Tissue Repair Models and Their Limits
Why healing in cells, rodents and tendon injury models does not equal human recovery, with meta-research, regulator label limits and peptide examples, and how to read a repair claim.
Content last updated . This page has not been medically reviewed.
Direct overview
Most claims that a peptide "heals" tendons, skin, gut or muscle come from experimental models: cells in a dish, tissue pieces, or injuries made on purpose in rodents. Models are essential for science. They are not people. This page explains what each model can and cannot show, so you can read a repair claim fairly. It is not advice to use any product.The ladder of evidence for a repair claim
| Level | What was studied | What it can show | What it cannot show |
|---|---|---|---|
| Cells in a dish | One cell type under controlled conditions | A mechanism, such as cell migration | Whether tissue heals in a body; in vitro models lack other tissue-matrix components [1] |
| Animal injury models | An injury made in a rodent or other animal | Whether a compound changes healing in that species and model | Human recovery, since no animal model is an exact replica of the clinical situation [1] |
| Early human studies | Small safety or pilot studies | Tolerability, signals | Efficacy |
| Randomized human trials | Comparison with placebo or standard care | Effects on outcomes people notice | Results for other conditions or products |
Why animal results often fail to carry over
A 2007 systematic review in the BMJ compared animal experiments with human trials for six treatments. Corticosteroids helped in head-injury animal models but not in the human trials; tirilazad improved infarct size in animals yet was associated with worse outcomes in stroke patients. The authors concluded that discordance may come from bias or from animal models failing to mimic the clinical disease. [2] That does not mean animal work is useless: some treatments, such as clot-busting drugs and bisphosphonates, did show effects in both. It means each result needs its own check against human data. [2]Healing models have particular problems
- Skin. Differences exist between animal and human skin healing, and responses to new therapies can differ. Reviews note a lack of standardization across wound models. [1] A review of mouse surgical models frames their benefits and limits for translation. [3]
- Tendon. A 2021 review states that no physiologically representative tendinopathy model exists to date, and that many lab models lack matrix and blood supply. [4]
- Scarring versus regeneration. Adult healing can leave scar tissue, while regeneration with minimal scarring happens in fetal skin and restricted tissues, so an animal that regenerates is not predicting an adult human's scar. [5]
Examples from the peptide field
- BPC-157 and musculoskeletal injury. A systematic review found 36 studies, 35 preclinical and 1 clinical (a retrospective study of knee pain in which 7 of 12 patients reported relief), and no clinical safety data. [7] A separate review reports only three human pilot studies, calls it investigational and says rigorous large trials are lacking. [8] Label: preclinical for tendon, ligament, muscle and bone claims; insufficient evidence in people. See BPC-157.
- Thymosin beta-4 as eye drops. A 9-patient phase 2 dry-eye trial reported improved signs and symptoms. [9] A phase 3 trial in neurotrophic keratopathy, with 18 analyzed participants, showed healing in 6 of 10 treated versus 1 of 8 on placebo, but the p-value (0.0656) did not reach the usual 0.05 threshold, and the company that developed the product helped design it. [10] Label: promising but limited, for eye drops only, not injected products. See TB-500 and thymosin beta-4.
How to read a repair claim
Ask in order: Which exact molecule and product? What was injured, in which species? Was there a comparison group? Were outcomes about function and pain, or only a lab marker? Who funded it, and has anyone else repeated it? Higher growth factor levels or more new vessels in tissue are biomarkers, not recovery. See biomarker changes vs meaningful health outcomes and how to evaluate peptide claims. This checklist helps you judge evidence; it cannot prove a product is safe.Frequently asked questions
If a peptide heals a rat tendon, will it heal mine?
Not necessarily. No animal model is an exact replica of human injury, and reviews of animal-to-human comparisons found some treatments helped animals but not people. [1][2]
Are animal studies worthless?
No. They guide which ideas to test, and some effects were seen in both animals and humans. They simply cannot stand in for human outcome trials. [2]
Is there a good lab model of tendon injury?
A 2021 review says no physiologically representative tendinopathy model exists yet. [4]
Is there an approved product for wound healing?
In the US, becaplermin gel is labeled for certain diabetic foot ulcers only. [6]
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
How to read the evidence labels
Labels used on this page: approved medicine, promising but limited, mixed, insufficient evidence, preclinical (cells or animals only), no good evidence of benefit and marketing claim.
Regulator labels show how narrow proven healing can be
Becaplermin gel (Regranex) is an FDA-labeled growth factor for lower-extremity diabetic ulcers that extend into tissue and have an adequate blood supply. The label states efficacy has not been established for pressure ulcers or venous stasis ulcers, and that effects on exposed joints, tendons, ligaments and bone have not been established in humans. [6] So even an approved repair product is not a general healer. Label: approved medicine (FDA, one indication).
Safety note
A model result is a reason to run a human trial, not a reason to self-treat an injury. A persistent injury needs assessment by a clinician. See peptides and tendon and ligament injury claims and talking with a health professional.
References
- Non-animal models of wound healing in cutaneous repair: In silico, in vitro, ex vivo, and in vivo models of wounds and scars in human skin. Wound Repair Regen 2017. PubMed 28120405. Read 2026-10-22.
- Comparison of treatment effects between animal experiments and clinical trials: systematic review. BMJ 2007. PubMed 17175568. Read 2026-10-22.
- Surgical approaches to create murine models of human wound healing. J Biomed Biotechnol 2011. PubMed 21151647. Read 2026-10-22.
- The Lack of a Representative Tendinopathy Model Hampers Fundamental Mesenchymal Stem Cell Research. Front Cell Dev Biol 2021. PubMed 34012963. Read 2026-10-22.
- Healing Mechanisms in Cutaneous Wounds: Tipping the Balance. Tissue Eng Part B Rev 2022. PubMed 34915757. Read 2026-10-22.
- U.S. Food and Drug Administration. Regranex (becaplermin) gel prescribing information (label effective 2026-05-08). FDA drug label record via openFDA. Read 2026-10-22.
- Emerging Use of BPC-157 in Orthopaedic Sports Medicine: A Systematic Review. HSS J 2025. PubMed 40756949. Read 2026-10-22.
- Regeneration or Risk? A Narrative Review of BPC-157 for Musculoskeletal Healing. Curr Rev Musculoskelet Med 2025. PubMed 40789979. Read 2026-10-22.
- Thymosin beta4 significantly improves signs and symptoms of severe dry eye in a phase 2 randomized trial. Cornea 2015. PubMed 25826322. Read 2026-10-22.
- 0.1% RGN-259 (Thymosin beta4) Ophthalmic Solution Promotes Healing and Improves Comfort in Neurotrophic Keratopathy Patients in a Randomized, Placebo-Controlled, Double-Masked Phase III Clinical Trial. Int J Mol Sci 2023. PubMed 36613994. Read 2026-10-22.
Education, not medical advice. This page is for general information. It does not diagnose or treat any condition and is not a substitute for care from a qualified health professional. Natural does not automatically mean safe. Ask a clinician or pharmacist before changing your care, especially if you take medication, are pregnant or breastfeeding, or have a health condition.