Abstract PR003: Multi-drug adaptive therapy protocol in vitro: An exercise in optimization for neuroblastoma

Abstract

Abstract From initiation, invasion, dissemination to local and distant progression, relapse, recurrence to metastasis – clinico-pathological development of cancer follows the ecological processes of a multi-species niche construction. Ecological interactions result in adaptive changes that are driven by Darwinian evolution. This can be observed in the heterogeneity of tumor cells and their differential response to therapy. Some tumor cells are sensitive to treatment and can be eliminated by therapy, while others are resistant and can survive and proliferate under therapy. Adaptive therapy is a novel approach that leverages the competitive dynamics between sensitive and resistant cells, by dynamically adjusting the therapy dose and schedule based on biomarkers of tumor burden. In this study, we propose a multi-drug adaptive therapy protocol that uses two drugs with orthogonal mechanisms of action, meaning that they target different pathways or processes in the cancer cells. We investigate how to optimize the switching time between the two drugs, based on the personalized response to therapy and the evolutionary trade-offs between resistance and fitness and describe a corresponding real-time in vitro protocol. We performed experiments using two human neuroblastoma cell lines, sknbe2c and lan1, which exhibit distinct resistance profiles to different therapies, have different genetic profiles and show different growth patterns. We chose the combination of cisplatin and vincristine as the drugs of interest, since they are part of the Rapid COJEC protocol, which is a standard multidrug regimen for treating high-risk neuroblastomas in children. We monitored the drug treatment effects using noninvasive real-time microscopy based on cellular holography, which allows us to capture the morphological and behavioral changes of the cells. We fitted the treatment response curves with a stochastic diffusion equation that incorporates evolutionary mechanisms such as speciation, cost-dependent predator-prey interaction and drift, to model the dynamics of drug resistance and sensitivity. Unlike conventional therapy, which follows a fixed schedule and dose, adaptive therapy uses biomarkers of tumor burden, such as imaging or circulating tumor cells, to guide the timing and intensity of treatment. We propose a stochastic model that predicts the optimal switching time between two drugs based on the personalized response to therapy, which is inferred from Bayesian updating of the model parameters. This results in a patient-specific response model that accounts for the uncertainty and variability in the data. Assuming that a reliable and sensitive measure of tumor burden is available, we demonstrate how our model can be used to design a customized multi-drug adaptive therapy protocol for each patient. We are currently testing the efficacy of our protocol over other conventional protocols, and we plan to translate it to patient derived xenograft mouse models of neuroblastoma in the future. Citation Format: Subhayan Chattopadhyay, Jenny Karlsson, David Gisselsson. Multi-drug adaptive therapy protocol in vitro: An exercise in optimization for neuroblastoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Translating Cancer Evolution and Data Science: The Next Frontier; 2023 Dec 3-6; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_2):Abstract nr PR003.