Dose Escalation

Dose escalation is how a first-in-human programme moves upward from a deliberately conservative starting point. In a single ascending dose study each new cohort receives a higher amount than the last, with a safety review between steps; a multiple ascending dose study repeats the exercise with repeated administration so accumulation can be observed. The starting point comes from animal toxicology, conventionally the no-observed-adverse-effect level scaled between species and divided by a safety factor.

The oncology tradition of the three-plus-three design, in which three participants are treated and three more added if one experiences a dose-limiting toxicity, remains widely used despite documented statistical inefficiency. The reason the conservatism exists is instructive: in the 2006 TGN1412 first-in-human trial, six healthy volunteers dosed on the same day suffered life-threatening cytokine release, after which staggered dosing and effect-based starting doses became standard practice in Europe.

What escalation produces is a tolerability boundary and a picture of how exposure grows with dose, not evidence that anything works. Its output feeds dose selection for later phases, where the aim is the smallest amount reaching the target exposure rather than the largest amount survivable.

Two misreadings are common. The first treats the top cohort of a phase 1 study as an established dose, when it is a ceiling explored in a handful of screened healthy volunteers over a short period, often with no efficacy measurement at all. The second, seen constantly around unapproved compounds, treats a published animal escalation range as though it transferred to humans by simple body-weight scaling, ignoring interspecies differences in clearance and immunogenicity.