P2X7 receptor antagonism with Brilliant Blue G promotes functional recovery in spontaneous canine spinal cord injury: a preliminary study in a large animal model

Abstract

Background:

Secondary injury cascades critically influence neurological outcomes after severe spinal cord injury (SCI), yet effective disease-modifying pharmacological interventions remain lacking. Extracellular ATP–mediated activation of the P2X7 receptor is a key driver of post-traumatic neuroinflammation, glial activation, and progressive tissue loss. While P2X7 receptor antagonism improves functional recovery in experimental rodent models of SCI, its translational relevance in naturally occurring large-animal injury has not been established.

Methods:

We conducted a prospective, uncontrolled, exploratory observational study in client-owned dogs serving as a naturally occurring large-animal model of severe SCI. Twenty non-ambulatory dogs with cervical or thoracolumbar SCI whose owners declined surgical decompression received intravenous Brilliant Blue G (BBG), a selective P2X7 receptor antagonist, at 10 mg/kg/day for seven consecutive days. Neurological function was assessed longitudinally using deep pain perception testing and the modified Frankel scale. Magnetic resonance imaging (MRI) was performed in a subset of cases.

Results:

Consistent neurological improvement temporally associated with BBG administration was observed across the cohort during follow-up. Eighteen dogs (90%) regained independent ambulation within 5 days to 15 months (median, 4 weeks). Among dogs presenting with complete injury equivalents (modified Frankel grade 5), 13 of 14 (92.9%; 95% CI, 66.1–99.8%) achieved independent ambulation. In cases with follow-up MRI, substantial functional recovery occurred despite persistent spinal cord compression. No adverse events attributable to BBG were observed.

Conclusion:

Systemic P2X7 receptor antagonism with BBG was feasible and well tolerated and was associated with substantial functional recovery in a naturally occurring large-animal model of severe SCI, even in the absence of surgical decompression. Although causality cannot be established in this uncontrolled study, the consistency of recovery patterns and the observation of improvement despite persistent mechanical compression support further investigation of P2X7 receptor inhibition as a disease-modifying therapeutic strategy for acute human SCI.

Introduction

Traumatic spinal cord injury (SCI) remains a devastating condition with limited therapeutic options. While early surgical decompression is considered the standard of care, neurological outcomes are often poor, particularly in patients presenting with complete sensorimotor paralysis (1, 2). Crucially, surgical intervention primarily addresses mechanical compression and does not directly target the secondary injury cascade that progresses over hours to days following the initial insult. To date, no pharmacological therapy has been definitively shown to modify disease progression or improve long-term functional recovery in either veterinary medicine or human medicine.

Secondary injury mechanisms play a critical role in determining neurological outcome following SCI. Cellular damage results in the release of extracellular ATP, which activates purinergic receptors expressed on neurons, oligodendrocytes, astrocytes, and microglia. Among these, the P2X7 receptor has been identified as a central mediator of neuroinflammation. Sustained activation of P2X7 receptors induces prolonged cation influx, mitochondrial dysfunction, inflammasome activation, microglial proliferation, and subsequent neuronal and oligodendroglial cell death (3, 4). In experimental rodent models of SCI, pharmacological inhibition or genetic deletion of the P2X7 receptor attenuates neuroinflammation, preserves white matter integrity, and improves locomotor recovery (3).

Translation of these findings to human SCI has been hindered by limitations inherent to experimental models. Most preclinical studies rely on artificial contusion or transection injuries in rodents, which fail to capture the clinical heterogeneity, compressive components, and temporal evolution of secondary injury observed in patients. Naturally occurring intervertebral disc herniation in dogs represents a clinically relevant large-animal model of acute compressive SCI, recapitulating key pathological and functional features of human disease under real-world clinical conditions (5). Importantly, this model allows evaluation of therapeutic strategies in the context of spontaneous injury, biological variability, and clinically meaningful functional outcomes.

Brilliant Blue G (BBG) is a selective P2X7 receptor antagonist that has demonstrated neuroprotective effects in experimental models of spinal cord injury (SCI) (3) and traumatic brain injury (TBI). Functional P2X7 receptors are conserved in canine neural tissues, and studies suggest that their function in dogs more closely resembles that of humans than rats (6). This characteristic positions the canine species as an appropriate translational model for evaluating P2X7-targeted interventions. However, the clinical feasibility and neurological outcomes associated with systemic P2X7 receptor inhibition in naturally occurring SCI have not yet been systematically investigated.

The objective of this study was to explore the feasibility, safety, and neurological outcomes associated with intravenous administration of BBG in naturally occurring canine spinal cord injury (a large-animal model of severe SCI) in which surgical decompression was not performed. Given the exploratory and uncontrolled nature of this investigation, the study was not designed to establish definitive therapeutic efficacy, but rather to generate translational hypotheses relevant to human spinal cord injury.

Materials and methods

Study design and subjects

This study is a prospective, uncontrolled, exploratory observational study conducted in a clinical veterinary setting between 2013 and 2025. Following informed consent, 20 client-owned dogs presenting with acute cervical or thoracolumbar spinal cord injury were enrolled. The inclusion criteria were: (1) a diagnosis of compressive spinal cord injury based on neurological examination and diagnostic imaging (radiography and/or MRI), (2) non-ambulatory status at the time of presentation, and (3) the owner’s refusal of surgical decompression. Cases with mild neurological deficits (Modified Frankel Grade 2 or lower) were excluded.

Neurological assessment

Neurological function was assessed using the modified Frankel scale (grades 0–5), which captures clinically meaningful functional states in naturally occurring, non-experimental SCI and facilitates comparison with historical human and veterinary cohorts. Deep pain perception (DPP) was evaluated by applying strong pressure to the digits and/or tail using forceps and observing for a conscious behavioral response. Purposeful reactions such as vocalization, head turning, or deliberate withdrawal were considered positive responses, whereas spinal reflexes alone were not. All neurological assessments throughout the study were performed by the same examiner (Y.S.), thereby minimizing inter-examiner variability. Although precise quantification of applied pressure was not performed, this approach is consistent with established standards in veterinary neurology and is the method used in previously published veterinary SCI studies cited in this manuscript.

Neurological grading

Neurological function was assessed using the modified Frankel scale (0–5).

• 0 Normal

• 1 Spinal pain only

• 2 Ambulatory paraparesis

• 3 Nonambulatory paraparesis

• 4 Paraplegia with deep pain perception

• 5 Paraplegia without deep pain perception

In this study, we employed the Modified Frankel Scale (MFS), a standard neurological assessment in veterinary medicine derived from the original Frankel Scale—the precursor to the globally recognized ASIA Impairment Scale (AIS). The present ‘Grade 5’ (loss of deep pain perception) represents an extremely severe neurological status that is broadly analogous to AIS Grade A in human clinical practice.

Treatment protocol

Brilliant Blue G (BBG) powder was dissolved in physiological saline and administered intravenously at a dose of 10 mg/kg/day for seven consecutive days using an infusion pump. Each administration lasted approximately one hour, with the exception of Case 1 (the first administration case), in which the dose was delivered via a 24-h continuous infusion. Case 1 was the first case treated in this series (2013), and at that time the optimal infusion rate had not yet been established; the 24-h continuous infusion was selected as a conservative approach to minimize potential adverse effects of rapid administration. The absence of adverse events in Case 1 informed the subsequent adoption of the 1-h infusion protocol for all remaining cases. No corticosteroids, nonsteroidal anti-inflammatory drugs, or other neuroprotective agents were administered during the treatment period at our institution. BBG was the first intervention administered at our clinic in all cases. However, three dogs (Cases 7, 13, and 19) were referred from other veterinary facilities where corticosteroids had been administered prior to presentation. Despite prior steroid treatment, all three dogs showed continued or progressive neurological deterioration and presented at our institution with Modified Frankel Grade 5 (Cases 13 and 19) or severe Grade 4 with only minimal unilateral deep pain perception (Case 7). The potential influence of prior corticosteroid administration on BBG efficacy cannot be fully excluded and is acknowledged as a limitation of this study.

Imaging

Magnetic resonance imaging (MRI) was performed before treatment in four dogs for which owner consent was obtained. In one of these dogs, a follow-up MRI examination was conducted 1 month after treatment. MRI was performed at referral institutions, including the Veterinary Medical Center of the University of Tokyo, under standard general anesthesia protocols, and imaging sequences included T1-weighted and T2-weighted images. Lesion assessment was primarily based on T2-weighted images This study was conducted in a clinical setting, and due to financial constraints and limitations in owner consent, magnetic resonance imaging (MRI) could not be performed in all cases. Consequently, a definitive etiological diagnosis of spinal cord injury could not be established in every dog included in the study.

However, plain radiography was performed in all cases and narrowing or morphological abnormalities of the intervertebral disc spaces consistent with the neurological localization were identified. Based on these findings, the suspected lesion sites were clinically estimated in all dogs.

Neurological evaluation and follow-up

In principle, neurological assessments were performed at presentation and at 1, 2, 4, 8, and 12 weeks (with extended follow-up up to 15 months in selected cases) after treatment. All examinations were performed by the same veterinarian. Ambulatory status, deep pain perception, postural reactions, and spinal reflexes were documented at each visit.

Gait recovery was additionally evaluated using serial video recordings taken by the clinician or owners (Supplementary Videos 1–5).

Statistical analysis

Analyses were descriptive. Exact two-sided 95% confidence intervals for binomial proportions were calculated using the Clopper–Pearson method. No hypothesis testing was performed.

Results

Clinical outcomes

The study population comprised 12 Miniature Dachshunds, 4 Toy Poodles, and 4 dogs of other breeds (Maltese, Pekingese, English Cocker Spaniel, and mixed breed). Ages ranged from 3 to 18 years. At initial presentation, all 20 dogs were classified as modified Frankel grade 3, 4 or 5 (Grade 3 was only 2 cases, making 18 cases for Grade 4/5), of which 14 lacked deep pain perception (DPP). The interval from injury onset to initiation of BBG treatment ranged from 1 to 20 days, with a median of 4 days, a mean of 5.7 days, and an interquartile range of 1–7 days (Table 1). None of the dogs included in this study exhibited clinically evident pain-related behaviors at the time of admission. However, in dogs with modified Frankel grade 5, the assessment of the presence or changes in pain is inherently problematic, as conscious perception of pain caudal to the lesion is by definition absent. Furthermore, the determination of pain in dogs remains fundamentally subjective. Only one case was reported by the owner to exhibit signs of pain 9 months post-treatment; this dog was prescribed a one-week course of NSAIDs, after which no further pain was observed and no additional treatment was required.

Background characteristics and outcomes for all cases are presented at the individual case level in Table 1. These include breed, sex, body weight, age, lesion location, modified Frankel grade at presentation (3–5), interval from onset to initiation of treatment, duration of follow-up (5 days to 15 months), and neurological grade and ambulatory status at the final evaluation.

Neurological improvement was observed in all dogs during the follow-up period. Furthermore, all patients—including one dog that died of an unrelated neoplastic disease during the follow-up—recovered spontaneous urination and defecation within the treatment period. Cases 1, 7, and 12 initially presented with clinically significant urinary dysfunction; however, Case 1 regained independent urination at 3 weeks, Case 7 at 2 weeks, and Case 12 at 3 weeks post-treatment.

Independent ambulation was regained in 18 dogs (90%), with the time to recovery ranging from 1 week to 15 months (median: 4 weeks). Among the 14 dogs presenting with modified Frankel grade 5 injury, 13 (92.9%; 95% confidence interval, 66.1–99.8%) regained independent ambulation. One dog showed partial neurological improvement but did not achieve ambulatory recovery.

Supplementary Videos 1–5 illustrate the clinical course of selected cases from the initial stage of the disease through neurological improvement.

Safety

Bluish discoloration of the feces was observed during BBG administration; however, no other adverse reactions, behavioral abnormalities, or clinically apparent organ toxicity were noted during or after treatment. All dogs tolerated the 7-day intravenous administration protocol well.

Supplementary information

Detailed case descriptions and sequential gait videos for Cases 1, 2, and 3 are provided in Supplementary Videos 1–3. These supplemental materials illustrate typical patterns of motor recovery observed across the cohort (Supplementary Videos 1–3).

Supplementary data: cases with MRI evaluation

Case 8—miniature dachshund, 3 years old

This dog presented with complete paraplegia and loss of deep pain perception at admission (modified Frankel grade 5). Continuous intravenous infusion of Brilliant Blue G (BBG) was initiated immediately following the diagnosis.

Neurological recovery after treatment was rapid; by two weeks, the dog had fully regained normal ambulation without pain (grade 0). Subsequently, the dog was able to run freely and exhibited no residual neurological deficits.

Given that improvement had been observed in all previous cases, we obtained the owner’s consent to perform a follow-up MRI after the recovery of ambulation, which was successfully conducted one month post-treatment. Despite the complete clinical recovery, the MRI findings demonstrated that the disc protrusion and spinal cord compression remained persistent and unchanged (Figures 1A,B; Supplementary Video 4-1 [pre] and Supplementary Video 4-2 [post]).

Figure 1

MRI of Case 19 (cervical IVDH) (Figure 2) and clinical course [Supplementary Video 5-1 (pre), Supplementary Video 5-2 (2 weeks after), Supplementary Video 5-3 (3 months after)].

Figure 2

Pre-treatment MRI revealed multiple disc herniations at C2–3, C3–4, and C4–5, associated with tetraparesis. Following BBG therapy, the dog regained independent ambulation within one month. Supplementary Video 5-1 demonstrates the pre-treatment status, while Supplementary Videos 5-2, 5-3 show marked clinical improvement at two weeks and three months post-administration, respectively.

As this case involved cervical spinal cord lesions, it was initially evaluated using the Texas Spinal Cord Injury Scale (TSCIS). The assessment yielded a gait score of 0, a proprioception score of 0, and a nociception score of 0 (noting that deep pain perception in the tail was not recorded), resulting in a total TSCIS score of 0/6. This score is consistent with the most severe category of spinal cord injury. However, to maintain methodological consistency with the other cases in this study, the modified Frankel scale was applied, and the case was classified as modified Frankel grade 5.

Discussion

This exploratory observational study evaluated the intravenous administration of Brilliant Blue G (BBG), a selective P2X7 receptor antagonist, in dogs with severe spinal cord injury (SCI) secondary to thoracolumbar or cervical intervertebral disc herniation (IVDH) whose owners declined surgical decompression. Although the uncontrolled study design precludes definitive conclusions regarding causality, meaningful neurological improvement was observed in all dogs during follow-up, particularly in those classified as modified Frankel grade (MFG) 5.

Dogs lacking deep pain perception (MFG 5) are generally considered to have a poor prognosis. While spontaneous recovery with conservative management alone has been reported in fewer than 10% of such cases, recovery rates of approximately 41–87% have been reported following surgical decompression (1, 2, 7, 8). In the present study, all 14 dogs (100%) presenting with MFG 5 showed neurological improvement, and 13 of these (92.9%) regained ambulatory function. Although the 95% confidence interval (66.1–99.8%) overlaps with previously reported success rates for surgical decompression, the achievement of such a high recovery rate without surgical intervention is noteworthy.

However, a recent study (9) reported neurological improvement in 96% of dogs with intact deep pain perception and 48% of those lacking it following conservative treatment alone. Therefore, the possibility that the recovery observed in this study reflects spontaneous improvement cannot be excluded. Nevertheless, the relatively high rate of functional recovery achieved without decompressive surgery supports the need for further evaluation of P2X7 receptor inhibition as a therapeutic target.

In Case 8 (see Supplementary material), persistent spinal cord compression remained evident on MRI even after complete clinical recovery. Despite the continued presence of a significant compressive lesion, the dog regained the ability to run freely, as demonstrated in Supplementary Video 4-2, and exhibited no signs of pain. Recent studies have reported that the volume of residual extruded disc material and the degree of spinal cord compression on postoperative MRI do not necessarily predict neurological outcomes (10), and favorable outcomes have also been documented in non-compressive injury models (11). Building upon these insights, our Case 8 highlights that meaningful functional recovery can occur even in the presence of persistent and significant mechanical compression. This observation underscores the potent neuroprotective potential of P2X7 receptor antagonism.

Furthermore, experimental research has demonstrated that P2X7 receptor inhibition suppresses microglia-dependent central neuroinflammatory signaling, which secondarily alleviates mechanical allodynia in neuropathic pain models (12). The absence of pain-related behaviors in Case 8, despite persistent disc extrusion, potentially aligns with the suppression of microglial activation rather than suggesting direct analgesic properties of BBG. While this single-case observation is hypothesis-generating, it suggests that factors other than the relief of mechanical compression may significantly contribute to functional restoration. Similarly, in Cases 7 and 19, neurological improvement was observed despite treatment initiation delayed by up to 20 days from injury onset, raising important questions regarding the therapeutic window of BBG in clinical settings.

If BBG contributed to the observed improvement, a plausible primary mechanism is the suppression of secondary injury pathways mediated by the P2X7 receptor. P2X7 receptor activation promotes the release of proinflammatory cytokines, glial activation, and neuronal death, contributing to long-term impairment. In experimental models, P2X7 inhibition has been shown to attenuate these processes and improve motor function (3). Furthermore, the recent identification of Muse cells (Multilineage-differentiating Stress Enduring cells) has shown that these endogenous pluripotent-like stem cells can migrate to injured spinal tissue and differentiate into neural lineages (13, 14). Although speculative, it is conceivable that BBG-mediated suppression of glial scar formation facilitated Muse cell migration, suggesting a potential synergistic mechanism between neuroprotection and endogenous repair.

Several limitations should be acknowledged, including the uncontrolled, non-blinded design, small sample size, and lack of MRI evaluation in all cases. Potential selection bias and heterogeneity in disease severity also exist. Furthermore, while no dogs exhibited clinically apparent progressive myelomalacia, it was not systematically excluded. Three dogs (Cases 7, 13, and 19) had received corticosteroids at referring institutions prior to BBG administration, and the potential influence of this prior treatment on outcomes cannot be fully excluded. Additionally, precise quantification of the pressure applied during deep pain perception testing was not performed; however, all assessments were conducted by the same examiner throughout the study period, minimizing inter-examiner variability. Accordingly, these results should be interpreted as exploratory, and controlled clinical trials incorporating standardized imaging and objective gait analysis are warranted. Nevertheless, given that dogs share greater physiological similarities with humans than rodents—particularly regarding P2X7 receptor function—the recovery observed in these naturally occurring clinical cases may provide a critical foundation for translating BBG therapy into human clinical practice.

Conclusion

In this preliminary observational study, systemic administration of the P2X7 receptor antagonist Brilliant Blue G (BBG) was feasible, well tolerated, and associated with substantial neurological recovery in dogs with naturally occurring severe spinal cord injury, even in the absence of surgical decompression. Notably, a high proportion of dogs presenting with complete injury equivalents, including those lacking deep pain perception, regained independent ambulation—an outcome that is generally considered unlikely under conservative management alone.

Although the uncontrolled design of this study precludes definitive conclusions regarding causality, the consistency of neurological improvement across cases, the absence of treatment-limiting adverse events, and the observation of functional recovery despite persistent radiological spinal cord compression collectively support the biological plausibility of P2X7 receptor antagonism as a disease-modifying strategy targeting secondary injury mechanisms. These findings align with experimental evidence implicating P2X7-mediated neuroinflammation in post-traumatic tissue loss and functional deterioration following SCI.

Naturally occurring canine spinal cord injury represents a clinically relevant large-animal model that captures the heterogeneity, compressive pathology, and temporal dynamics of human SCI more faithfully than traditional rodent models. Within this context, the present findings provide translationally meaningful evidence that modulation of purinergic signaling may alter the clinical course of severe SCI.

Taken together, this study supports further investigation of P2X7 receptor inhibition in controlled preclinical studies and early-phase clinical trials for acute human spinal cord injury, with an emphasis on defining therapeutic windows, biomarkers of response, and objective functional outcomes. Targeting secondary injury cascades through purinergic modulation may represent a critical step toward disease-modifying pharmacotherapy for spinal cord injury.

References

• 1.

  LaitinenOMPuertoDA. Surgical decompression in dogs with thoracolumbar intervertebral disc disease and loss of deep pain perception: a retrospective study of 46 cases. Acta Vet Scand. (2005) 46:79–85. doi: 10.1186/1751-0147-46-79,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 2.

  ImmekeppelARuppSDemierreSRentmeisterKMeyer-LindenbergAGoessmannJet al. Investigation of timing of surgery and other factors possibly influencing outcome in dogs with acute thoracolumbar disc extrusion: a retrospective study of 1501 cases. Acta Vet Scand. (2021) 63:30. doi: 10.1186/s13028-021-00596-w,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 3.

  PengWCotrinaMLHanXYuHBekarLBlumLet al. Systemic administration of an antagonist of the ATP-sensitive receptor P2X7 improves recovery after spinal cord injury. Proc Natl Acad Sci USA. (2009) 106:12489–93. doi: 10.1073/pnas.0902531106,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4.

  MarcilloAFrydelBBramlettHMDietrichWD. A reassessment of P2X7 receptor inhibition as a neuroprotective strategy in rat models of contusion injury. Exp Neurol. (2012) 233:687–92. doi: 10.1016/j.expneurol.2011.06.008,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5.

  LevineJMLevineGJPorterBFToppKSNoble-HaeussleinLJ. Naturally occurring disk herniation in dogs: an opportunity for pre-clinical spinal cord injury research. J Neurotrauma. (2011) 28:675–88. doi: 10.1089/neu.2010.1645,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6.

  RomanSCusdinFSFonfriaEGoodwinJAReevesJLappinSCet al. Cloning and pharmacological characterization of the dog P2X7 receptor. Br J Pharmacol. (2009) 158:1513–26. doi: 10.1111/j.1476-5381.2009.00425.x,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7.

  AikawaTFujitaHKanazonoSShibataMYoshigaeY. Long-term neurologic outcome of hemilaminectomy and disk fenestration for treatment of dogs with thoracolumbar intervertebral disk herniation: 831 cases (2000-2007). J Am Vet Med Assoc. (2012) 241:1617–26. doi: 10.2460/javma.241.12.1617,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8.

  LangerhuusLMilesJ. Proportion recovery and times to ambulation for non-ambulatory dogs with thoracolumbar disc extrusions treated with hemilaminectomy or conservative treatment: a systematic review and meta-analysis of case-series studies. Vet J. (2017) 220:7–16. doi: 10.1016/j.tvjl.2016.12.008,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9.

  KhanSJefferyNDFreemanP. Recovery of ambulation in small, nonbrachycephalic dogs after conservative management of acute thoracolumbar disk extrusion. J Vet Intern Med. (2024) 38:2603–11. doi: 10.1111/jvim.17149,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10.

  AuffretVCastelAJuetteTFinckC. Residual volume of extruded disc material and residual spinal cord compression measured on postoperative MRI do not predict neurological outcomes in dogs following decompressive surgery for thoracolumbar intervertebral disc extrusion. Vet Radiol Ultrasound. (2024) 65:76–86. doi: 10.1111/vru.13325,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11.

  TogawaGLewisMJDevathasanD. Outcome in paraplegic dogs with or without pain perception due to thoracolumbar fibrocartilaginous embolic myelopathy or acute non-compressive nucleus pulposus extrusion. Front Vet Sci. (2024) 11:1406843. doi: 10.3389/fvets.2024.1406843,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12.

  HeWJCuiJDuLZhaoYDBurnstockGZhouHDet al. Spinal P2X(7) receptor mediates microglia activation-induced neuropathic pain in the sciatic nerve injury rat model. Behav Brain Res. (2012) 226:21933684:163–70. doi: 10.1016/j.bbr.2011.09.015

  • CrossRef
  • Google Scholar

• 13.

  KajitaniTEndoTIwabuchiNInoueTTakahashiYAbeTet al. Association of intravenous administration of human muse cells with deficit amelioration in a rat model of spinal cord injury. J Neurosurg Spine. (2021) 34:648–55. doi: 10.3171/2020.7.SPINE20293,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14.

  TakahashiYKajitaniTEndoTNakayashikiAInoueTNiizumaKet al. Intravenous administration of human muse cells ameliorates deficits in a rat model of subacute spinal cord injury. Int J Mol Sci. (2023) 24:14603. doi: 10.3390/ijms241914603,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar