Pharmacological intervention for chronic phase of spinal cord injury

Abstract

Spinal cord injury is an intractable traumatic injury. The most common hurdles faced during spinal cord injury are failure of axonal regrowth and reconnection to target sites. These also tend to be the most challenging issues in spinal cord injury. As spinal cord injury progresses to the chronic phase, lost motor and sensory functions are not recovered. Several reasons may be attributed to the failure of recovery from chronic spinal cord injury. These include factors that inhibit axonal growth such as activated astrocytes, chondroitin sulfate proteoglycan, myelin-associated proteins, inflammatory microglia, and fibroblasts that accumulate at lesion sites. Skeletal muscle atrophy due to denervation is another chronic and detrimental spinal cord injury–specific condition. Although several intervention strategies based on multiple outlooks have been attempted for treating spinal cord injury, few approaches have been successful. To treat chronic spinal cord injury, neural cells or tissue substitutes may need to be supplied in the cavity area to enable possible axonal growth. Additionally, stimulating axonal growth activity by extrinsic factors is extremely important and essential for maintaining the remaining host neurons and transplanted neurons. This review focuses on pharmacotherapeutic approaches using small compounds and proteins to enable axonal growth in chronic spinal cord injury. This review presents some of these candidates that have shown promising outcomes in basic research (in vivo animal studies) and clinical trials: AA-NgR(310)ecto-Fc (AXER-204), fasudil, phosphatase and tensin homolog protein antagonist peptide 4, chondroitinase ABC, intracellular sigma peptide, (-)-epigallocatechin gallate, matrine, acteoside, pyrvate kinase M2, diosgenin, granulocyte-colony stimulating factor, and fampridine-sustained release. Although the current situation suggests that drug-based therapies to recover function in chronic spinal cord injury are limited, potential candidates have been identified through basic research, and these candidates may be subjects of clinical studies in the future. Moreover, cocktail therapy comprising drugs with varied underlying mechanisms may be effective in treating the refractory status of chronic spinal cord injury.

Introduction

Spinal cord injury (SCI) is a major cause of health loss with long-term disability. The global survey indicates 20,635,000 individuals were living with SCI in 2019 (GBD Spinal Cord Injuries Collaborators, 2023). From 1990 to 2019, global incidence (from 595,000 to 909,999), prevalence (from 11,367,000 to 20,635,000), and years of life lived with disability (from 3,749,000 to 6,201,000) remarkably increased (GBD Spinal Cord Injuries Collaborators, 2023). Considering large differences in individual numbers between incidence and prevalence, SCI patients who were not recovered and entered the chronic phase continuously increased.

SCI is an intractable traumatic injury involving axonal disruption at the injured site, which interrupts signal transmission from the central nerve to the muscle and from the peripheral sensory organ to the central nerve. Axonal regrowth and reconnection to target sites are the most commonly faced yet most challenging issues in SCI. Post-SCI timeline is approximately classified into acute (at most 1 week after injury), subacute (1–2 weeks after injury), and chronic phases (4 weeks after injury). A limited therapeutic window, if any, is available in the early phase and cases of mild injury. However, as the SCI stage progresses to the chronic phase, the recovery of motor and sensory functions becomes impossible owing to multiple reasons. Factors that inhibit axonal growth include the accumulation of activated astrocytes (Li et al., 2020), chondroitin sulfate proteoglycans (CSPGs) (Andrews et al., 2012), myelin-associated proteins (Schwaiger et al., 2023), inflammatory microglia (Freyermuth-Trujillo et al., 2022), and fibroblasts (Cooper et al., 2018) in the lesion site (called as glial scar) and progression of histological cavity in the lesion center (Ellingson et al., 2010). Additionally, skeletal muscle atrophy due to denervation is a detrimental chronic SCI-specific condition (Biering-Sørensen et al., 2009; Léger et al., 2009). Several intervention strategies based on multiple viewpoints, such as bioactive substances (neurotrophic factors, small molecule compounds, and exosomes) (Hu et al., 2023), biomaterial transplantation (hydrogel, inorganic nanomaterial, and biomaterials) (Hu et al., 2023), cell transplantation (embryonic stem cells, induced pluripotent stem cells, neural stem cells/neural progenitor cells, mesenchymal stem cells, and olfactory ensheathing cells) (Hu et al., 2023), physical stimulation (electric stimulation, light stimulation, magnetic stimulation, and ultrasound stimulation) (Hu et al., 2023) have been attempted for treating SCI.

Treating chronic SCI may require the introduction of neural cells or tissue substitutes into the cavity area to induce axonal growth because the remaining host neurons and transplanted neurons require the stimulation of axonal growth activity by extrinsic molecules. However, few studies have reported satisfactory efficacy in the approaches for treating chronic SCI. No promising treatments for chronic SCI are established. Major recent challenges to the chronic phase of SCI in animal models used stem cell transplantation (Hashimoto et al., 2023). In addition to supplying neuronal cells in the injured areas, enhancing axonal growth and synaptic connection is necessary for functional recovery in severe conditions of chronic SCI. The potential of pharmacotherapy for axonal growth is worth to be focused although reports of mono-pharmacotherapy are very limited. This review focuses on the pharmacotherapeutic approaches to induce axonal growth, which were conducted by treatments with small compounds and proteins in chronic SCI. Other approaches, such as cell transplantation, biomaterial transplantation, and physical stimulation, have been well summarized in other reviews (Ahuja et al., 2017; Hashimoto et al., 2023; Hu et al., 2023).

Search Strategy and Selection Criteria

Studies cited in this narrative review published until April 2024 were searched on PubMed using the following keywords: spinal cord injury (SCI), chronic phase, pharmacotherapy, pharmacological drug, drug intervention, clinical study, placebo, clinical trials, axon growth, axonal regeneration, and neuronal circuit. All references cited in this review were screened according to the following criteria and searched until April 14, 2024. Full articles in English; pharmacotherapy; experimental studies in vivo; chronic phase of SCI studies; incomplete and complete SCI studies; basic study and clinical studies. Registered clinical studies were searched in ClinicalTrials.gov (https://clinicaltrials.gov/) and jRCT (https://jrct.niph.go.jp/).

Pharmacotherapy of Chronic Spinal Cord Injury: Basic Study

Blocking axonal growth inhibition signaling

AA-NgR(310)ecto-Fc

Myelin-associated inhibitors expressed by oligodendrocytes include Nogo-A, myelin-associated glycoprotein (MAG), oligodendrocyte myelin glycoprotein (OMgp), and ephrin-B3 (McKerracher et al., 1994; Chen et al., 2000; GrandPré et al., 2000; Kottis et al., 2002; Benson et al., 2005). Nogo-A, MAG, and OMgp bind to the leucine-rich repeat of the Nogo-A receptor (NgR1) (Fournier et al., 2002; Barton et al., 2003; McGee et al., 2003). Ligand binding to NgR1 results in the downstream activation of the small GTPase RhoA and Rho-associated protein kinase (ROCK), which alters actin dynamics. This signaling induces growth cone collapse and blocks axonal growth (Duffy et al., 2009; Schwab et al., 2014). To block NgR1 signaling and promote neurological recovery following central nervous system (CNS) injury, a soluble decoy fragment of NgR1—termed NgR(310)ecto-Fc or AXER-204—has been developed (Maynard et al., 2023; Figures 1 and 2). The decoy consists of an NgR1 ligand-binding domain fused to the Fc region of IgG1. AA-NgR(310)ecto-FC is a function-blocking soluble NgR protein fragment. The cDNA encoding the first 310 amino acid residues of rat NgR1 was cloned, expressed in Chinese hamster ovary cells, and purified. “AA” indicates that two Cys residues at positions 266 and 309 of the 310-amino acid ecto protein were mutated to Ala (Wang et al., 2011).

Figure 1.

Structures of small compounds and sequences of peptides and proteins.

Female Sprague–Dawley rats (250–270 g, 11–12 weeks old) were subjected to T7 moderate contusion injury. Twelve weeks after inflicting the contusion injury, AA-NgR(310)ecto-Fc treatment was initiated using an intracerebroventricular cannula connected to an osmotic minipump (Table 1). Each pump was filled with 2.25 mg of AA-NgR(310)ecto-Fc (0.29 mg/kg per day) or 2.25 mg of IgG from rat serum, which was delivered at a rate of 2.5 μL/h for 4 weeks. During the treatment period (12 weeks), two osmotic minipumps filled with the same quantity of AA-NgR(310)ecto-Fc or rat IgG were used and switched every 4 weeks. Immediately before intracerebroventricular (i.c.v.) treatment, the Basso, Beattie, and Bresnahan (BBB) score (Basso et al., 1995) was evaluated and found to be 7.75 ± 0.10 at 12 weeks, indicating that the majority of rats were capable of hindlimb movement but not weight support. During the 12-week treatment phase, BBB score versus time in the IgG group increased from the pre-treatment value of 7.75 ± 0.10 to 8.10 ± 0.20 post-treatment, which was not a significant improvement, whereas the BBB scores of the AA-NgR(310)ecto-Fc group improved significantly from 7.75 ± 0.10 to 8.65 ± 0.20, indicating a significant difference between the groups (Wang et al., 2011). Thus, AA-NgR(310)ecto-Fc treatment of chronic spinal contusion improves neurological recovery, particularly with respect to open field locomotion. Although AA-NgR(310)ecto-Fc-induced change in BBB score was minor (a maximum score of 21), AA-NgR(310)ecto-Fc-treated rats showed remarkably increased weight-bearing capacity in the open field and improved hindlimb foot movement. The serotonin-positive raphespinal tracts were slightly but significantly increased in the caudal region of the lesion center following AA-NgR(310)ecto-Fc infusion.

Table 1.

Pharmacotherapies of chronic SCI: basic study

Category	Treatment	Animal	Injury	Treatment start	Dose	Treatment period	Outcome		Reference
							Motor function	Axon extension
Blocking axonal growth inhibition signaling	NgR(310)ecto-Fc	Sprague–Dawley rat (11–12 wk)	T7, Moderate contusion	12 wk after injury	0.29 mg/kg/d, continuous i.c.v. infusion	12 wk	BBB: Improved	Raphespinal tract: Extended (caudal site)	Wang et al., 2011
		African green monkey	C5/C6 right hemisection	1 mon after injury	0.10–0.17 mg/kg/d, L3, continuosu intrathical infusion	4 mon	Hindlimb walking score: Improved	Corticospinal tract: Extended (caudal site)	Wang et al., 2020
	Fasudil	Wistar rat (8–10 wk)	T9, Contusion	4 wk after injury	180 μg/d, L3–L4, continuosu intrathical infusion	2 wk	BBB: No effect	Corticospinal tract: No extension	Nishion et al., 2006
	PTEN antagonist peptide 4	Sprague–Dawley rat (250–300 g)	C2, Hemisection	8 wk after injury	1 mg/d, s.c.	21 d	Diaphragm motor function: very slightly improved	Bulbospinal rostral Ventral Respiratory Group axon: Extended (caudal site)	Cheng et al., 2021
	Chondroitinase ABC	Sprague–Dawley rat (200–250 g)	C2, Lateral hemisection	12 wk after injury	20 U/mL × 0.25 μL, intraparenchymal injection	Once	Hemidiaphragm paralysis: Improved	Not determined	Warren et al., 2018
	Intracellular Sigma Peptide	Sprague–Dawley rat (280±20 g)	C2, Lateral hemisection	12 wk after injury	500 μg, s.c.	60 d	Forelimb locomotor score: Improved, IBB rating scale: Improved	5-HT-positive fibers: Increased	Milton et al., 2023
Neuroprotection	(-)-Epigallocatechin-3-gallate	Sprague–Dawley rat (11–12 wk)	T9–T10, Contusion	12 mon after injury	20 mg/kg/h, continuous i.v.	6 wk	BBB: Improved	GAP-43-positive area: Increased	Renno et al., 2014
Stimulating axonal growth activity	Matrine	ddY mouse (8 wk)	T10–T11, Contusion	28 d	100 μmol/kg/d, p.o.	154 d	BMS: Improved, BSS: Improved	NF-H-positive axon: Increased (caudal site), Presynaptic density on motor neuron: Increased (caudal site)	Tanabe et al., 2019
	Acteoside	ddY mouse (10 wk)	L1, Contusion	30 d	0.1 mg/hindlimb, i.m., 3 times/wk	62 d	BMS: Improved, TMS: Improved	Raphespinal tract: Extended (caudal site), Presynaptic density on motor neuron: Increased (caudal site)	Kodani et al., 2019
	Pyrvate kinase M2	ddY mouse (8 wk)	T13, Contusion	32 d	1 ng/mL in CSF, continuous i.c.v. infusion	28 d	BMS: Improved, TMS: Improved, Vertical cage scale: Improved	Raphespinal tract: Extended (caudal site)	Kikuchi et al., 2020
	Diosgenin	ddY mouse (8 wk)	L1, Contusion	31 d	0.1 μM in CSF, continuous intrathical infusion	56 d	BMS: Improved, TMS: Improved	Raphespinal tract: Extended (lesion center)	Nakano et al., 2021

5-HT: Serotonin; BBB: Basso, Beattie, and Bresnahan score; BMS: Basso Mouse Scale; BSS: Body Support Score; CSF: cerebrospinal fluid; GAP-43: growth associated protein 43; i.c.v.: intracerebroventricular; i.m.: intramuscular; i.v.: intravenous; IBB: Irvine, Beatties, and Bresnahan rating scale; NF-H: neurofilament heavy chain; NgR: Nogo receptor; PTEN: phosphatase and tensin homolog protein; SCI: spinal cord injury; TMS: Toyama Mouse Score.

In a non-human primate study, female African green monkeys (4.2–7.2 kg) were subjected to C5/C6 right hemisection injury (Wang et al., 2020; Table 1). One month after SCI, an intrathecal catheter connected to an osmotic minipump was placed at the L3 vertebrate level. The treated dose of human NgR(310)ecto-Fc was 0.10–0.17 mg/kg per day, and a 4-month administration was continued. At 6 months after treatment (1 month after the end of administration), the hindlimb walking score was significantly increased in the NgR(310)ecto-FC group compared with the control group. Glia scar size (C5/C6 level), microglia numbers (C5/C6 level), raphespinal serotonergic axons (T3/T4 level), corticospinal tracts (CST) (C1 level) were not different between control and NgR(310)ecto-Fc groups. However, CST fibers in the caudal part of the lesion center (T3/T4 level) significantly increased in NgR(310)ecto-Fc groups. The NgR(310)ecto-Fc treatment increased CST fibers intact side as well as the injured side, suggesting greater ramification of CST fibers in the denervated grey matter. NgR(310)ecto-Fc gets entered clinical studies of spinal cord injury.

Chondroitinase ABC

CSPGs are extracellular matrix molecules and one of the axonal growth inhibitory factors. Within a few days after SCI, CSPGs increase within and adjacent to the forming glial scar. CSPGs induce growth inhibition by binding and activating several transmembrane receptors, including protein tyrosine phosphatases (PTP) σ, leukocyte common antigen related, NgR1, and NgR3 (Sami et al., 2020). Intracellularly, activation of PTPσ and leukocyte common antigen related by CSPGs activate RhoA/ROCK signaling and inactivate Akt and Erk pathways (Figure 2). However, the two receptors use distinct pathways downstream of RhoA/ROCK, Akt/GSK-3β, and Erk signals to mediate the inhibition of axon growth by CSPGs. CSPGs also can bind NgR1 and NgR3 to inhibit axon growth. Intracellularly, NgR functions are mediated mainly by the Rho pathway.

Figure 2.

Schematic of blocking axonal growth inhibition signaling.

Black arrows indicate stimulating, and red bars indicate inhibiting. Blue arrows mean ligand stimulation to NgR1 and Ng3 by MAG, Nogo-A, and OMgp. Yellow structures indicate plasma membranes. Created with Microsoft® PowerPoint® for Microsoft 365 MSO. 4E-BP: Eukaryotic translation initiation factor 4E binding protein 1; ChABC: chondroitinase ABC; CSPG: chondroitin sulfate proteoglycan; GSK3β: glycogen synthase kinase 3β; ISP: Intracellular Sigma Peptide; LAR: leukocyte common antigen related; MAG: myelin-associated glycoprotein; mTOR: mammalian target of rapamycin; NgR(310)ecto-Fc: soluble decoy fragment of NgR1; NgR1: Nogo receptor 1; NgR3: Nogo receptor 3; OMgp: oligodendrocyte myelin glycoprotein; PAP4: PTEN antagonist peptide 4; PIP2: phosphatidylinositol-4,5-bisphosphate; PIP3: phosphatidylinositol(3,4,5)-trisphosphate; PTEN: phosphatase and tensin homolog protein; PTPσ: protein tyrosine phosphatases σ; RhoA: ras homolog family member A; ROCK: Rho-associated protein kinase; S6K: ribosomal protein S6 kinase I.

The glycosaminoglycan side chains, especially those containing 4-O-sulfated CS-E (Miller and Hsieh-Wilson, 2015; Pearson et al., 2018) are known to bind with the highest affinity to PTPσ. The CSPG effect is greatly decreased by enzymatic digestion using the bacterial enzyme Chondroitinase ABC (ChABC; Figure 1).

Lateral cervical level hemisection was inflicted at C2 in Sprague–Dawley rats (200–250 g, female; Table 1). Twelve weeks after SCI, ChABC (from Proteus vulgaris, 20 U/mL × 0.25 μL) or saline was injected at C4–C5. Four weeks after the ChABC injection, hemidiaphragm paralysis was significantly improved (Warren et al., 2018).

Intracellular sigma peptide

The use of ChABC has not been applied clinically because of the heat lability of a bacterial enzyme which, in addition, must be injected intraparenchymally into the injured spinal cord, which can further traumatize an already damaged spinal cord. Intracellular Sigma Peptide (ISP) is a mimetic of the PTPσ wedge domain that contains a trans activator of transcription domain that facilitates membrane penetration (Figure 2).

Lateral cervical level hemisection was inflicted at C2 in Sprague–Dawley rats (280 ± 20 g, female) (Milton et al., 2023; Table 1). Twelve weeks after SCI, ISP (500 μg, subcutaneously, daily) or saline was treated for 60 days. To stimulate forelimb utilization, daily swimming was added from 2 weeks after ISP treatment. Swimming took place 5 days a week and was continued until the end of ISP treatment. Systemic ISP treatment significantly improved forelimb locomotor scores. In the case of Irvine, Beatties, and Bresnahan rating scale, evaluating cereal eating ability, and monotreatment by ISP showed slight improvement. However, combination with 4-methylumbelliferone (perineuronal net (PNN) inhibitor) treatment (per os) significantly enhanced Irvine, Beatties, and Bresnahan score. The role of PNN is to limit plasticity within the CNS. The PNN is activity-dependent and forms around the soma and proximal neurites. Therefore, these behavioral data suggest that modulating CSPGs in the scar and especially in the PNN as well as in their receptor simultaneously is advantageous to the recovery of precision digit function long after SCI.

ISP gets into clinical trials. NVG-291 (=ISP, from NervGen Pharma) is being evaluated in a Phase 1b/2a clinical trial. It is a placebo-controlled proof-of-concept trial (NCT05965700) to evaluate the efficacy of NVG-291 in two separate cohorts of individuals with cervical spinal cord injury: chronic (1–10 years post-injury) and subacute (10–49 days post-injury) are ongoing (https://nervgen.com/cto/).

Fasudil

Fasudil (molecular weight: 291.37) is a Rho-kinase inhibitor. Nogo receptors bound by CNS/myelin-associated inhibitors such as Nogo-A, MAG, and OMgp have no intracellular domains (Figures 1 and 2). Three different membrane proteins have been shown to form a trimolecular complex with the Nogo receptor, namely, the p75 receptor, Troy, and LINGO-1 (Mueller et al., 2005). Both trimeric receptor complexes, namely, Nogo receptor/p75/LINGO-1 or Nogo receptor/Troy/LINGO-1, stimulate the RhoA/Rho-kinase pathway when myelin-associated inhibitors bind to the receptor complex. Additionally, CSPGs activate the RhoA/Rho-kinase pathway (Mueller et al., 2005). CSPGs are secreted from astrocytes and accumulate inside the glial scars after SCI, leading to the inhibition of axonal growth (Ohtake and Li, 2015). These findings led to the idea that the inhibition of RhoA/Rho-kinase is key to axonal growth.

Contusion injury was inflicted at T9 in Wistar rats (8–10-week-old, male; Table 1). Four weeks after SCI, Fasudil, which is a Rho-kinase inhibitor, was administered via a silicone tube inserted into the subarachnoid cavity through L3–4. The tube was connected to an osmotic minipump containing 15 μg/μL of Fasudil or saline. The infusion rate was 12 μL/d or 180 μg/d of Fasudil for 14 days. The total quantity of Fasudil administered was approximately 2.5 mg/rat. The BBB scores were evaluated up to 12 weeks after SCI (for 8 weeks from the start of Fasudil delivery). No significant difference was observed between the control and Fasudil groups in terms of BBB scores and CST extension (Nishio et al., 2006). Moreover, Fasudil infusion for 14 days immediately after SCI increased motor function and CST growth (Nishio et al., 2006). Any clinical studies of Fasudil for chronic SCI have not been reported.

Phosphatase and tensin homolog protein antagonist peptide 4

Enhancing signaling through the phosphatidylinositol 3-kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR) pathway is recognized as a mechanism capable of inducing axon growth and regeneration (Akram et al., 2022). Deletion of the PTEN, which antagonizes PI3K activity, induces limited regeneration in CST and rubrospinal-tract axons after SCI (Liu et al., 2010; Danilov and Steward, 2015). Since inhibiting PTEN activity is an identified actionable target capable of inducing axon growth, several challenges were performed in both acute and chronic conditions of SCI as follows (Stewart et al., 2023). The PTEN knockdown was effective in SCI mice in the case of conditional knockdown at the acute phase (knockout immediately after SCI). For chronic phase SCI, PTEN knockout was induced at 12 weeks later of the T9 crush injury by AAV-Cre retrograde delivery to descending tracts of loxP-PTEN mice. Hindlimbs function evaluated by BBB score was not improved by PTEN knockout at the chronic phase. Axonal densities of raphespinal serotonergic tracts and CST were not increased by PTEN knockout. In addition, PTEN knockout at the chronic phase decreased the number of motor cortex neurons, indicating the long-term toxic effect of PTEN-KO on neurons in the motor cortex. In a subgroup analysis of severely injured mice at knockout starting, BBB was improved although the functional gains with PTEN-KO began to return to pre-treatment levels 9 weeks post-treatment.

Since genetic manipulation is not feasible for treating patients, Li’s group designed and prepared PTEN antagonist peptides (Ohtake et al., 2014). PTEN protein consists of 403 amino acid residues and multiple activity domains, including the phosphatidylinositol-4,5-bisphosphate (PIP2) binding motif and the protein tyrosine phosphatase, C2, and C-terminal tail regions. The PIP2 binding motif at the N-terminus serves a dual role of localizing PTEN to the membrane and regulating its activity. The ATP-binding motifs associate PTEN subcellular localization and its tumor-suppressor activity. The C-terminal tail contains the PDZ domain-binding sequence and multiple phosphorylation sites and contributes to PTEN stabilization and targeting for phosphorylation by microtubule-associated serine/threonine kinases. To selectively block PTEN function with a pharmacological approach, five PTEN antagonist peptides (PAP1–5) were designed by targeting the PIP2, ATP-type A or B, PDZ, and C-terminal tail regions of PTEN (Ohtake et al., 2014). To facilitate peptide access into cells, the trans activator of transcription sequence (GRKKRRQRRRC) was induced at the C-terminus of all the PAPs. Among designed five peptides, PAP1-PAP4 enhanced axon lengths in dorsal root ganglion cells and cerebellar granular neurons. PAP4 most significantly inhibited PTEN signaling in the spinal cord and cerebral cortex in T7 hemisection mice when subcutaneously injected from 2 to 16 days after SCI (Figures 1 and 2). Hindlimbs movement evaluated by Basso Mouse Scale (BMS) and grid walk test was significantly improved by PTEN antagonist peptide 4 (PAP4) injection.

In chronic phase SCI experiment, female Sprague–Dawley rats (250–300 g) received C2 hemisection (Cheng et al., 2021; Table 1). At 8 weeks following the C2 hemisection, PAP4 was subcutaneously every 12 hours for 21 days at 1 mg/d per rat. At 20 weeks after SCI, rats underwent intra-medullary injection of the anterograde tracer, AAV2-mCherry. At 23 weeks after SCI, compound muscle action potential recordings were then measured. At 24 weeks after SCI, diaphragm electromyography was performed. PAP4 treatment very slightly improved diaphragm motor function. Bulbospinal rostral Ventral Respiratory Group axons extended to the caudal area beyond the lesion center by PAP4 treatment. PAP4 did not affect the morphological and functional innervation of phrenic motor neurons to the diaphragm. Any clinical studies of PAP4 for chronic SCI have not been reported.

Neuroprotection

(–)-Epigallocatechin gallate

(–)-Epigallocatechin gallate (EGCG) (molecular weight: 458.372) is a natural polyphenol and a main constituent of green tea extract (Figure 1). It exerts a neuroprotective effect through its antioxidant and anti-inflammatory properties (Lee et al., 2000, 2023a, b; Surh et al., 2001; Aktas et al., 2004; Choi et al., 2004; Mandel et al., 2005). EGCG may exert its neuroprotective effect through an underlying molecular mechanism linked to its anti-inflammatory and antioxidant activity, and it regulates autophagy and lipid metabolism (Payne et al., 2022).

A contusion injury was inflicted at the T9–T10 region in Sprague–Dawley rats (11–12 weeks of age, female; Table 1). EGCG treatment was initiated 12 months after SCI onset via continuous intravenous infusion for 36 hours using a syringe pump at the rate of 32 μL/h or 20 mg/kg dosage adapted. Motor functional evaluation showed that the BBB score was significantly improved by approximately 4 points at 2–6 weeks after EGCG infusion (Renno et al., 2014). In the case of chronic SCI, tactile allodynia, and thermal nociceptive hyperalgesia were improved by EGCG treatment. Any clinical studies of EGCG for chronic SCI have not been reported.

Stimulating axonal growth activity

Matrine

Matrine (molecular weight: 248.364) is a lupin alkaloid that occurs abundantly in the roots of Sophora flavescens Aiton (Li and Wang, 2004; Figure 1). Matrine promotes the axonal growth of cultured cortical neurons under inhibitory conditions (Tanabe et al., 2016). Matrine enhances the functional recovery and extension of raphespinal serotonin-positive tracts beyond the lesion site in mice with acute SCI (Tanabe et al., 2018). Matrine has shown efficacy in mice with chronic SCI (Tanabe et al., 2019).

Matrine is an activator of the chaperone function of extracellular heat shock protein 90 (HSP90), which was identified as the direct target molecule of matrine in inducing axonal growth and ameliorating SCI (Tanabe et al., 2018; Figure 3). Matrine stimulates axonal growth in cultures of mouse cortical neurons seeded on CSPG-coated dishes. Matrine-induced axonal growth is completely diminished by HSP90 neutralizing antibody (Tanabe et al., 2018). Additionally, neurite outgrowth (without CSPG coating) is reduced by FITC-geldanamycin, which is an impermeable HSP90 inhibitor that targets extracellular HSP90. These findings indicate that the endogenous activity of extracellular HSP90 contributes to neurite outgrowth and that extracellular HSP90 mediated matrine-induced axonal growth on CSPG. A variety of small molecule compounds that interact with HSP90 have been identified, but almost all of them are functional inhibitors of HSP90 (Verma et al., 2016). In contrast, matrine is the activator of the HSP90 chaperone function. HSP90 folding requires ATP binding and ATPase activity. Several HSP90 inhibitors such as geldanamycin inhibit HSP90 by preventing HSP90–ATP interactions (Pearl and Prodromou, 2006). As matrine increases HSP90 chaperone activity, it may increase HSP90 ATPase activity as well and thus promote the folding function. HSP90 through its chaperone activity stabilizes, activates, and/or renatures various client proteins, which impacts the signaling pathways associated with those client proteins (Taipale et al., 2010). Therefore, further investigation of matrine-induced HSP90 activation may provide a novel pharmacological approach other than promoting axonal growth to regulate signaling.

Figure 3.

Schematic of stimulating axonal growth signaling.

Black arrows indicate stimulating, and purple arrows indicate secretion from myocytes in case of PKM2, and upregulation of activities in other molecules. Yellow structures indicate plasma membranes. Created with Microsoft PowerPoint. 1,25D₃-MARRS: Membrane-associated rapid response steroid-binding receptor; Gal-1: galectin-1; HSP90: heat shock protein 90; MEK1: mitogen-activated protein kinase kinase 1; PI3K: phosphatidylinositol 3-kinase; PKA: protein kinase A; PKC: protein kinase C; PKM2: pyruvate kinase 2; SPARC: secreted protein acidic and rich in cysteine; VCP: valosin-containing protein.

Mice (ddY, 8 weeks old, female) were subjected to contusion injury at the T10–T11 level (Table 1). Oral treatment with matrine was initiated 28 days after the onset of injury. The dosage was 100 μmol/kg/d (24.84 mg/kg/d). Continuous matrine treatment for 154 days significantly improved hindlimb motor function, as evidenced by the improved BMS (Basso et al., 1995) and BSS (Teshigawara et al., 2013). Matrine treatment significantly enhanced axonal densities at the rostral and caudal parts 2 mm from the lesion center. Similarly, the presynaptic density terminating at motor neurons in the ventral horn of the lumbar spinal cord also significantly increased. Any clinical studies of matrine or HSP90 for chronic SCI have not been reported.

Acteoside

Acteoside (also known as Verbascoside, Kusaginin; molecular weight, 624.59) is a phenyl ethanolic glucoside and the main component of the Cistanchis Herb (Figure 1). Acteoside exerts anti-inflammatory (Jing et al., 2015; Nam et al., 2015), antioxidant (Lecci et al., 2014), and neuroprotective (Wang et al., 2012; Wu et al., 2014) effects. Kodani et al. (2019) studied its myokine stimulating activity when injected into skeletal muscles (Figure 3).

Mice (ddY, 10 weeks old, female) were subjected to contusion injury at the T11 level (Table 1). Thirty days after SCI onset, acteoside was injected intramuscularly into the right and left biceps femoris muscles (three times per week, 0.1 mg/limb). On day 15 after administering the first dose, the motor function of the hind limbs showed significantly improved BMS and Toyama Mouse Score (Kodani et al., 2019). Histological evaluation performed after 62 days of drug administration showed that the density of serotonin-positive raphespinal tracts and presynaptic density of motor neurons significantly increased at the caudal part 2 mm away from the lesion center. Additionally, skeletal muscle atrophy of hind limbs in the chronic-phase SCI was significantly ameliorated.

The comprehensive identification of acteoside treatment-induced myokines from cultured mouse hindlimb myocytes revealed PKM2 as a new myokin that is transferred to the CNS via blood circulation (Kodani et al., 2019). Nevertheless, the molecular mechanism underlying acteoside stimulation that leads to the synthesis and secretion of PKM2 in the muscles remains unknown.

A clinical study of Cistanchis Herb extract for the chronic phase of cervical spondylotic myelopathy is ongoing in Japan (Trial ID: jRCTs041200080). Any other clinical studies of acteoside or Cistanchis Herb for chronic SCI have not been reported.

Pyrvate kinase M2

Pyrvate kinase M2 (PKM2) is a multifarious protein (Figure 1). In its canonical role as a glycolytic enzyme, PKM2, in its tetramer form, catalyzes the last step of glycolysis, which is the dephosphorylation of phosphoenolpyruvate to pyruvate. In its dimer or monomer form, PKM2 localizes in a variety of sites that are nuclear, mitochondrial, and extracellular. In mice with chronic SCI, PKM2 played a new role as a myokine, and secreted PKM2 from the skeletal muscle reached the central nervous system via blood circulation (Kodani et al., 2019; Figure 3). Therefore, Kikuchi et al. (2020) investigated the effects of i.c.v. infusion of PKM2 on motor function in mice with chronic SCI.

Mice (ddY, 8 weeks old, female) were subjected to contusion injury at the T13 level (Table 1). Thirty-two days later, recombinant mouse PKM2 was i.c.v. infused for 28 days via an osmotic minipump (Kikuchi et al., 2020) at the rate of 0.11 μL/h. Artificial cerebrospinal fluid was used as the vehicle. The final concentration of infused PKM2 in mouse cerebrospinal fluid was adjusted to 1 ng/mL. On day 8 after initiating PKM2 treatment, the BMS, Toyama Mouse Score (Shigyo et al., 2014), and vertical cage scale evaluations showed that motor function of the hind limbs improved significantly compared with that observed after vehicle infusion. Additionally, the density of the serotonin-positive raphespinal tracts increased significantly in the caudal part 2 mm away from the lesion center. A significant increase in the termination of the raphespinal tract on motor neurons was confirmed using retrograde transsynaptic tracing by injecting wheat germ agglutinin into the sciatic nerve (Kikuchi et al., 2020).

Drug affinity-responsive target stability analysis identified valosin-containing protein (VCP) as a direct target of PKM2 (Figure 3). Extracellular PKM2 binds to the VCP located on the plasma membrane (Kikuchi et al., 2020). VCP has many functions including ER-associated protein degeneration (Kakizuka, 2015), cell division, organelle biogenesis, nuclear envelope formation, and protein degradation via the ubiquitin–proteasome system (Ju and Weihl, 2010). As these cellular activities of VCP are associated with ATPase, ATPase activity was evaluated in the presence of recombinant VCP. In vitro, ATPase activity increased in a concentration-dependent manner as VCP concentration increased. CB-5083 (100 and 1000 nM), which is a potent and selective VCP ATPase inhibitor, significantly inhibited VCP-induced ATPase activity. Treatment with mouse recombinant PKM2 (1 ng/mL) increased the axonal length in cultured mouse cortical neurons, and the axonal extension was completely diminished by co-treatment with CB-5083 (100 nM). In vivo, CB-5083 i.c.v. co-infusion with recombinant PKM2 completely inhibited PKM2-elicited recoveries of motor function and raphespinal tracts in mice with chronic SCI (Kikuchi et al., 2020). Any clinical studies of PKM2 or VCP for chronic SCI have not been reported.

Diosgenin

Diosgenin (molecular weight: 414.63) is a steroidal sapogenin and a component of yam (Figure 1). Diosgenin exerts potent axonal growth activity in cerebral cortical and hippocampal neurons both in vitro and in vivo (Tohda et al., 2012; Yang and Tohda, 2023a). In a mouse model of Alzheimer’s disease, diosgenin treatment induced direct axonal regrowth (Yang and Tohda, 2023a).

Mice (ddY, 8 weeks old, female) were subjected to contusion injury at the L1 level (Table 1). Thirty-one days after the onset of SCI, diosgenin treatment was initiated using an intrathecally inserted tube connected to an osmotic minipump. Diosgenin dose was maintained at 0.1 μM in the cerebrospinal fluid during the 56-day administration period because in vitro data indicated that this was the optimal diosgenin dose for axonal growth. The motor function of the hind limbs was significantly improved in BMS, Toyama Mouse Score, and vertical cage scales (Nakano et al., 2021). Moreover, diosgenin intrathecal infusion significantly increased the density of serotonin-positive raphespinal tracts at the lesion center. In vitro, diosgenin treatment almost completely increased the axonal lengths of medullary and spinal cord neurons seeded on CSPG coating. Receptors, second messengers, and downstream-regulated proteins involved in diosgenin signaling were identified (Figure 3). Diosgenin binds directly to 1,25D3-MARRS and recruits protein kinase C, protein kinase A, mitogen-activated protein kinase kinase 1, and PI3K (Tohda et al., 2012). The 1,25D3-MARRS-mediated signal upregulates SPARC (Yang and Tohda, 2023a) and Gal-1 (Yang and Tohda, 2023b) and downregulates HSC70 (Yang and Tohda, 2018), leading to axonal growth. Any clinical studies of diosgenin for chronic SCI have not been reported.

Pharmacotherapy of Chronic Spinal Cord Injury: Clinical Study

Blocking axonal growth inhibition signaling

AA-NgR(310)ecto-Fc

NgR(310)ecto-Fc is also known as AXER-204, and it blocks endogenous NgR1 binding and prevents subsequent growth-inhibitory signaling. NgR(310)ecto-Fc contains amino acids 27–310 of human NgR1 and was expressed stably in Chinese hamster ovary cells (Wang et al., 2014, 2015, 2020; Figure 1). Cys266 and Cys309 were mutated to alanine residues to prevent disulfide scrambling between the NgR1 and Fc regions and improve stability (Weinreb et al., 2010; Wang et al., 2011, 2014). AXER-204 was produced under Good Manufacturing Practice conditions.

A study on AXER-204 included patients (aged 18–65 years) with chronic (over 12 months after the injury) traumatic cervical SCI from six rehabilitation centers in the USA (Maynard et al., 2023). During part 1 of the open-label study, AXER-204 was delivered as single intrathecal doses of 3, 30, 90, or 200 mg to each set of four patients with AIS grades A, B, C, or D, respectively (Table 2). The intrathecal injection was administered on day 1, and cerebrospinal fluid was collected on days 8 and 29. Neurological examination and questionnaires were administered on day 29. No severe adverse effects were observed in patients treated with AXER-204. The study showed that among the 17 patients who received single doses of 30, 90, or 200 mg of AXER-204, five (29%) patients showed an increase in the baseline International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) upper extremity motor score (UEMS) by ≥ 3 points on day 29. None of the six patients who received 3 mg of AXER-204, which was designed to be sub-efficacious, showed an increase in the score.

Table 2.

Pharmacotherapies of chronic SCI: clinical study

Category	Treatment	Study style	SCI severity	Time after injury of subjects	Number of subjects	Dose	Treatment period	Outcome	Adverse effects	Reference
Blocking axonal growth inhibition signaling	NgR(310)ecto-Fc	Open-label	ASIA impairment scale: A, B, C, D	Over 12 mon (mean 51.5 mon)	24	3, 30, 90, 200 mg/person, intrathical infusion	29 d	ISNCSCI UEMS: Increased at 3 mg < 30–200 mg		Maynard et al., 2023
		Double-blind randomized, placebo-controlled		Over 12 mon (mean 107.8 mon)	27	200 mg/person, intrathical infusion	169 d	ISNCSCI UEMS: No effect GRASSP prehensionperformance: No effect SCIM III self-care: No effect	Similar to placebo	Maynard et al., 2023
		Subgroup analysis	ASIA impairment scale: B, C, D + No previous treated		11			ISNCSCI UEMS: Improved		Maynard et al., 2023
Neuroprotection	Granulocyte-colony stimulating factor (G-CSF, filgrastim)	Double-blind randomized, placebo-controlled (Ph III)	ASIA impairment scale: B, C, D	Over 6 mon	114	300 μg/d, s.c.	7 d	ASIA motor score: Improved Pinpick sensory score: Improved IANR-SCIFRS: Improved SCIM-III: Improved	Treatable neuropathic pain (12.5%) and spasticity (1.8%)	Derakhshanrad et al., 2018
Blocking voltage-dependent potassium channel	Fampridine-SR	Single-blind randomized, placebo-controlled (Ph III)	ASIA impairment scale: B, C, D	Over 18 mon	212	25 mg × 2/d, p.o.	18 wk	Ashworth score: No effect SGI: No effect	High treatment-related TEAEs	Cardenas et al., 2014
					203	25 mg × 2/d, p.o.	18 wk	Ashworth score: No effect SGI: No effect	High treatment-related TEAEs	Cardenas et al., 2014

ASIA: American Spinal Injury Association; CSF: colony-stimulating factor; GRASSP: Graded and Redefined Assessment of Strength, Sensibility and Prehension; IANR-SCIFRS: International Association of Neurorestoratology-Spinal Cord Injury Functional Rating Scale; ISNCSCI: International Standards for Neurological Classification of Spinal Cord Injury; p.o.: per os (oral); s.c.: subcutaneous; SCI: spinal cord injury; SCIM: Spinal Cord Independence Measure; SGI: Subject Global Impression of treatment; TEAE: treatment-emergent adverse events; UEMS: upper extremity motor score.

During part 2, which was performed as a randomized double-blind placebo-controlled study, 14 patients with AIS grades A, B, C, or D were administered AXER-204 as a single 200 mg intrathecal dose, and 13 patients were administered a placebo (Table 2). The single intrathecal injection was administered on day 1. Cerebrospinal fluid collection and functional evaluation were performed until day 169. No statistically significant differences were observed between the AXER-204 and placebo groups based on the ISNCSCI UEMS or other efficacy outcome measures across all patients. However, a post-hoc subgroup analysis was performed in patients with incomplete injury (AIS grades B, C, or D) and who were not previously treated with AXER-204 in part 1. The ISNCSCI UEMS improved by 4 points and the ISNCSCI total motor score improved by 9 points in the AXER-204 group, whereas these measures remained unchanged in the placebo group. The incidence of treatment-related adverse events was similar between the placebo and AXER-204 groups. Pharmacokinetic measurements revealed high AXER-204 concentration in the cerebrospinal fluid with minimal systemic exposure and no accumulation during repeat dosing. The concentration in cerebrospinal fluid in the 200 mg dosing group exceeded that observed in animal efficacy studies, which reported successful stimulation of axonal growth and recovery of neurological functions (Wang et al., 2014, 2020). I think that larger-scale studies should be conducted for patients with incomplete injury as a next step.

Neuroprotection

Granulocyte-colony stimulating factor

Granulocyte-colony stimulating factor (G-CSF) is a 19.6-kDa glycoprotein (Figure 1). It is a cytokine that is used in pharmaceutical preparations for the treatment of chemotherapy-induced neutropenia. G-CSF has been approved by the Food and Drug Administration to treat neutropenia and mobilize bone marrow-derived hematopoietic stem cells for transplantation (Mehta et al., 2015). Several types of formulations of recombinant human G-CSF have been already used in medicine, such as Filgrastim, Lenograstim, Nartograstim, and Pegfilgrastim.

The mechanism underlying G-CSF-mediated functional improvement in SCI involves neuroprotection. In animal models of acute SCI, G-CSF mobilizes bone marrow-derived cells to the injured spinal cord, where it directly suppresses neuronal apoptosis and the death of oligodendrocytes, protects myelin, and suppresses the expression of inflammatory cytokines (Koda et al., 2007; Nishio et al., 2007; Aschauer-Wallner et al., 2021). G-CSF also exerts neuroprotective effects via angiogenesis during the subacute phase of SCI (Kawabe et al., 2011).

In a 6-month double-blind parallel randomized placebo-controlled phase III study, 120 patients with incomplete chronic traumatic SCI (over 6 months after the injury) evaluated as B, C, or D based on the AIS were allocated to the placebo or 300 μg/d of G-CSF (filgrastim) group (Table 2). Treatment was done by subcutaneous administrations for 7 consecutive days (Derakhshanrad et al., 2018). The functions were evaluated before intervention and at 1, 3, and 6 months. At the endpoint, 56 patients (93.3%) in the G-CSF group and 58 (96.6%) in the placebo group completed the study. After 6 months of intervention, the mean improvement in the American Spinal Injury Association motor score in the G-CSF group was significantly higher than that in the placebo group. The sensory scores for light touch and pinprick were significantly increased following G-CSF treatment as were scores evaluated based on The International Association of Neurorestoratology-Spinal Cord Injury Functional Rating Scale and Spinal Cord Independence Measure-III.

Blocking voltage-dependent potassium channel

Fampridine-sustained release

Fampridine-SR is a “sustained release” (SR) tablet formulation of 4-aminopyridine (Figure 1). Animal studies have indicated that 4-aminopyridine restores action potential conduction in damaged nerve fibers by blocking voltage-dependent potassium channels (Tang and Kocsis, 1985; Shi et al., 1997). In January 2010, dalfampridine (another name for fampridine) was approved by the United States Food and Drug Administration to improve walking speed and distance in MS patients. Fampridine in Europe was conditionally approved for marketing by the European Medicines Agency in July 2011 and fully approved for marketing in 2017, and is now available in Germany, the United Kingdom, France, and other countries (Guo et al., 2016).

For chronic SCI, following two multicenter randomized double-blind placebo-controlled phase III studies were performed for 6 months. A total of 212 and 203 patients with incomplete chronic traumatic SCI (over 18 months after the injury) between C3 and T10 and rated B, C, or D according to the AIS were included (Cardenas et al., 2014; Table 2). Placebo or fampridine-SR (twice a day, dose of 25 mg) was administered orally. Both studies consisted of a 2-week single-blind placebo run-in, a 2-week titration, 12 weeks of stable dosing, a 2-week downward titration, and 2 weeks of untreated follow-up. The primary outcomes were the Ashworth score, averaged over the bilateral knee flexors and extensors, and a 7-point Subject Global Impression of treatment (1 = terrible; 7 = delighted). The Ashworth scale in both studies was not significantly different between placebo and fampridine-SR groups, although numerically, the scale in the fampridine-SR group was better than that in the placebo group. The Subject Global Impression of treatment scores in both studies were not significantly different between the placebo and fampridine-SR groups. No significant differences were observed in total motor scores between treatments in either study. Subscale analyses indicated a significantly better upper extremity subscale with fampridine-SR in one study and better erectile and orgasmic functions with fampridine-SR in the other. Additionally, the fampridine-SR group showed a higher frequency of treatment-related treatment-emergent adverse events, such as urinary tract infection, hypertonia, dizziness, pain, and constipation, than the placebo group. The effectiveness of 4-AP in improving function depends primarily on the condition of the spinal tract. If the axons are already damaged and demyelinated, 4-AP would not be effective. Therefore, the effect of fampridine-SR treatment on chronic-phase SCI was limited.

Others

Riluzole

Low concentrations of riluzole specifically and potently block the persistent Na⁺ current and inhibit chronic spasticity in SCI rats (Brocard, 2016). Currently, riluzole for treating spasticity in patients with chronic traumatic spinal cord injury is ongoing as a phase 1b/2b adaptive multicenter randomized controlled RILUSCI trial (NCT02859792) (Cotinat et al., 2023).

Potential of combinatory treatments

Chronic phase of SCI is a very severely progressed and untreatable stage. Several combinatory treatments showed additive or synergistic efficacy in the chronic phase. Hepatocyte growth factor-releasing scaffold enhanced motor function in T10 complete transection rats when combined with transplantation human induced pluripotent stem cell-derived neural stem/progenitor cells (Hashimoto et al., 2023). Transplantation of human neural progenitor cells biased toward an oligodendrogenic fate induced functional recovery in T7 compression rats when combined with sustained delivery of ChABC (Nori et al., 2018). Rehabilitation is also quite supportive and facilitates motor recovery for monotherapies. Treatment with ChABC (Shinozuka 2016), cell transplantation (Shibata et al., 2016), or sema3A inhibitor (Zhang et al., 2014) in rodent chronic SCI showed better efficacy when combined with rehabilitation. The molecular mechanism of rehabilitation underlying the beneficial effect has not yet been fully elucidated, but muscle activity induces the release of several anti-inflammatory myokines and neurotrophic myokines (Vints et al., 2023).

Potential pharmacological interventions introduced in this review (Figure 4) have not been combined with each other. Combinations of drugs in different categories may result in additive or synergistic effects. Although blocking axonal growth inhibition signaling has been much attracted focus in the SCI field, it is a comparatively passive strategy. Not only blocking inhibitory environments, but also actively stimulating axonal growth is quite important to gain satisfactory effects. Axonal growth by combinatory treatments probably becomes the key focus to overcome the limitation of repairing neural relays.

Figure 4.

Dosing routes of drugs introduced in this review.

Created with Microsoft PowerPoint. ChABC: Chondroitinase ABC; EGCG: (–)-epigallocatechin gallate; Fampridine-SR: Fampridine sustained release tablet; G-CSF: granulocyte-colony stimulating factor; ISP: Intracellular Sigma Peptide; NgR(310)ecto-Fc: soluble decoy fragment of NgR1; PAP4: PTEN antagonist peptide 4.

Conclusion

Studies on drug therapies that show functional recovery in patients with chronic spinal cord injury remain limited. Currently, only a few drugs have entered clinical studies after intense basic research. More candidate compounds and proteins showing efficacies in basic research are expected to reach the clinical studies stage. A number of conditions should be considered carefully for appropriate design, such as doses, dosing routes, treatment durations, patients’ conditions, monitoring pharmacokinetics, and so on.

Cocktail therapy that combines multiple drugs with different underlying mechanisms may be the most effective approach to treat the refractory status of chronic SCI. For that, getting evidence and narrowing appropriate conditions in basic studies are necessary.

Data availability statement:

Not applicable.