Registered but Never Reported: The Trials That Vanish

The rule

Under the Food and Drug Administration Amendments Act of 2007, sponsors of applicable trials are required to report their results directly onto ClinicalTrials.gov within one year of completion. The first trials covered by the Final Rule became due to report in January 2018¹.

That is a legal obligation with a deadline, not a courtesy.

How often it is met

A cohort study of every applicable trial due to report under the rule, using data extracted in September 2019, found the following.

4,209 trials were due to report results.

1,722 of them — 40.9%, 95% CI 39.4 to 42.2 — reported within the one-year deadline.

2,686 — 63.8%, 95% CI 62.4 to 65.3 — had results submitted at any point.

The median delay from primary completion date to submission was 424 days, 95% CI 412 to 435 — 59 days past the legal requirement.

Compliance had not improved since July 2018¹.

Two associations from the same study are worth carrying: industry sponsors were more likely to be compliant than non-industry, non-US-Government sponsors (odds ratio 3.08, 95% CI 2.52 to 3.77), and sponsors running large numbers of trials were far more likely to be compliant than smaller ones (odds ratio 11.84, 95% CI 9.36 to 14.99)¹.

The pattern is not what most readers expect. Small, occasional sponsors are where results go quiet.

What that looks like on one compound

Ipamorelin is a useful case because its registered history is short enough to read in full.

A ClinicalTrials.gov search for ipamorelin returned 3 studies in August 2026. Two of them are completed phase 2 trials.

NCT00672074 — a phase 2, double-blind, placebo-controlled study in post-operative ileus. Actual enrolment 117, 19 sites, quadruple masking, primary completion December 2009, last updated April 2017. No results posted as of August 2026².

NCT01280344 — a phase 2 double-blind placebo-controlled dose-finding study for recovery of gastrointestinal function after bowel resection. Actual enrolment 320, 45 sites, quadruple masking, four arms, primary completion June 2013, study completion May 2014, last updated April 2017. No results posted as of August 2026³.

Two properly designed, adequately sized, masked, multi-site trials. 437 people were randomised between them. Neither has a results table on the record.

There is a second thing to notice. Both trials studied gastrointestinal motility after surgery. Neither studied growth hormone secretion, body composition, or anything resembling the reasons this compound circulates today.

FDA's own note on the substance points at the same literature from the other direction: in the agency's description of bulk drug substances nominated for use in compounding, ipamorelin acetate is listed with the observation that a study published in the literature identified serious adverse events including death when ipamorelin was administered intravenously for improving gastric motility, and that FDA has not identified safety-related information for certain other injectable routes⁴.

A phase 3 with 700 participants

ARISE-3, NCT03937882, was a multi-centre randomised double-masked placebo-controlled phase 3 of a thymosin beta-4 ophthalmic solution in dry eye. Actual enrolment 700. Twenty sites. Primary completion 8 November 2020, study completion 7 October 2021, record last updated May 2022. No results posted as of August 2026⁵.

Alongside it, NCT02597803, a phase 2/3 in the same indication with 317 participants and a June 2016 primary completion, also shows no posted results as of August 2026⁶.

Whatever those 1,017 participants experienced, the registry does not say. That is the single largest fact about this compound's late-stage record, and it is invisible to anyone reading only the trials that did report — including the 18-patient phase 3 that our TB-500 page covers.

Posted is not the same as published

This is where a careless version of this argument goes wrong, so it is worth stating plainly.

NCT00123253 — the pivotal phase 3 of tesamorelin in HIV-associated lipodystrophy, 412 participants — also shows no posted results on ClinicalTrials.gov⁷.

That trial was published in the New England Journal of Medicine in 2007, reporting a 15.2% decrease in visceral adipose tissue against a 5.0% increase on placebo, with the trial's own registry number cited in the paper⁸.

So "no results posted" means the results are not in the registry's structured results section. It does not mean the trial vanished. Before drawing any conclusion, search the NCT number in the literature as well as on the record. Our tesamorelin page works through what that programme did establish.

Why the silent ones are not a random sample

If unreported trials were a coin flip, they would cancel out and the published literature would still be roughly right.

They are not a coin flip. A trial that supports a programme gets written up, presented and cited. A trial that does not is easier to leave alone — nobody is required to be enthusiastic about it, and the deadline for posting it goes unenforced.

The consequence is directional: the visible evidence for any compound is, on average, more favourable than the complete evidence. That is not a claim about any particular missing trial. It is a claim about what survives.

How to check, in three steps

Search the compound on ClinicalTrials.gov and read the whole result set, not the first page. Note the total.

For every completed or terminated record, check the has-results flag and the primary completion date. Anything more than a year past completion with nothing posted is a gap.

Search each of those NCT numbers in PubMed. If a publication exists, the trial reported through a different door. If nothing comes back, the trial is genuinely dark.

Three searches, and you know more about a compound's evidence base than its marketing will ever tell you. The rest of the reading is in how to read a record.