What Is a Pivotal Trial? Spotting the FDA Approval Study

You read a headline: “Phase 3 trial succeeds.” The stock jumps 40%. You buy in. Six months later, no FDA filing. The stock drifts back down. What happened?

The trial wasn’t pivotal. It was a Phase 3, but not a registrational study. The company ran it for strategic or exploratory reasons, and the data, however positive, wasn’t designed to support an FDA approval. Understanding the difference between a pivotal trial and a regular Phase 3 is the difference between trading a real catalyst and chasing noise.

The Solution

A pivotal trial is a registrational study: a trial designed to serve as the primary basis for FDA approval. Not all Phase 3 trials are pivotal. You can tell the difference by checking ClinicalTrials.gov, company pipeline slides, and the trial’s design characteristics. Pivotal trial success leads to an NDA/BLA filing, which leads to a PDUFA date, which is a stock-moving catalyst. Non-pivotal Phase 3 success is a pipeline signal, not a filing event.

Step 1: Understand what makes a trial “pivotal”

The FDA requires that a pivotal trial be adequate and well-controlled per 21 CFR 314.126. That regulation specifies characteristics including:

• A clear statement of the trial’s objectives
• A design that permits a valid comparison with a control
• Adequate measures to minimize bias (randomization, blinding)
• Well-defined and reliable methods for assessing treatment response
• A pre-specified primary endpoint

A pivotal trial is the study a company submits to the FDA as the primary evidence of efficacy. It’s the trial whose data the FDA reviews to decide: approve or reject.

Step 2: Know the two-trial standard (and its exceptions)

Historically, the FDA prefers two adequate and well-controlled trials to establish substantial evidence of effectiveness. But the agency accepts a single registrational trial in several situations:

• Rare diseases: where patient populations are too small for two large trials
• Oncology: where a single large, well-designed trial with compelling results can suffice
• Situations where one trial provides particularly strong evidence (large sample size, multicenter, strong endpoint)

This matters for investors. A company running a single registrational trial in a rare disease faces less statistical risk of replication failure than a company that needs two trials. But a single-trial approval still requires strong data. One trial, one shot.

Step 3: Check ClinicalTrials.gov for registrational signals

Go to ClinicalTrials.gov and search the trial by NCT number or company name. Look for these indicators:

• Primary purpose: Treatment (not observational or supportive)
• Masking: Double-blind (or at least single-blind with a control arm)
• Allocation: Randomized
• Trial title: Often includes “pivotal” or “registrational” in the official title
• Enrollment target: Large enough to power a primary endpoint (typically hundreds to thousands)

For a full walkthrough of the site, see our ClinicalTrials.gov guide.

Step 4: Check company pipeline slides

Companies explicitly label which trials are registrational. In investor presentations and 10-K filings, pipeline slides show trial phase and whether the trial is intended to support a filing. If the slide says “pivotal,” “registrational,” or “NDA/BLA-supporting,” it’s the real deal. If it says “Phase 3” without those qualifiers, dig deeper before treating the readout as a filing catalyst.

Step 5: Use real examples to calibrate

Cogent Biosciences (COGT): The PEAK trial (NCT05208047) is the registrational trial for bezuclastinib + sunitinib in gastrointestinal stromal tumors (GIST). It’s a Phase 3, randomized, open-label, multicenter trial designed to support an NDA. The FDA accepted the NDA with Priority Review and a PDUFA date of November 30, 2026. Registrational trial success directly translated to a filing and a PDUFA date: a stock-moving catalyst. Read the full Cogent bezuclastinib GIST PDUFA analysis for the investment thesis.

Cogent also ran the SUMMIT trial (non-advanced systemic mastocytosis) and the APEX trial (advanced systemic mastocytosis) as separate registrational trials for separate indications. Same drug, three registrational studies, three potential NDAs. That’s how a single molecule can generate multiple catalysts.

Stoke Therapeutics (STOK): The EMPEROR trial (NCT06872125) is the registrational Phase 3 for zorevunersen in Dravet syndrome. It’s a randomized, sham-controlled trial of 162 patients. When EMPEROR reads out in Q3 2027, the data will determine whether Stoke can file an NDA. This is a single-registrational-trial design in a rare disease. One trial, one shot. The full Stoke zorevunersen Dravet analysis covers the mechanism and risk profile, and the four-year open-label durability data tracks the long-term extension.

Step 6: Understand the accelerated approval shortcut

There’s an exception to the “large Phase 3 registrational trial” rule. Under the FDA’s Accelerated Approval pathway, a company can seek approval based on a surrogate endpoint: a lab measurement or biomarker that’s reasonably likely to predict clinical benefit. This can allow a single Phase 2 trial to serve as the registrational trial if the surrogate endpoint is accepted.

Real example: Sarepta’s Elevidys (DMD gene therapy) was approved based on dystrophin protein production as a surrogate endpoint, not a functional improvement trial. That’s accelerated approval from a non-Phase 3 registrational study. See our accelerated approval guide for the full pathway and its risks.

Common mistakes

Assuming all Phase 3 trials lead to filings. They don’t. Some Phase 3 trials are exploratory, supportive, or designed to expand the label after an initial approval. Only registrational trials lead to NDA/BLA filings. Verify before trading.

Ignoring trial design flaws. A registrational trial that’s open-label, single-arm, or lacks a control arm is harder for the FDA to accept as adequate and well-controlled. The PEAK trial was open-label. The FDA accepted it, but the design was a noted risk factor. Always check whether the trial design meets the 21 CFR 314.126 standard.

Treating Phase 2 data as registrational. Phase 2 data can support accelerated approval in special cases, but for standard approval, Phase 2 is a hypothesis-generating step. Only about 33% of Phase 2 drugs reach Phase 3. Don’t trade Phase 2 data as if it were registrational. See our clinical trial phases explainer for the full transition rates.

Final checklist

• Verify the trial is registrational on ClinicalTrials.gov (randomized, controlled, treatment purpose)
• Check company pipeline slides for “pivotal” or “registrational” labeling
• Determine whether it’s a single-registrational-trial or two-trial design
• Check if accelerated approval with a surrogate endpoint is in play
• Map registrational trial success to NDA/BLA filing to PDUFA date
• Understand the trial endpoints being measured, because they define what “success” means

A pivotal trial is the trial that matters for FDA approval. Everything else is supporting data. When you see “Phase 3 data” in a headline, your first question should be: is this the registrational study, or is it noise? The answer determines whether you’re trading a catalyst or chasing a bounce.