How to Read ClinicalTrials.gov: Catch Missed Endpoints

A biotech company announces “positive Phase 2 results” and the stock jumps 30%. You read the press release: “meaningful clinical benefit” and “statistically significant improvement in key secondary endpoints.” You buy. Three months later, the Phase 3 trial fails and the stock drops 60%. What happened? The press release buried that the trial missed its primary endpoint. The “positive” results were from a secondary analysis the company highlighted to cover the miss. If you had checked ClinicalTrials.gov before buying, you would have seen the primary endpoint and known the press release was spinning a failure.

The solution: ClinicalTrials.gov is the free public database where every registrational clinical trial must be registered before it starts. It contains the pre-specified primary endpoint, the enrollment status, the expected completion date, and the trial arms. Companies cannot retroactively change these fields without leaving a trace. If you can read ClinicalTrials.gov, you have an edge over investors who only read press releases.

Step 1: Understand what ClinicalTrials.gov is

ClinicalTrials.gov is maintained by the National Library of Medicine at the NIH. Under FDAAA 801 (Section 801 of the FDA Amendments Act of 2007), the trial sponsor must register an “applicable clinical trial” on ClinicalTrials.gov within 21 days of enrolling the first patient. The registration must include the primary endpoint, secondary endpoints, enrollment target, and expected completion date before enrollment begins.

This matters because the primary endpoint is locked in before the trial starts. If a company registered a trial with “dose response on CDR-SB at Week 76” as the primary endpoint, that is the metric the trial was designed to test. If the trial misses that endpoint and the company highlights a different metric in the press release, the ClinicalTrials.gov registration is your evidence that the headline result was not the trial’s actual goal. Full requirements are on the ClinicalTrials.gov FDAAA 801 policy page.

Step 2: Search by company, drug, condition, or NCT number

The search bar on the ClinicalTrials.gov homepage accepts company names, drug names, disease conditions, and the unique NCT number assigned to each trial. If you have the NCT number from a press release, enter it directly. If not, search by company or drug name and filter by Phase.

Real example: search “Biogen diranersen” and the top result is NCT05399888, the CELIA Phase 2 trial. The registration page lists the official title, sponsor, phase, enrollment (416 participants), and the primary outcome measure. The full trial registration is at NCT05399888.

Another example: search “Stoke Therapeutics zorevunensen” to find the EMPEROR Phase 3 trial registration for Dravet syndrome. See our Stoke zorevunensen analysis for how that trial fits the investment thesis.

Step 3: Read the key fields investors care about

Six fields matter most for evaluating a trial before you trade on its results:

1. Primary Outcome Measure. The endpoint the trial was designed to test, pre-specified and locked before enrollment. If the press release does not clearly state whether the primary endpoint was met, this field tells you what the company committed to. For Biogen’s CELIA trial, the primary outcome was “dose response for change from baseline on the CDR-SB at Week 76.” The trial missed it.
2. Secondary Outcome Measures. Additional endpoints the company will report, but not the basis for trial success. Press releases often lead with secondary endpoints when the primary fails. If the headline focuses on a secondary measure, check whether the primary was met.
3. Enrollment (Actual vs Estimated). The registration lists an estimated enrollment target and, once the trial finishes, the actual enrollment. If actual enrollment is much lower than estimated, the trial may be underpowered and less likely to replicate in Phase 3.
4. Estimated Primary Completion Date. The date the company expects to finish collecting primary endpoint data. Usually more accurate than the “Q3 2026” guidance companies give in investor presentations, because the ClinicalTrials.gov date is updated as the trial progresses.
5. Arms. Lists what each group receives. If there is no placebo or active comparator arm, the trial is single-arm and the results are harder to interpret. Single-arm trials are common in oncology accelerated approvals but carry higher risk because there is no control to compare against.
6. Sponsor. The company running the trial. Collaborators are additional partners. This tells you who owns the economics and who is paying for the trial.

Step 4: Catch the press-release spin trick

Here is the specific move companies use. They register a trial with a broad, ambitious primary endpoint. The trial misses the primary. The company then highlights a secondary endpoint or a subgroup analysis in the press release and frames the result as “positive.” Investors who only read the press release never realize the primary endpoint failed.

Real example: Biogen’s CELIA trial for diranersen in Alzheimer’s disease. The primary endpoint was dose response on CDR-SB at Week 76. The trial did not show a dose response: the lowest dose (60 mg) produced the largest cognitive benefit, while higher doses showed smaller effects. Biogen’s press release led with “first study to show reduction in tau pathology and cognitive benefit” and highlighted the 26% cognitive slowing at 60 mg. Both statements are true. But the trial missed its primary endpoint, and the dose-response inversion is unexplained. An investor who read only the press release would think the trial succeeded. An investor who checked ClinicalTrials.gov would see that the primary endpoint was dose response, and that it missed. See our diranersen CELIA analysis for the full breakdown.

The same pattern shows up across biotech. Companies register one primary endpoint, miss it, and then lead the press release with a secondary endpoint that happened to look good. ClinicalTrials.gov is how you catch it.

Step 5: Track status changes before the press release

Trial status moves through a sequence: Recruiting → Active, not recruiting → Completed → Results posted. The change from “Recruiting” to “Active, not recruiting” often signals enrollment has finished, which means primary endpoint data is being collected. This can happen weeks before any press release. When the status flips to “Completed,” topline data is usually imminent. Some investors monitor these status changes as a timing signal, because companies must update ClinicalTrials.gov but may delay the press release until they have framed the results.

Step 6: Know the limitations and check the revision history

ClinicalTrials.gov is not perfect. Not all trials are registered promptly despite FDAAA 801, and enforcement has historically been inconsistent. International trials may have partial data. Results posting can lag by months. Some companies update the primary outcome measure after the trial starts. The revision history is your friend. Click the “History of Changes” tab on any trial page. If a company changed the primary endpoint mid-trial, that is a red flag.

Step 7: Cross-reference with the press release

Before you trade on a trial readout, open the ClinicalTrials.gov registration and find the primary outcome measure. Then open the press release and find where the company reports the primary endpoint result. If they match and the primary was met, the result is real. If the press release does not clearly address the primary endpoint, or leads with a secondary endpoint, check whether the primary failed. Our guide to reading clinical trial press releases walks through how to find the primary endpoint, p-value, and confidence interval in the press release itself. For a broader explanation of endpoints, see our clinical trial endpoints guide.

Common mistakes

• Trading on the press release without checking the registration. The press release is a marketing document. The ClinicalTrials.gov registration is the pre-specified trial design. If they disagree, the registration is the truth.
• Not checking the revision history. Companies can change the primary endpoint mid-trial, and the change is tracked. A mid-trial endpoint change is a red flag.
• Ignoring enrollment actuals. If actual enrollment is well below the estimated target, the trial is underpowered and the result is less reliable, even if the headline number looks good.
• Forgetting that single-arm trials have no control. Without a placebo or comparator arm, you cannot tell whether the drug caused the result. Single-arm trials are hypothesis-generating, not confirmatory.

Final checklist

• NCT number found from press release or company IR page
• Primary outcome measure identified on ClinicalTrials.gov
• Press release checked: does it clearly state whether the primary endpoint was met?
• Enrollment actuals vs estimated compared
• Trial arms reviewed: is there a control group?
• Revision history checked for any mid-trial endpoint changes
• Status tracked: Recruiting, Active not recruiting, Completed, Results posted
• Estimated primary completion date noted on the calendar

For the broader framework on how biotech catalysts drive stock prices, see our How to Invest in Biotech Stocks guide.