Potential of Allogeneic Adipose-Derived Stem Cell–Hydrogel Complex for Treating Diabetic Foot Ulcers

The pathophysiologic relationship between diabetes and impaired wound healing is complicated. Attenuated activities of cells that play a key role in wound healing contribute to the impairment of tissue restoration in diabetic ulcers. Keratinocytes and fibroblasts isolated from diabetic foot ulcers show lower proliferative potential and attenuated growth factor production (1). Therefore, there is considerable interest in the treatment of diabetic foot ulcers with biological dressings and/or tissue-engineered products.

Mesenchymal stem cells (MSCs) may hold great promise for treating diabetic wounds because they have advantages as allogeneic and autologous cells. MSCs demonstrate low levels of immunity-assisted rejection with the ability to divide without apoptosis (2,3). Even after 20 or 30 cycles of cell doubling in culture, they still retain their initial stem cell properties (4).

Bone marrow stroma has been a main source of MSCs. Previous studies performed by our group have demonstrated that bone marrow–derived MSCs can synthesize higher amounts of collagen, fibroblast growth factor, and vascular endothelial growth factor in vitro than dermal fibroblasts. Furthermore, they showed greater activity in terms of granulation tissue formation, epithelialization, and angiogenesis in vivo, indicating their potential use in accelerated wound healing (2,5,6). However, there have been no commercial drugs that contain bone marrow–derived MSCs to treat diabetic foot ulcers.

Recently, adipose-derived stem cells (ASCs) have been demonstrated to be one of the main sources of MSCs (7–10). Kato et al. (11) have shown that allogeneic transplantation of an ASC combined with artificial skin can accelerate wound healing in Zucker diabetic fatty rats in vivo. McLaughlin and Marra (12) demonstrated that ASC sheets can increase wound healing compared with untreated controls in vivo. Furthermore, ASCs showed greater activity in terms of epithelialization, angiogenesis, and secretion of growth factors (11). Moreover, a bioengineered dermal substitute comprised of allogeneic ASCs has been commercialized to help diabetic wound healing. The purpose of this study was to examine the potential of hydrogel-based allogeneic ASC sheets for treating diabetic foot ulcers.

This was a randomized, comparator-controlled, single-blind, parallel-group, multicenter study in which patients with diabetic foot ulcers were recruited consecutively from four centers in Korea. This trial was registered with ClinicalTrials.gov (clinical trial reg. no. NCT02619877). This clinical study protocol and informed consent document were also approved by the appropriate Institutional Review Board for each participating center and the Food and Drug Administration (FDA) of Korea (study code ALLO-ASC-DFU-201).

The major inclusion criteria were as follows: age between 18 and 80 years; patients with type 1 or type 2 diabetes; longer than 4 weeks for the history of ulcer at screening; wound size between 1 and 25 cm²; and wound depth of Wagner grade 1 and 2. Additional criteria were as follows: blood flow around the ulcer was detectable by a Doppler test, ankle-brachial index range of >0.7 to <1.3, or transcutaneous oxygen pressure >30 mmHg. Key exclusion criteria included a change in wound size of >30% within 1 week from screening, wound infection, being HIV positive, HbA_1c >15%, and postprandial blood glucose level >450 mg/dL.

Once patients were enrolled, they were randomized into one of two treatment groups. Randomization schedules were stratified according to a clinical center using a permuted-block method with a block size of four to six through SAS.

Between November 2015 and October 2016, 24, 11, 12, and 12 patients with diabetic foot ulcers, respectively, from the Korea University Guro Hospital, Eulji General Hospital, Asan Medical Center, and the Yonsei University Severance Hospital were included in this clinical trial. Of these 59 subjects, 5 subjects in the control group were excluded before the study because of the withdrawal of consents and the ineligibility of criteria such as small wound size (<1 cm²). Therefore, the assessment of a modified intention-to-treat (ITT) set was conducted on data available from the remaining 54 subjects. Five subjects in the treatment group and five subjects in the control group were further excluded before completion of the study because of the occurrence of adverse events and protocol violations such as violation of the timeline. Of the remaining 44 subjects, 25 were assigned to the treatment group and 19 were assigned to the control group. Three subjects in the treatment group and two subjects in the control group were excluded because of violations of eligibility criteria such as high serum glucose levels (>450 mg/dL) during the study period. Therefore, assessment of the modified per-protocol (mPP) set was conducted on data available from the remaining 39 subjects who completed the study (Fig. 1). Demographics and wound characteristics of subjects in the two groups at baseline are shown in Table 1.

Allogeneic ASC sheet (ALLO-ASC-Sheet; Anterogen, Seoul, South Korea) is a 5 × 5 cm hydrogel sheet containing allogeneic ASCs. Briefly, ASCs were obtained from the subcutaneous fat tissue of healthy donors who provided informed consent before undergoing liposuction. Obtained adipose tissues were rinsed with PBS (Hyclone, Logan, UT) in DMEM/Ham’s F-12 (DMEM; Hyclone) containing 0.025% type I collagenase (Invitrogen, Carlsbad, CA) for 80 min at 37°C according to the manufacturer protocol. The top lipid layer was removed, and the remaining liquid portion was centrifuged at 300g for 10 min at 4°C. The stromal vascular fraction was collected and cultured in DMEM to obtain the required number of ASCs. ASCs seeded onto the hydrogel matrix were cultured until the number of ASCs reached about 1 × 10⁶ cells/sheet. They were stored frozen at −80°C according to the undisclosed manufacturer protocols.

ASCs were characterized by their expression of stromal cell–associated markers such as CD10, CD13, CD29, CD44, and CD90. These cells were negative for the expression of hematopoietic stem cell–associated markers (CD34 and CD45). The genomic stability of ASCs was evaluated by karyotyping analysis. A series of additional efficacy release tests were performed for the final product, including confirmation of the cell count and assessment of cell viability. ASCs were subjected to a series of quality controls to ensure their purity, safety, and potency that were approved by the FDA of Korea.

In the treatment group, the ulcer was cleaned to remove dirt and other debris with 3% hydrogen peroxide followed by saline solution. The allogeneic ASC sheet, which was stored at −80°C, was thawed at room temperature for 1–2 min, tailored according to the wound size and shape, and applied directly to the wound bed as a primary dressing. It was then covered with a polyurethane foam (Mepilex; Molnlycke Health Care, Gothenburg, Sweden) as a secondary dressing. The graft and polyurethane foam were left on the wound for 7 days. Dressing changes were scheduled at weekly intervals. If found necessary, the patient returned to the hospital two to three times per week after application of the graft so that the wound could be examined. Only the secondary dressing was changed. In the control group, all conditions including management for diabetic foot ulcers were set up to be identical to those for the treatment group. A meshed polyurethane film with silicone adhesive (Mepitel; Molnlycke Health Care) was applied as a primary dressing over the wound. It was then covered with the same secondary dressing used for the treatment group. Visits and dressing changes for the control group were scheduled in the same way as for the treatment group.

All patients with ulcers on weight-bearing sites or sites otherwise subjected to pressure when wearing shoes had pressure off-loaded using foam dressings with a hole on the ulcer site and footwear with cushioning insoles.

Wounds were evaluated weekly until the 12th week visit. When the ulcer was completely healed before the 12th week visit, treatment was stopped but visits were continued as scheduled until week 12 to evaluate long-term safety. Wound evaluation was performed in a single-blinded fashion. Patients did not know whether or not their wounds had been treated with ASC sheets. However, wound evaluators were aware of the method of treatment. Photographs were captured at baseline, follow-up, and the last visit using standardized photographic equipment.

The mPP population was used for primary and secondary efficacy analysis in this study. The modified ITT population was used for the safety analysis and summaries that included all subjects who were treated with ASC sheets and had visited the clinic at least once.

Primary efficacy criterion was the percentage of subjects who achieved complete wound closure within the 8-week study period. Secondary efficacy criteria were the proportion of subjects who achieved complete wound closure within the 12-week study period, the time required for complete wound closure in patients who achieved complete healing within 12 weeks, the rate of wound size reduction from baseline, and the proportion of subjects by Wagner grade at each week. In addition, post hoc analyses were performed to evaluate the efficacy of the treatment and prognostic factors such as wound depth and location. Wound size was determined using a Visitrak Digital Planimetry Wound Measurement System (Smith & Nephew, Hull, U.K.).

For the safety study, information on adverse events, adverse drug reactions, serious adverse events, and serious adverse drug reactions was collected at each visit. Safety was also monitored at indicated timepoints by evaluating adverse event reports, laboratory assessments, and vital signs. A systematic immunologic study was conducted to detect anti-HLA panel reactive antibodies at baseline, the 1st week, and the 12th week after ASC treatment.

Ulcer recurrence and adverse events of healed ulcers were monitored for 2 years after healing by reviewing medical records and/or through telephone interviews. This follow-up study was registered with ClinicalTrials.gov (clinical trial reg. no. NCT03183804) and the FDA of Korea (study code ALLO-ASC-DFU-202).

Based on previous studies (13,14), the number of subjects was calculated in this study. The weighted complete healing was the sum of number of subjects with complete healing in each study divided by the sum of the number of subjects used in each study. The weighted complete healing was 0.81 in the treatment group and 0.45 in the control group. The following values were used to calculate the number of subjects: type I error (α) = 5%, power of test = 80%, complete healing rate in the treatment group = 0.81, and complete healing rate in the control group = 0.45. Where each complete healing rate was p_t for the treatment group and p_c for the control group, the two-sided null hypothesis and alternative hypothesis were as follows: null hypothesis, p_t = p_c; alternative hypothesis, p_t ≠ p_c.

In this study, Z_0.975 was 1.96 and Z_0.80 was 0.84. Therefore, the number of subjects in each group could be calculated as follows (15):

Considering a dropout rate of broadly 10%, the number of subjects in each group was 26.97 (= 24.28/(1 − 0.10)), and a sample size of 54 (= 27 × 2) randomly assigned subjects was required. Based on our experience from previous clinical trials for treating diabetic foot ulcers, ∼10% of subjects experienced unexpected events related to diabetes-related complications and high-risk comorbidities. Therefore, we added another 10% of subjects, bringing the total number of subjects enrolled in this study to 59.

Pearson χ² test was used to evaluate the proportion of subjects who achieved complete wound closure and the safety set. Log-rank test and independent t test or Wilcoxon rank sum test were used to evaluate the time required for complete closure and the rate of wound size reduction, respectively. The median time to complete closure was also estimated using the Kaplan-Meier method. A P value <0.05 was considered to be statistically significant.

The proportion of subjects achieving complete wound closure at week 8 in the treatment group was 73% (16 of 22 subjects). It was 47% (8 of 17 subjects) in the control group (P = 0.102).

For results of secondary efficacy criteria, the proportion of subjects achieving complete wound closure at week 12 in the treatment group was 82% (18 of 22 subjects) and 53% (9 of 17 subjects) in the control group (P = 0.053). The mean time required for complete wound closure was 40.8 ± 5.3 days in the treatment group and 51.2 ± 3.9 days in the control group. The Kaplan-Meier median time to complete wound healing was 28.5 days for the treatment group and 63.0 days for the control group (P = 0.033) (Fig. 2). The rate of wound size reduction at week 1 was 49.6 ± 25.7% in the treatment group and 23.0 ± 32.2% in the control group (P = 0.007). The rate of wound size reduction was also statistically significant between the two groups at 9 of 12 study weeks (Fig. 3). Post hoc analyses were performed to further explore other prognostic factors such as baseline wound depth and location that might affect the efficacy of the treatment (Fig. 4 and Tables 2 and 3).

Incidence rates of adverse events are shown in Table 4. None of these events were related to study dressings. There were no significant differences between the two groups. The incidence of serious adverse event was similar between the treatment and control groups. Cellulitis on untargeted sites, paresthesia, uncontrolled diabetes, and cardiac arrest occurred as serious adverse events. However, none of these serious adverse events was considered to be related to the treatment. No other serious adverse events were found in either group. No clinically meaningful changes from baseline clinical and laboratory parameters, including serum chemistry, hematology, urinalysis, and vital signs, occurred in any of these patients.

In the systematic immunologic study, 16 subjects (11 in the treatment group and 5 in the control group) showed anti-HLA antibodies >1,500 fluorescence intensity values at the 1st week and the 12th week. Among these subjects, three (27%) in the treatment group and one (20%) in the control group showed a slight elevation of antibodies at the 12th week. However, we observed no obvious clinical signs of rejection such as erythema, local inflammatory signs, or visible signs of necrosis, as described in previous studies (16,17), in these subjects.

Ulcer recurrences and adverse events in healed ulcers were successfully monitored for 2 years for 16 subjects in the treatment group and 8 subjects in the control group. We reviewed the medical records of subjects who attended routine appointments for seven subjects in the treatment group and six subjects in the control group. Nine subjects in the treatment group and two subjects in the control group were interviewed over the telephone. Two subjects in the treatment group and no subjects in the control group showed ulcer recurrence 6 months after completion of the trial, but these ulcers that recurred eventually healed. We noted no adverse events related to wound dressings in either group for 2 years after completion of the trial.

The development of advanced wound-healing technology has triggered the use of cells to overcome the limitations of conventional methods. Various commercially available allogeneic cell-scaffold complexes have been developed and widely used for treating diabetic foot ulcers (18–21). Recently, a new bioengineered dermal substitute composed of cultured autologous fibroblasts seeded on a hyaluronic acid sheet (Hyalograft 3D; ChaBio & Diostech, Seoul, South Korea) was developed. However, in several cases, autologous primary cells are not accessible or are unavailable in sufficient numbers because they may not have enough proliferative capacity for treatment in patients with diabetes due to cellular senescence in the presence of high glucose concentrations (22–24).

Therefore, stem cells may hold great promise for addressing the need for viable cell sources, and MSCs have attracted much attention in the bioengineering field. The therapeutic potential of allogeneic ASCs has been widely explored (11,12,25–27). However, most of these studies are in vitro or in vivo studies. Our study might be the first report demonstrating the efficacy of a commercially available allogeneic ASC in a scaffold for treating diabetic foot ulcers. Cultured ASCs are derived from an unrelated allogeneic donor. They can be stockpiled. A cell bank enables large-scale production, and an ASC sheet can be stored at −80°C. It is stable for 12 months. The allogeneic ASC sheet can maintain cell viability and potency up to 1 year and sustain its specifications during transport and storage. The sheet is available immediately and is convenient to use for physicians and patients. Unlike autologous ASCs, no biopsy is necessary. Patients do not need to wait for cells to be processed. The ASC sheet is nearly ready-made when the patient arrives at the clinic.

Prior to this phase II clinical trial study, a phase I trial was carried out to evaluate early safety of the treatment. For the phase I trial, five subjects were enrolled. There was no adverse event. Complete wound closure was achieved for all subjects within 8 weeks (unpublished observations). This phase II clinical trial involving allogeneic ASC products also showed promising results. At the 8th and 12th weeks after the treatment, more subjects in the treatment group had complete wound healing compared with those in the control group. Statistically, there were borderline differences. Other parameters demonstrated statistically significant differences. In detail, the mean time to complete wound healing was shorter for the treatment group compared with that for the control group (P = 0.033). The rate of wound size reduction was also statistically significant between the two groups at 9 of 12 weeks of study period. The results of post hoc analyses indicated that more subjects with Wagner grade 2 in the treatment group achieved complete wound closure compared with those in the control group.

The results of the 2-year follow-up study showed that ulcers recurred in two subjects in the treatment group 6 months after completion of the trial, although their ulcer locations were the toe tip and the plantar foot, areas that are vulnerable to pressure. Therefore, ulcer recurrence might not have been related to the ASC treatment. In addition, we observed no adverse events related to the treatment for 2 years after completion of the study.

The exact mechanism of action of allogeneic ASCs remains unknown. They might be able to exert their effects through several mechanisms, including the release of various growth factors such as vascular endothelial growth factor, hepatocyte growth factor, transforming growth factor-β1, insulin-like growth factor-1, epithelial growth factor, and keratinocyte growth factor to promote angiogenesis, collagen synthesis, and epithelialization (11). ASCs may produce extracellular matrix proteins, induce proliferation of native fibroblasts, protect cells from a noxious environment to aid in wound healing, and enhance the regeneration of new tissue. In addition, ASCs have an anti-inflammatory effect by inhibiting T-cell activation, thereby inhibiting the cellular signaling pathway of immune cells and resulting in decreased inflammatory molecules such as tumor necrosis factor-α and interferon-γ. A combination of these processes is likely to be involved in their action mechanisms (11,27,28).

Diabetic foot ulcers present a difficult treatment problem because the pathophysiology of the condition involves multiple factors such as peripheral neuropathy, peripheral vascular disease, repetitive trauma or pressure, and superimposing foot infection. However, many diabetic foot ulcers are delayed or fail to heal despite standard treatment such as debridement, infection control, pressure off-loading, and lower extremity revascularization because of attenuated activities of cells contributing to wound healing. In patients with diabetes, delayed wound healing may result in serious complications such as osteomyelitis and major/minor amputation (29–32); the longer the ulcer persists, the greater the possibility that serious complications will develop in the patient that can lead to hospitalization, and delayed wound healing may increase medical, economic, and social burdens. Therefore, efforts to reduce the time to complete wound closure are important for treating diabetic foot ulcers. In the current study, subjects in the treatment group showed significantly faster complete wound closure than did those in the control group (P = 0.033). This study suggested that the time to complete wound closure could be reduced simply by applying allogeneic ASC sheets in patients with diabetes. This is the key highlight of our study.

In this study, the follow-up duration was 12 weeks because many previous similar clinical trials for diabetic wound healing also used 12 weeks for the follow-up duration (13,14,18,32–35). For precise assessment of the efficacy of the allogeneic ASC treatment, we excluded external confounding factors that could have affected the treatment such as vascular insufficiency and infection. Therefore, 12 weeks might be enough to assess the time to complete wound closure in patients with diabetes after standardized diabetic foot ulcer management. It is important to emphasize that allogeneic ASC treatment must be used along with other standard principles of diabetic foot ulcer management, including debridement, infection control, pressure off-loading, and revascularization (21,36). Without adhering to these important principles, adding an active adjunctive modality is unlikely to result in improved healing rates. Patients with diabetic foot ulcers who do not exhibit significant signs of wound healing despite good metabolic control, acceptable vascularity (transcutaneous oxygen pressure >40 mmHg), adequate pressure off-loading, and absence of infection might be good responders to allogeneic ASC treatment.

The current study has some limitations. First, our study had all the limitations inherent in a phase II trial. For example, the sample size was relatively small although the number of subjects was determined by power calculation. Although acknowledging the implications of multiplicity adjustments was important for helping to interpret the trial results, we did not consider correcting the additional significance level using adjustment for α inflation because multiplicity adjustments may not be necessary in exploratory trials (37,38). Among the parameters of this study, differences in the proportion of subjects with complete wound closure were statistically at borderline. Furthermore, the percentage of subjects with a history of amputation, a risk factor for replication, in the treatment group was more than two times that in the control group. On the contrary, more older subjects were enrolled in the control group compared with the treatment group. Therefore, a phase III trial study should be followed with large sample size of subjects to establish the effect of allogeneic ASCs with certainty. Second, wound evaluation was performed in a single-blind fashion. Wound evaluators knew whether or not patients were in the ASC-treated group, which might have affected the results. However, this was unavoidable because the wound evaluators could discriminate between ASC sheets and polyurethane films. Therefore, it was difficult to conduct a double-blind design in this study. To increase transparency, our design included careful checks of the extent to which evaluators were blinded to group allocation. Third, further studies are necessary to determine the optimal interval for ASC application. The application of ASCs more frequently than once a week might reduce the time to healing.

In summary, the results of this study showed that an allogeneic ASC sheet might be effective and safe to treat nonischemic diabetic foot ulcers without infection.

Acknowledgments. The authors thank all patients and health care professionals who contributed to make the trial possible. In addition, the authors thank the investigators for their commitment, time, and effort.

Funding and Duality of Interest. This study was supported by grants from Anterogen (Seoul, South Korea) and the Ministry of Health & Welfare, Republic of Korea (grant HI16C1037). This research also was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. K.-C.M. had direct access to original data, interpreted the data, critically revised the draft of the manuscript, wrote the final version of the manuscript, and approved the final manuscript. H.-S.S. had direct access to original data; contributed to data recording; identified, treated, and monitored study participants; critically revised the draft of the manuscript; and approved the final manuscript. K.-B.K. had direct access to original data, contributed to data analyses and handling, critically revised the draft of the manuscript, and approved the final manuscript. S.-K.H. coordinated the work; had direct access to original data; contributed to data recording; interpreted the data; identified, treated, and monitored study participants; critically revised the draft of the manuscript; and wrote and approved the final version of the manuscript. K.-W.Y. and J.-W.L. identified, treated, and monitored study participants; had direct access to original data; contributed to data recording; critically revised the draft of the manuscript; and approved the final manuscript. M.-H.K. was a researcher and supervised laboratory analyses of this study. K.-C.M. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at PRS Korea 2018: the 76th Congress of the Korean Society of Plastic and Reconstructive Surgeons and the 2nd Asian Forum for Fat & Stem Cells, Seoul, South Korea, 9–11 November 2018; the 6th World Congress for World Association for Plastic Surgeons of Chinese Descent, the 2018 Asian Pacific Plastic and Reconstructive Surgery Forum, and the 2018 Annual Meeting of Taiwan Society of Plastic Surgery, Taipei, Taiwan, 29 November to 1 December 2018; and the 8th Congress for the Korea Association of Stem Cell and Tissue Regeneration, Seoul, South Korea, 2 December 2018.