Peptide basics

What “no registered trial” does and does not mean

This page explains how to interpret a registry search result. It is not about any particular compound, and it takes no position on whether any compound works or is safe.

What a registry is

ClinicalTrials.gov is a data bank operated by the National Library of Medicine. A responsible party — the sponsor, or a designated principal investigator — files a record describing a planned study before or shortly after it begins, and is required in defined circumstances to return later and submit summary results.¹²

The obligation has edges. The 2016 final rule implementing Title VIII of the Food and Drug Administration Amendments Act sets out which trials of drug, biological and device products the requirement covers, and took effect on 18 January 2017. Studies falling outside that definition may be registered voluntarily, and often are, but nobody is compelled.¹²

Two further pressures fill the register beyond what the law demands. Member journals of the ICMJE will not consider a clinical trial for publication unless it was registered before the first participant was enrolled, in a registry meeting the WHO's criteria. And the WHO Registry Network maintains a portal that searches across registries in many jurisdictions at once.³⁴⁸

The four things a zero can mean

1. Nobody has run a registered human study

The plainest reading, and usually the right one for compounds that have never had a commercial sponsor. There was no trial, so there is no record.

2. The study predates the requirement

Registration became mandatory for covered trials in January 2017, and the registry's earliest posted study records date from 1999. Research conducted in the 1980s and 1990s left no registry trace by design. This is not hypothetical: it is the entire explanation for at least one FDA-approved peptide with an empty registry row.¹²

3. The study is registered under a different name

Compounds in development are routinely registered under a sponsor code. NCT02637284 registers its intervention as Bepecin, with BPC-157 and PCO-02 in the other-names field; a search that reads only the primary intervention name discards it. Any count of registered studies is therefore a count under a stated matching rule, and a count without a stated rule is not interpretable.⁹

4. The study is in a registry you did not search

The WHO Registry Network's primary registries cover Australia and New Zealand, Brazil, China, Korea, the European Union, India, Cuba, Germany, Iran, the United Kingdom, Japan, Lebanon, Thailand, Africa, Peru and Sri Lanka. There is no Russian primary registry on that list, which matters for the several peptides whose literature is largely Russian. Searching only ClinicalTrials.gov is a defensible default and an incomplete census.⁴⁸

The two conclusions a zero cannot support

It does not mean the compound does nothing

This is the statistical point Altman and Bland made in 1995 and it has not aged: absence of evidence is not evidence of absence. A registry silence is produced by the absence of a filing, and an absent filing is caused by funding, commercial incentive, jurisdiction and date at least as often as by anything about the molecule. Nothing about an empty row licenses the sentence “it has been shown not to work”.⁷

It does not mean the gap is a technicality

The mirror-image error is more common in the marketing of these compounds. If the registry is empty, then whatever is being asserted about human effects rests on something other than a registered trial, and the reader is entitled to know what. Preclinical work, mechanism, anecdote and inference are all real things; none of them is a human trial, and describing the difference as a paperwork problem is a way of not answering.

Registration is not results

The opposite trap catches readers who find a registration and stop. A record is a plan. It states what a sponsor intended to do, and it is filed before the answer exists.²

Whether the answer ever arrives is a measurable question and the measurements are not flattering. A 2020 cohort study in the Lancet examined compliance with the legal requirement to report results on ClinicalTrials.gov and found a substantial share of due trials had not reported. Earlier, a cross-sectional analysis in the BMJ found that fewer than half of a cohort of NIH-funded trials had been published in a peer-reviewed journal within 30 months of completion.⁵⁶

How to state it honestly

The wording that survives scrutiny is boring and specific. Not “there is no evidence”, which overclaims. Not “it is unproven”, which smuggles in an expectation of eventual proof. Instead: as of this date, a search of this registry, under this matching rule, returned no study naming this compound as an intervention.²

Every clause in that sentence is doing work. The date, because registries change weekly. The registry, because there are others. The matching rule, because a fuzzy search returns a different and larger number. And “naming this compound as an intervention”, because a mention in a title is not administration to a person.⁴⁹