Local Transplantation of G-CSF-Mobilized CD34 + Cells in a Patient with Tibial Nonunion: A Case Report

Abstract

Although implantation of crude bone marrow cells has been applied in a small number of patients for fracture healing, transplantation of peripheral blood CD34⁺ cells, the hematopoietic/endothelial progenitor cell-enriched population, in patients with fracture has never been reported. Here, we report the first case of tibial nonunion receiving autologous, granulocyte colony stimulating factor mobilized CD34⁺ cells accompanied with autologous bone grafting. No serious adverse event occurred, and the novel therapy performed 9 months after the primary operation resulted in bone union 3 months later without any symptoms including pain and gait disturbance.

Introduction

Whereas most fractures typically heal, a significant proportion (5–10%) of fractures fail to heal and result in delayed union or persistent nonunion (17,24). Nonunion of the long bone is a common problem that can be disabling. Treatment may require multiple operative procedures, prolonged hospitalization, and years of disability until a union is obtained or an amputation is performed. Among several factors resulting in this failure, severe skeletal injuries consisting of fractures with a compromised blood supply have a high risk for leading to either delayed unions or established nonunions. An essential requirement for healing such intractable fractures is to restore the local blood flow, which has traditionally been accomplished through complex vascular procedures or soft tissue transfers with adequate blood supply (8,10,21).

Recent progress in human embryonic and adult stem cell research has been reported in various fields, and bone formation and regeneration has received much attention as a target for regenerative medicine because of the capacity of stem cells to self-renew and differentiate into various types of adult cells or tissues (1,22,28). Adult human peripheral blood (PB) CD34⁺ cells contain intensive endothelial progenitor cells (EPCs) as well as hematopoietic stem cells (HSCs) (3). Tissue ischemia and cytokine mobilize EPCs from BM into PB, and mobilized EPCs specifically home to sites of nascent neovascularization and differentiate into mature endothelial cells (vasculogenesis) (2,25). Therapeutic potential of BM-derived CD34⁺ cells for neovascularization in hindlimb, myocardial, and cerebral ischemia has been demonstrated in both preclinical and clinical studies (12,16). Interestingly, recent reports indicate that BM-derived CD34⁺ cells are capable of differentiating into osteogenic as well as hematopoietic and vasculogenic lineages (4,6,7,15,26). We and other groups reported that fracture induces mobilization of EPCs from BM into PB and incorporation of the circulating EPCs into the fracture site (13,14,19). We first demonstrated that systemic infusion of human circulating CD34⁺ cells into immunodeficient rats with nonhealing fracture contributes to morphological and functional fracture healing by enhancing vasculogenesis and osteogenesis (18). In addition, we attempted local transplantation of CD34⁺ cells with atelocollagen gel, a bioabsorbable scaffold, in the same animal model and demonstrated the similar effect at the lower dose compared with the systemic administration (20). Considering the essential scarcity of EPCs in adult human, this preclinical outcome provided us with a realistic strategy for the future clinical application.

Based on these scientific evidences, we here report the first clinical case of tibial nonunion treated with autologous, granulocyte colony stimulating factor (G-CSF) mobilized CD34⁺ cells with atelocollagen scaffold immediately after the autologous bone grafting from iliac crest.

Case Report

A 42-year-old male presented himself at our hospital complaining of tibial delayed union with pain at the fracture site and disability of life. He had had a closed tibial fracture and been treated by open reduction and internal fixation with plate fixation at another hospital 9 months before the initial presentation at our hospital. During the 9 months, fracture site failed to heal in spite of treating with low intensity pulsed ultrasound device [Sonic Accelerated Fracture Healing System (SAFHS), Teijin Ltd., Japan]. At the time of presentation, he complained of moderate pain and tenderness at the fracture site causing disability of weight-bearing gait. He was clinically diagnosed as a nonunion according to the 1988 FDA Guidance Document Definition requiring 9 months' duration of the nonunited fracture with no evidence of progressive healing over the previous 3 months (27). Anteroposterior and lateral radiographs led to diagnosis of noninfected bone defect type nonunion showing no bridging of four cortical sides. The radiographs also revealed no apparent instability at the fracture site and absence of radiolucency around screws. The radiological findings were supported by three-dimensional (3D) computed tomography (CT) (Fig. 1).

Figure 1.

Preoperative radiograph and three-dimensional (3D)-computed tomography (CT). Anteroposterior radiograph (left panel) led to a diagnosis of noninfected defect type nonunion showing no bridging of four cortical sides. This finding was supported by 3D-CT (right panel).

We obtained an informed consent from the patient for participating in a phase I/IIa clinical trial regarding transplantation of G-CSF mobilized CD34⁺ cells in patients with nonunion. The clinical study protocol conformed to the Declaration of Helsinki and was approved by the ethics committees of the participating hospitals, Institute of Biomedical Research and Innovation (#08-01) and Kobe University Hospital (#735). After the subject eligibility was confirmed, the patient was registered as the first participant in the clinical trial. He received subcutaneous administration of G-CSF (5 μg/kg per day for 5 days) to mobilize EPCs from BM. Leukoapheresis (AS.TEC204; Fresenius HemoCare, Bad Homburg, Germany) was performed to harvest PB mononuclear cells (MNCs) on day 5. The apheresis product number was 2.85 × 10¹⁰ cells and the frequency of CD34⁺ cells in the apheresis product was 0.67% by fluorescence-activated cell sorting (FACS) analysis using CD34-specific monoclonal antibodies (Becton, Dickinson and Company, San Jose, CA). The apheresis product was kept at a concentration of 2 × 10⁸ cells/ml in autoplasma at 4°C overnight (18 h) until the magnetic separation of CD34⁺ cells was started. CD34⁺ cells (1.30 × 10⁸) were isolated by the CliniMACS system consisting of a CliniMACS Instrument, CD34 reagent, phosphate-buffered saline (PBS)/EDTA buffer, and tubing set (Miltenyi Biotec, Bergisch Gladbach, Germany). Purity of the isolated CD34⁺ cells was 92.5% by FACS analysis and the cell viability was 98.3%. The sorted CD34⁺ cells were also positive for the following endothelial lineage surface markers: CD133, c-Kit, and CD31 (94.8%, 87.2%, and 99.3%, respectively). Immediately after the magnetic cell sorting, a predefined dose (5 × 10⁵ cells/kg) of CD34⁺ cells, which was determined by the preclinical study (20), was dissolved in 5 ml of atelocollagen gel (final concentration 1.5%) (KOKEN, Tokyo, Japan), which was used as a bioabsorbable scaffold for retaining the cells at the transplanted site.

Cell transplantation and bone grafting was performed under general anesthesia. Following refreshing fibrous tissue at the nonunion site and the surrounding cortical bone and grafting autologous cancellous bone from iliac crest, CD34⁺ cells dissolved in atelocollagen gel were locally administered into the fracture site (bone defect site) using an injection needle under fluoroscopic control (Fig. 2). Replacement of the original plate was not performed because of no apparent instability at the fracture site and absence of radiolucency around screws.

Figure 2.

Intraoperative findings. Following refreshing fibrous tissue at the nonunion site and the surrounding cortical bone and grafting autologous iliac bone, CD34⁺ cells dissolved in 5 ml of atelocollagen gel were locally administered into the fracture site (bone defect site).

The patient was allowed to gait with partial weight bearing at 6 weeks and with full weight bearing at 12 weeks after the operation. Twelve weeks after the treatment, the patient had no pain complaint with full weight-bearing gait. Anteroposterior and lateral radiograph provided diagnosis of achieved union showing the bony bridging in three of four cortical sides. 3D-CT also supported the radiographical findings (Fig. 3). Taken together, the patient met the criteria of radiographical and clinical union as the primary end point in this treatment at 12 weeks. Six months after the treatment, the patient had no symptoms relating to the fracture and the combined therapy of cell transplantation and bone grafting. He could gait with full weight bearing. No serious adverse events relating to G-CSF administration, leukoapheresis, and cell transplantation occurred during the observation period.

Figure 3.

Postoperative radiograph and 3D-CT. Twelve weeks after the treatment, anteroposterior radiograph (left panel) provided diagnosis of achieved union showing the bony bridging in three of four cortical sides. The radiographical diagnosis was supported by 3D-CT (right panel).

Discussion

To the best of our knowledge, this is the first clinical report of transplantation of autologous, G-CSF mobilized, and purified CD34⁺ cells in a patient with tibial nonunion. The cell therapy combined with autologous iliac bone grafting successfully achieved bone union, which was confirmed by clinical symptoms, radiograph, and 3D-CT as early as 3 months after the treatment. As for the safety evaluation in the first case, there were no serious adverse events for which a causal relationship to the cell therapy could not be denied. In a recent clinical trial in our institution using autologous, G-CSF mobilized CD34⁺ cells in 17 patients with critical limb ischemia (12), no serious adverse events occurred although mild to moderate events relating to G-CSF and leukoapheresis were frequent during the 12-week follow up. Safety, feasibility, and efficacy of this cell-based therapy in patients with nonunion/delayed union would be evaluated after completing this phase I/IIa clinical trial.

Several research groups have demonstrated the usefulness of local transplantation of total BM cells for fracture healing (5,9). Hernigou et al. reported that in 88% of patients with noninfected nonunions of the tibia, bone union was achieved by percutaneous grafting of autologous total BM cells accompanied with the external fixation or cast immobilization (9). Quarto et al. are the first to report the clinical effectiveness and usefulness of BM mesenchymal stem cells (MSCs) associated to a porous ceramic for large long-bone defects (23). Compared with transplantation of purified CD34⁺ cells, crude BM cell or BMMSC therapy does not require the time and cost for magnetic cell sorting. However, our group recently reported that intramyocardial transplantation of human G-CSF mobilized total MNCs represents a possible risk of severe hemorrhagic myocardial infarction through the excessive inflammation induced by abundant infiltration of hematopoietic cells (11). Infusion of the crude BM cells might cause similarly unfavorable event in the case of fracture. Further preclinical/clinical studies would be warranted to compare the feasibility, safety, and efficacy between the various modalities for bone repair.

In conclusion, harvest, isolation, and transplantation of autologous, G-CSF mobilized CD34⁺ cells was first performed in a patient with nonunion/delayed union. Both clinical and radiological healing of the fracture was achieved 12 weeks after the cell therapy and bone grafting. No serious adverse events occurred during the 12-week follow-up. These promising outcomes encourage the early completion of this phase I/IIa clinical trial. Efficacy of the cell therapy would be further elucidated in a randomized controlled clinical trial with an appropriate control group receiving G-CSF only or placebo in the future.