For the better part of a decade, a handful of us have argued that a diabetic foot ulcer is less an acute wound than a chronic, relapsing, life-limiting disease — one that behaves a lot like cancer. Five-year mortality after a DFU rivals many solid tumors. “Healed” is not “cured”; it is remission. We surveil, we don’t discharge. That reframing changed how we counsel patients and how we structure follow-up.
Windy Cole and colleagues have now taken the analogy one important step further. If we’re going to think about these wounds like cancer, they ask, why not test our therapies like cancer, too?
Their manuscript in the Journal of Wound Care lays out a master-protocol basket trial for hard-to-heal wounds — an architecture lifted straight from oncology, where it has produced a run of tissue-agnostic drug approvals (pembrolizumab for mismatch-repair–deficient tumors; larotrectinib and entrectinib for NTRK-fusion cancers, among others). The logic in cancer is simple and powerful: if a drug targets a molecular alteration shared across many tumor types, you can evaluate it across all of them under one protocol rather than running a dozen siloed trials.
Cole et al. make the parallel case for wounds. Diabetic foot ulcers and venous leg ulcers are sorted by cause, but they share a common pathobiology — impaired angiogenesis, stalled inflammation, and biofilm that won’t quit. A therapy that hits those shared mechanisms should, in principle, work across aetiologies. So instead of two separate, from-scratch trials — two buckets — they propose one master protocol with two parallel baskets (Basket A: DFUs; Basket B: VLUs) sharing infrastructure, standardized care, blinding, and central adjudication, while each basket is still analyzed on its own terms.
The first tenant of this framework is a non-thermal bioactive therapy — cold atmospheric plasma — chosen because its proposed actions (antimicrobial, anti-biofilm, immunomodulatory, pro-angiogenic) map neatly onto those shared barriers. Adults whose wounds fail to budge during a standardized two-week run-in are randomized within each basket to standard care plus active therapy or standard care plus a convincingly engineered sham. The primary endpoint is the one that matters to patients and payers alike: complete closure by 12 weeks, confirmed at two consecutive visits and centrally adjudicated from standardized photography.
The statistical engine is where it gets interesting. Rather than the usual even split, the design randomizes 2:1 in favor of active therapy (60 active, 30 concurrent controls per basket; 180 total) and then leans on Bayesian power priors to borrow strength from historical controls — anchored to the classic Margolis meta-analysis of standard-care DFU healing (~24% closure at 12 weeks). Discounted at α₀ = 0.40, that borrowing is worth roughly 60 additional “virtual” control patients per basket, letting the trial expose more people to the investigational therapy without surrendering rigor. A basket is declared a winner only if the posterior probability that the therapy beats standard care by at least 10 percentage points exceeds 0.975 — and the trial is sized to detect a larger, 20-point effect that clears that bar by a comfortable margin.
To their credit, the authors don’t oversell it. They’re refreshingly candid about the seams. Borrowing a DFU-derived historical prior for the VLU basket is an assumption, not a fact, so they pre-specify a sensitivity analysis with no borrowing at all for VLUs. Each basket is analyzed independently, which means the design does not formally test whether the treatment effect is consistent across aetiologies — a cross-basket EXNEX analysis is included, but explicitly as exploratory. And unlike oncology, where a basket is anchored to a crisp molecular target, wound biology is messier and less discretely targetable. That messiness is the whole reason the sensitivity analyses are baked in rather than bolted on.
What I like most is that this paper lets the metaphor grow up. We borrowed remission and survivorship from oncology to change the conversation at the bedside. Cole and colleagues are borrowing the trial machinery to change how we generate the evidence — with the FDA’s own Complex Innovative Design pilot, its master-protocol guidance, and its recent embrace of Bayesian methods providing the regulatory runway. If it works, the same skeleton could be reused to test the next dressing, growth factor, or cell therapy across wound types instead of reinventing a trial each time. That’s exactly the kind of shared, efficient infrastructure the Diabetic Foot Consortium and the rest of us have been circling.
It’s also an on-brand reminder that the best ideas come from putting unlike things — and unlike people — in one room: a wound-care clinician-scientist and her collaborators from the wound-research world, oncology trial design, and Bayesian statistics, out of which comes a scalable template for wounds that have resisted almost everything else. Nicely done, Dr. Cole and team!
Cole W, Tunyiswa Z, Docter M. Borrowing from the cancer playbook: a master protocol basket trial for hard-to-heal wound therapeutics targeting shared pathobiology. J Wound Care. 2026;35(Sup7a):S23–S31. https://doi.org/10.12968/jowc.2026.0203
#ActAgainstAmputation #WoundCare #WoundHealing #DFU #VLU #DiabeticFoot #LimbSalvage #LimbPreservation #BasketTrial #MasterProtocol #ClinicalTrials #BayesianMethods #ColdAtmosphericPlasma #Remission #Survivorship #Cancer #FDA @ALPSLimb
Leave a Reply