Optimal dose selection in oncology biologics development: Need for a paradigm shift.

Abstract

3093 Background: In oncology, the recommended phase 2 dose (RP2D) is typically selected from the first-in-patient (FIP) dose escalation and then carried forward to late stage trials. The RP2D is often informed by limited data from a relatively small number of patients. Further dose optimization in late stage trials is rare as these trials often evaluate only one dose/dosing regimen. Traditionally oncology development relied on establishing an MTD with the expectation this dose would ultimately provide an optimal benefit/risk profile. However, for biologics there are clear examples suggesting a need to reconsider this approach to dose selection during drug development. The present review attempts to evaluate the dose selection approaches for approved oncology biologics to identify opportunities for improved dose optimization of future oncology drugs. Methods: The dose selection steps for oncology biologics approved by the USFDA from 2010 to 2020 were reviewed. The primary focus of the review was RP2D selection based on early clinical data, doses tested in pivotal trials, exposure-response and dose justification for the proposed dose, and post-marketing requirements (PMR) or post-marketing commitments (PMC) related to dose optimization. Results: The dose selection process for a total of 22 biologics was analyzed. Among these, 17 (77%) did not identify MTD during the FIP trial, and 6 (27%) tested more than one dose in the registration trials. In the initial approval, 2 (10%) had PMR or PMC for further dose optimization. For majority of the biologics the dose justification was based on 1) receptor occupancy 2) biomarker response 3) early clinical activity and 4) target efficacious concentration based on xenograft data. The majority of biologics had shown a shallow exposure-response relationship in safety or efficacy end points which partly can be attributed to the common practice of testing single dose level in the registrational trial. The dose recommended in the FDA label appears to be highly correlated to the RP2D determined in FIP studies. Conclusions: Oncology biologics have heavily relied on pharmacodynamic, early signs of efficacy, and safety data from the limited number of patients in the FIP dose escalation to identify the dose for pivotal trials. In most cases, there have been no dose ranging or optimization based on the final therapeutic outcome. While integration of relevant biomarker, PK, and PD data for optimal dose selection is an important aspect, there is a need to change the paradigm to consider more robust dose optimization, especially in early stage development. Dose ranging assessments incorporating therapeutically relevant endpoints and adequate sample size should be considered, ideally prior to pivotal trials.