(375b) An Experimentally-Driven Quantitative Model to Optimize Oncolytic Adenovirus Cancer Treatment

Replication-selective adenoviruses replicate in cells containing certain mutations, motivating their use in targeted gene therapy. An example is ONYX-015, which aimed to target and preferentially replicate in p53-defective tumor cells, potentially offering improved treatment options for certain cancers. Although recent studies demonstrate that ONYX-015 affects a variety of signaling pathways not necessarily involving p53, the adenovirus continues to be studied in a number of clinical trials.

The efficacy of adenovirus mediated therapy depends on three general processes: (i) cell entry via the coxsackievirus and adenovirus receptor, CAR, (ii) viral replication, and (iii) virus-mediated cell death. Although ONYX-015 is designed to preferentially target cancer cells, experimental data suggests that CAR is down-regulated in highly malignant cells, hindering ONYX-015's ability to infect. Pharmaceutical intervention into the Raf-MEK-ERK pathway via the CI1040 MEK inhibitor has shown to counter this effect by up-regulating CAR expression. MEK inhibition, however, causes G1 cell cycle arrest thereby stunting viral replication and consequently virus-induced cancer cell death. Through the data-driven modeling of cancer cells subject to CI1040 and ONYX-015, we aim to characterize and predict system dynamics, providing a means to optimize the efficacy of oncolytic adenovirus cancer treatment by manipulating the timing of drug treatment and infection. Preliminary studies have supported a population based (cellular level) deterministic model that highlights sub-cellular virus-host dynamics as components necessary for accurate and predictive simulations. Without the resolution offered by multiscale models, computer simulations fail to characterize the sub-cellular phenomena that give rise to complex cellular response. As a result, we hypothesize that increased cell death due to the combinatorial effect of CI1040 and ONYX-015 is better described by a model that accounts for specific virus-host dynamics.

Validation of simulated results should provide a means to refine the current cellular level ODE model to include relevant sub-cellular phenomena that better characterizes both the mechanistic and dynamic behavior of adenovirus cancer treatment. Pending successful test of model predictions, our goal is to elucidate optimization strategies that could offer practical and effective means for minimizing cancer growth.