Improving Dose-Optimization Processes Used in Oncology Drug Development to Minimize Toxicity and Maximize Benefit to Patients

Abstract

This review highlights strategies to integrate dose optimization into premarketing drug development and discusses the underlying statistical principles. Poor dose optimization can have negative consequences for patients, most commonly because of toxicity, including poor quality of life, reduced effectiveness because of inability of patients to stay on current therapy or receive subsequent therapy because of toxicities, and difficulty in developing combination regimens. We reviewed US Food and Drug Administration initial approvals (2019-2021) of small molecules and antibody-drug conjugates for oncologic indications to determine the proportion with a recommended dosage at the maximum tolerated dose or the maximal administered dose, to characterize the use of randomized evaluations of multiple dosages in dose selection, to describe the frequency of dose modifications at the recommended dosage, and to identify case examples that highlight key principles for premarket dose optimization during drug development. Herein, we highlight major principles for dose optimization and review examples of recent US Food and Drug Administration approvals that illustrate how investigation of dose- and exposure-response relationships and use of randomized dose trials can support dose optimization. Although there has been some progress, dose optimization through randomized dose evaluation in oncology trials is not routinely conducted. Dose optimization is essential to ensure that patients receive therapies which maximize efficacy while minimizing toxicity.

A multistate modeling and simulation framework to learn dose–response of oncology drugs: Application to bintrafusp alfa in non‐small cell lung cancer

“…The relationships between dose and multiple clinical end points (e.g., ORR, PFS, and OS) were bridged and quantified in one modeling practice. Moreover, the simulation step provides insight into the “learning” of the dose–response relationship, which is not a statistical testing problem, but rather a quantification of the probability of achieving better outcomes with one dose versus another based on integration of all available data 38 . Together with clinical safety data and associated exposure‐safety relationships, the framework described here can be powerful as a component of a Totality of Evidence approach to dose optimization for oncology therapies that is based on a robust characterization of dose/exposure‐response relationships 38‐40 .…”

High-dose short-term osimertinib treatment is effective in patient-derived metastatic colorectal cancer organoids

“…Another important aspect to take into account is that we nd really high concentrations in vitro that might be di cult to reach in patient tumours. Although it may be simplistic to directly compare these to MC and ITC, respectively, future research should indicate whether and how in vitro exposure can be extrapolated to feasibly attainable drug levels in the clinical setting-informing the design of phase I/II studies [55,56], including the exploration of alternative ways to boost plasma and potentially tumour concentrations [57].…”

Model‐Informed Selection of the Recommended Phase III Dose of the Inhibitor of Apoptosis Protein Inhibitor, Xevinapant, in Combination with Cisplatin and Concurrent Radiotherapy in Patients with Locally Advanced Squamous Cell Carcinoma of the Head and Neck

“…Model-based approaches are planned to confirm the selected xevinapant dose/regimen once data from these phase III studies become available (Figure 1). Viewed from a broader perspective and in the context of the growing recognition of the importance of dose optimization in oncology drug development, 33,[37][38][39][40] these analyses illustrate the value of quantitative clinical pharmacological contextualization for dose and regimen selection. Importantly, they illustrate the value of a holistic integration of preclinical and clinical PK, PD, safety, and efficacy data and associated E-R relationships, with a totality of evidence approach, 41 in which confidence can be gained from consistency across multiple approaches and data sources integrated in a mechanism-informed manner through modeling and simulation.…”