Clinical Trial Results: What They Mean

A trial press release usually leads with one number. That number is the trial's endpoint: the result the study used to judge the treatment. Different endpoints answer very different questions. The U.S. Food and Drug Administration (FDA) is clear about which is which.

Learn four of them and most headlines become readable.

FDA defines overall survival as "the time from randomization until death from any cause," and calls it "the most reliable cancer endpoint." In plain language, FDA describes it as "the average length of time patients are alive after the start of treatment."

Its strengths are that it is "easily and precisely measured" and "generally based on objective and quantitative assessment." Death is not a matter of interpretation.

Its weaknesses are practical. Overall survival requires "longer follow-up," "includes noncancer deaths," and "may be affected by switch-over of control to treatment or subsequent therapies." That last one matters. If the comparison group later gets the new drug anyway, the gap between groups can narrow for reasons unrelated to the drug.

Objective response rate is "the proportion of patients with tumor size reduction of a predefined amount and for a minimum time period." FDA's plain-language version is "the percentage of patients whose cancer shrinks or disappears after treatment." It is the sum of partial and complete responses, and it does not count stable disease.

The appeal is speed. Response rate can be "generally assessed earlier and with smaller sample size compared with survival studies."

The catch is real too. FDA notes that response "definitions vary among studies" and that it "may not always correlate with survival." A tumor can shrink without a person living longer.

Progression-free survival is "the time from randomization until objective tumor progression or death, whichever occurs first." FDA's plain-language version: "the average length of time after the start of treatment in which a person is alive, and their cancer does not grow or spread."

A close cousin, time to progression, "does not include deaths." FDA treats progression-free survival as the preferred regulatory endpoint of the two.

Its limitation is the one to remember. FDA says it is "potentially subject to assessment bias, particularly in open-label studies," and that it "may not always correlate with survival." Progression is defined by measurements on scans, and measurements involve judgment. That is why FDA advises that tumor assessments "generally should be verified by central reviewers blinded to study treatments."

Tumor response is usually measured with RECIST, which FDA calls "a standard way to measure how well a patient responds to treatment. It is based on whether tumors shrink, stay the same, get bigger, or spread to a new part of the body."

Results are often summarized as a hazard ratio. FDA recommends reporting "the hazard ratio (HR) obtained from a Cox proportional hazard model as a summary measure, along with its 95% confidence interval."

A hazard ratio compares how fast an event, such as death or progression, is happening in one group against the other across the study period.

• A ratio of 1 means the two groups looked the same.
• Below 1 favors the new treatment. A hazard ratio of 0.75 means the event was happening about a quarter less often in that group over the study.
• Above 1 favors the comparison group.

Two cautions. First, a hazard ratio is a comparison of rates across a whole population. It is not your personal odds. Second, always look at the confidence interval beside it, the range the true value is likely to sit in. FDA looks at whether "a 95% confidence interval for the overall survival HR excludes a clinically relevant threshold." A range that crosses 1 means the result is compatible with no difference at all.

FDA also warns that "use of early or immature overall survival data (i.e., relatively few deaths) can cause high uncertainty in treatment effect estimates." Early survival numbers move.

There is a legal deadline, and it is more generous than most people assume.

Under federal rules, results information "must be submitted no later than 1 year after the primary completion date of the applicable clinical trial." The primary completion date is when the last participant was measured for the main endpoint, not when the trial finally closes.

That year can stretch legally. For a drug not yet approved for any use, or seeking approval for a new use, submission may be delayed until 30 days after FDA approves it, refuses it, or the application is withdrawn, with an outer limit of two years from the certification. Sponsors can also request an extension for "good cause," which the Director may grant, and a denial can be appealed within 30 days.

Publication in a medical journal is a separate process with no deadline at all. So a trial you joined in 2024 may post summary results in 2027 and appear in a journal later still. Ask your study team when results are expected and whether the site will notify you.

Every registered trial carries an identification number beginning with NCT, assigned in the ClinicalTrials.gov registry. Write yours down at enrollment. It is the most reliable way to find your study later.

The National Cancer Institute's advanced trial search includes a Trial ID field, which it describes as "an identification (ID) number that makes it easy to find a specific trial." If the number is lost, you can also search by drug name, phase, lead organization, or investigator.

Endpoint results describe averages across people who met that trial's eligibility criteria. You may differ in stage, prior treatment, organ function, or biomarker status. Any of those can change what the number means for you.

Take three questions to the person who knows your case: were people like me in this trial, does this result change what you would recommend for me, and what would you want to see before it does?

• Which endpoint was the primary one, and did it meet its target?
• What was the hazard ratio, and what was the confidence interval around it?
• Was overall survival measured, and is it mature yet?
• Was the trial open-label, and was there independent blinded review of scans?
• How do the side effects compare, not just the benefit?
• Were people with my stage and prior treatment included?

• U.S. Food and Drug Administration — Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics.
• U.S. Food and Drug Administration — Patient-Friendly Language for Cancer Clinical Trials.
• U.S. Food and Drug Administration — Approaches to Assessment of Overall Survival in Oncology Clinical Trials.
• Electronic Code of Federal Regulations — 42 CFR 11.44, Deadline for submission of clinical trial results information.
• National Cancer Institute — Help Using the Clinical Trials Search Form.

Helpful next pages include What Are Endpoints in a Clinical Trial?, The Phases of Clinical Trials, What Is Randomization in a Clinical Trial?, Clinical Trial vs. Standard Treatment, Reporting Side Effects in a Clinical Trial, Decentralized Clinical Trials: What Patients Should Ask.