Analysis and modeling of cancer drug responses using cell cycle phase-specific rate effects

Abstract

A major challenge to improving outcomes for patients with cancer is the identification of effective therapeutic strategies that can prevent tumor cell proliferation. Here we sought to gain a deeper understanding of how anti-cancer agents modulate cell cycle progression in HER2+ breast cancer, a disease subtype that accounts for 20% of all breast cancers. We treated HER2+ breast cancer cells with a panel of drugs and tracked changes in cell number and cell cycle phase, which revealed drug-specific cell cycle effects that varied across time. This suggested that a computational model that could account for cell cycle phase durations would provide a framework to explore drug-induced changes in cell cycle changes. Toward that goal, we developed a linear chain trick (LCT) computational model, in which the cell cycle is partitioned into subphases that faithfully captured drug-induced dynamic responses. The model inferred phase-specific drug effects and independent modulation of cell cycle phases, which we confirmed experimentally. We then used our LCT model to predict the effect of unseen drug combinations that target cells in different cell cycle phases. Experimental testing confirmed several model predictions and identified combination treatment strategies that may improve therapeutic response in patients with HER2+ breast cancer. Overall, this integrated experimental and modeling approach opens new avenues for assessing drug responses, predicting effective drug combinations, and identifying optimal drug sequencing strategies.