An evaluation of designs for Phase I/IIa dose-finding studies in Tuberculosis.

Early development of tuberculosis (TB) treatments often separates dose-finding in healthy volunteers (Phase I) and safety and early activity in patients (Phase IIa). The BTZ-043 study (NCT04044001) is a recent study that combined these two objectives in patients in a single study. In this work, we describe and compare three different design options which consider safety and activity endpoints differently in the dose escalation process. In simulations we show that the design that incorporates information about activity together with safety in the dose escalation process allows more precise estimation of the optimal dose and leads to a higher power on average in selecting at least one suitable dose at the end of the study compared to the design that considers only the safety endpoint for dose escalation.