A Seamless Phase 1/2 Clinical Trial Designed to Expedite Oncology Drug Development
*Jeff Wetherington, GSK 

Keywords:

Our goal within oncology clinical statistics is to expedite the progression of assets through the portfolio by reducing the time interval between FTiH and Commit to Medicines Development (C2MD) whilst increasing the value of the information leveraged into the C2MD decision making process. Historically the C2MD decision to progress a drug candidate into phase 3 confirmation for a chosen tumor histology is based upon evidence generated in the cohort expansion phase (i.e., phase 2) from a single dose identified in phase 1 on the basis of 3 - 6 patients’ worth of safety data at the MTD. Rarely are results from other doses formally integrated into the candidate progression decision making process. This development paradigm often results in expensive post-marketing commitments requiring sponsors to perform clinical trials to evaluate lower doses for the small percentage of oncology drug candidates gaining market approval. One of the key contributors to identifying a single dose for phase 3 highly likely to be confirmed as both safe and effective is skillful clinical trial design. In collaboration with Berry Consultants, we have developed a seamless phase 1/2 trial design to efficiently optimize dose selection with respect to safety and efficacy. The essence of the design allows for simultaneous implementation of expansion cohorts evaluating safety and efficacy against target tumor histologies while dose escalation evaluating safety is still ongoing. Once a dose is cleared for safety and assuming there is an additional PK/PD rationale, patients in the expansion cohorts are allocated to the newly cleared dose in a randomized fashion. Allocation to each expansion cohort continues until such time as enrollment to that dose is halted either for futility or sufficient evidence of benefit. Ultimately, the C2MD decision is based upon a formal integration of results gleaned from all doses. A comprehensive evaluation of this design through simulation has demonstrated a reduction in the time to C2MD; a significant improvement in the ability to identify a safe and efficacious dose at the end of the cohort expansion phase; and substantial improvement in “program power” – the probability of confirming in phase 3 a safe and effective dose.