Advancing Dose Optimization: Innovations and Practices
Chair: Victoria Chang, PhD (JBS | BeiGene)
Speaker: Ruitao Lin, PhD (The University of Texas MD Anderson Cancer Center)
Title: DEMO: Bayesian Adaptive Dose Exploration-Monitoring-Optimization Design Based on Short, Intermediate, and Long-term Outcomes
Abstract: While phase 1-2 designs provide a methodological advance over phase 1 designs for dose finding by using both clinical response and toxicity, a phase 1-2 trial still may fail to select a truly optimal dose. This is because early response is not a perfect surrogate for long term therapeutic success. To address this problem, a generalized phase 1-2 design first uses a phase 1-2 design’s components to identify a set of candidate doses, adaptively randomizes patients among the candidates, and after longer follow up selects a dose to maximize long-term success rate. In this paper, we extend this paradigm by proposing a design that exploits an early treatment-related, real-valued biological outcome, such as pharmacodynamic activity or an immunological effect, that may act as a mediator between dose and the clinical outcomes including tumor response, toxicity, and survival time. We assume multivariate dose-outcome models that include effects appearing in causal pathways from dose to the clinical outcomes. Bayesian model selection is used to identify and eliminate biologically inactive doses. At the end of the trial, a therapeutically optimal dose is chosen from the set of doses that are acceptably safe, clinically effective, and biologically active to maximize restricted mean survival time. Results of a simulation study show that the proposed design may provide substantial improvements over designs that ignore the biological variable.
Speaker: Yong Zang, PhD (Indiana University)
Title: Great Wall: A Generalized Phase I/II Dose Optimization Design for Drug Combination Trials with Survival Endpoint
Abstract: Most phase I/II clinical trial designs assume that selecting the optimal dose based on early outcomes will also lead to maximum long-term survival benefits. However, this assumption is not true in many clinical scenarios, often due to high rates of relapse following initial response. To address this problem, we propose the Great Wall design, a general approach for dose optimization in drug-combination trials. The Great Wall design employs a “divide-and-conquer” algorithm to address the issue of partial order of toxicity and leverages early outcomes to eliminate dose combinations that are excessively toxic or less efficacious. It constructs a candidate set of the promising dose combinations using the mean utility method balancing the toxicity and early efficacy outcomes. The patients assigned to the candidate set are followed to collect the survival outcomes and the final optimal dose combination is then selected to maximize the survival benefit. A simulation study confirmed the desirable operating characteristics of the Great Wall design, compared with other conventional phase I/II designs for drug-combination trials. The Great Wall design is modular and can be tailored to suit various clinical contexts. It is particularly valuable in situations where investigators intend to follow patients for extended periods to assess survival outcomes.
Speaker: Brad Carlin, PhD (PharmaLex)
Title: Bayesian Adaptive Dual Objective Phase I-II Designs for Personalized Dose-Finding with Combination Therapies
Abstract: We describe a Phase I-II design for testing a new cancer drug, where interest lies in its value as both a monotherapy, and in combination with a second drug. Phase I begins with a monotherapy “run-in” period that can model efficacy alone, or use a clinical utility index to trade off safety and efficacy, and can capture correlation among the two competing endpoints. Our bivariate dosing model employs Bayesian Optimization (BO) over a bivariate Gaussian process approximation, providing smooth and efficient estimation over the two-dimensional dosing grid. Ultimately, the trial identifies a recommended dose region, from which the two or three doses can be selected for Phase II comparison with the optimal monotherapy dose and placebo. We evaluate both stages of our design using simulation, where in Phase I we study the probability of correct dose selection and a related root mean squared error (RMSE) criterion, while in Phase II we return to the traditional benchmarks of Type I error and power. The proposed design appears to satisfy modern Project OPTIMUS-inspired regulatory guidelines for Phase I-II oncology trials, while offering improved efficiency, flexibility, and interpretability. We also offer an illustration of the approach in a non-cancer setting, where a sponsor is interested in a design for the development of an intraocular implant that combines two topical agents. In this setting, response heterogeneity is expected to exist with respect to a key binary covariate (a particular characteristic of the lens of the eye), and we seek a design that accommodates this feature.
Speaker: Xiaojiang Zhan, PhD (Servier Pharmaceutical)
Title: A Bayesian Latent-Subgroup Phase I/II Platform Design to Co-Optimize Doses in Multiple Indications
Abstract: The US Food and Drug Administration (FDA) launched Project Optimus to reform the dose optimization and dose selection paradigm in oncology drug development, calling for the paradigm shift from finding the maximum tolerated dose to the identification of optimal biological dose (OBD). Motivated by a real-world drug development program, we propose a master-protocol-based platform trial design to simultaneously identify OBDs of a new drug, combined with standards of care or other novel agents, in multiple indications. We propose a Bayesian latent subgroup model to accommodate the treatment heterogeneity across indications and employ Bayesian hierarchical models to borrow information within subgroups. At each interim, we update the subgroup membership and dose-toxicity and efficacy estimates, as well as the estimate of the utility for risk-benefit tradeoff, based on the observed data across treatment arms to inform the arm-specific decision of dose escalation and de-escalation and identify the optimal biological dose for each arm of a combination partner and an indication. The simulation study shows that the proposed design has desirable operating characteristics, providing a highly flexible and efficient way for dose optimization. The design has great potential to shorten the drug development timeline, save costs by reducing overlapping infrastructure, and speed up regulatory approval.
