Bruna Nakayama¹, BS; Brock A. Liden², DPM, FAPWCA; Kylie Griffey¹, BS; *Brianna Tramelli¹, BS; Thomas Serena¹, MD

¹SerenaGroup, Inc., Norwell, MA 02061, USA; ²Wound, Ankle, Foot, and Leg, Circleville, OH, USA

* Correspondence: btramelli@serenagroups.com

ABSTRACT

Background:

Chronic diabetic foot ulcers (DFUs) remain difficult to resolve despite standard wound care. Real-world evidence describing complete wound closure and the temporal pattern of wound resolution may provide clinically relevant information regarding treatment performance outside randomized clinical trials. This study compared wound closure outcomes among patients with DFUs treated with ON101 plus standard wound care with an external historical standard-of-care (SOC) cohort.

Methods:

This multicenter retrospective observational study included 20 DFUs from 19 US patients treated with ON101 plus standard wound care and 40 de-identified DFUs from a nonconcurrent historical SOC cohort. The primary endpoint was complete wound closure by Week 12.

Secondary and exploratory outcomes included serial closure rates, Week 4 percent area reduction (PAR), achievement of ≥50% PAR at Week 4, and time to complete closure. Week 4 PAR was analyzed among evaluable wounds, with a missing-as-nonresponder sensitivity analysis. Between-cohort comparisons were performed for continuous outcomes and categorical outcomes. Time to complete closure through Week 12 was evaluated using time-to-event methods. In the ON101 cohort, the positive predictive value (PPV) of achieving ≥50% PAR at Week 4 for subsequent complete wound closure by Week 12 was also assessed as an exploratory prognostic measure.

Results: 

Complete wound closure by Week 12 was achieved in 11 of 20 ON101-treated DFUs (55.0%; 95% CI, 34.2%–74.2%) compared with 5 of 40 DFUs (12.5%; 95% CI, 5.5%–26.1%) in the external historical SOC cohort (p=0.0012). In an exploratory available-case analysis, mean Week 4 percent area reduction (PAR) was 86.0 ± 23.6% in the ON101cohort versus 8.6 ± 84.0% in the historical SOC cohort (p=0.00006). Among DFUs with evaluable Week 4 PAR, 9 of 10 ON101-treated DFUs (90.0%) achieved ≥50% PAR compared with 13 of 40 historical SOC DFUs (32.5%); however, Week 4 PAR was evaluable in only half of the ON101-treated DFUs. Kaplan–Meier analysis demonstrated earlier accumulation of complete closure events in the ON101 cohort through Week 12 (log-rank p<0.0001).

Conclusions:

In this US real-world retrospective comparative study, treatment with ON101 plus standard wound care was associated with a higher observed Week 12 closure rate, while exploratory available-case analyses suggested greater early wound-area reduction. The consistency of the Week 12 closure finding with serial closure assessments and time-to-event analysis provides supportive real- world evidence of an earlier observed pattern of wound resolution. Because the comparator was nonconcurrent and baseline differences were present, the findings should be considered exploratory and warrant confirmation in prospective controlled studies.

Keywords: diabetic foot ulcer; ON101; real-world evidence; retrospective observational study; standard wound care

Diabetes represents a substantial and growing public health burden.^1,2 In 2022, the estimated cost of diagnosed diabetes in the US was $412.9 billion USD, including $306.6 billion in direct medical costs and $106.3 billion in indirect costs and associated conditions.³ Further modeling suggests that, as the population affected by diabetes continues to expand, the associated annual medical and societal burden could exceed $622 billion by 2030.⁴ This disease is strongly associated with the development of chronic lower-extremity wounds,^5,6 with diabetic foot ulcers (DFUs) as one of the most serious complications of diabetes.^7,8 DFUs are a leading cause of infection, hospitalization, and lower extremity amputation.^8,9 DFU arises from multiple factors, including peripheral neuropathy, reduced blood flow, impaired immune response, and ongoing inflammation, all of which interfere with the normal wound-healing process.^7,10 Chronic wounds frequently become stuck in the inflammatory phase of healing, characterized by impaired angiogenesis, abnormal extracellular matrix remodeling, and diminished keratinocyte migration.^11,12 These biological disturbances significantly delay tissue repair and contribute to the persistence of non-healing ulcers.^11–13

The standard of care (SOC) management for DFUs involves a combination of wound cleansing, surgical debridement, infection control, moisture balance, and mechanical off-loading of the affected foot,^14,15 components that are commonly described within the TIME framework, which includes tissue management, infection control, moisture balance, and advancement of the epithelial edge.^15,16 However, despite these well- established approaches, healing outcomes remain limited with SOC alone.^17–19 Published studies reported that only 24–33% of DFUs achieve complete closure within 12–20 weeks with SOC alone.6,^17–20 The 5-year mortality rate for patients with DFUs stands at 30.5%, but increases to approximately 70% following a major amputation.^21,22 In a study evaluating risk factors for DFUs in 4,709 people, 977 (20.7%) required hospitalization, and among them, 669 (68.5%) worsened to require amputations,²⁴ highlighting the need for therapies that enhance the wound environment and wound characteristics.

Meaningful care in DFUs should not be limited to complete closure alone but should also include measurable improvements in wound bed condition. Maintenance of an optimal moist wound environment is a critical determinant of successful tissue repair.²⁵ Adequate hydration supports keratinocyte migration, promotes angiogenesis, and facilitates extracellular matrix deposition while preventing desiccation and secondary tissue damage.^25,26 Topical formulations capable of maintaining this environment may therefore support wound closure.^25,26

The topical agent (ON101, Oneness Biotech Co., Ltd., Taiwan) evaluated in this study is a water-based, semi-viscous matrix formulation designed for direct application to the wound bed.^20,23 The formulation contains proprietary Centella asiatica extract and Plectranthus amboinicus (Mexican mint extract)²⁷ which creates an interface layer over the wound surface, helping to maintain a moist microenvironment conducive to granulation and re-epithelialization.^28–30 Thereby making the topical matrix that is a water-based, semi-viscous and semi-solid agent by physically maintaining moisture balance and supporting wound-healing processes.^20,23

Accordingly, this retrospective comparative study evaluated complete wound closure in patients with chronic DFUs treated with ON101 plus standard wound care in routine clinical practice in the US using an external historical SOC cohort as a nonconcurrent comparator. The primary comparative endpoint was complete wound closure by Week 12. Week 4 percent area reduction (PAR) was evaluated as an exploratory measure of early wound response before the definitive closure assessment. Additional supportive and exploratory analyses examined achievement of the ≥50% Week 4 PAR threshold, serial complete-closure rates, time to complete closure, wound-bed characteristics, and the relationship between observed early PAR response and subsequent Week 12 closure. This framework was intended to characterize both the definitive clinical outcome of complete wound closure and the temporal pattern of wound progression preceding that outcome.

MATERIALS AND METHODS

STUDY DESIGN

This study was a multicenter retrospective observational analysis conducted in the US to evaluate the clinical effectiveness of ON101 in the management of chronic DFUs. The analysis included patients treated with ON101 in conjunction with standard wound care and an external historical SOC comparator cohort comprising 40 de-identified DFU records derived from previously collected SOC clinical trial data. The historical SOC cohort was nonconcurrent with the ON101 cohort, and treatment allocation between the two cohorts was not randomized within the present study. Cohort flow diagram is shown in Figure 1. The 40 historical SOC records were selected retrospectively by the participating physician after follow-up of the ON101 cohort had occurred.  The total number of potentially eligible records in the historical source population was not available for the present analysis, and no contemporaneous screening log or prespecified record-selection algorithm was available. The comparator records therefore cannot be considered a documented consecutive series.  Considering that record selection occurred retrospectively, clinical outcomes may have been known at the time of selection; accordingly, potential outcome-informed selection cannot be excluded.

Figure 1. Flow of ON101-treated patients and diabetic foot ulcers into the analytic cohort. No prospective screening log was maintained.

The ON101 cohort was evaluated under a post-marketing retrospective research protocol (VRA003A, Version A, dated January 25, 2024) using clinical data collected during routine wound care. Protocol VRA003A was finalized before retrospective record identification, selection, and data extraction. The cohort included patients treated with ON101 at participating U.S. sites between July 2023 and January 2025, with clinical follow-up available through January 2025. The ON101 cohort included three participating centers, consisting of two clinics and one hospital-based practice, located in Ohio and Wisconsin, USA. The three sites contributed 8 DFUs from 8 patients, 6 DFUs from 6 patients, and 6 DFUs from 5 patients, respectively. Sites were selected as a convenience sample of collaborating clinicians experienced in chronic wound management who used ON101 in routine post-marketing practice and agreed to provide the protocol-specified clinical information. At two sites, clinicians identified potentially eligible records using the protocol-defined criteria; at the third site, records were identified after Oneness provided training on the same eligibility criteria and data-collection requirements. No formal protocol-specific clinical training was provided for ON101 application, wound assessment, or routine patient care; the product training at the third site addressed retrospective case identification and the protocol-defined data requirements. The cases were not identified through a formally documented consecutive screening process; inclusion was based on protocol eligibility and availability of the required clinical documentation. No additional centralized standardization of treatment or wound-assessment procedures was implemented beyond the protocol-defined retrospective data requirements. Photography and wound-area measurement were performed using each site's routine equipment and documentation methods.

Considering that this was a retrospective post-marketing study, potentially eligible cases were identified from existing ON101-treated clinical records rather than through prospective screening, and a separate prospective screening log was not maintained. A total of 19 patients with 20 DFUs were identified as meeting the protocol-defined eligibility criteria and having the clinical documentation required for analysis. No identified eligible ON101-treated cases were subsequently excluded; all 19 patients with 20 DFUs comprised the analytic cohort. Outcome-dependent case selection cannot be completely excluded because wound outcomes may have been known for some retrospectively identified records. The data cutoff for the present analysis was January 2025; no clinical outcomes occurring after this date were included. The historical SOC dataset was not assembled prospectively under Protocol VRA003A, and equivalence of all protocol eligibility criteria could not be established retrospectively. Instead, comparability was evaluated criterion by criterion using variables available in the historical dataset. Criteria for which corresponding historical information was unavailable or incomplete were classified as not verifiable rather than assumed to have been met.

For the ON101 cohort, patient demographics, ulcer characteristics, treatment exposure, wound-bed characteristics, wound-area progression, and wound closure outcomes were assessed where available.  
Comparative analyses between the ON101 and historical SOC cohorts were conducted for predefined wound outcomes using variables available in both data sources. Baseline characteristics of the historical SOC cohort were summarized using aggregate-level descriptive statistics in accordance with data-use restrictions, whereas comparative outcome analyses were based on the available de-identified data. Given the nonconcurrent historical comparator design, between-cohort findings were interpreted as exploratory associations rather than causal estimates of treatment effects.

Table 1. Comparison of Key Design and Outcome-Ascertainment Characteristics of the ON101 and Historical SOC Cohorts.

ETHICAL APPROVAL AND PATIENT CONSENT

The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and applicable regulatory requirements for retrospective research involving human subjects. The study protocol (Version A, dated 25 Jan 2024) received approval from the WCG Institutional Review Board (IRB) under Study Number 1372521. The IRB granted a waiver of informed consent under 21 CFR 50.22, as the  research involved minimal risk and relied exclusively on previously collected clinical data. All patient data were fully de-identified before analysis to ensure confidentiality and compliance with HIPAA privacy requirements.

STUDY POPULATION

The comparative study population comprised 60 DFUs, including 20 DFUs from 19 patients treated with ON101 and 40 DFU records from the external historical SOC comparator cohort.

The ON101 cohort included 20 DFUs from 19 patients treated during routine wound care in the US. All patients were adults with diabetes and had ulcers below the malleolus present for at least four weeks before the treatment initiation. The mean (SD) age was 59.6 ± 13.6 years; 14 patients (73.7%) were male and 5 (26.3%) were female. Target ulcers were characterized according to the Wagner grade system, plantar or non- plantar location, and the presence of infection. Detailed patient demographics and clinical characteristics of the ON101 cohort are summarized in Table 2.

Table 2. Baseline Patient and Ulcer Characteristics of the ON101 Cohort.

The external historical SOC comparator cohort comprised 40 de-identified DFU records from a single outpatient clinical site in Ohio. The comparator data were provided for the present analysis in de-identified form and were summarized using aggregate-level descriptive statistics in accordance with applicable data-use restrictions. Mean (±SD) age was 60.6 ± 13.9 years among 36 patients with available age data, and sex was available for all 40 records, of which 28 (70.0%) were male and 12 (30.0%) were female. Baseline clinical characteristics included a mean HbA1c of 8.1 ± 1.8%, a mean baseline wound area of 5.1 ± 4.0 cm², and a mean ulcer duration of 22.2 ± 14.5 weeks among the 23 records with available duration data. These aggregate summaries are provided to characterize the historical comparator cohort without disclosure of individual-level baseline data. No formal between-cohort baseline comparison table was generated.

ELIGIBILITY CRITERIA

Eligibility criteria for the ON101 cohort are presented in Table 3. For the external historical SOC cohort, comparability with the ON101 eligibility criteria was assessed using the variables available in the historical dataset. Age was available for 36 of 40 historical records, and ulcer duration was available for 23 of 40 records.

Availability of other eligibility-related variables is reported where applicable. Eligibility criteria that could not be verified because the corresponding historical variable was unavailable or incomplete were not assumed to have been met. For efficacy analyses, the protocol-defined evaluable population included patients who received ON101 for at least four weeks, including those who achieved complete ulcer healing at any visit during the treatment period, and who had at least one post-treatment target-ulcer assessment.

Table 3. Inclusion and Exclusion Criteria for the ON101 Cohort.

TREATMENT PROCEDURE

ON101 was applied to the ulcer twice daily to cover the wound surface for up to 16 weeks or until complete wound closure. All patients received SOC for the management of the DFU, including cleansing, debridement when needed, infection control, off-loading, and appropriate dressings. No other topical agents or advanced wound products were used during treatment. Systemic antibiotics were used when clinically indicated. Other topical antibiotics, topical steroids, antiseptic soaps, growth factors, grafts, or investigational treatment were not applied to the target ulcer during the study period.

For the external historical SOC cohort, treatment consisted of standardized wound care treatment, including sharp debridement, offloading, dressings, infection management, and systemic risk factors management, such as glycemic control, comparable to the standard wound care in ON101 cohort.

DATA COLLECTION AND WOUND ASSESSMENT

Clinical data were retrospectively extracted from the medical records of patients and recorded in standardized case report forms (CRFs). The following categories of data were collected:

• Demographics and medical history

• Sex, height, body weight

• Smoking habits

• Relevant medical history within three months before treatment initiation

• Clinical assessments

• Area of the serial ulcer

• Wound bed characteristics

• Ulcer duration before treatment

• Off-loading device usage

• Photographic documentation of the target ulcer

• Laboratory and physiological parameters

• White blood cell count (WBC) and C-reactive protein (CRP) within one week before treatment

• HbA1c, albumin, creatinine, and hemoglobin within three months before treatment

• Vascular perfusion assessments using ankle-brachial index (ABI), transcutaneous oxygen measurement (TcPO₂), or Doppler ultrasound

• Treatment exposure

• Documentation of concomitant medications

• Record of prior therapies within four weeks before treatment

• Outcome Measures

The primary objective of this study was to compare complete wound closure by Week 12 between the ON101-treated cohort and the external historical SOC comparator cohort. For the ON101 cohort, the time origin for outcome assessment was from the first/baseline visit, with wound assessments performed weekly.
Complete wound closure was defined as complete re-epithelization without drainage or dressing requirement assessed by the investigators. For the historical SOC cohort, the corresponding time origin was from the first/baseline visit, with wound assessments performed weekly. For both cohorts, complete wound closure was defined as complete re-epithelialization without drainage or dressing requirement and was assessed by the treating physician/investigator based on routine clinical documentation. No protocol-required confirmatory assessment at a subsequent visit or blinded central adjudication of wound closure was performed. Week 12
was used as the aligned comparative time point for both cohorts.

Week 4 PAR was evaluated as an exploratory endpoint and was calculated from the change in wound area from baseline to the Week 4 assessment. This endpoint was used to quantify early wound-area response before the primary Week 12 closure assessment. An exploratory analysis examined the association between achieving
≥50% PAR at Week 4 and subsequent complete wound closure by Week 12; this analysis was descriptive and was not intended to validate ≥50% PAR as a prognostic threshold in this dataset. Supportive and exploratory endpoints included the proportion of DFUs achieving ≥50% PAR at Week 4, cumulative complete wound closure at Weeks 4, 6, 8, and 12, and time to complete wound closure. Wound-bed characteristics were evaluated as supportive descriptive outcomes and included granulation tissue coverage, slough, eschar, and re-epithelialization where available. Considering that these characteristics were not consistently captured using harmonized definitions and measurement scales across the ON101 and historical SOC datasets, wound-bed findings were not interpreted as independent comparative efficacy outcomes.

Exploratory subgroup analyses examined Week 12 complete wound closure according to ulcer location (plantar versus non-plantar). These analyses were considered hypothesis-generating, and potential effect modification by ulcer location was further assessed using formal interaction tests.

STATISTICAL ANALYSIS

Categorical variables were summarized as counts and percentages, and continuous variables as means with standard deviations (SDs) and medians with interquartile ranges (IQRs). Between-cohort comparisons were performed using Fisher’s exact test for categorical variables and the Mann–Whitney U test for continuous
variables. The DFU was the primary unit of analysis. A total of 20 ON101-treated DFUs were contributed by 19 patients, one patient contributed two DFUs; therefore, complete independence of all ulcer-level observations could not be assumed. Baseline characteristics were summarized descriptively without formal hypothesis
testing because individual-level baseline data were not uniformly available for the external historical SOC cohort.

The primary endpoint was complete wound closure by Week 12. Complete-closure proportions at Weeks 4, 6, 8, and 12 were presented as counts and percentages with Wilson score 95% CIs calculated without continuity correction. Between-cohort differences in closure proportions were assessed using Fisher’s exact
test.

Week 4 PAR was analyzed as an exploratory early-response outcome and was summarized among evaluable wounds using the mean and standard deviation, and the median and interquartile range. Between-cohort distributions were compared using the Mann–Whitney U test. Wounds were considered evaluable if they had either documented complete closure on or before Week 4, or observed baseline and Week 4 wound-area measurements. Documented complete closure on or before Week 4 was assigned a PAR of 100% as an endpoint-derivation rule. Accordingly, Week 4 PAR values did not uniformly represent paired baseline and Week 4 wound-area measurements, because wounds that had already achieved complete closure were assigned a PAR of 100%. The proportion of evaluable DFUs achieving Week 4 PAR ≥50% was compared between cohorts using Fisher’s exact test, with the odds ratio and corresponding 95% confidence interval reported.

Due to the absence of Week 4 PAR data for 10 of the 20 ON101-treated DFUs, a sensitivity analysis was performed in which wounds without evaluable Week 4 PAR were classified as nonresponders for the binary PAR ≥50% endpoint. This rule was applied only to the binary responder analysis; no continuous PAR values
were imputed.

Complete-closure proportions at Weeks 4, 6, and 8 were evaluated as supportive longitudinal outcomes. Time to complete wound closure through Week 12 was analyzed using the Kaplan–Meier method, and the ON101 and historical SOC cohorts were compared using the log-rank test. DFUs without documented closure
by Week 12 were censored at their last available wound assessment or at Week 12, whichever occurred first. Closure documented after Week 12 was administratively censored at Week 12. The cumulative probability of complete closure was presented as 1 − S(t). Kaplan–Meier median time to closure was reported when estimable; otherwise, it was reported as not reached. Observed time to closure among DFUs that achieved closure was summarized separately using the median and IQR. 

The prognostic relationship between observed early wound response and subsequent Week 12 closure was explored descriptively within the ON101 cohort. Among DFUs with evaluable Week 4 PAR, the positive predictive value of achieving PAR ≥50% for subsequent complete closure by Week 12 was calculated. A separate sensitivity analysis classified DFUs without evaluable Week 4 PAR as nonresponders and summarized the resulting association using an odds ratio and Fisher’s exact test.

Wound-bed characteristics and ulcer-location subgroup findings were considered supportive or exploratory analyses.

Propensity scores were estimated using baseline wound area and plantar ulcer location. The historical comparator dataset did not contain complete individual-level information for several clinically important prognostic variables, therefore, the propensity-score model was intentionally limited to covariates available with sufficient completeness and overlap across cohorts. Overlap weighting was applied, and covariate balance was assessed using standardized mean differences, with an absolute standardized mean difference of <0.10 indicating acceptable balance. Covariate balance before and after weighting, propensity-score overlap, and effective sample sizes are reported in further datasets that can be provided upon author request.

Effect modification by ulcer location was evaluated using formal interaction tests. For complete wound closure by Week 12, a logistic regression model including treatment group, ulcer location, and a treatment-by-location interaction term was fitted. For time to complete closure, a Cox proportional hazards model containing
the corresponding interaction term was used. Interaction p-values, rather than differences between subgroup-specific p-values, were used to assess heterogeneity of the observed treatment-group association. The Week 12 complete wound closure comparison was the primary analysis. All other analyses were considered
supportive or exploratory. No adjustment for multiplicity was performed; therefore, p-values from secondary, supportive, subgroup, sensitivity, and prognostic analyses were interpreted descriptively and were not considered confirmatory.

No continuous missing values were imputed, and available-case denominators were used for the Week 4 PAR analysis. Documented complete closure on or before Week 4 was assigned a PAR of 100% as an endpoint-derivation rule rather than as imputation. For the binary Week 4 PAR ≥50% sensitivity analysis only, wounds
without evaluable Week 4 PAR were classified as nonresponders. No interpolation or last-observation-carried-forward procedure was applied. All tests were two-sided, with P<0.05 considered statistically significant.

RESULTS

The comparative analysis included 60 DFUs, comprising 20 DFUs treated with ON101 and 40 DFUs from the external historical SOC comparator cohort. Comparative outcomes were assessed through Week 12, with additional descriptive follow-up of the ON101 cohort beyond Week 12 where available.

TREATMENT EXPOSURE AND CONCOMITANT WOUND CARE

Detailed longitudinal information regarding ON101 adherence, cumulative treatment exposure, frequency of debridement, duration of offloading, systemic antibiotic use, temporary treatment interruptions, discontinuations, and other concomitant wound interventions could not be reconstructed reliably and consistently across the retrospective records. These variables were therefore not summarized quantitatively. 

PRIMARY ENDPOINT: COMPLETE WOUND CLOSURE AT WEEK 12

By Week 12, complete wound closure was observed in 11 of 20 DFUs (55.0%) in the ON101 cohort and 5 of 40 DFUs (12.5%) in the historical SOC cohort, corresponding to an absolute risk difference of 42.5 percentage points (95% CI, 17.6–62.9) and a risk ratio of 4.40 (95% CI, 1.77–10.94; Fisher’s exact p=0.0012). The observed Week 12 closure rates are shown in Figure 2.

In an exploratory propensity-score overlap-weighted sensitivity analysis of the full analytic cohort (ON101, n=20; historical SOC, n=40), estimated Week 12 complete closure rates were 55.3% and 12.0%, respectively,  corresponding to an adjusted risk difference of 43.2 percentage points (95% bootstrap CI, 10.6– 71.5) and an adjusted risk ratio of 4.59 (95% bootstrap CI, 1.55–26.46). Adequate balance was achieved for both covariates included in the propensity-score model, with post-weighting absolute standardized mean differences <0.10.

Figure 2. Comparison of Complete Wound Closure Rate at Week 12 between the ON101 and Historical
SOC Cohorts.