss percentage or swelling ratio after immersing the scaffold into normal saline or PBS in vitro over a period of time to express the degradation rate of the scaffold. Alternatively, some studies after implanting the scaffold subcutaneously over a period of time roughly estimate the percentage of degradation by gross observation. However, gross observation or simply weighing are not absolute to the cell environment. For example, if small corruptions had happened inside the lumen of scaffold, it might be a critical situation for cells in the scaffold even if no obvious degradation could be observed. Therefore, we should evaluate the degradation rate of scaffold in detail using a microscope in order to effectively express the situation of cells, but none of the 39 studies have done the evaluation under a microscope.

The biocompatibility of nerve conduit is the key indicator for in vivo implantation. An in vitro evaluation of nerve conduit with actual cells is necessary before having implantation to assess survival and viability of cells after seeding onto fabricated nerve conduit. Six studies did not evaluate in vitro biocompatibility (Table 1). Cells used by the other studies include PC12 cells, SCs, DRG cell lines, RSC96 cell line, A-172 cell line, bone marrow stromal cells (BMSCs), adipose-derived stem cells (ASCs), SCTM41 cell line, cortical neurons, neural stem cells (NSCs) and induced pluripotent stem cells–neural crest stem cells (iPSCs-NCSCs). Furthermore, biocompatibility tests provide researchers to observe the tolerability of cells and whether the scaffolds are suitable for in vivo implantation.

Additional Factors

Apart from providing topological support for facilitating nerve regeneration, researchers believe that the additional chemical or biological factors, such as neurotrophic factors, drugs, bio-functional peptides or cells-seeded on conduit, can boost the regeneration further (Tables 2 and 3). Most of the studies used single factors, such as cells only or drugs only; whilst 3 studies^48–50 used combination strategies, combining cells and neurotrophic factors or drugs.

Cells

Cell therapy acts as a nerve repair strategy which creates a favorable environment in the peripheral nervous system.⁶⁰ Mesenchymal stem cells (MSCs) are the most popular candidate for providing biological support for nerve regeneration (Table 2). The sources of MSCs were mostly bone marrow cavity (BMSCs),^30,49 adipose tissue (ADSCs, ASCs)^31,48 and exfoliated deciduous teeth pulp (SHED).³⁴ Researchers have demonstrated that MSCs can replace lost neurons and non-neuronal cells by differentiating them into the neural lineage or by providing trophic support for the repair process.⁵⁹ Thus MSCs, with their ease of accessibility, fast proliferation rate, differentiation potential and inherent ability to secret nerve growth factor,⁶¹ are a suitable cell source for the cell-based treatment of peripheral nerve injury. Neural differentiation is better to be performed during cell culture to ensure that the stem cells will grow into neural lineage, whereas only Hu et al³¹ have clearly stated that they have performed neural differentiation. Furthermore, recent studies have indicated that as a gene manipulation strategy, the regulation of microRNA (miRNA) expression levels has been useful for facilitating stem cell neuronal differentiation and neurogenesis, with profound value for nerve regeneration.⁶² Hu et al³¹ have identified an original approach to stem cell transformation based on temporally sequential treatment with Fibroblast Growth Factor 2 (FGF2) and miR-218 to promote ASCs neuronal differentiation.⁶³

SCs are essential for nerve regeneration, as they are responsible to form myelin in the peripheral nervous system, secrete neurotrophins and produce extracellular matrix molecules to facilitate axonal outgrowth and elongation.⁶⁴ MSCs are believed to be differentiated to SC-like cells, however, this could be complicated and unstable. As a result, several groups of researchers^32,35,48 tend to seed SCs directly onto the conduit, rather than attempting to turn stem cells into SCs. As shown in Table 2, the cell source of SCs is mainly from sciatic nerves.

Sources for cells were mainly from rats, as the nerve graft would be implanted to a rat nerve defect model. Only Farzamfar et al⁴⁰ and Pereira dos Santos et al³⁴ used stem cells from human instead of rats, without applying any immunosuppressant, which may bring immunological rejected concerns to the studies. Nevertheless, 3rd–5th passage of cells had been used, and cell numbers varied from 1.42 × 10⁴ to 2 × 10⁶ cells in each conduit. Moreover, cell fate tracking is important as researchers should have a better and clearer understanding about the condition of their seeded cells on the conduit after in vivo implantation. It is important to know whether the seeded cells have successfully been turned into neural lineage, or they provide environmental support without turning into SCs. Disappointedly, only Ma et al³⁵ have engineered the seeded cells for cell fate tracking.

Although incomplete, they manage to demonstrate how their seeded SCs interacted with their fabricated nerve conduit, after 14 days of surgery. They used fluorescence protein for tracking the seeded SCs, to measure the viability of SCs after implantation, and they found that SCs seeded on PFTBA fabricated nerve conduit produced the highest viability of SCs while comparing to only PCL fabricated nerve conduits.

The best strategy for nerve regeneration should be supported by the results of in vivo implantation in an animal nerve defect model, which will be discussed later.

Neurotrophic Factors (NFs)

NFs that have been used for nerve regeneration include brain-derived neurotrophic factor (BDNF),^37,39,50, nerve growth factor (NGF),^37,41,49 neurotrophin-3 (NT-3),³⁹ platelet-derived growth factor (PDGF)³⁹, vascular endothelial growth factor (VEGF)⁴¹ and platelet-rich plasma (PRP).⁴⁷ These NFs were carried by gelatin nanoparticles or solution, hydrogel, chitosan nanoparticles, as well as directly incorporated onto the electrospun fibers with different release profile (sustained, gradual or accumulated release), as shown in Table 3.

BDNF is important in NSCs proliferation, differentiation and directional migration.⁶⁵ Adding exogenous BDNF enhances myelination, whereas the removal of endogenous BDNF inhibits the formation of mature myelin internodes.⁶⁶

NGF plays a vital role in nerve cell growth, differentiation, regeneration and neurotransmitter homeostasis,⁶⁷ and prevents neuronal degeneration in animal models with encouraging results.⁶⁸ NGF has been widely employed to maintain neuronal differentiation and promote the proliferation of pheochromocytoma cells.⁶⁹ Moreover, the axonal growth of the neural cells was significantly enhanced in a dose-dependent manner of NGF.⁷⁰

NT-3 is known to proliferate fibroblasts to help the nerve regeneration in a rat model.⁷¹ NT-3 usually works in combination with BDNF to promote axonal regeneration and behavioural outcomes.⁷² The combined secretion of NT-3 and BDNF usually occurs in vivo for cellular proliferation before the differentiation stage.⁷³

PDGF acts as a mitogen and protects against neuronal degeneration, which induces the neurogenesis in the post-mitotic stage, where the differentiation stage would start to dominate the proliferative states.⁷⁴ With a sustained release manner from conduit, the later appearance of PDGF would further extend the neurogenesis process,⁷⁵ which creates a better environment for the nerve regeneration process.

VEGF is capable of increasing vascular permeability, endothelial cell proliferation, migration, survival and angiogenesis.⁷⁶ It has been shown to stimulate SCs invasion and neovascularization.⁷⁷

PRP, the platelet-enriched compartment of centrifuged whole blood, contains a high concentration of various proteins and growth factors (GFs) such as PDGF, VEGF, transforming growth factor (TGF-β1), epidermal growth factor (EGF), insulin-like growth factor (IGF) and basic fibroblast growth factor (bFGF).⁷⁸ Due to its abundant contents, it is usually used as a complementary treatment for periodontal tissue disease, dental implants, burns, dermatology, musculoskeletal system disruption, as well as nerve regeneration.⁷⁹

Bio-Functional Peptides

The ability of MAP to confer cell adhesion, cell proliferation and tissue regeneration is due to its unique adhesive property and biocompatibility, thus making it an alternative to conventional synthetic polymer-based nerve conduits.⁸⁰ Interestingly, it is capable of the genetic incorporation of various bio-functional peptides,⁸¹ such as the bio-functional peptide that derived from ECM, which is IKVAV. The bio-functional peptide plays crucial roles in tissue regeneration through diverse integrin-mediated pathways.⁸² Cheong et al⁴² found that incorporating bio-functional peptide into a material, especially MAP, can intrinsically enhance its regenerative capacity, which maintains the effectiveness of a peptide for prolonged period.

Apart from that, short peptide fragments with cell-specific functions can mimic the functions of growth factors and significantly improve the biological function and specificity of materials. Bioactive RGI peptide derived from BDNF plays an important role in motor neuron outgrowth.⁸³ Peptide KLT simulates the functional segment of VEGF which as stated earlier is important in angiogenesis. Peptide KLT acts as an analog and is the only synthetic peptide that can activate the VEGF receptor.⁸⁴ A combination of KLT and RGI can potentially promote microcirculation reconstruction and recovery of motor function.⁴⁴

Drugs

Drugs that have been used by studies include methylcobalamin (MeCbl),^36,43 citalopram hydrobromide (CTL),³⁸ gabapentin (GBP),⁴⁰ melatonin (MLT),⁴⁵ lignin,⁴⁶ citicoline (CTC)⁴⁷ and curcumin.⁴⁸ These drugs were carried by gelatin nanocarriers or particles, chitosan nanoparticles, as well as water/ethanol (3:7) coagulated or directly incorporated with the electrospun fibers, and result in different release profiles (sustained, gradual or accumulated release), as shown in Table 3.

MeCbl is one of the active forms of vitamin B12 homologues. It is delivered favorably to nerve tissues and effective in promoting nerve regeneration and neuronal cell survival.⁸⁵ It promotes neurite outgrowth and neuronal survival in cerebellar granule neurons and DRG neurons, which improves functional and electrophysiological results in a rat sciatic nerve transection model.⁸⁶ High concentrations of MeCbl are likely to be necessary to maximize effectiveness in promoting nerve regeneration.

CTL is a blood–brain barrier permeable bioactive molecule, selective serotonin reuptake inhibitor antidepressant, as well as a widely used drug for many mental disorders such as depression and anxiety.⁸⁷ It promotes neuronal cell proliferation and enhances neuroplasticity.⁸⁸

GBP is an anticonvulsant drug used as an analgesic to control neuropathic pain.⁸⁹ It is able to enhance the nerve remyelination after chronic constriction of the sciatic nerve, suggesting its application for nerve regeneration.⁹⁰

MLT is an indoleamine that has neuroprotective and anti-inflammatory effects that protects cells from reactive oxygen species and reactive nitrogen species.⁹¹ It is effective in scavenging free radicals, inhibiting apoptosis, preventing scar formation and cytoskeletal remodeling.⁹² It might be an anti-inflammatory agent as it can regulate expression of proinflammatory cytokines and transcription factors related to antioxidant response such as NFκB and Nrf2.⁹³ Thus, loading MLT on nerve conduit may help to reduce the inflammatory response after surgery.

Lignin has antioxidant, antimicrobial and stabilizer properties. Therefore, it can act as a scavenger of oxygen free radical for sustaining the reactions initiated by oxygen radicals, and at the same time promotes neural cell viability, proliferation and differentiation.⁹⁴

Phosphatidylcholine is the primary component of neuron membrane, while CTC is a choline donating intermediate in the biosynthesis of phosphatidylcholine. As a result, CTC is beneficial for cognitive impairment of diverse etiology, treatment of cerebrovascular disease, Parkinson’s disease (PD) and head trauma,⁹⁵ which exhibited neuroprotective activities. Intraperitoneal and topical administration of CTC may enhance the anatomical regeneration and functional recovery of PNI.⁹⁶

Curcumin has a great potential ability in the regeneration of injured peripheral nervous system (PNS)⁹⁷ and protect dorsal root ganglion.⁹⁸ It is effective in the proliferation and differentiation of SCs,⁹⁹ which significantly decrease the apoptosis of SCs and dramatically increase the number of myelinated axons in the injured sciatic nerve.¹⁰⁰

Others

Chen et al⁴⁵ adopted Iron (II, III) oxide-magnetic nanoparticles (Fe₃O₄-MNPs) as an additional factor to nerve conduit. Fe₃O₄, with certain superparamagnetic properties, is capable of inducing axonal extension or direct neurite outgrowth under an external magnetic field without any side effects.¹⁰¹ Fe₃O₄ nanoparticles can also enhance neurite outgrowth by activating the mitogen activated protein kinase signaling pathway.¹⁰²

Jahromi et al⁵⁰ employed gold nanoparticles (AuNPs) which may enhance the cell–material interactions in terms of cell adhesion, proliferation, differentiation, stimulate axonal elongation and sprouting axons.¹⁰³

By all possible means, the best strategy for nerve regeneration should be supported by the results of in vivo implantation in an animal nerve defect model, which will be discussed later.

Preclinical Model

An animal defect model is the most important part for in vivo investigation. The size and species of the animals may reflect the level of studies, as studies from rats, rabbits, monkeys or humans are hugely different. Almost all of the studies adopted Lewis rats, SD rats or Wistar rats. Only 1 out of 39 studies³⁷ adopted New Zealand white rabbits for in vivo nerve regeneration study. With respect to the anatomical site of the investigation, 38 employed sciatic nerve defect, and only 1 employed median nerve²¹. The defect size and duration of implantation were at the range of 7–15 mm and 4–48 weeks, respectively (Table 4).

Although almost all of the studies created a segmental nerve defect. One study crushed the nerve from sciatic notch³⁶ and another clamped the sciatic nerve by micro clamp³⁹. For both of these studies, researchers utilized tools to dysfunction the nerve, then wrapped the electrospun fiber sheet around the injured site, to serve as a nerve “conduit”.

Outcomes

Functional Assessments

The functional outcomes, brief information and data of those 39 in vivo peripheral nerve repair studies are listed in Table 4. Statistically, p < 0.05 and N.S. indicate significant difference and not significant respectively, and are important and symbolic to a randomized study. However, some studies^22,32,49 discussed their results without providing enough statistical comparison. Thus, in this review, only those quantitative results with statistical analysis will be focused and discussed in detail.

Studies will be compared within groups, whereby the groups will be divided according to the duration of their study. In summary: 2 studies in the group of 60 days, 6 studies in the group of 8 weeks, 18 studies in the group of 12 weeks, and 2 studies in the group of 14 weeks, total 4 groups to be compared; those 12 studies which are not belonging to these 4 groups are not comparable as their animal model or duration of implantation are unique and different with others, thus they were listed as a reference. Besides, the experimental groups within a study have been standardized as: autograft group (A), normal group (sham operation, without surgery; P), negative control group (injured nerve without treatment; N) and other treatment groups (B and C). The details of each group are labeled in Table 4.

During the functional assessments or examinations, researchers should always compare the defect site to the normal site of the model. There are some terms that are common in functional assessment of an animal nerve defect model: sciatic functional index (SFI) for motor functional analysis; hot plate latency (HPL) and paw withdrawal latency (PWL) for sensory recovery analysis; muscle weight measurement for muscle atrophy analysis after surgery; Masson Trichrome (MT) staining for muscle and collagen fiber staining; and amplitude of compound muscle action potential (CMAP), nerve conduction velocity (NCV) and time latency of compound muscle action potential (TL) that belong to the electrophysiological examination, which correspond to the area and size of axons and level of myelination.

For the 2 studies in the group of 60 days, they are not comparable as Naseri-Nosar et al³⁸ did not compare their experimental groups with the autograft group while Farzamfar et al⁴⁰ did. Interestingly, it seems like GBP truly provided its beneficial effect on enhancing remyelination and hence promoted the nerve regeneration as Farzamfar et al⁴⁰ showed us the great performance of the experimental group which have no significant difference to the autograft group only after 60 days of implantation. Although there was no autograft group to be compared in the study by Naseri-Nosar et al³⁸, their result stated that the functional recovery of experimental group with CTL was significantly better than without CTL.

Within the group of 8 weeks, only Zhang et al¹⁷ who adopted aligned PHBV/PEO fibers, showed significantly greater SFI in the nanofibers group compared to the autograft group. However, they only evaluated SFI, thus, it should not be overgeneralized to say that their strategy is better. Nonetheless, both Song et al¹³ and Moharrami Kasmaie et al²⁷ who adopted aligned PPY/PLCL and PCL/PLGA fibers, respectively, have shown that no significant difference between experimental groups and autograft group. However, Kaka et al³⁰, and Cheong et al⁴² and Zheng et al²⁹ have shown inferior results in the experimental group, even with the addition of BMSCs or bio-functional peptides. This may due a short duration of 8 weeks which may not be enough for effective functional recovery.

Within the group of 12 weeks, Zhang et al¹⁴, Sun et al¹⁹, Ma et al³⁵ and Jahromi et al⁴⁸ have shown no significant difference between experimental groups and the autograft group. As expected, Ma et al³⁵ who adopted oxygen releasing PFTBA fibers with seeded-SCs, and Jahromi et al⁴⁸ who employed the combination strategy of aligned fibrin fibers with SCs and curcumin, have shown better results than the other 16 studies in the group of 12 weeks. Jahromi et al⁵⁰ using the combination strategy of aligned PLGA fibers with r-ADSCs, AuNPs, BDNF and alginate showed significant difference between experimental groups, in which the group with all additional factors “seeded” on the conduit was significantly better than the group with PLGA fibers only. However, they did not compare their experimental groups with the autograft group.

For the 2 studies in the group of 14 weeks, they are not comparable as Lopez et al²¹ did not compare their experimental groups with the autograft group while Farzamfar et al⁴⁰ did. However, it could be said that these 2 studies did not have satisfying results. Lopez et al²¹ failed to show better performance of nanowrap compared to autograft. Farzamfar et al⁴⁰ adopted human unrestricted somatic stem cells (hUSSCs) from umbilical cord to be seeded on nerve conduit. The result of the autograft group was still significantly better than the experimental group, especially the value of SFI and muscle weight loss percentage. This is probably due to the immunosuppressive effect as they employed stem cells from human source without applying any immunosuppressant. However, we should correspond the functional and histological results together to give a better conclusion, which will be discussed in a later part.

It is known that animals may have tendency to self-harm (autotomy) due to the neuropathic pain and inflammations after surgery. Administration of pain-killer or antibiotics may help to alleviate the inflammatory symptoms. However, only 7 out of 39 studies^{14,25,27,32,45,46,49} have reported the application of pain-killer or antibiotics after surgery to avoid serious inflammation and other adverse effects. These drugs were administered only for short durations post-operatively and not the entire duration of the study.

Gene and Protein Expression and Morphometric Assessments

As a matter of fact, the more the number of axons, the larger the CAMP; the larger the diameter of axons, the faster the NCV. Thus, there would be a strong relationship between functional and morphometric outcomes.

According to Table 5, only 1 out of the 39 studies did not provide any histological outcomes³⁸, while the others have adopted hematoxylin and eosin (H&E) staining, toluidine blue (TB) staining and transmission electron microscopes (TEM) for morphometric assessments, and various gene and protein expression for visualization of regenerative axons, SCs or inflammatory markers.

Studies will be compared by groups whereby the groups will be divided according to the duration of study. There are 2 groups: 5 studies in the group of 8 weeks and 16 studies in the group of 12 weeks, total 2 groups to be compared. Nineteen studies were not included in either of these 2 groups as they were not comparable due to lack of statistical analysis, incomplete assessments or unique methodology. The experimental groups within a study have been standardized as: autograft group (A), normal group (sham operation, without surgery; P), negative control group (injured nerve without treatment; N) and other treatment groups (B and C). The details of each group are as labeled in Table 4.

Within the group of 8 weeks, although Zhang et al¹⁷ has shown significantly better functional outcomes to other groups, they did not provide the statistical significance of their histological study, thus was excluded from the discussion here. With respect to functional outcomes, both Song et al¹³ and Moharrami Kasmaie et al²⁷ have shown no significant difference of nerve regeneration between experimental groups and the autograft group whilst Zheng et al²⁹ again showed that their nerve conduit was still inferior to autograft. Although Kaka et al³⁰ did not compare their results with autograft, they showed that the number and area of regenerated nerve fibers and blood vessels were larger with the presence of BMSCs. Cheong et al⁴² had shown that functionalization with bio-functional peptides resulted in higher S100 (marker for SCs) and NF200 (marker for regenerated axons) fluorescence intensity relativity and number of myelinated axons in the experimental group compared to the autograft group. Both the studies from Kaka et al³⁰ and Cheong et al⁴² showed us the potential of the strategies of stem cells and bio-functional peptides to be used in effective nerve regeneration.

For the group of 12 weeks, Zhang et al¹⁴, Jing et al¹⁸, Ma et al³⁵ and Samadian et al⁴⁷ have shown no significant difference between experimental groups and autograft group. Apart from the 4 above-mentioned studies, the regenerated nerves from other studies were still inferior than autograft, even with additional factors or combination strategies. Zhou et al¹² and Jahromi et al⁵⁰ have shown us that nerve conduit with aligned nanofibers or additional factors brought the superior nerve regeneration results to the one without aligned nanofibers or additional factors, which provided encouragement to the application of electrospun nanofibers or additional factors for the development of nerve guidance conduit for peripheral nerve regeneration.

We may observe the degradation of materials of nerve conduit through H&E staining as well, in which the residual materials can be recognized from the staining results. Nerve conduit should provide sufficient support to the nerve during the regeneration process. Hence it should not be fully degraded before the nerve regeneration has completed. Zhou et al¹² have mentioned that residual PPY polymers could still be observed after 12 weeks of implantation, which indicated that the polymer is suitable for nerve conduit fabrication as it will be gradually degraded throughout the nerve regeneration process.

Inflammation that occurs after a surgery will further be compounded by the presence of foreign component into the body. It can be observed by H&E staining, or by inflammatory protein markers expression. Xia et al⁴¹ observed more inflammatory cells in the conduit group compared to the autograft group, although SCs proliferation, distribution, arrangement of nerve fibers and formation of new blood vessels in both groups were similar. Interestingly, Chen et al⁴⁵ have mentioned that MLT & Fe₃O₄-MNPs actuated the translation of macrophage phenotype from pre-inflammatory (M1) to anti-inflammatory (M2), which showed us the potential of adding anti-inflammatory drugs to the nerve conduit to reduce pain in animal models.

Although not being mentioned in Table 5, from the histological results, it was observed that the regenerative axons and SCs were able to grow parallel and along the alignment of electrospun fibers. Besides, higher density of regenerated nerve fibers, axons and SCs, less inflammation could also be observed in the aligned morphology. Conversely, random-oriented fibers showed lower density of nerve fibers and more inflammation. Thus, electrospun aligned fibers not only guide regeneration of axons and SCs in an orderly fashion, they also aid nerve repair in a healthier environment.

Biasness of Study

In order to look for alternatives of autograft for treatment of PNI, we should always compare our fabricated nerve conduit with the autograft, to give more appropriate evaluation findings, and to reveal how much the conduit is inferior to or close to the autograft. Besides a clearly stated sample size or capacity, such as the number of animals to be used in the research, it is important to indicate the transparency of a research study. Moreover, as mentioned above, statistical significance is extremely important and symbolic to a randomized study. It indicates the statistical relationship between randomized subjects, and visualizes the results through a simple index, allowing readers to understand the results of comparison and efficacy of specific nerve conduit intuitively. Furthermore, for studies using cells for additional factors, the passage number and number of cells to be seeded on conduit should be clearly stated. It is because cells at different developing stage may present different performances, which may have a huge impact on the studies.

However, according to Tables 2, 4 and 5, it was found that some of the studies: did not compare the efficacy of nerve conduit with autograft; did not clearly state their sample size, passage number of cells and numbers of cells to be seeded on conduit; and lastly the statistical analysis has not been done; which those “non-transparent” studies may lead to the lack of objectivity, and the decrease of reliability.

Studies are not without limitations. However, these limitations may drive the development of future studies, to fabricate better nerve conduits. Although some studies have fabricated nerve conduits that are reported to be at par with autograft, in general most were found to be inferior to autograft, even with added cellular or non-cellular factors. This may be due to the difficulty to mimic the unique microstructure of native nerves by conduits, and the lack of integrated delivery systems for growth factors or facilitating cells that limit the nerve repair function of the nerve conduits. The majority of the researches were of short duration, resulting in insufficient time for recovery of the injured nerve which may also contribute to a poorer outcome of the nerve conduit.

Conclusion

Electrospinning techniques produce nerve conduit that consists of aligned fibers which have been shown to produce better results compared to random fibers. Additionally, the alignment of the nerve conduit may help the orientation and attachment of additional factors such as cells, NFs, drugs and bio-functional peptides, which further extended the potentials and usage of a nerve conduit (Figure 2).

After a series of comparison, the most popular material for electrospun nerve conduit was PCL and PLLA. These two materials have shown better results compared to others and can be considered as the more suitable materials for fabrication of a nerve conduit. From Table 1, we found that conductivity of a nerve conduit was not significant to improvement in nerve regeneration and, thus, not a necessary indicator to be tested.

Based on Tables 1–5, the paradigm of development of electrospun nerve conduit has been switched from simple nano- or micro-fibers conduit to functionalization of a conduit using combination factors. Attempts are made to maximize the advantages of additional factors, with more complicated and complete strategy, to complement the deficiency of only a single factor. Although most of these studies with additional factors have shown better outcome compared to those without, they are still unable to demonstrate a superior result compared to autograft. The addition of stem cells, NFs, drugs or bio-functional peptides although optional, have potential to accelerate the nerve repairing procedure, and provide an environmental guidance for nerve regeneration. Furthermore, functionalization with antibiotics would be advantageous in a setting of an open wound in preventing surgical site infection and improving wound healing.¹⁰⁵

Further Study

Studies on development of electrospun nerve conduit are still at the preclinical phase. There are many potentials awaiting to be explored, and further studies should continue until the best strategy has been found. There are other drugs or materials such as graphene and boron nitride that are used on other nerve conduit which have not been explored in electrospun conduits.^106,107 Aside from basic evaluation of the performance of in vivo implanted nerve conduit, researchers should perform molecular and cellular study to elucidate the effects of nerve conduit on facilitating better fascicle alignment, decreasing scarring, reducing inflammation and protecting fascicles from mechanical deformations.

There are only a few current literatures to show that electrospun nerve conduits are superior to autograft and further researches manipulating various parameters may not alter this fact. Researchers should consider accepting the fact that tissue engineered nerve conduit may never be at par with autograft and commence clinical trials where we will have a better understanding of its action in the human body.¹⁰⁸ The ability to avoid harvesting a donor nerve, thus, preventing donor site morbidity may ameliorate any disadvantages of a nerve conduit.

The logistic aspect of the commercialized product must also be considered. Stem cells harvested from the patient have to be factored in and this undoubtedly would affect the cost. Furthermore, the need to culture cells will also affect timing of surgery from the initial injury which could inherently impact the outcome of nerve regeneration.