A clinical trial phase is a stage of human testing, and every stage answers one question before a drug moves ahead. Phase 1 asks whether a treatment is safe enough to keep studying. The later phases ask whether it works and whether the result holds up in large groups. The ladder is long by design: an MIT-led analysis of 406,038 trial entries found that cancer drugs succeed only 3.4 percent of the time (Biostatistics, 2019).
The clinical trial phases explained here follow the Investigational New Drug route that the FDA supervises for drugs and biologics. One late-stage shortcut, measuring a stand-in result instead of survival, gets its own treatment in this explainer on surrogate endpoints. More coverage of the evidence base sits on the site’s research hub. Devices and vaccines follow different review paths inside the same agency.
What does Phase 1 actually test?
Safety and dosage, tested in a small group. The FDA’s patient guide puts Phase 1 at 20 to 100 healthy volunteers, or people who have the condition, over several months (FDA, 2018). Cancer drugs break the pattern, because oncology Phase 1 studies enroll patients who have that cancer instead. A few dozen people can surface common side effects and a workable dose range.
Size is the point at this stage. Twenty people cannot prove a drug works. Twenty people can show what a safe dose looks like, and the FDA assigns exactly that job to Phase 1 (FDA, 2018).
Why do promising drugs stall in Phase 2?
Proving a drug works is the hardest gate. Phase 2 studies enroll up to several hundred patients for a few months to two years, and the FDA is blunt about the limits: these studies are not large enough to show whether the drug will be beneficial (FDA, 2018). The agency’s own funnel shows about 70 percent of drugs moving on after Phase 1, but only about 33 percent clearing Phase 2.
Attrition at this stage is the norm. The 2019 Biostatistics analysis of 21,143 compounds across 406,038 trial entries put oncology success at 3.4 percent, well under the 5.1 percent reported in earlier industry-curated samples (Wong, Siah and Lo, 2019). The same dataset found higher success odds when trials used biomarkers to select patients. Study length runs from several months to two years here, and longer is not the same as stronger.
What makes Phase 3 the decisive test?
Scale, duration and comparison. Phase 3 studies enroll 300 to 3,000 patients and run for one to four years, and the FDA designs them to show whether a product offers a treatment benefit to a specific population (FDA, 2018). Larger and longer studies are also the likeliest place for rare or long-term side effects to appear.
Approval rides on this stage, because Phase 3 supplies most of the safety data a reviewer reads. The FDA expects 25 to 30 percent of drugs to move on from here, which means most candidates die within sight of the market (FDA, 2018). The funnel is intentional, because the later a failure arrives, the more the testing costs. Phase 4 keeps the questions coming after approval, with post-market studies enrolling several thousand volunteers (FDA, 2018).
Conclusion: what the phase system is really for
A phase is a question, and the ladder exists so each question gets asked in the right order: safety in dozens of volunteers, then proof in thousands (FDA, 2018). Most drugs do not finish the climb, and the 2019 read of 406,038 trial entries explains why headlines about early results deserve patience (Biostatistics, 2019). The phase number attached to a headline is a measure of how much testing stands behind the claim. The stages move slowly on purpose, and this is information, not medical or insurance advice, for anyone weighing a trial decision. For related coverage, see Cohort Study or Randomized Trial? Why the Design Behind a Finding Changes the Claim.
