Drug development

What Is a Primary Clinical Trial Endpoint?

A primary endpoint is the main outcome a clinical trial is designed to measure, chosen before the trial starts. Whether it is met usually decides whether the trial counts as a success.

Key points

  • The primary endpoint is the single outcome, or small set of outcomes, a trial is designed and sized to evaluate. It must be specified before the results are known.
  • Secondary endpoints provide supporting information but generally cannot rescue a trial that missed its primary endpoint.
  • Clinical endpoints measure how patients feel, function, or survive. Surrogate endpoints, such as a lab value, stand in for them.
  • Surrogate endpoints can speed development, but some have badly misled: drugs have improved a surrogate while increasing deaths.
  • When reading trial news, check whether the reported success was on the prespecified primary endpoint.

Every well-designed clinical trial is built around one central question, such as "Does this drug lower the risk of heart attack?" or "Does it help patients with this cancer live longer?" The primary endpoint is the measurement that answers that question. Understanding it is the single most useful skill for reading clinical trial news.

The definition

An endpoint is an outcome measured in a clinical trial to judge the effect of a treatment. The primary endpoint is the one the trial is specifically designed to evaluate. It is used to calculate how many participants the trial needs, and its result usually determines whether the trial is declared a success.

International guidelines on the statistics of clinical trials (ICH E9) and FDA guidance on multiple endpoints both emphasize the same principle: the primary endpoint must be specified in advance, in the trial protocol and statistical analysis plan, before anyone has seen the results.

Why it has to be chosen in advance

If researchers could pick their main outcome after seeing the data, they could almost always find something that looked positive. A trial measuring 20 different outcomes will, by chance alone, often show an apparent effect on one of them even if the drug does nothing. Committing to a primary endpoint beforehand prevents this kind of cherry-picking. It is why a trial that "missed its primary endpoint" is generally considered to have failed, even if other measurements looked good.

Primary, secondary, and exploratory endpoints

Type Role Example in a heart-failure trial
Primary The main question; decides success Time to cardiovascular death or hospitalization for heart failure
Secondary Supporting evidence; tested only in a prespecified order Change in a quality-of-life score; all-cause death
Exploratory Generates hypotheses for future trials Changes in blood biomarkers

Secondary endpoints matter, but they are usually tested in a prespecified order and only after the primary endpoint has succeeded. A positive secondary result in a trial that missed its primary endpoint is treated as a hypothesis, not as proof.

Clinical endpoints and surrogate endpoints

Endpoints come in two broad kinds.

Clinical endpoints directly measure how a patient feels, functions, or survives. Examples include death, stroke, hospitalization, pain scores, and the ability to walk a set distance.

Surrogate endpoints are measurements that stand in for a clinical endpoint because they are expected to predict it. Examples include blood pressure, LDL cholesterol, blood sugar control (HbA1c), viral load, and tumor shrinkage.

Surrogate endpoints can make trials smaller and faster, because a lab value changes long before a heart attack or death would occur. The FDA publishes a table of surrogate endpoints that have been the basis of drug approvals.

The FDA-NIH BEST glossary groups surrogate endpoints by how well they have been shown to predict clinical benefit:

  • Validated surrogate endpoints are supported by strong evidence that an effect on the surrogate predicts clinical benefit. They can support traditional approval.
  • Reasonably likely surrogate endpoints have strong biological reasoning and some clinical support, but are not fully validated. They can support accelerated approval, which requires confirmatory trials afterward.
  • Candidate surrogate endpoints are still being evaluated.

When surrogates mislead

Surrogate endpoints have occasionally pointed in exactly the wrong direction:

  • CAST (1989). Irregular heartbeats after a heart attack were known to predict sudden death, and the drugs encainide and flecainide suppressed those irregular beats. In the Cardiac Arrhythmia Suppression Trial, however, patients given those drugs had a higher rate of death from arrhythmia than patients given placebo.
  • Torcetrapib (2007). This drug raised "good" HDL cholesterol by 72% and lowered "bad" LDL cholesterol by 25% after a year. In a large trial, it also increased the risk of cardiovascular events (hazard ratio 1.25) and death from any cause (hazard ratio 1.58).

Both drugs did exactly what the surrogate measured. Neither helped patients. A drug can affect a surrogate through one pathway while harming patients through another, which is why confirmatory trials on clinical outcomes remain important.

Composite endpoints

Many trials use a composite endpoint that combines several events, such as "cardiovascular death, heart attack, or stroke." Composites let trials collect enough events to reach a conclusion with fewer patients. The trade-off is interpretation: if a composite succeeds mainly because of its least serious component, the headline result may overstate the benefit. Good trial reports show the result for each component separately.

Reading endpoint claims in trial news

When a company announces results, ask:

  1. Was the primary endpoint met? If the announcement leads with secondary or subgroup results, the primary may have failed.
  2. Is the endpoint clinical or surrogate? If surrogate, is it validated for this disease?
  3. Did the endpoint change between the registered protocol and the announced results? Trial registrations on ClinicalTrials.gov list the primary outcome measures and record changes over time.
  4. How large was the effect, not only whether it was statistically significant? See What Is Statistical Significance in Clinical Trials?

On Spectral Genomics, clinical drug trial results show each trial's registered primary endpoint, whether it was met, and whether that result is peer-reviewed or company-reported.

Sources

  1. FDA-NIH Biomarker Working Group. BEST (Biomarkers, EndpointS, and other Tools) Resource.Regulator
  2. U.S. Food and Drug Administration. Table of Surrogate Endpoints That Were the Basis of Drug Approval or Licensure.Regulator
  3. U.S. Food and Drug Administration. Guidance: Multiple Endpoints in Clinical Trials.Regulator
  4. Cardiac Arrhythmia Suppression Trial (CAST) Investigators. Preliminary report: effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after myocardial infarction. New England Journal of Medicine. 1989;321(6):406–412.Peer-reviewed
  5. Barter PJ, et al. Effects of torcetrapib in patients at high risk for coronary events. New England Journal of Medicine. 2007;357(21):2109–2122.Peer-reviewed
  6. International Council for Harmonisation. E9: Statistical Principles for Clinical Trials.Regulator

This article is for education. It is not medical advice and not investment advice. Figures are drawn from the sources above and describe historical averages, not the prospects of any particular drug. See our editorial standards and disclaimer.