Many diabetic foot ulcer (DFU) therapies look promising in the laboratory and in early studies, then struggle to show clear benefit in randomized trials. Three problems explain most of the gap: how well control arms heal, which endpoint is chosen, and how the trial is built. This article explains each and what developers can do about them.
Short answer: Diabetic foot ulcer clinical trials fail mainly because control arms heal at meaningful rates, endpoints may not capture durable benefit, and designs leave too much variability. Standard care alone heals roughly a third of ulcers within 12 to 24 weeks, so a new therapy must show a clear gain over a well-run control, not just biological activity.
Why Diabetic Foot Ulcer Clinical Trials Are Difficult
A DFU is not one disease. Two patients with the same ulcer area can have very different prospects. Healing depends on:
- Ulcer size and duration: Larger, older ulcers heal less often.
- Perfusion: Ischemia limits healing regardless of treatment.
- Infection: Active infection, including bone involvement, changes both risk and care.
- Neuropathy and pressure: Loss of protective sensation allows repeated injury.
- Glycemic control and comorbidity: Kidney disease, smoking and nutrition all matter.
- Offloading and adherence: Pressure relief works only if patients use it.
- Prior therapy: Patients who failed earlier treatments differ from newly diagnosed ones.
This variability widens the spread of outcomes in every arm. When spread is wide, a small or loosely controlled trial cannot separate a real treatment effect from chance. A 2025 scoping review of 527 DFU trials in Frontiers in Endocrinology found that most were small, many were open-label, and most did not report a follow-up period. Those features make single results hard to trust.
The International Working Group on the Diabetic Foot (IWGDF) reached a similar view in its 2023 wound-healing guideline: evidence certainty is improving for some interventions but remains poor overall, and the authors called for "not more, but better quality trials" (IWGDF, 2024).
The Control-Arm Problem in DFU Trials
Control groups in DFU trials are not untreated. They receive standard care: sharp debridement, offloading, infection management, vascular assessment and revascularization where appropriate, moisture-balancing dressings and regular follow-up. That care works for many patients. For the standard-care landscape, see Diabetic Foot Ulcer Treatment in US: Current Options, Gaps and What Comes Next.
Pooled data show how much:
- A 1999 meta-analysis of neuropathic DFU control arms found 24.2% healed by 12 weeks and 30.9% by 20 weeks (Margolis et al., 1999).
- A 2020 pooled analysis of 16 randomized trials reported about 33% healing at 12 weeks and 43% at 20 weeks (Parks et al., 2020).
- A 2025 meta-analysis of 32 trials estimated about 33% healing over 12 to 24 weeks (Coye et al., 2025).
Why control arms improve
Several forces push control-arm healing up and shrink the apparent treatment effect:
- Better care inside trials. Protocols standardize debridement and offloading, and frequent visits improve adherence.
- Regression to the mean. Patients are enrolled when an ulcer looks worst, so measurements tend to improve on their own.
- Natural variability. Some ulcers heal without any added therapy.
- Run-in periods. Excluding ulcers that shrink during a standard-care run-in removes fast healers, which lowers control healing but also changes the population.
A practical example
Suppose a trial assumes 25% of control ulcers will heal and plans for a treated arm at 45%. If standard care is delivered well and 40% of controls heal, a treated arm that reaches 50% is a 10-point difference, not 20. A study sized for the larger gap is then underpowered for the smaller one, and a therapy with a real but modest effect is read as a failure. This example uses hypothetical numbers.
The lesson is that a therapy must beat an adequately designed control, not merely show activity. Early-phase estimates drawn from weak or historical controls often overstate what a later trial will find.
Endpoint Mistakes That Cause DFU Trial Failure
FDA's guidance on chronic cutaneous ulcers recommends complete wound closure, confirmed at two consecutive visits about two weeks apart, as the principal efficacy measure, with follow-up of at least three months after closure. It states that partial healing may be used in early studies if prespecified, but it is not sufficient as a Phase 3 primary endpoint (FDA, 2006). The guidance sets no fixed trial duration.
Common DFU trial endpoints: strengths, limitations and best use
| Endpoint | Strengths | Limitations | Best use |
|---|---|---|---|
| Complete closure at a prespecified time | Clear, patient-relevant, matches FDA's stated primary measure | Needs enough duration and sample size; sensitive to control-arm healing | Pivotal efficacy trials |
| Percent wound-area reduction | Continuous, efficient, early signal | Partial healing may not translate to closure; measurement error | Phase 2 signal, mechanism checks |
| Time to complete healing | Uses timing as well as closure; handles variable follow-up | Needs clear rules for censoring amputations, deaths, dropouts | Supportive or co-primary analysis |
| Sustained healing / durability | Shows closure lasts | Requires longer follow-up and more cost | Confirmatory follow-up phase |
| Recurrence | Captures long-term benefit | Needs large, long trials; depends on footwear and adherence | Long-term follow-up studies |
| Amputation | Highly patient-relevant | Rare events over short windows; very large samples | Large outcomes trials, composites |
| Infection outcomes | Relevant to safety and complications | Definitions vary; separate from healing | Safety and secondary analyses |
| Patient-reported outcomes / quality of life | Reflects patient experience | Rarely blinded; limited validated DFU tools | Secondary, supportive |
| Composite endpoints | Raise event counts | Mix outcomes of unequal importance; hard to interpret | Outcome-oriented designs, prespecified |
Why area reduction alone is not enough
Wound-area reduction is a useful early signal. It is continuous, so it needs fewer patients, and it can show whether a mechanism is doing something. But a smaller ulcer is not a closed ulcer, and the 2006 FDA guidance does not treat partial healing as sufficient for Phase 3. A trial that relies on percent area reduction at a short time point can look positive while closure is unchanged.
Why complete healing is demanding
Complete closure matters to patients and regulators, but it is a binary outcome. Binary outcomes need more participants than continuous ones, longer follow-up, and clear rules for amputations, deaths and discontinuations. Amputation and recurrence matter even more to patients, yet events are fewer, so trials need to be larger and longer. The 2025 scoping review found amputation was rarely the primary endpoint.
Matching the endpoint to the program
The right endpoint depends on mechanism, population, setting and regulatory goal. A therapy meant to speed granulation may be read at an early timepoint in Phase 2, then tested for confirmed closure in Phase 3. A program aimed at preventing recurrence needs an entirely different time horizon.
Trial-Design Failures in the Diabetic Foot Ulcer Pipeline
Design flaws can erase a real effect or create a false one. In the diabetic foot ulcer pipeline, the same issues recur.
Common DFU trial-design risks and mitigation strategies
| Design risk | Why it matters | Mitigation |
|---|---|---|
| Underpowered study | Misses real but modest effects | Power for a realistic effect given control healing |
| Weak randomization or allocation concealment | Selection bias | Central randomization, concealed allocation |
| No blinding or assessor blinding | Bias in measurement and care | Placebo or vehicle where feasible; blinded central assessment |
| Baseline imbalance (area, duration, perfusion, infection) | Confounds results | Stratify or minimize on key prognostic factors |
| Inconsistent wound measurement | Noisy area and closure data | Standardized digital imaging, central review |
| Short follow-up | Misses durability and recurrence | Confirm closure; follow after closure |
| Dropout and missing data | Biases effect estimates | Retention plans; prespecified handling |
| Protocol deviations, unequal co-interventions | Unequal background care | Standardize and audit care across arms |
| Single-center design | Limited generalizability | Multicenter with independent oversight |
| Post-hoc endpoint changes or analyses | Inflates false positives | Prespecify endpoints, subgroups and sensitivity analyses |
Two points deserve emphasis. First, blinding is hard: the 2025 scoping review found many trials were open-label and relatively few were placebo-controlled, partly because credible placebos and ethical limits on withholding care complicate the design. Where blinding is impractical, FDA's guidance points to blinded third-party assessment of wound status (FDA, 2006).
Second, standard care itself is often poorly defined. A 2024 Journal of Wound Care paper on DFU trials for advanced therapies noted ill-defined standard of care and the use of dressings and offloading that were not standard. It described one trial that required a four-week run-in with 30 days of standard care, confirmed debridement and total contact casting for plantar ulcers (Serena et al., 2024).
Failure is not hypothetical. In 2016, a Phase 3 trial of a topical PDGF spray enrolled 252 patients with chronic DFUs, compared it with placebo spray on top of standard care for up to 20 weeks, and missed its primary endpoint of complete closure. That older landmark shows how a growth-factor mechanism with preclinical support can still fall short against a well-run control.
How to Design a More Robust DFU Clinical Trial
No design removes all risk, but the following framework addresses the failure modes above.
- Define a narrowly characterized population. Specify ulcer size and duration range, perfusion thresholds, infection status and prior therapy.
- Standardize background care across arms. Define debridement, offloading and dressing protocols, and audit adherence.
- Select a primary endpoint aligned with benefit and regulators. For pivotal work, confirmed complete closure is the stated FDA measure.
- Use blinded, centralized, standardized wound assessment. Digital imaging and independent reading reduce measurement bias.
- Predefine handling of amputations, deaths, discontinuations, recurrence and missing data.
- Power for a realistic effect. Base sample size on contemporary control-arm healing, not optimistic assumptions.
- Follow long enough. Confirm closure and observe durability and, where feasible, recurrence.
- Use multiple sites and independent oversight. A data monitoring committee protects integrity.
- Prespecify analyses. Lock endpoints, subgroups and sensitivity analyses before unblinding.
- Balance innovation with feasibility. The most rigorous design is useless if it cannot enroll or finish.
What This Means for Novel Diabetic Foot Ulcer Therapeutics
The pipeline is broad. Developers are working on biologics and growth factors, cell-based and tissue products, gene-based approaches, devices such as oxygen and negative-pressure therapies, and topical small molecules (Exploration of Drug Science, gene-based therapeutics review). Each class faces the same trial problems, with its own twists: biologics and cell products carry manufacturing and handling complexity, devices are hard to blind, and combination products multiply the number of variables. For how the therapy classes compare, see chronic wound therapy classes.
For novel diabetic foot ulcer therapeutics, especially first-in-class wound healing candidates, the translational path needs to be explicit. Preclinical wound-closure data in animal models are a starting point, not an endpoint. They do not tell a sponsor how much a candidate will add on top of modern standard care in people. Claims about mechanism must eventually be tested with clinically meaningful endpoints, in a population chosen to show an effect.
ZegeneX is working in this area. It is developing Zegen-X5, an investigational, preclinical-stage, shelf-stable topical small-molecule candidate, as part of its focus on first-in-class wound-healing therapeutics (novel wound-healing mechanism). Zegen-X5 has not been approved by the US FDA, and it has not been tested in human DFU trials. See the clinical development roadmap.
Key Takeaways
- Control arms in DFU trials often heal about a third of ulcers within 12 to 24 weeks, which narrows the room for a treatment effect.
- Confirmed complete closure is FDA's stated primary measure; percent area reduction is an early signal, not a substitute.
- Amputation, recurrence and quality of life matter, but need larger, longer trials.
- Weak blinding, poor wound measurement and undefined standard care weaken many trials.
- Power calculations should use realistic control-arm healing, not optimistic assumptions.
- Prespecified analyses and clear handling of missing data protect against false conclusions.
- A promising mechanism does not guarantee clinical success; design and execution decide it.
Frequently Asked Questions
Why do diabetic foot ulcer clinical trials fail?
Mainly because control arms heal at meaningful rates, endpoints do not capture durable benefit, and designs are underpowered, unblinded or too short. Variable populations and unequal background care add noise.
Why do control groups heal in DFU trials?
They receive standard care, including debridement, offloading and infection control, and are seen often. Regression to the mean and natural variation also play a role. Pooled control arms healed about 24% to 33% by 12 weeks across analyses.
What is the best primary endpoint for a DFU clinical trial?
For pivotal trials, FDA's guidance recommends confirmed complete wound closure. The best choice still depends on the mechanism, population and regulatory goal.
Is wound-area reduction a valid DFU trial endpoint?
It is a useful early signal in Phase 2, but FDA's guidance does not consider partial healing sufficient as a Phase 3 primary endpoint.
How long should a diabetic foot ulcer clinical trial follow patients?
FDA sets no fixed duration, but recommends follow-up of at least three months after closure. Trials aiming to show recurrence or amputation benefit need longer.
What is the diabetic foot ulcer pipeline?
It is the set of DFU therapies in development, including biologics, cell and tissue products, gene-based approaches, devices and small molecules, at stages from preclinical to Phase 3.
How can sponsors improve DFU trial design?
Define the population tightly, standardize care, use blinded central wound assessment, power for realistic control healing, prespecify analyses and follow patients long enough.
This article is for informational purposes only and is not medical or investment advice.
Sources
- FDA, Chronic Cutaneous Ulcer and Burn Wounds: Developing Products for Treatment (June 2006)
- IWGDF 2023 wound-healing guideline, Diabetes Metab Res Rev 2024;40(3):e3644
- Serena, Cole, King. A model for diabetic foot ulcer clinical trials on advanced therapies. J Wound Care 2024
- Margolis et al., Diabetes Care 1999;22(5):692
- Parks et al., Diabetes 2020 (supplement abstract); Coye et al., Wound Repair Regen 2025
- Frontiers in Endocrinology 2025, scoping review of 527 DFU trials
- GEN News, Phase III failure of a topical PDGF program (2016)
- Exploration of Drug Science, Emerging targets and gene-based therapeutics in DFU
- ClinicalTrials.gov (registry of clinical studies)
Translational Innovation from the University of Pittsburgh
ZegeneX Therapeutics is advancing first-in-class small-molecule regenerative wound healing therapeutics licensed from the University of Pittsburgh to eliminate avoidable diabetic amputations worldwide.