As a prerequisite for marketing approval, regulatory agencies typically require that a sponsor conducts a randomized controlled trial, where the control group is either a placebo, an active control, or a standard of care control arm. There are, however, situations in which a placebo control may be unethical or where neither an active control nor standard of care exists. In these scenarios, potential solutions include constructing a historical control arm using real-world data (RWD) or using an in silico method to construct a virtual patient control arm.
In in silico trials, investigational drugs are tested in virtual patients using sophisticated computational modeling and simulation techniques. In silico approaches may be used to evaluate drug candidates prior to testing in humans or to serve as an external synthetic control arm in single-arm prospective human trials.