(T-140) Modeling the frequency of toxicity-induced dose modifications and their impact on conventional exposure-response analysis in targeted therapy

Yanguang (Carter) Cao, n/a – Associate Professor, Division of Pharmacotherapy and Experimental Therapeutics, University of North Carolina at Chapel Hill

Objectives: Dose modification is ubiquitous across small-molecular targeted therapy in cancer treatment, largely due to their toxicity [1,2]. Among oncology drugs approved by the US Food and Drug Administration (FDA) between 2010-2020 for treating hematologic and solid tumor cancers, the median rates of dose reduction, interruption, and discontinuation rates were 28%, 55%, and 10%, respectively, and those rates could be higher in everyday clinical use [2]. Such high incidence of intercurrent events of dose modification significantly impacts the exposure-response (E-R) relationships, critical for assessing drug safety and effectiveness [3]. Specifically, dose modification can impact the predictions of actual drug exposure (e.g. acute or integrated drug concentrations in plasma), leading to biased E-R analysis. Therefore, the objective of the study is to assess the accuracy of the traditional E-R analysis, and to investigate potential alternative methods accounting for dose modification in targeted therapy that enhances the accuracy of E-R analysis.

Methods: A simulation platform considering dose modification was developed using R and the mrgsolve package, and duvelisib, a PI3K inhibitor approved to treat chronic lymphocytic leukemia and small lymphocytic lymphoma, was used as a model drug to perform the simulations. A population PK (popPK)-adverse event (AE) model was constructed, and dynamic simulations based on various and frequent AE-induced dose modifications were performed to assess the clinical relevance using results from the FDA multidisciplinary review for duvelisib, as well as to explore potential alternative methods for E-R analysis.

Results: Dynamic trial simulation using the popPK-AE model to implement dose modification strategies predicted the rates of dose reduction, dose interruption, and dose discontinuation to be 16%, 51%, and 18%, demonstrating the clinical relevance of the model. In addition, our simulations revealed a distortion of the E-R relationship due to dose modification towards a more significant one, compared to traditional E-R analysis assuming 100% adherence, indicating the reported non-significant E-R relationships are likely overshadowed by exposure overprediction. We proposed a more refined method to reconstruct the E-R relationship incorporating reported dose modification information more accurately, such as by using dose intensity-normalized exposure.

Conclusions: Our findings suggest a potential approach to include dose modification to improve the confidence in E-R analyses, therefore supporting robust assessment of drug safety and effectiveness.

Citations: [1] Roda, D., Jimenez, B., Banerji, U.. Are Doses and Schedules of Small-Molecule Targeted Anticancer Drugs Recommended by Phase I Studies Realistic?. Clin Cancer Res 1 May 2016; 22 (9): 2127–2132.

[2] McCabe C, Bryson E, Harvey RD. Dose derivation of oral anticancer agents: tolerability in late phase registration trials. Eur J Cancer. 2020;138(suppl 2):S51. https://event.eortc.org/ena2020/wp-content/uploads/sites/17/2020/10/EJC-138S2-ENA-2020-abstracts.pdf. Accessed May 12, 2024.

[3] Guidance for Industry: Exposure-Response Relationships — Study Design, Data Analysis, and Regulatory Applications. U.S. Food and Drug Administration (last updated Aug 24, 2018).