Chapter 31 — Key Takeaways
The one-sentence version
Veterinary medicine is real medicine with real evidence, and that evidence answers veterinary questions — so the failure is never the animal data, it is the transfer.
The frame
Veterinary medicine has the full apparatus. Approved drugs, dedicated regulators (in the U.S., the FDA's Center for Veterinary Medicine), controlled trials, post-approval pharmacovigilance, and a substantial peer-reviewed literature. In some respects — production-animal sample sizes, dietary and housing control — it is better powered than its human counterpart.
Peptides are genuinely part of that pharmacopeia. GnRH analogs for reproductive management, insulin for diabetic dogs and cats, desmopressin, synthetic ACTH analogs as diagnostic probes, and an approved ghrelin-receptor agonist for appetite stimulation.
Every one of those is scoped. Named species, named indication, labeled dose. The evidence is excellent evidence for the population it was gathered in, which is exactly why it does not travel.
The five ways translation breaks
| # | Mechanism | Standing example |
|---|---|---|
| 1 | Receptors differ in sequence and distribution | rodent thyroid C-cells express GLP-1 receptors far more abundantly than human ones (Ch 8); porcine insulin differs from human by one residue and matches canine exactly (Ch 11) |
| 2 | Clearance differs with body size | metabolic rate scales sublinearly with mass; half-life and exposure do not correspond |
| 3 | Dose scaling is not linear with mass | standard conversion normalizes to body surface area; naive mg/kg conversion overestimates, ~12× from mouse |
| 4 | Models differ from the diseases they name | transected tendon ≠ overuse tendinopathy (Ch 17); ob/ob mouse = leptin deficiency, not common obesity (Ch 13) |
| 5 | Lifespans differ by roughly fortyfold | a two-year rodent study is most of a life; two years in a human is 2.5% of one |
Plus two that ride along with (4): route (intraperitoneal injection and gavage have no human equivalent) and subject characteristics (young, uniform, pathogen-free, unmedicated animals).
The portable tool: four questions
When you meet "it has been used in animals for years":
- In whom? — a horse is not a person, and neither is a rat.
- For what? — growth promotion is not injury recovery.
- Measured how? — weight gain is not function, and no animal reports pain or wellbeing.
- And collected how? — veterinary and laboratory use frequently generates no systematic outcome data at all.
And, following from the fourth: compared to what? Clinical experience has no control group.
A track record is not a dataset unless somebody was recording.
Things worth not getting wrong
A ban is not an efficacy finding. Racing and sporting authorities prohibit substances for uncertainty, animal welfare, unapproved regulatory status, competitive integrity, and administrability. "They wouldn't ban it if it didn't work" upgrades a rating on the strength of a risk-management document.
A large literature can be entirely silent. Livestock growth research measures average daily gain, feed conversion ratio, and carcass composition. It contains no measurement of strength, function, injury, recovery, or wellbeing — and its safety endpoints protect a food consumer from trace residue, which is the inverse of an injected bioactive exposure. Call this endpoint mismatch.
"Used in animals" ≠ "used in laboratory animals." BPC-157's record is preclinical research, not veterinary practice. It has no labeled indication in any species, and never had a human development program (Ch 17).
"TB-500" is not thymosin β4. The parent is 43 residues; the marketed compound is a short fragment (Ch 18). Citing the parent's literature for the fragment is a molecule error stacked on top of a species error.
Predictive validity is the only one that counts. Face validity (it looks like the condition) and construct validity (it arises from the same mechanism) are not sufficient. A model built to be tractable is often built to be different.
The regulatory asymmetry
Veterinary regulation supplies a licensed prescriber, a labeled product with a named species and indication, a manufacturing standard, an adverse-event channel, withdrawal periods for food animals, and legal accountability. Extra-label use is a conditioned exception — licensed veterinarian, valid veterinarian-client-patient relationship, clinical necessity, no adequate approved alternative, established withdrawal period — not a loophole.
The gray market cites veterinary use and omits the veterinarian. It borrows the credibility of a system while operating outside every one of its controls, and labels the product "not for human consumption" while implying the opposite.
The two-way street
Insulin's development involved dogs. Exenatide came from a Gila monster (Ch 4). The first orally active ACE inhibitor descends from pit viper venom; an approved intrathecal analgesic from a cone snail; salmon calcitonin outperforms the human molecule in humans. Leptin was found in a mouse — the model told the truth about the mechanism and misled about the application. Veterinary practice adopts human drugs constantly, often extra-label.
The argument is about the direction of inference, not the value of animal science. Preclinical work earns a compound the right to be tested in people. It does not earn it a conclusion.
Ratings issued in this chapter
| Claim | Rating |
|---|---|
| GnRH agonists for reproductive management in approved veterinary species | ✅ |
| TB-500 for human tissue repair, on the basis of veterinary use | ❌ |
| GHRPs for human performance benefit, on the basis of livestock research | ❌ |
| "Used in animals for years, therefore safe in humans" (the inference itself) | ❌ |
The ✅ is not decoration. It demonstrates that the machinery rates the fit between evidence and claim, not the species the study was run in.
Dossier progress
Field 2 now carries a species translation block for every compound whose support is animal or veterinary: species, model, endpoint, duration, route, dose-scaling basis, molecule identity — then two written sentences naming what human question the evidence does address and what it does not. Appendix C has the workbook pages.