Chair: James Rogers, PhD (Metrum Research Group)
Co-Chair: Nicole Li, PhD (BeiGene)
Abstract: The FDA’s Project Optimus initiated a new era to “Reform the dose optimization and dose selection paradigm in oncology”. Various workshops have been hosted on this topic in addition to FDA’s guidance on “Optimizing the Dosage of Human Prescription Drugs and Biological Products for the Treatment of Oncologic Diseases” since January 2023. The historical paradigm for dose selection based on the maximum tolerated dose may lead to doses and schedules of molecularly targeted therapies that are inadequately characterized before initiating registration trials. In this session, our speakers will present their journey on dose optimization in oncology via designs, simulations, and practical considerations. The session as a whole will illustrate the range of considerations from early to late phase development that are essential to getting the right dose in the right patient at the right time.
Speaker: Heng Zhou, PhD (Merck & Co.)
Title: Design Strategies and Considerations for Oncology Dose Optimization
Abstract: The U.S. Food and Drug Administration launched Project Optimus with the aim of shifting the paradigm of dose-finding and selection toward identifying the optimal biological dose that offers the best balance between benefit and risk, rather than the maximum tolerated dose. However, achieving dose optimization is a challenging task that is considerably more complicated than identifying the maximum tolerated dose, both in terms of design and implementation. The presentation will provide a comprehensive review of various design strategies for dose-optimization trials, including phase 1/2 and 2/3 designs, and highlight practical consideration as well. Case examples will be presented for better demonstration. A Bayesian optimal phase 2 design for jointly monitoring efficacy and toxicity will be also discussed, focusing on its application to dose-optimization trials.
Speaker: Huan Cheng, PhD (BeiGene)
Title: A simulation study comparing backfilling and randomized expansion in oncology dose optimization
Abstract: Backfilling strategies have gained increasing interests in oncology clinical trials, particularly with the growing emphasis on dose optimization by health agencies. Backfilling can be implemented during the dose escalation phase, potentially accelerating the development process. Non-randomized backfilling is often favored because of its ease of implementation, though it may introduce imbalances in some important prognostic factors. Another common approach for dose optimization is a two-stage design, where the first stage follows a traditional dose escalation design to identify MTD, followed by a randomized expansion of two or more doses to determine the optimal dose. However, this two-stage approach typically requires a relatively larger sample size and longer time to complete compared to the backfilling approach. In this study, we explored different approaches to select optimal doses and conducted simulations to compare the performance of backfilling and randomized expansion in terms of optimal dose selection rate and other metrics. We further discussed how to address the issues of potential unbalanced prognostic factors which may contribute to difference in efficacy that is not attributed to dose, when using backfilling approach, to improve the optimal selection rate.
Speaker: Jonathan French, PhD (Johnson & Johnson)
Title: Dose Selection for Novel Modalities in Oncology: Science vs Empiricism
Abstract: Drug development in oncology is increasingly focused on targeted therapies and novel modalities, such as cell-based therapies, T-cell redirecting antibodies, and antibody-drug and radionuclide conjugates. Dose selection for these new modalities can be challenging due to the multifaceted nature of the dosing strategies, including the incorporation of step-up dosing to improve safety, and the possibility of non-monotonic dose-response relationships. With these therapies, we often develop a significant understanding of the mechanism of action through in vitro and preclinical experiments prior to entering clinical studies. Mechanistic pharmacokinetic-pharmacodynamic (PK-PD), physiologically-based pharmacokinetic (PBPK), and quantitative systems pharmacology (QSP) models aim to integrate this information with the relevant knowledge of the disease, therapeutic mechanisms, and pharmacology. In this talk, I will discuss case studies where PK-PD, PBPK, and QSP modeling have been used to inform the selection of dosing regimens and argue that, for novel modalities, dose selection in oncology should routinely be supported with mechanistically-informed modeling.
Speaker: Dan Polhamus, PhD (Metrum Research Group)
Title: Model-Based Dose Optimization in the Context of Adaptive Risk Mitigation: A Case Study with Tisotumab-Vedotin
Abstract: Tisotumab vedotin (TV) is a tissue factor-directed antibody-drug conjugate (ADC) that is approved in the United States at a dose of 2.0 mg/kg every 3 weeks (Q3W) for adult patients with recurrent or metastatic cervical cancer following disease progression on or after chemotherapy. TV exposure is positively correlated with ocular adverse events (OAEs), which have been identified as prespecified adverse events of interest in TV clinical studies. An eye care plan (ECP) including prophylactic steroids, vasoconstrictor eye drops, lubricating eye drops throughout treatment, and cold packs during infusion was introduced with the goal of reducing Grade >=2 OAEs. Clinicians were instructed to reduce dosing for patients who continued to experience OAEs, or AEs of several other body systems. The goal of this analysis was to investigate alternative dosing regimens in other cancer types under the presence of dose modification and the ECP.
A total of 757 patients who received TV monotherapy or combination (with either carboplatin, bevacizumab, or pembrolizumab) across seven clinical studies were used to develop discrete-time Markov models (DTMM) to characterize exposure–response (E–R) relationships of exposures of both ADC and the microtubule‐disrupting agent, monomethyl auristatin E, on the incidence, severity, and longitudinal time course of grade ≥2 OAEs in patients with advanced solid tumors. Dynamic simulations from this DTMM leveraged models of other events related to dose modification and patient selection to characterize occurrence of OAEs in virtual target populations. Using these simulations, alternative dosing regimens were evaluated across several tumor types (tumors of the ovary, cervix, endometrium, bladder, prostate, esophagus, squamous cell carcinoma of head and neck (SCCHN), and non-small cell lung cancer (NSCLC)). This work demonstrates the use of the simulation platform in the evaluation of various regimens in SCCHN.
