The leading edge: Emerging neuroprotective and neuroregenerative cell‐based therapies for spinal cord injury

Abstract Spinal cord injuries (SCIs) are associated with tremendous physical, social, and financial costs for millions of individuals and families worldwide. Rapid delivery of specialized medical and surgical care has reduced mortality; however, long‐term functional recovery remains limited. Cell‐based therapies represent an exciting neuroprotective and neuroregenerative strategy for SCI. This article summarizes the most promising preclinical and clinical cell approaches to date including transplantation of mesenchymal stem cells, neural stem cells, oligodendrocyte progenitor cells, Schwann cells, and olfactory ensheathing cells, as well as strategies to activate endogenous multipotent cell pools. Throughout, we emphasize the fundamental biology of cell‐based therapies, critical features in the pathophysiology of spinal cord injury, and the strengths and limitations of each approach. We also highlight salient completed and ongoing clinical trials worldwide and the bidirectional translation of their findings. We then provide an overview of key adjunct strategies such as trophic factor support to optimize graft survival and differentiation, engineered biomaterials to provide a support scaffold, electrical fields to stimulate migration, and novel approaches to degrade the glial scar. We also discuss important considerations when initiating a clinical trial for a cell therapy such as the logistics of clinical‐grade cell line scale‐up, cell storage and transportation, and the delivery of cells into humans. We conclude with an outlook on the future of cell‐based treatments for SCI and opportunities for interdisciplinary collaboration in the field.

families. 1,2 Direct lifetime costs of care range from $1.1 to $4.7 million per person not including lost wages and productivity. 2 Rapid delivery of specialized medical and surgical care has significantly reduced mortality; however, long-term functional recovery remains limited. [3][4][5][6] Cellbased therapies have emerged as an exciting strategy to neuroprotect and regenerate the injured cord through multiple mechanisms such as immunomodulation, paracrine signaling, extracellular matrix (ECM) modification, and lost cell replacement. 7,8 Herein, we summarize the most promising preclinical and clinical cell therapies, adjunct strategies to enhance transplant success, as well as key translational considerations such as sex and age. Throughout, we emphasize the fundamental biology of stem cells, critical features in the pathophysiology of spinal cord injury and provide meaningful discussions on the strengths and limitations of each therapeutic approach.

| Epidemiology
The epidemiology of SCI is an important consideration when designing clinical trials. Traumatic SCI is more common in males (79.8%) than females (20.2%). Most injuries are cervical (60%) followed by thoracic (32%) and lumbosacral (9%). 9 There is a bimodal age distribution with one peak occurring from 15 to 29 years of age and a second, smaller but growing peak, occurring after age 50. 10,11 High-energy motor vehicle collisions (MVCs) and sports-related injuries disproportionately affect younger individuals. Low-energy trauma, such as falls, are more common in those over 60 years old where underlying degenerative spinal conditions, such as degenerative cervical myelopathy, are more prevalent. 11,12 Interestingly, MVCs account for a declining majority (38%) of SCIs in North America, 9 whereas falls are increasing and account for 31% of injuries followed by sports-related impacts at 10% to 17%. 11,12 2 | PATHOPHYSIOLOGY 2.1 | Acute injury and the postinjury milieu The initial traumatic event causes permeabilization of cell membranes, ion and small molecule dysregulation, and ischemia due to damage to the sensitive microvascular supply. 13,14 Together, these events initiate a secondary injury cascade which generates further permanent damage ( Figure 1A). Over several hours, progressive edema and hemorrhage cyclically add to the harsh postinjury milieu. The compromised blood-spinal cord barrier (BSCB) exposes the vulnerable cord to inflammatory cells, vasoactive peptides, and cytokines such as tumor necrosis factor and interleukin-1β. 16 Ongoing cell death releases DNA, ATP, and K + into the microenvironment; microglia respond by secreting additional pro-inflammatory cytokines and promoting the infiltration of large numbers of macrophages, neutrophils, and nearby microglia. This activates astrocytes and endothelial cells which further secrete factors such as BMPs, TGF-β, and Notch activating ligand, Jagged. Activated phagocytes can clear myelin debris within the injury but also produce oxygen free radicals (eg, O2 − , peroxynitrite and hydrogen peroxide) and cytotoxic by-products which generate additional cell death through lipid peroxidation, protein oxidation, and DNA damage. 17,18 Extracellular glutamate accumulates as neurons die and astrocytes' reuptake capacity is lost. 19,20 This leads to excitotoxic cell death of the remaining neurons through NMDA, kainate, and AMPA receptor overactivation combined with ATP-dependent ion pump dysfunction and subsequent sodium dysregulation ( Figure 1B). 21,22 At a systemic level, poor respiratory function can cause hypoxia whereas loss of sympathetic innervation to the vasculature can result in profound hypotension. Combined with the impaired autoregulatory capacity of the cord, this can contribute to ongoing ischemia for days to weeks postinjury. 23 The multiple causes of acute and subacute cell death in this injury cascade represent important targets for cell-based neuroprotective approaches.

| Barriers to recovery
In the intermediate-chronic phase, acute inflammation subsides and the cord undergoes alterations in ECM composition, attempts at remyelination, and remodeling of neural networks. 24 Although this can result in limited recovery, multiple barriers to local circuit and long-tract regeneration persist.
Neuroglial cell death and degeneration in the early phase disrupts the cord's structural framework and leads to ex vacuo formation of microcystic cavitations containing extracellular fluid with thin bands of connective tissue. 25 These cavities coalesce into larger collections which lack substrate for directed axonal regrowth and regenerative cell migration. 26,27 Additionally, oligodendrocytes are susceptible to necrotic and apoptotic cell death. The denuded axons they leave behind cannot utilize rapid saltatory conduction and are particularly susceptible to nonfunctional electrogenesis which further contributes to poor recovery. 28 Significance statement F I G U R E 1 Pathophysiology of traumatic spinal cord injury. "(a) The initial mechanical trauma to the spinal cord initiates a secondary injury cascade that is characterized in the acute phase (that is, 0-48 hours after injury) by oedema, haemorrhage, ischaemia, inflammatory cell infiltration, the release of cytotoxic products and cell death. This secondary injury leads to necrosis and/or apoptosis of neurons and glial cells, such as oligodendrocytes, which can lead to demyelination and the loss of neural circuits. (b) In the subacute phase (2-4 days after injury), further ischaemia occurs owing to ongoing oedema, vessel thrombosis and vasospasm. Persistent inflammatory cell infiltration causes further cell death, and cystic microcavities form, as cells and the extracellular architecture of the cord are damaged. In addition, astrocytes proliferate and deposit extracellular matrix molecules into the perilesional area. (c) In the intermediate and chronic phases (2 weeks to 6 months), axons continue to degenerate and the astroglial scar matures to become a potent inhibitor of regeneration. Cystic cavities coalesce to further restrict axonal regrowth and cell migration." Republished with permission from Ahuja et al 15 Early after injury, astrocytes also proliferate within the perilesional zone and tightly interweave an irregular mesh of processes to sequester the injured region ( Figure 1C). Resident neural stem and progenitor cells surrounding the central canal can also differentiate to astrocytes and contribute to this astrogliosis. The astrocytes, pericytes, and ependymal cells in the region generate dense deposits of chondroitin sulfate proteoglycans (CSPGs), NG2, and tenascin which form the fibrous component of the glial scar. [29][30][31][32] Although literature exists supporting the beneficial aspects of scar, the balance of evidence suggests that chronic scarring potently inhibits axonal regeneration and neurite outgrowth by acting as a physical barrier and tightly binding transmembrane protein tyrosine phosphatase receptors. [33][34][35] Furthermore, CNS myelin-and neuron-associated ligands, such as  Table 1, and completed and ongoing clinical trials are summarized in Tables 2 and 3, respectively.

| Cell source
MSCs, SCs, and OECs can all be harvested from an adult allogeneic source to generate standardized stocks depending on the success of proliferation. MSCs, SCs, and OECs can also be derived directly from the patient to avoid post-transplant immunosuppression. 84

| Neural stem cells
NSCs are tripotent, self-renewing cells which have attracted great interest as they can potentially replace the neurons, oligodendrocytes, and astrocytes lost after injury. 88,[95][96][97] During embryological development, NSCs are found throughout the neural tube where they acquire unique identities based on their position and temporal exposure to patterning morphogens. [98][99][100][101] In adults, they are found in a more limited number of regions such as the subventricular zone in the brain [95][96][97] and around the central canal in the spinal cord. [102][103][104][105] There are two distinct NSC populations that can be isolated from the adult spinal cord: (a) primitive NSCs (pNSCs) and (b) the definitive NSCs (dNSCs) they give rise to ( Figure 2). 106 recovery. The underlying mechanism may be a combination of remyelination, local immunomodulation, trophic factor secretion, and provision of a physical scaffold to support growing axons. 69,70,115,116 The cells also possess a favorable secretome, 117

| Growth factors
To support graft survival, growth factors (eg, PDGF, EGF, and IGF-1), neurotrophins (eg, BDNF, NT3, NGF), and anti-inflammatory agents (eg, minocycline) have all been successfully delivered via intrathecal injections and pumps. 63,[164][165][166] Unique biomaterials have also been engineered to gradually deliver key factors to support grafts. 167  been successfully engineered to express bFGF, 168 HGF, 169 NT3, 170 BDNF, 171 and GDNF 172 in vivo for various applications. SCs have also been transduced to overexpress BDNF and NT3 simultaneously. 173 Similarly, safe and highly efficient methods of engineering human iPSCs, ESCs, and NSCs are currently being developed.

| Rehabilitation
An often overlooked method of promoting endogenous trophic factor release and long-term cell survival noninvasively is rehabilitation. Physical rehabilitation, with or without electroceutical augmentation, is an integral component of the care plan for individuals with SCI; however, it is underrepresented in preclinical trials. Whether the rehabilitation entails forced treadmill training, free swimming, or task-specific tests such as forelimb reaching, the functional benefits can be significant. 174 In addition to enhancing cardiorespiratory and musculoskeletal function, treadmill locomotor training has been found to enhance transplanted NSC survival by more than fivefold through increased IGF-1 signaling. 175 This finding underscores the value of multimodality, interdisciplinary care in SCI.

| Biomaterials
Biomaterials can enhance cell-based approaches for SCI in several ways. Scaffolds derived from either natural or synthetic polymers have been implanted within the lesion cavity to bridge the gap to serve as a substrate for axonal growth and cell migration. [176][177][178][179][180] InVivo Therapeutics' Neuro-Spinal Scaffold is a porous bioresorbable polymer scaffold shown to promote appositional healing, white matter sparing, and normalization of intraparenchymal tissue pressure in preclinical models of SCI. 181 A recent case study at 6 month follow-up from a patient enrolled in the clinical trial (NCT02138110) reported no adverse effects related to acute scaffold implantation. 182 Scaffolds can also provide a physical substrate for seeded cells and provide directional guidance for axons. Moreover, biomaterials can also be used as vehicles to deliver cells and release growth factors to aid in graft cell survival, integration, and differentiation. Injectable in situ polymerizing hydrogels can deliver cells and factors directly into a lesion site with less invasive surgical interventions. For example, a polymer blend of hyaluronan/methylcellulose (HAMC) is injectable, in situ gelling, biodegradable, and noncytotoxic. 183 HAMC modified with PDGF-AA, to enhance graft survival and oligodendrocyte differentiation of cotransplanted rat brain-derived NSCs, 184,185 promoted host oligodendrocyte sparing and improved fine motor function. 186 Further modification of the HAMC hydrogel with RGD peptide promoted the survival, integration, and differentiation of human iPSC-derived OPCs. 187 Another approach has been the use of fibrin scaffolds which have been shown to promote the survival of transplanted stem cells after SCI and, when codelivered with growth factors, have been used to direct differentiation and enhance recovery. 188,189 QL6 is an exciting peptide biomaterial which self-assembles to form a lattice-like structure at physiological temperatures. QL6 injected with NSCs improved graft survival, reduced glial scarring and inflammation, and improved forelimb function in cervical models of SCI. [190][191][192]

| Galvanotaxis
Electrical fields (EFs) are a physical environmental cue present within living tissue. During development, multipotent cells rely on these fields for appropriate migration and differentiation. If the fields are disrupted, severe defects can result. 193 EFs have also been shown to guide cells in adults after injury. 194,195 Therapeutic galvanotaxis (the directed migration of cells in an electric field) exploits the electrosensitivity of cells to promote migration using externally applied EFs. 196 This has been shown to be feasible with SCs, 197 NSCs, 198 and many other cell types. 199 Both endogenous and transplanted NSCs, but not their differentiated progeny, have been shown to migrate with transcranial direct current electrical stimulation; however, directed migration will require further optimization to establish the ideal current, voltage, phase, lead placement, and timing of EF application. 200,201 4.5 | Disrupting the glial scar The glial/CSPG scar is well established in chronic injury and limits axon regeneration through the lesional/perilesional region ( Figure 1C).  Another exciting approach is to inhibit the association of the protein tyrosine phosphatase σ receptor with its CSPG ligand. 210 An example is Intracellular Sigma Peptide (ISP) which is administered subcutaneously after injury, crosses the BSCB, and results in significant axonal regrowth within the injured cord. 34 Recently, ISP has also been shown to indirectly immunomodulate when combined with leukocyte antigen-related receptor blockade. 211 5 | TRANSLATING STEM CELL THERAPIES

| Immunorejection
Another important translational hurdle is understanding and overcoming potential immunorejection within the CNS. The extent and temporality of cell graft rejection within the human CNS is currently unknown. Additionally, the cell source (eg, autologous, allogenic, genetically modified, etc.) may significantly affect downstream graft survival in humans. 215,216 Enhancing endogenous cell proliferation is one strategy to avoid immunorejection; however, endogenous cell pools may still be limited and optimal methods to drive their differentiation and migration have not yet been established. 217 Recently, genetic techniques which modify major histocompatibility complexes and CD47 have been shown to generate immune-evasive iPSC lines. 218 This may become an important strategy in the future to protect grafts from immune cells.

| Storage and transport
Conducting clinical trials or treatments across multiple sites requires a coordinated storage and transportation approach capable of accommodating international shipping delays and unexpected package handling conditions. For example, a study of human MSCs found that cells stored at 2 C to 8 C were sensitive to 25 Hz vibrations, 227 leading to cell death and increases in MSC marker expression such as CD29 and CD44. 228 Similarly, temperature and repeated freeze-thaw cycles have a significant effect on cell viability. Cryoprotectant toxicity, rapid osmotic shifts, ice crystal formation, and activation of apoptosis are key mechanisms underlying cell death during the process. 229 Early studies achieved human cryopreserved pluripotent cell survival rates of only 30% or less. 230 As a result, other cold storage techniques have been developed such as vitrification, the rapid cooling of cells in high concentrations of cryoprotectants to inhibit ice formation. These were found to provide >75% survival but add technical complexity and may be limiting in the production of large-scale banks. 229 Furthermore, the storage solution (eg, DMSO, polyethylene glycol, etc.) and additives (eg, ROCK-inhibitor, trehalose, poly-L-lysine, etc.) exert their own effects on the survival and differentiation of stored cells. 229 Therefore, each cell type and cell line requires optimization to establish ideal conditions for a functional and reliable supply chain. profound benefits for quality of life making this an important population for inclusion in cell-based studies. It is, however, important to balance these inclusion benefits with potential risks. The cervical SCI population can be more expensive to study in trial as hospital stay, treatment, and rehabilitation costs are higher. 2 Additionally, limiting studies to highly specific study regions can make recruitment more challenging.

| Patient selection
Another key consideration is transplant timing. Due to differences in physiology, injury etiology, cord architecture, patient comorbidities, and a host of microenvironmental cell signals, the optimal timing for transplant in animal models likely differs from humans. 15 Furthermore, even with established infrastructure, there is often a lead time associated with delivering banked cells (eg, allogeneic therapy) or generating a cell line (eg, autologous therapy). 231

| Delivery techniques
Adapting delivery techniques is a key facet of translating cell therapies. Systemic or intrathecal treatments avoid many challenges by using existing, well-established medical techniques; however, the distribution of cells is poorly controlled. For intraparenchymal treatments ( Figure 3), trials are increasingly utilizing standardized, tightly controlled delivery systems which must be scaled for human doses, provide high reliability, allow sterilization, and have undergone regulatory approval as a medical device. 232 The two main classes of injectors currently in use are tablemounted and spine-mounted systems.

| OUTLOOK
The multifaceted pathophysiology of SCI and the complexities of neural repair and regeneration necessitate novel approaches to treatment.
Cell-based therapies continue to be very attractive and hopeful strategies for repair of SCI and the rapid pace of innovation continues to increase as our understanding of fundamental cell biology deepens.
We predict that as the timing, dose, and delivery of adult-and pluripotent stem cell-derived treatments are optimized, increasing numbers of cell-based therapies will be translated to humans. It is highly likely that successful approaches will integrate strategies to enhance and support cells, such as genetic engineering, biomaterials, galvanotaxis, and scar degradation to maximize clinical outcomes. Ongoing preclinical and clinical trials highlight the excitement and tremendous progress that has been made in the field and underscore the importance of the collaborative work being conducted by researchers, clinicians, stakeholders, and funding agencies worldwide.

DATA AVAILABILITY STATEMENT
Data sharing is not applicable to this article as no new data were created or analyzed in this study.