Among industry trials that gave a stop reason, 6 in 10 now cite a sponsor or business decision, up from 3 in 10
A Priority Review analysis of 18,422 terminated trials on ClinicalTrials.gov. Among industry trials that gave a stop reason, 63.5% of those ended in 2025 cited a sponsor or business decision (407 of 641), up from 31.5% in 2016 (107 of 340). Enrollment fell from 31.8% to 12.3% on the same basis.
Machine-summarized from the primary sources listed below; not independently reported. How we use automation
When a company stops a clinical trial early, the registry record carries a line that says why. The answer has changed. Among industry-sponsored trials that ended early in 2025 and gave a reason, 63.5% (407 of 641) cite a sponsor or business decision and 12.3% (79) cite enrollment. In 2016 the two were level, at 31.5% (107 of 340) and 31.8% (108). Every share below is a share of the trials that gave a reason, unless the text says otherwise.
Those figures come from our analysis of every interventional study marked terminated on ClinicalTrials.gov with a completion date from January 1, 2016 through October 31, 2026: 18,422 studies, pulled November 14, 2026. The rules, the counts for every year and the check sample are in the data and method file.
For a research site, that line is the public account of why a study ended. Of the 641 industry trials stopped in 2025 that gave a reason, about 1 in 4 (151) gave a business reason of four words or fewer.
How the reasons were counted
The rule behind the registry defines the field as "a brief explanation of the reason(s) why the clinical trial was stopped" (2016 final rule). The clinical trial sponsor writes it as free text. We sorted each reason into groups with keyword rules, and one reason can fall into several groups. Of the 17,652 trials with completion dates from 2016 through 2025, 16,906 gave a reason.
As a check, 150 randomly drawn reasons were read against their labels in two separate passes, and the less favorable count is given here. The enrollment label was right for 55 of the 59 reasons it marked and missed 5 others. The sponsor-decision label was right for 20 of the 26 it marked, with every arguable case counted as wrong, and missed 2.
Across those ten years, among trials that gave a reason:
- 40.4% cite enrollment, accrual or recruitment (6,828 trials)
- 18.3% cite a sponsor or business decision (3,090)
- 10.0% cite funding (1,687)
- 9.4% cite efficacy or futility (1,589)
- 8.9% cite COVID-19 (1,499)
- 3.7% cite safety (623)
The enrollment figure is close to older work that used a different method. A 2015 study that read each reason by hand, for 905 terminated trials with posted results, put insufficient accrual at 39% of them and "unspecified business decision/strategic reason" at 77 trials (Williams and colleagues, PLoS One).
The industry shift
The ten-year average hides a split by sponsor type. The share of industry-sponsored terminated trials that gave a reason and whose reason cites a sponsor or business decision, by completion year:
- 2016: 31.5% (107 of 340)
- 2018: 30.9%
- 2020: 36.7%
- 2022: 47.8%
- 2024: 58.5%
- 2025: 63.5% (407 of 641)
On the same basis, enrollment went the other way: 31.8% in 2016, 26.0% in 2019, 24.5% in 2022 and 12.3% in 2025 (79 of 641).
Short business reasons also became more common. In 2016, 34 of the 340 industry reasons were a business reason of four words or fewer. In 2025, 151 of 641 were, and 61 of those read "Sponsor decision" and nothing more.
Part of the rise is a change in mix. Phase 1 trials, where business reasons are most common, were 87 of the 340 industry reasons in 2016 and 296 of the 641 in 2025. Late-stage trials show the same turn, though. In industry phase 3 trials, sponsor or business decisions rose from 25.8% of the trials that gave a reason for 2016 through 2019 to 51.1% for 2022 through 2025. Efficacy or futility held steady, at 27.4% and 26.2%.
Industry also stopped a larger share of the trials it closed out. Terminated trials were 11.8% of industry trials that ended in 2016 (467 of 3,947 terminated or completed), 17.4% in 2023 (846 of 4,870) and 15.9% in 2025 (652 of 4,097).
Industry ends a bigger share of its trials early than in 2016; other sponsors end a smaller one
Terminated trials as a share of terminated plus completed trials, by completion year
Source: The Priority Review analysis of ClinicalTrials.gov, records pulled November 14, 2026. Completed-trial counts come from the same API with overall status Completed.
Other sponsors: enrollment first, then money
For universities, hospitals, government bodies and other lead sponsors outside industry, enrollment is still the reason given most often among the trials that gave one: 52.5% in 2016 (504 of 960) and 49.9% in 2025 (437 of 875). Funding grew most, from 10.9% (105 of 960) to 19.3% (169 of 875).
At least 16 of the 2025 reasons say the NIH, the National Cancer Institute, USAID or a federal funder ended a grant or the study. One reads: "Funding was terminated by NIH in 2025 with a statement that the research no longer meets agency priorities" (NCT07017595).
At non-industry sponsors enrollment still leads, and funding is the reason that grew
Terminated trials with a lead sponsor outside industry: share whose stated reason falls in each group, by completion year
In 2020, 35.7% of non-industry reasons cited COVID-19, and the other groups dip that year.
Source: The Priority Review analysis of ClinicalTrials.gov terminated trials, records pulled November 14, 2026. Shares are of trials that gave a reason, by completion year; a reason can fall in more than one group.
What a slow-enrolling trial leaves behind
Trials that cite enrollment and report an actual enrollment count, 6,767 of them from 2016 through 2025, had enrolled 363,015 participants before they stopped. Half enrolled 13 or fewer. In 3,022 of them, 10 or fewer people took part.
What remains uncertain
The shares leave out trials that gave no reason. For industry, 340 of 467 terminated trials gave one in 2016 and 641 of 652 in 2025, so the 2016 and 2025 shares rest on different portions of all terminated trials. The field holds the sponsor's own account, and the registry does not record whether slow enrollment, weak data or cost sits behind a "sponsor decision". Keyword rules mislabel some reasons. Records for 2025 and 2026 are still being updated, and the file does not measure whether some kinds of reason are updated sooner than others; for January through October 2026 the industry figures are 61.7% and 11.2% (data and method file).
Sources
- The Priority Review: data and method, reasons trials stopped, 2016 to 2026
- ClinicalTrials.gov API: interventional studies marked terminated, completion dates January 1, 2016 through October 31, 2026 (pulled November 14, 2026)
- Federal Register: Clinical Trials Registration and Results Information Submission, final rule (September 21, 2016)
- Williams, Tse, DiPiazza and Zarin: Terminated Trials in the ClinicalTrials.gov Results Database, PLoS One 2015 (Europe PMC full text)
- ClinicalTrials.gov: study record NCT07017595
- Cancer.gov dictionary: clinical trial sponsor