Professor University of Texas MD Anderson Cancer Center Houston, Texas, United States
Objectives: Drug combination trials have received growing interest in oncology due to the potential of drug combination to mitigate resistance and improve response to treatment. Many dose-finding designs for drug combination trials have been proposed without systematic incorporation of pharmacokinetic / pharmacodynamic (PK/PD) information. We aim to develop a novel Bayesian design that incorporates PK/PD information for identifying the maximum tolerated dose combination (MTC) for phase I drug combination trials.
Methods: We extend a recently developed semi-mechanistic dose-finding (SDF) model framework for a single drug [1,2] to drug combinations via the use of a Bliss model-based link function to model drug-drug interaction [3]. We propose Bayesian joint modeling of the PK, PD, and dose-limiting toxicity (DLT) outcomes. We perform extensive simulation studies to evaluate the operating characteristics of the proposed design in the setting of a motivating phase I/II trial of a synthetic anti-tubulin agent and an anti-neoplastic alkylating agent in the treatment of small cell lung cancer, where the DLT is mainly non-overlapping kidney toxicity and neurotoxicity associated with each drug, respectively.
Results: Our simulation studies show, when the underlying PK/PD mechanisms are reasonably understood, that the proposed design on average outperforms some common phase I trial designs for drug combinations, including the logistic model designs [4,5] and Bayesian optimal interval design for combination treatments (BOIN-Comb) [6], in terms of the percentage of correct selection of MTC and average number of patients allocated to MTC, under a variety of scenarios. The proposed design also yields similar safety profile to the comparator designs. A sensitivity analysis suggests that the performance of the proposed design is also robust to prior specification for the parameters in the link function.
Conclusions: By incorporating PK/PD information into the SDF model framework, the proposed design improves its performance upon the existing designs for phase I drug combination trials.
Citations: [1] Su X, Li Y, Müller P, Hsu C-W, Pan H, Do K-A. A semi-mechanistic dose-finding design in oncology using pharmacokinetic / pharmacodynamic modeling. Pharm Stat 21(6):1149-1166, 2022. [2] Yang C, Li Y. An extended Bayesian semi-mechanistic dose-finding design in oncology using pharmacokinetic and pharmacodynamic information. Stat Med 43(4):689-705, 2024. [3] Ashford JR. General models for the joint action of mixtures of drugs. Biom 37(3):457-474, 1981. [4] Riviere M-K, Yuan Y, Dubois F, Zohar S. A Bayesian dose-finding design for drug combination clinical trials based on the logistic model. Pharm Stat 13(4):247-257, 2014. [5] Mozgunov P, Knight R, Barnett H, Jaki T. Using an interaction parameter in model-based phase I trials for combination treatments: A simulation study. Int J Environ Res Public Health 18(1):345, 2021. [6] Lin R, Yin G. Bayesian optimal interval design for dose finding in drug-combination trials. Stat Methods Med Res 26(5):2155-2167, 2017.
