Chronic, non-healing wounds cause major morbidity, risk of amputation and death and high healthcare costs, but therapeutic development is hindered by repeated translational failures from preclinical to clinical translation. Based on the classic four-phase approach to wound healing (Hemostasis, inflammation, proliferation and regeneration, and maturation and remodeling), this paper critically synthesizes and evaluates the epistemic assumptions that underlie the selection of animal models for the study of chronic wound healing, with a particular emphasis on the question of clinical translatability and plausibility of human models. The analysis uses narrative comparative analyses of in vivo animal models (rodents, pigs, rabbits, guinea pigs, zebra fish), human wound models (skin grafts, skin, cutaneous, cuticle, excision model), and platform in vitro and ex vivo (in vitro, 3D skin and histocultures, computational simulation). Rodents, despite their practicality and genetic tractability (including chemically induced and db-db diabetic models), wounds heal primarily by contraction due to “panniculus carnosus”, which risks misleading closure endpoints compared to human re-epithelialization; the rat tail model is noted for its reduction in contraction over prolonged periods. Pigs provide the closest overall architectural (anatomical) similarities in skin (thickness, sparse follicles, structures of the rete ridge), but are still limited by cost, accommodation, infection risk, differences in sweat glands, and the persistence of wounds of complete thickness. The rabbit ear model is unique in that it suppresses contraction by splinting the cartilage, allowing for a more human-like re-epithelialization and reproducible hypertrophic scarring and ischaemic site-specificity with respect to the agent and genetic factors. In disease models, especially in diabetes, short induction-to-wound intervals do not recapitulate long-term comorbidities (neuropathy, vascular disease) and in vitro 2D and 3D systems lack immunological, neural and vascular dynamics and are not well predictive in vivo. The bottom line is that no single model is enough; controlled human models, which are robust, well validated, and, despite ethical and practical limitations, remain the ultimate reference for moving from simple screening to more complex, disease-relevant systems, with the support of the emerging vascularized 3D constructs, microfluidics, and 3D bioprinting. Keywords Chronic models ;Wound Healing ;Animal models; Rodents ;Rabbits; Pig ;Genetic Mouse ; Impaired wound healing ;Translation
Navneet Soni (Sun,) studied this question.