Clinical Study of NeuroRegen Scaffold Combined with Human Mesenchymal Stem Cells for the Repair of Chronic Complete Spinal Cord Injury

Abstract

Regeneration of damaged neurons and recovery of sensation and motor function after complete spinal cord injury (SCI) are challenging. We previously developed a collagen scaffold, NeuroRegen, to promote axonal growth along collagen fibers and inhibit glial scar formation after SCI. When functionalized with multiple biomolecules, this scaffold promoted neurological regeneration and functional recovery in animals with SCI. In this study, eight patients with chronic complete SCI were enrolled to examine the safety and efficacy of implanting NeuroRegen scaffold with human umbilical cord mesenchymal stem cells (hUCB-MSCs). Using intraoperative neurophysiological monitoring, we identified and surgically resected scar tissues to eliminate the inhibitory effect of glial scarring on nerve regeneration. We then implanted NeuroRegen scaffold loaded with hUCB-MSCs into the resection sites. No adverse events (infection, fever, headache, allergic reaction, shock, perioperative complications, aggravation of neurological status, or cancer) were observed during 1 year of follow-up. Primary efficacy outcomes, including expansion of sensation level and motor-evoked potential (MEP)-responsive area, increased finger activity, enhanced trunk stability, defecation sensation, and autonomic neural function recovery, were observed in some patients. Our findings suggest that combined application of NeuroRegen scaffold and hUCB-MSCs is safe and feasible for clinical therapy in patients with chronic SCI. Our study suggests that construction of a regenerative microenvironment using a scaffold-based strategy may be a possible future approach to SCI repair.

Introduction

Spinal cord injury (SCI) often leads to severe disability in patients owing to the loss of voluntary motor function and sensation below the injury level. According to incomplete statistics, approximately 40 people per million sustain a new SCI worldwide every year¹. Most patients with SCI are young males with injuries caused by motor vehicle crashes, falls, and violent acts². As well as having a marked impact on personal, social, and professional lives, SCI can lead to considerable financial, emotional, and psychological burdens for patients and their families³. Although current clinical therapeutic approaches for SCI, including operative decompression, injured lesion stabilization, and rehabilitation, improve overall patient outcome, there is no effective treatment for the neurological deficits that result from the powerful regeneration-inhibitive pathophysiological microenvironment and inherently inadequate regenerative ability of the central nervous system axons⁴.

In the subacute and chronic phases of SCI, many reactive cells, including hypertrophic astrocytes, oligodendrocyte precursors, microglia, and meningeal fibroblasts, aggregate at the injury site and secrete multiple inhibitory molecules. This ultimately leads to formation of scar tissue at the central part of the spinal cord⁵. Scar tissue is a physical and chemical barrier to axon regeneration because very few axons can penetrate its core. The main inhibitory molecules expressed by reactive cells in glial scar formation include chondroitin sulfate proteoglycans, semaphorin 3 proteins, and ephrin tyrosine kinases, which exert their inhibitory effects by binding to the axon cell surface or antagonizing trophic factors and cell adhesion molecules that are essential for axonal regeneration^6-8. For patients with chronic SCI, precise identification and safe resection of scar tissue without damage to sensation or motor function may help to eliminate the inhibitory effect of glial scarring on nerve regeneration.

Biomaterial scaffolds can provide a structural platform and bridge the gap in an SCI to facilitate axonal growth across the injury lesion. In addition to providing physical support and spatial guidance, biomaterials also can be used to deliver stem cells and functional biomolecules to construct a favorable microenvironment at the site of the SCI^4,9. Many scaffold types, with different spatial structure, mechanical strength, and biodegradability, have been developed for SCI repair in animal models. Their therapeutic effects are promising; however, no commercial biomaterial product has yet been approved for treating patients with SCI worldwide. Because collagen is a type of extracellular matrix with excellent biocompatibility and biodegradability, we previously developed a linearly ordered collagen scaffold, NeuroRegen, which promoted axonal growth along collagen fibers and inhibited scar formation in animals with SCI¹⁰. During the past few years, NeuroRegen scaffold functionalized with multiple functional molecules, such as neurotrophic factors, antagonizing proteins, and antibodies, has been implanted into rats and dogs with SCI. Recovery of motor function was observed in the animals, accompanied by de novo neurogenesis, electrical conductivity, and remyelination^11-14. The NeuroRegen scaffold product standard is now established, and a third-party safety evaluation by the China Food and Drug Administration (CFDA) has been completed, which lays the foundation for the clinical study of NeuroRegen scaffold in patients with SCI.

Cell-based therapies for SCI aim either to directly replace or repair the damaged cells themselves or to promote reconstruction indirectly by secreting factors conducive to a regenerative microenvironment¹⁵. Transplantation of multiple cell types has been reported for SCI repair in animals and human beings; the cell types studied include mesenchymal stem cells (MSCs), olfactory ensheathing cells (OECs), oligodendrocytes, Schwann cells, and neural stem cells¹⁶. MSCs can promote SCI recovery by producing numerous neurotrophic factors and immune cytokines. MSCs exhibit a broad degree of plasticity that enables them to differentiate into multiple cell types, and they secrete cytokines and growth factors that promote immunosuppression, inhibit gliosis and apoptosis, and enhance angiogenesis, axon sorting, and myelination in animals^17,18. Clinical trials have evaluated the safety and efficacy of MSC transplantation in patients with acute and chronic SCI since 2005^19-29.

In this study, eight patients with complete chronic SCI were enrolled to evaluate the safety and efficacy of NeuroRegen scaffold implantation with human umbilical cord MSCs (hUCB-MSCs). The scar tissue of the patients was identified using intraoperative neurophysiological monitoring and was surgically resected to eliminate its inhibitory effect on nerve regeneration. To inhibit re-formation of scar tissue and facilitate axonal growth across the lesion, NeuroRegen scaffold was implanted into the resection sites. The scaffold was preloaded and soaked with hUCB-MSCs to reduce diffusion of the cells from the SCI site and promote reconstruction of the SCI microenvironment. During 1 year of subsequent clinical observation, no adverse effects were observed, and partial recovery of sensation and motor function were detected. This suggests that the combined application of NeuroRegen scaffold and MSCs is safe and feasible for clinical therapy in patients with chronic SCI.

Materials and Methods

Patient Selection

The experiment was designed in accordance with the Declaration of Helsinki, approved by the ethics committee of the Affiliated Hospital of Logistics University of Chinese Armed Police Forces (CAPF), and registered on the National Institutes of Health (NIH) database (ClinicalTrials.gov NCT02352077). Written informed consent to participate in the study was obtained from each patient prior to commencement of procedures. Patients were fully aware of the treatment process and possible adverse outcomes.

Eight patients (seven men and one woman) with a mean age of 31.5 years were enrolled in the study. All enrolled patients had chronic (2-36 months postinjury) and complete [American Spinal Injury Association (ASIA) classification A] cervical or thoracic lesions. The main inclusion and exclusion criteria of the study were as follows:

Inclusion criteria:

• 1.Male or female, 18-65 years old;
• 2.Complete SCI at the cervical or thoracic level (C5-T12);
• 3.ASIA A classification with no significant improvement since the injury was sustained;
• 4.Provided signed informed consent;
• 5.Able and willing to regularly visit the hospital and participate in follow-up during the study.

Exclusion criteria:

• 1.Serious complications;
• 2.History of life-threatening allergic or immune-mediated reaction;
• 3.Clinically significant abnormality in routine laboratory examination;
• 4.Lactating or pregnant;
• 5.Participated in any other clinical trial within 3 months before enrollment;
• 6.Poor compliance with or difficultly completing the study requirements.

Preparation, Scanning Electron Microscopy Analysis, and Safety Evaluation of NeuroRegen Scaffold Before Implantation

NeuroRegen scaffolds were prepared from bovine aponeurosis using a modification of previously reported methods¹⁰. Briefly, fresh bovine aponeuroses were harvested from a local slaughter house (Wanlifa, Beijing, P.R. China) and rinsed with cold distilled water several times, and then residual muscles, connective tissue, and fat were carefully removed. Tri(n-butyl) phosphate hypertonic solution (Sigma-Aldrich, St. Louis, MO, USA) and enzymes (Sigma-Aldrich) were used to remove residual fat, cellular components, and soluble proteins. The samples were then repeatedly rinsed to completely remove the residual agents and freeze dried. For structural analysis, the NeuroRegen scaffolds were sputter coated with gold and observed by scanning electron microscopy (SEM; S-3000N SEM; Hitachi, Tokyo, Japan). Allergen detection and assays for acute toxicity, cytotoxicity, subchronic toxicity, intradermal irritation, genetic toxicity, hemolytic toxicity, and degradation were performed by the National Institute of Food and Drug Control according to the Chinese Criterion for Medical Devices GB16886 to evaluate the biological safety of the NeuroRegen scaffold.

Human Umbilical Cord-MSC Isolation, Culture, and Characterization

hUCB-MSCs were isolated from Wharton's jelly from the umbilical cord by methods described previously³⁰. The procurement of human tissues was approved by the ethics committee of the Affiliated Hospital of Logistics University of Chinese Armed Police Forces (CAPF), and written informed consent was obtained from the mothers. Briefly, the cords were dissected, and the blood vessels were removed. The remaining tissues were cut into small pieces and digested with collagenase (Sigma-Aldrich) for 18h, followed by digestion with 0.25% trypsin (Sigma-Aldrich). A 100-μm filter was used to remove undigested tissue from the cell suspension. Cells were seeded in Dulbecco's modified Eagle's medium (Gibco, Grand Island, NY, USA) containing KnockOut^™ serum supplement (Gibco), 2 mM glutamine (Gibco), and 50 U of penicillin/streptomycin (Gibco). Cells were cultured in a humidified atmosphere with 5% CO₂ at 37°C. After 3 days of culture, the medium was replaced to remove nonadherent cells. Cells were passaged when they reached confluence. Microbiological and cytogenetic safety was ensured throughout the preparation process. Cells were characterized by flow cytometry with antibodies against CD45, CD34, CD11b, CD19, HLA-DR, CD105, CD73, and CD90. For morphological analysis, hUCB-MSCs seeded on NeuroRegen scaffold were fixed in phosphate-buffered saline (PBS) containing 1% glutaraldehyde (Sigma-Aldrich) and 4% formaldehyde (pH 7.2; Sigma-Aldrich), dehydrated with ethanol, and then dried in a critical point drier (Hitachi) for coating with gold. The images were captured under a SEM.

Scar Tissue Resection Under Neural Electrophysiology Monitoring and Implantation of NeuroRegen Scaffold and hUCB-MSCs

The patient was placed under general anesthesia in a prone position. A midline skin incision was made, followed by paravertebral muscle dissection, and laminectomy. Adhesion outside the injured dura was removed under an operating microscope. A midline durotomy was performed followed by sharp dissection of the post-traumatic adhesions between the spinal cord surface and dura. The injured area consisted of two spinal cord stumps separated by scar tissue. The texture of scar tissue is different from normal spinal cord structure, and the middle of the scar may contain necrotic tissue or cavities. The boundaries between normal spinal cord and scar tissue are generally determined according to neural electrophysiology³¹. Somatosensory-evoked potentials (SSEPs) and motor-evoked potentials (MEPs) were used to determine the rostral and caudal ends of the scar tissue, respectively. Electromyography stimulation electrodes (XLTEK^®; Natus^®, Oakville, Canada) were placed near the rostral side of the SCI site, and recording electrodes were placed in the scalp. Normal SSEP response indicated that the stimulation electrodes were positioned on normal spinal cord tissue; lack of SSEP response indicated scar tissue. For caudal scar tissue, the stimulation electrodes were placed near the caudal end of the SCI, and the recording electrodes were placed into the external anal sphincter. If an MEP response was detected, this indicated that the stimulation electrodes were positioned on normal spinal cord tissue; lack of MEP response indicated scar tissue. After the rostral and caudal ends of the scar tissue were detected, the glial scar was carefully removed under the operating microscope. Next, 4 × 10⁷ hUCB-MSCs were loaded onto NeuroRegen scaffolds and incubated for 5 min to ensure cell adherence to the scaffolds. The NeuroRegen scaffolds functionalized with hUCB-MSCs were then implanted in the SCI lesion in the area from which the glial scar had been removed. The dura was closed with absorbable sutures.

Rehabilitation Program and Patient Follow-Up

Patients underwent rehabilitation for 6 months after surgery in the form of a clear, constant, and regular program that included physical training strategies to encourage motor and sensation function and muscle strength recovery. Before surgery and at each follow-up appointment (1, 3, 6, and 12 months after surgery), complete clinical and neurological evaluations and ASIA scale assessment were conducted. MEP testing was carried out using scalp stimulation for evaluation of muscle response. For determining evoked motor potentials of patients, the stimulation electrodes were positioned on the scalp, and the recording electrodes were placed into target muscles including rectus abdominis, then single-pulse stimulation (160 V) was applied to produce neuroelectrical signals. The response positions of the evoked motor potential were recorded.

Results

Characterization of NeuroRegen Scaffold and hUCB-MSCs

hUCB-MSCs had a fibroblast-like morphology (Fig. 1A) and expressed the typical MSC markers CD105, CD73, and CD90, but not CD34, CD45, CD11b, CD19, or HLA-DR (Fig. 1B). NeuroRegen scaffold was composed of multiple longitudinally arranged white collagen fibers (Fig. 1C). SEM analysis of the fiber morphology demonstrated many tiny channels along the aligned fibers (Fig. 1D). The biodegradability and biological safety of NeuroRegen scaffold were evaluated by authorized thirdparty inspection, the CFDA, and were shown to meet the Chinese Criterion for Medical Devices GB16886 regarding absence of allergens and biological toxicity. The SEM data demonstrated that hUCB-MSCs adhered to the NeuroRegen scaffold, which would decrease diffusion of cells from the injury site after implantation (Fig. 1E).

Figure 1.

Morphology and characterization of human umbilical cord mesenchymal stem cells (hUCB-MSCs) and NeuroRegen scaffold. (A) MSC morphology visualized by phase-contrast microscopy. Scale bar: 100 μm. (B) Flow cytometry analysis of cell surface markers with antibodies against CD34, CD45, CD11b, CD19, HLA-DR, CD73, CD90, and CD105. (C) NeuroRegen scaffold. (D) NeuroRegen scaffold morphology visualized by scanning electron microscopy (SEM). Scale bar: 200 μm. (E) Morphology of hUCB-MSCs adhered to NeuroRegen scaffold, visualized by SEM. Scale bar: 100 μm.

Patient Information

Eight patients (seven men and one woman) with complete SCI who met the inclusion criteria were enrolled in our study. Subjects had chronic traumatic SCI with a mean duration of approximately 13.4 months (range: 2-36 months). The mean age of participants was 31.5 years (range: 28-41 years). All were classified as ASIA grade A and had injuries in the cervical or thoracic segments of the spinal cord. Scar size was estimated by MRI analysis, and the boundaries of the normal spinal cord and scar tissue were determined by neural electrophysiology monitoring during the surgery. Scar length ranged from 1.4 to 5 cm. Glial scar length varied between patients and did not directly correlate with the location of injured segments or time after SCI, but it might be correlated with individual physical condition differences and injury severity (Table 1).

Table 1.

Patient Demographic and Clinical Features

Patient ID	Sex	Age	Time After Injury (Months)	Level of Injury	Length of Scar Tissue (cm)	ASIA Grade
1	M	28	36	C7	1.4	A
2	M	31	9	T10	5.0	A
3	M	24	11	C6	1.3	A
4	M	32	2	T7	3.5	A
5	M	41	16	T5	3.8	A
6	F	27	15	C5	2.3	A
7	M	34	7	T4	5.0	A
8	M	35	11	C6	2.0	A

Safety and Primary Efficacy of NeuroRegen Scaffold and MSC Implantation for Patients with SCI

After identifying the rostral and caudal limits of the glial scar and surgically resecting the scar tissues to remove molecules that inhibit axon regeneration, we implanted MSC-loaded NeuroRegen scaffold to bridge the gap in the spinal cord, inhibit the re-formation of scar tissues, and construct a microenvironment favorable to SCI regeneration. The safety of implanting NeuroRegen scaffold with hUCB-MSCs after scar tissue resection was evaluated. No early adverse events related to NeuroRegen scaffold or MSC transplantation (infection, high fever, headache, allergic reaction, shock, or perioperative complications) or late adverse events (aggravation of neurological status or cancer) were observed during 1 year of follow-up visits, which indicates the safety of scar resection and scaffold implantation.

During 1 year of neurological function observation after scaffold implantation, we found that 62.5% of subjects demonstrated expansion of sensation level (Fig. 2A), and three patients with cervical lesions showed increased finger flexibility (Table 2). We also analyzed MEPs in patients and found marked expansion of the MEP-responsive area in 87.5% of patients (Fig. 2B), which suggests partial recovery of neurological function. Two patients reported defecation sensation, although without achieving sphincter control. Increased stability and trunk equilibrium in the sitting position were reported in four patients, but no improvement was observed in the ASIA classification of the patients. Autonomic dysfunction is a common clinical consequence of SCI; we detected autonomic neural function recovery, such as enhanced skin sweating below the injury level, in some patients after the treatment (Table 2).

Figure 2.

Change in sensation level (A) and motor-evoked potential (MEP)-responsive area (B) in patients 1 year after treatment. The y-axis indicates spinal cord segment from T1 to L3. L, left side of the patient; R, right side of the patient. Green bars represent sensation level (A) or MEP-responsive area (B) before treatment. Red bars represent expansion of sensation level (A) or MEP-responsive area 1 year after treatment (B). Blue bars represent the level at which there was no sensation (A) or MEP response (B) 1 year after treatment.

Table 2.

Summary of Neural Function Change of the Patients 1 Year Treatment

Patient ID	Change of Sensation Level	Defecation Sensation	Change of Motor Function	Motor-Evoked Potential Recovery	Trunk Equilibrium Improvement	Enhanced Sweating Below the Injury Site
1	Yes	Yes	No	No	No	Yes
2	Yes	No	No	Yes	No	Yes
3	No	No	Yes	Yes	Yes	Yes
4	Yes	No	No	Yes	No	Yes
5	No	No	No	Yes	No	Yes
6	No	No	Yes	Yes	Yes	No
7	Yes	Yes	No	Yes	Yes	Yes
8	Yes	No	Yes	Yes	Yes	No

Discussion

SCIs are life-changing events that may lead to substantial disability. Conventional clinical treatments have no significant effects on neurological recovery in chronic SCI. In this study, NeuroRegen scaffold loaded with hUCB-MSCs was surgically transplanted into the injury site in patients with SCI to construct a favorable local microenvironment for regeneration. To our knowledge, this is the first clinical trial to examine the safety and efficacy of implanting scaffolds with hUCB-MSCs in chronic complete SCI.

The safety of implanting NeuroRegen scaffold with hUCB-MSCs was examined during 1 year of follow-up visits. No early adverse events (infection, fever, headache, allergic reaction, shock, or perioperative complications) or late adverse events (aggravation of neurological status or cancer) were observed, which indicates that the scaffold, hUCB-MSCs, and the scar resection surgery are safe. The scaffolds used in this study are biodegradable collagen scaffolds whose biological safety has been examined in rats and dogs with SCI during the past few years^11-13. Before this clinical study commenced, the scaffolds were evaluated by authorized third-party inspection, the CFDA, and were shown to meet the Chinese Criterion for Medical Devices GB16886 regarding absence of allergens and biological toxicity. Previous studies of the application of MSCs to treat SCI in experimental animal models and clinical trials have demonstrated the safety of this approach^16,32. In this study, the transplanted MSCs were obtained according to established protocols, and their purity was assessed with commercial antibodies. Microbiological and cytogenetic safety was ensured throughout the preparation process. Consistent with previous clinical MSC transplantation studies, no adverse events were reported after cell transplantation in our study. The safety of scar resection based on neural electrophysiology monitoring was also demonstrated in our previous study. The immunostaining results confirmed that no nerve fibers or neural cells were falsely resected, and no neurological deterioration or spine instability was detected after scar resection³¹. Further follow-up of the patients is required to determine the long-term safety of the scaffold and MSC implantation after scar resection.

The potential benefits of implanting NeuroRegen scaffold with hUCB-MSCs in patients with complete chronic SCI in terms of motor and sensory improvements were observed during the 1-year follow-up. SCI can lead to loss of sensation below the injury site because of damage to ascending neuronal circuitry. It has been reported that improvement in sensation is the first sign of recovery in patients with SCI³³. In our study, expansion of sensation level in five patients and the feeling of defecation in two patients indicate partial recovery of sensation, which may be attributed to the regeneration of ascending axons. Furthermore, we observed expansion of the MEP-responsive area in 87.5% of the patients. Although no improvement in ASIA classification was observed in patients, the increase in finger flexibility in three patients with cervical lesions and the increase in trunk stability in half of the patients were found. We propose that, taken together, these data likely indicate partial motor function recovery after the treatment. Some studies reported spontaneous recovery of some complete SCI patients^34,35. However, we believed the implantation of scaffold with hUCB-MSCs should play a beneficial role in the recovery of patients for the following reasons: first, the spontaneous recovery rate of chronic complete SCI patients was relatively low. A previous study reported that only 5.6% complete SCI patients could achieve obvious spontaneous recovery from 1 to 5 years after SCI³⁶. Second, our pre-clinical studies in animals demonstrated that NeuroRegen scaffold could inhibit scar formation and promote axon growth along the collagen fibers^10,13. Third, some studies reported that the vast majority of SCI recoveries occurred in the first 3 months after injury³⁷. Patients enrolled in our trial were ASIA A classified with a mean duration of SCI approximately 13.4 months and no obvious improvement in neurological function since injury. However, after implantation of scaffold with hUCB-MSCs, the continuous improvement of motor or sensation function was found in some patients during the 1-year follow-up. Especially, the recovery of patients with SCI more than 1 year might be related with scaffold–MSC implantation. A randomized controlled study of a larger sample is needed to clarify the importance of scaffold with MSCs for SCI therapy in the future.

Autonomic dysfunction, such as abnormal blood pressure and heart rate, sweating, and temperature dysregulation, is a common clinical consequence of SCI. In our study, enhanced sweating below the injury level was observed, indicating that autonomic nervous system function was partly restored. The recovery of motor and sensation function is regarded as key index to evaluate the therapeutic effects of SCI. Previous studies of MSC transplantation also revealed differing extents of recovery of sensation and motor function in patients with SCI³². However, it is difficult to improve ASIA classification for most patients of chronic complete SCI. Yazdani et al. demonstrated that no change in ASIA classification was seen in chronic patients with autologous Schwann cell and bone marrow mesenchymal stromal cell (BM-MSC) transplantation²⁶. The ASIA grade increased in 30.4% of the acute and subacute treated patients, whereas no significant improvement was observed in the chronic treatment group with autologous BM cell transplantation²¹. Similarly, incomplete SCI patients treated with hUCB-MSCs had improved motor function, but there was no response to treatment among patients with complete SCI²⁵. Although we also did not observe the improvement of ASIA classification, the obvious expansion of the MEP-responsive area in 87.5% of patients really bring us hope for patients' motor score change in the near future. The expansion of sensation level after cell transplantation was reported in some clinical studies, which is consistent with our data. Cristante et al. demonstrated that patients showed recovery of somatosensory-evoked response to autologous BM-MSC infusion after 2.5 years of follow-up³⁸. Jarocha et al. reported that a 15-year-old girl, treated with multiple rounds of autologous BM nucleated cells and MSCs, expanded her sensation level from T1 to L3³³. They also reported improved superficial sensation in 40% of SCI children after autologous BM nucleated cell transplantation²³. It is difficult to evaluate which of these cell-based therapies for SCI is superior because of differences in the cell number, delivery route, therapeutic window, and rehabilitation program, but use of a scaffold in SCI treatment would greatly increase the efficiency of delivery of cells to injury sites and provide the potential to regulate cell behavior, such as proliferation and differentiation, by scaffold functionalization in the future.

Rehabilitating patients with complete chronic SCI is a complicated and challenging task. It includes not only central nerve function regeneration but also recovery of the peripheral nervous system and muscle, bone, and joint strength³⁹. More efforts should be devoted to developing a comprehensive strategy for constructing a local microenvironment that promotes neural regeneration and improving overall motor function in future patients.

There are several possible explanations for sensation and motor recovery in patients in this study. First, NeuroRegen scaffold is composed of longitudinally arranged fibers, which could bridge lesion gap and inhibit re-formation of glial scar tissue. Multiple tiny channels on the fibers could induce cell adhesion and guide axonal growth along the fibers. Our previous study revealed that when the scaffold was transplanted into rat and canine models of complete SCI, it guided neurite outgrowth along its fibers and decreased glial scar formation^10,11,13. Second, NeuroRegen scaffold also acted as a vehicle to deliver MSCs to modify the microenvironment. It has previously been reported that MSCs exhibit a broad degree of plasticity, enabling them to differentiate into multiple cell types, and that they secrete cytokines and growth factors that promote immunosuppression, inhibit gliosis and apoptosis, and enhance angiogenesis, axon sorting, and myelination^17,18. Traditionally, MSCs were intraspinally or intrathecally transplanted into patients; therefore, the therapeutic efficiency was variable because of rapid diffusion of cells from the SCI site. When MSCs are loaded on the NeuroRegen scaffold, they attach to the surface of the collagen fibers, which may decrease diffusion of the MSCs from the implantation site. The primary recovery of sensation and motor function in our study could be partially attributed to paracrine effects of MSCs retained at the injury site. Previously, NeuroRegen scaffolds were used to deliver autologous bone marrow mononuclear cells (BMMCs) to repair SCI patients³¹, but the individual difference of fresh isolated autologous BMMCs from patients might affect the therapeutic outcome. In this study, hUCB-MSCs were cultured by established methods, and cells were relatively homogeneous for the expression of similar cell surface protein. Compared to the previous study, where there was recovery of SSEPs from the lower limbs of the patients after implantation of NeuroRegen scaffold loaded with autologous BMMCs, we found NeuroRegen scaffold loading with hUCB-MSCs promoted the expansion of sensation level and MEP-responsive area and increased finger activity and defecation sensation. Third, surgical glial scar resection is a physical approach to eliminating the inhibitory effect of glial scarring on nerve regeneration in chronic SCI. Many studies have regulated the gliotic scarring response by reducing specific glial scar molecules to promote axon regeneration and functional SCI recovery. For example, enzymatic digestion of the glycosaminoglycan side chains of chondroitin sulfate proteoglycans by chondroitinase ABC stimulates axonal regeneration after SCI⁴⁰, and functional blocking of semaphorin receptors by antibodies promotes sensory axon regeneration⁸. Animals with clonal deletions of ephrin tyrosine kinase exhibit reduced glial scar response⁴¹. We believe that the surgical glial scar resection approach has the potential for further improvements based on the multiple molecules and complicated microenvironment of glial scars⁴². Molecules aiding axon regeneration could be preserved or supplemented during or after surgical resection to achieve better patient recovery in the future. In addition, the physical rehabilitation based on the transplantation of NeuroRegen scaffold with hUCB-MSCs also plays an important role for neurological function recovery of chronic complete patients. Some studies revealed that long-term locomotor training could improve motor function in partial SCI patients^43,44. Our study did not evaluate the priority of physical rehabilitation, scaffold implantation, and MSC transplantation for SCI therapy. However, we found scaffold–MSC implantation combined rehabilitation can improve the neurological function in some patients who had no obvious improvement by physical rehabilitation before enrollment. This suggests that scaffold and MSC implantation might be crucial for the therapeutic outcome.

Conclusions

We report for the first time the safety and efficacy of implanting NeuroRegen scaffold and hUCB-MSCs in patients with chronic complete SCI. Our preliminary results indicate the feasibility of combined application of NeuroRegen scaffold and hUCB-MSCs for clinical therapy of chronic SCI and suggest that construction of a regenerative microenvironment using a scaffold-based strategy may be a possible future approach to SCI repair.