FDA leans toward one pivotal trial

The Food and Drug Administration is making a change that sounds technical and is anything but. For decades, drug makers were taught to expect two “adequate and well-controlled” studies before a new drug could win approval. In February 2026, FDA Commissioner Marty Makary and senior regulator Vinay Prasad said the agency’s new default will be one such trial, as long as it is backed by confirmatory evidence. They framed it as the end of the “two-trial dogma,” not a tweak around the edges (ajmc.com, medicalxpress.com). That sounds like a break with history. It is, but not a clean one. The two-trial expectation was never written into law as rigidly as people often assume. The 1997 Food and Drug Administration Modernization Act already allowed “substantial evidence” of effectiveness to come from a single trial plus confirmatory evidence, and FDA’s own 1998 and 2019 guidance documents said exactly that. The agency’s December 2019 draft guidance explained that the standard had “generally been interpreted” as two trials, while also acknowledging that one trial could be enough in the right setting (fda.gov, fda.gov). So the real shift is not that the FDA discovered a new legal power. It is that the agency is turning an exception into the starting point. A separate 2023 draft guidance focused entirely on how one adequate and well-controlled clinical investigation can support approval when paired with confirmatory evidence. That document was aimed at spelling out what kind of backup counts, from mechanistic evidence to data in related populations or indications. The February 2026 announcement takes that logic and makes it the agency’s public posture for new drug review (fda.gov, pharmaceuticalcommerce.com). In one sense, the FDA is admitting what has already been happening. Many drugs were already getting through on one pivotal study, especially in oncology and rare disease. A JAMA Network Open analysis of approvals from 1995 through 2017 found that approvals were increasingly based on fewer pivotal trials over time, alongside broader use of expedited programs. Another JAMA Network Open study found that more than half of novel drugs approved in 2020 were supported by a single pivotal trial. The evidence bar had been drifting for years before the agency said so out loud (jamanetwork.com, jamanetwork.com). That drift matters because one big trial is not the same thing as two independent tries. Replication protects against luck, hidden bias, and fragile results that look impressive once and then vanish. The FDA’s answer is that modern trials can be larger, cleaner, and more statistically disciplined, and that other forms of evidence can help confirm the result. Sometimes that is true. But “confirmatory evidence” is a broad category, and broad categories are where standards get soft. The agency is plainly choosing speed now and promising that rigor can be recovered through design, analytics, and follow-up (clinicaltrialsarena.com, biopharmadive.com). That pushes the real test of this policy past the moment of approval. If more uncertainty is tolerated before a drug reaches the market, then more burden falls on postmarket surveillance, required confirmatory studies, adverse event reporting, and real-world evidence systems that are supposed to catch what the preapproval package missed. The FDA has been moving in that direction for years. Now it is doing so more openly, with a policy that begins in a journal essay and ends in the pharmacy aisle (jamanetwork.com, biospace.com).