Department of Mathematics and Statistics
The College of New Jersey
April 26, 2016
Tumor treatment with oncolytic viruses and dendritic cell vaccines: hierarchical model development and the robustness of optimal treatment protocols
A common challenge in mathematical biology is developing a model that is sufficiently detailed to describe the complexity of a given biological dataset, without making the model so complex that it becomes challenging to understand from a mathematical perspective. In this talk, I will explain how we hierarchically developed, fit and studied the structure of parameter space in a set of mathematical models designed to understand tumor response to a relatively novel virotherapeutic/immunotherapeutic treatment protocol. This protocol involves the use of an oncolytic virus (OV) that is engineered to selectively replicate inside of tumor cells, which eventually results in tumor cell lysis. Beyond being able to lyse tumor cells, these OVs can be enhanced to act as vectors to deliver immune-boosting molecules to the tumor site. These non-enhanced and enhanced OVs have been tested in melanoma-bearing mice as a stand-alone treatment, and in combination with a second immunotherapy protocol that utilizes dendritic cell (DC) injections. After identifying the best-fit parameters in the dynamical systems developed to hierarchically fit this data, we optimized treatment with three doses of enhanced OVs and three doses of DCs in different regions of dosing space. Bootstrapping of the experimental data allowed us to study the robustness of the predicted optimal treatment protocols. Through this analysis, we identified an ideal region of dosing space that ensures a robust anti-tumor response across the bootstrap samples. Thoughts on how this modeling approach can be extended to personalizing optimal treatment protocols will also be discussed.