Open-Label Extension Study

An open-label extension follows the randomised phase of a trial and offers continued or newly started active treatment to participants who completed it, with everyone aware of what they are receiving. Its formal purpose is long-term exposure data. Regulatory expectations set out in ICH E1 for medicines intended for chronic use call for a substantial number of patients exposed for six months and a smaller number for a full year, and extensions are the usual way of accumulating that exposure.

The design appears throughout metabolic and endocrine development. Extensions of the liraglutide and semaglutide obesity programmes ran for years beyond their randomised phases, and related withdrawal designs made the counterpoint visible: in STEP 4, participants randomised to switch to placebo after an initial period of active treatment regained weight while those continuing did not, and after STEP 1 ended participants regained roughly two thirds of the weight lost within about a year.

An extension is a good source of one thing and a poor source of another. It can reveal adverse events too rare or too slow to appear in a 12-month randomised phase, show whether tolerance or antibody formation develops, and describe whether an effect persists on continued treatment. It cannot establish efficacy, because there is no concurrent control, no blinding, and no protection against the natural course of the condition.

The population itself is the trap. Everyone in an extension completed the randomised phase, meaning they tolerated the drug and chose to continue, so the cohort is enriched for responders and for people without early side effects. A safety rate calculated there underestimates what happens in an unselected population, and an efficacy figure quoted from it describes survivors rather than starters.