Why Most Peptide Evidence Is Preclinical
Most peptides have animal data and no human trials because trials cost hundreds of millions of dollars and only patent-holders fund them. Natural or off-patent peptides never get tested, and that gap is not going to close on its own.
An explainer. The cost and failure-rate figures come from published analyses of drug development; the patent and regulatory points are general pharmaceutical economics applied to the peptides graded on this site.
- Covers
- the cost of clinical trials, why natural peptides are not patented, how compounding and research-chemical markets fill the gap, what preclinical evidence can and cannot say
The pattern
Read enough pages on this site and a pattern appears. Semaglutide and tirzepatide have trials with tens of thousands of people. BPC-157, with decades of animal research and a plausible mechanism, has no human trial at all. That is not because BPC-157 is less promising. It is because nobody has a financial reason to test it.
This page explains the reason, because understanding it changes how you read the animal data.
What a trial program costs
Getting a drug from discovery to approval takes roughly a decade and, counting the failures a company absorbs along the way, costs on the order of a billion dollars. The failure rate is the driver: only about one in ten compounds that enter human trials ever reaches approval, and each failure was paid for (Hay et al., 2014). Success rates have not improved much over time (Dowden & Munro, 2019).
A Phase III program alone, the large trials that approval requires, runs to hundreds of millions. No university, foundation, or research-chemical vendor can pay that. Only a pharmaceutical company can, and a company pays only when it expects to earn the money back.
The patent problem
Earning it back requires exclusivity: a patent that stops competitors from selling the same molecule for the years it takes to recoup the cost. A natural peptide sequence, a fragment of a protein the body already makes, is very hard to patent as a composition. Anyone can synthesize it, and the moment it were approved, anyone could sell it.
So the peptides that get trials are the modified ones. Semaglutide is GLP-1 with an amino acid swap and a fatty chain, patentable and worth billions. Tesamorelin is GHRH with a protective cap. CJC-1295 with DAC was a patentable modification, which is why it got a human trial and unmodified GHRH fragments did not. The unmodified peptides, BPC-157, TB-500, KPV, MOTS-c, sit in what drug developers call the valley of death: promising enough to study in animals, unfundable to study in people.
Peptides as a class are a growing share of new drugs (Muttenthaler et al., 2021), and almost all of that growth is in engineered molecules with patents attached.
The regulatory maze
The approval path adds friction of its own. The FDA requires a full program for each indication, so a peptide with five plausible uses needs five sets of trials. Compounding pharmacies once offered a side door, preparing peptides on prescription without approval, and the FDA has been closing it, restricting BPC-157 and others from compounded use. The research-chemical market, selling the same molecules “not for human consumption,” fills the gap without generating any evidence, since selling for research produces no trials.
Other countries run different systems, which is why Semax and Selank are prescription drugs in Russia and research chemicals everywhere else, and why thymosin alpha-1 is approved in 35 countries and not the United States. Approval abroad does not transfer.
What animal evidence can and cannot tell you
It can tell you a peptide does something in a living body, roughly how the effect scales with dose, whether it is acutely toxic, and what mechanism is involved. Extensive, consistent animal data from several labs is a real signal.
It cannot tell you the human dose, the human side effects, whether the effect survives the differences between a rat and a person, or what years of use do. About half of animal results fail to predict the corresponding human result (Hackam & Redelmeier, 2006). AOD-9604 is the peptide world’s own example: half the fat gone in mice, nothing meaningful in a 300-person trial.
How to hold this
Two mistakes are common. One is dismissing animal data as worthless, which it is not; it is where every drug starts. The other is treating it as proof, which it is not; it is where most drugs stop. The accurate position is uncomfortable: a peptide with good animal data and no human trial might work in people, might not, and nobody knows. Using it means accepting that, with eyes open. See how peptides are studied for the full ladder and how to read peptide claims for the checking method.
Will it change?
Slowly, if at all. A few things could move it: government or philanthropic funding of trials for unpatentable compounds, which exists in small amounts; new formulations or delivery methods that create patentable versions of old peptides, which is how sermorelin became CJC-1295; and cheaper trial designs. None is close to closing the gap for the peptides most people ask about. For the foreseeable future, “no human trial” will stay the honest description of most of this field.
Frequently Asked Questions
Why doesn’t a university run the trial?
Universities run early, small trials. A Phase III program costs more than most universities’ entire research budgets for a year. Without a company to pay, it does not happen.
Isn’t the lack of trials itself suspicious, as if companies know it does not work?
Usually not. The absence reflects the patent economics, not a hidden negative result. That said, when a company did run a trial and dropped the compound, as with AOD-9604 and CJC-1295, that history is worth knowing.
Does approval in Russia or China count?
It counts as human data, often small and rarely repeated in the West. It is more than nothing and less than a Western Phase III.
Which peptides have escaped the valley of death?
The modified, patented ones: semaglutide, tirzepatide, tesamorelin, PT-141. Sermorelin, briefly. Nearly everything else on this site is still in it.
References
- Hay M, et al. (2014). Clinical development success rates for investigational drugs. Nat Biotechnol. PubMed
- Dowden H, Munro J. (2019). Trends in clinical success rates and therapeutic focus. Nat Rev Drug Discov. PubMed
- Muttenthaler M, et al. (2021). Trends in peptide drug discovery. Nat Rev Drug Discov. PubMed
- Hackam DG, Redelmeier DA. (2006). Translation of research evidence from animals to humans. JAMA. PubMed