#1208 1998 · Pfizer · Pharmaceuticals
Pfizer worked backward from a trial side effect to find the disease its drug really cured
the problem
Pfizer's angina compound barely lowered blood pressure in trials, the one thing it was funded and built to do
background
In the mid-1980s, Pfizer chemists in Sandwich, England synthesized UK-92,480 (later named sildenafil), a compound designed to widen blood vessels around the heart and treat angina by inhibiting the PDE5 enzyme. Drug development runs forward by design: define the target disease, screen or design a molecule against it, then run trials that either confirm or kill the compound. In early 1990s Phase I and II trials, sildenafil failed its own brief — its effect on angina was too weak and its interaction with nitrate heart medication too dangerous to justify further cardiac development.
By the normal rules of pharma R&D, that result ends the project: the compound gets shelved and the budget gets written off as a sunk cost. But trial physicians and nurses had also been logging an unrelated, unrequested side effect volunteers kept reporting — spontaneous erections — data that had no home in a study measuring blood pressure and was almost dismissed as noise rather than signal.
what everyone would do
Standard pharma practice said to kill the compound: the trial data showed a clear failure against the funded target, nitrate interactions posed real cardiac risk, and continuing to chase angina with a weak signal would have burned more capital for the same result.
what they saw
A side effect is an effect with no matched disease yet. The usual order — pick the disease, then find a molecule — runs just as well in reverse: pick the effect, then find the disease it already treats.
the move
Instead of closing the file, the clinical team led by physician Ian Osterloh took the side-effect reports seriously as data and worked backward from the observed physiological effect to identify what condition it actually treated, rather than continuing to search forward for a cardiac use. Recent science on nitric oxide's role in erections gave them a mechanism; Pfizer redirected the entire trial program toward erectile dysfunction, running fresh studies against the new target instead of the original one.
why it works
Forward drug development is expensive because most candidate molecules fail against their intended target, and each new disease hypothesis requires new trials from scratch. Sildenafil had already cleared safety and dosing in humans; the only missing piece was the correct target, and the trial itself had already generated the evidence for that target as a side effect of testing something else. Reading backward from the observed physiological response to its cause let Pfizer skip years of the normal discovery funnel, because the compound-to-effect link the entire industry works to establish was already sitting, unused, in the safety logs.
the payoff
FDA approved sildenafil as Viagra for ED on March 27, 1998, turning a shelved heart drug into one of pharma's biggest launches.
where it breaks
It only works when the anomalous effect is safe, reproducible, and mechanistically explainable rather than a one-off artifact, and when the company culture treats unplanned trial data as a lead instead of noise to be filtered out before the report reaches decision-makers. It also requires willingness to enter a new, sometimes stigmatized therapeutic category with new regulatory and marketing risk, which is a real internal battle many companies choose not to fight.
what came after
Viagra became the textbook case for drug repurposing from unplanned trial side effects, and sildenafil later found a third life as Revatio for pulmonary arterial hypertension, entering commercial use for two conditions unrelated to the one it was originally designed and funded to treat.
references
- [1]Sildenafil: from angina to erectile dysfunction to pulmonary hypertension and beyondNature Reviews Drug Discovery, 2006pmc.ncbi.nlm.nih.gov