(673f) Reaction-Diffusion Modeling of Nanocarrier Cocktails for Cancer Therapeutics | AIChE

(673f) Reaction-Diffusion Modeling of Nanocarrier Cocktails for Cancer Therapeutics

Authors 

Kevrekidis, I. G. - Presenter, Princeton University
Kavousanakis, M., National Technical University of Athens
Sofou, S., Johns Hopkins University
Despite the fact that cancer death rates are declining due to early detection, cases of advanced metastatic and/or recurrent solid cancers still have no cure. Current approaches use combinations of therapeutic agents, but unfortunately the vast majority of patients fail to reach a durable response. One of the major reasons for this failure is the uneven distribution of therapeutics within the tumor volume; significant tumor areas are not exposed to therapeutics, therefore their killing action in these areas is limited. In order to address such limitations, we have recently discovered that the delivery of therapeutics is more efficient using two different types of carriers, each preferentially acting on a different region of the tumor. (i) Drug-labeled-antibodies that can effectively kill the tumor perivascular regions and (ii) tumor-responsive liposomes that release their highly-diffusing drug loading in the interstitium, thus enabling the deeper penetration of therapeutics within the tumor.

In order to suggest an optimum combination of the two different carriers we adopt a continuum-level modeling approach that can simulate efficiently the various undergoing processes, namely, carrier transport, carrier binding, drug release and drug transport. Our model is inspired by reaction-diffusion models that have been successfully applied to describe and optimize the performance in heterogeneous catalysis, a classical chemical engineering process that bears great similarities with our system of interest. Solid tumors are composite porous materials hindering the uniform transport of therapeutics throughout their entire volume. The killing action of these therapeutic agents, involve their binding on active sites, and release of their drug-content that will concomitantly kill the cancer cells. Transport and kinetic parameters of our model are estimated by comparing simulations against experimental measurements. Finally, our experimentally informed model can predict the best possible scenario (combinations of the two carriers) for given tumor sizes, target markers expression levels, and cell packing densities by employing model machine learning and data mining techniques.