Combination dose finding trials for targeted/immuno oncology agents in early development: Lessons learned.

Abstract

e13595 Background: Combination therapy is commonly evaluated in clinical development of oncology investigational agent (IA). An IA may be combined with an approved agent (AA), or with another IA. The aim of this landscape analysis was to 1) summarize and evaluate the different dose finding study design options used in early clinical development for oncology targeted/immuno therapy combinations; 2) propose best-fit application scenarios for each trial designs. Methods: The data sources included 1) FDA approved oncology combination therapies from 2010 to 2021; 2) registered early-stage oncology combination trials in the US with protocol available at clinicaltrial.gov from 2015 to mid-2021; 3) Pfizer internal programs. This analysis focused on “IA + IA” and “IA + AA” combinations. We extracted data on modality type, target, starting dose rationale, trial design and approved dose (if applicable) of the combination. Starting dose considerations and pros/cons of different trial designs were assessed. Results: Both external and internal in scope trials were summarized, including 8 approved non-Pfizer oncology combinations, 32 clinicaltrial.gov registered external trials and 14 Pfizer internal programs. Most surveyed combination trials included an AA. Selection of starting doses and dose escalation design for the combination were the two critical factors influencing the efficiency of clinical development (time, cost and resources). Overlapping toxicities are key concerns resulting in more conservative combination starting doses. The combination dose finding was generally performed with one of the following designs: sequential design, parallel (staggered) design, healthy volunteer first-in-human single ascending dose before first-in-patient combination and monotherapy lead-in (intra-patient “crossover”). A comparison of the characteristics of these dose finding designs indicate that the choice of a particular design should consider factors including exposure-response relationships for efficacy and safety of both monotherapy and combination, the benefit/risk to trial participants, study population, etc. We proposed application scenarios for each trial designs by incorporating the above considerations and analyzing the collected trial examples in the database. The dose finding for combinations between two IAs should consider additional factors including the relative contributions of the two agents to efficacy and safety, the anticipated minimal dose/exposure level for efficacy, and potential pharmacokinetic/pharmacodynamic drug-drug interactions. Conclusions: Multiple design options for dose finding exist for oncology targeted/immuno therapy combinations involving IA. The choice of a suitable design should consider properties and available data of each agents involved, trial efficiency, as well as benefit/risk of trial participants and target indication.