Chronic wounds in diabetic patients are one of the costliest quietly growing burdens in medicine. Roughly 15 to 25 percent of people with diabetes will develop a foot ulcer in their lifetime, and the associated amputation and mortality rates are well documented. Despite that scale of need, the therapeutic pipeline for diabetic wound healing has been thin, and the candidates that do progress often stumble between promising rodent data and mixed clinical results.
Rodent diabetic wound models, whether streptozotocin induced or db/db genetic, have been the default preclinical platform for years. They are tractable, inexpensive, and mechanistically informative. They are also, in several important ways, not like human skin. For programs approaching clinical development, the argument for adding porcine data has grown less controversial and more necessary.
The translational case for porcine wound models rests on anatomy. Pig skin is structurally and functionally similar to human skin across five parameters that matter for wound biology. Thickness, hair follicle content, pigmentation, collagen composition, and lipid composition all align between pig and human in ways that rodent skin, with its loose architecture, absent sweat glands, and reliance on contraction rather than re epithelialization, simply does not.
That alignment has practical consequences. Dressings, biomaterials, and topical pharmacology behave differently on pig skin than on rodent skin because the substrate is doing different things. A hydrogel that performs well in a mouse excisional wound may fail to adhere or release properly on pig tissue, and the reverse is also true. For candidates whose mechanism depends on tissue level interactions with skin architecture, the porcine model provides a translational check that rodent data alone cannot.
MD Biosciences runs a streptozotocin induced diabetic pig wound model, with diabetes verified through serial blood glucose measurements before wound creation. Full thickness excisional wounds are placed on the pig back, allowing within animal comparisons across treatment conditions and controls. Standard of care comparators anchor the preclinical data to a meaningful clinical reference.
The endpoint panel is what makes the model a serious translational tool. Wound closure rates are tracked longitudinally, typically with complete closure observed in well performing treatment arms around day 28. Granulation tissue thickness is measured histologically, with MD Biosciences' AI histology platform enabling quantitative comparison across groups. Blood vessel density is quantified through CD31 immunohistochemistry, providing an angiogenesis readout that maps directly onto the neovascularization requirements of human chronic wound healing.
Inflammatory dynamics are captured through multiplex cytokine analysis, including TNF alpha, IL 8, MCP 1, and VEGF A. This is where the diabetic pig wound model has produced some of its most consequential data. Schirmer et al. (2021, Advanced Science) used this model to characterize a starPEG GAG hydrogel wound contact layer designed to selectively scavenge pro inflammatory chemokines. The hydrogel produced complete wound closure by day 28, doubled granulation tissue thickness relative to Adaptic, reduced TNF alpha and IL 8 significantly, and preserved pro regenerative growth factors. The pig data supported a clear mechanistic story, and the dressing has continued to advance in development.
Wound healing is a morphology driven biology, and histology is the endpoint that matters most for regulators. The challenge has always been that manual quantification is slow, subjective, and hard to reproduce across studies. MD Biosciences uses an AI platform to quantify CD31 for blood vessel density, Herovici staining for collagen I versus collagen III differentiation, and intraepidermal nerve fiber density via anti PGP9.5. The platform produces continuous quantitative outputs rather than categorical scores, which changes what sponsors can ask from a wound healing study.
Collagen phenotype is particularly relevant for chronic wound programs. Herovici staining differentiates young collagen III, which predominates in early healing, from mature collagen I, which reflects remodeling and scar maturation. For candidates targeting scarless healing or remodeling kinetics, the ratio itself becomes an actionable endpoint.
The diabetic pig model is the flagship, but it is not the only pig wound platform worth considering. MD Biosciences also runs an aged pig wound model, which captures the impaired healing phenotype characteristic of elderly populations. Burn wound and incisional wound models extend the platform into other clinical scenarios. For sponsors working on products aimed at specific patient populations, matching the preclinical model to the clinical demographic produces preclinical data that regulators and clinicians can more readily interpret.
The pattern that has worked for wound healing programs is layered. Rodent models remain useful for early mechanism work, candidate screening, and establishing dose ranges. Porcine models enter when translational confidence becomes the question, particularly around formulation performance, dressing behavior, biomarker trajectories, and histology that a clinician would recognize.
For diabetic wound healing specifically, the argument for pig data is stronger than for many indications, because the biology of impaired healing depends so heavily on tissue architecture and cellular milieu. A candidate that works in a db/db mouse may or may not behave similarly in a substrate that resembles the human foot ulcer, and that uncertainty is expensive to resolve later.
Sponsors moving into IND enabling work for chronic wound indications typically find the planning question is not whether to run pig data, but how to integrate it efficiently with the rodent and in vitro work already underway. The model has earned its place in the package for a reason, and the published outcomes continue to validate that position.
— — —
MD Biosciences operates diabetic, aged, burn, and incisional pig wound models with integrated AI histology, cytokine profiling, and biomarker analysis. For questions about wound healing study design, contact neuro@mdbiosciences.com.