Computational and theoretical fluid mechanics modeling for transport in dense tumors

Intratumoral perfusion has long been recognized as a critical issue in both clinical diagnosis and therapy of dense solid cancerous tumors.

In such context, a first-principles mechanics-based model that can quantify perfusion in the intratumoral extracellular spaces as a function of the tumor vasculature shapes and the fiber packing fraction in the stroma – can launch new avenues in cancer diagnosis and care.

With that vision, the proposed project will integrate computational fluid dynamics (CFD) tracking with theoretical fluid mechanics analysis to generate an in silico tumor uptake modeling framework, that can operate over a wide parametric space. The test geometries will be based off computed tomography (CT) scans of human pancreatic tumors implanted in mice.

The project will also design supplementary physical experiments in microfluidic setups and artificial tumor spheroids to benchmark and validate the proposed in silico approach.

With use of mean continuum-level transport frameworks such as Darcy's Law and Starling model still a go-to resort for basic fluids modeling of intratumoral uptake quantification, the proposed CFD-informed advanced theoretical fluid mechanics approach to parameterize tumor perfusion based on tumor geometry and intratumoral stress will constitute this project's key contribution to the literature.

The long-term objective of the project is as follows: in a clinical setting, the packing fraction of the fiber bundles inside the stroma can be readily assessed from image-processing the CT-slices from a tumor, it being logarithmically proportional to the intratumoral stress.

The proposed numerical-theoretical model will be designed such as to project the tumoral uptake as a function of the packing fraction.

This can trigger new diagnostic/therapeutic solutions with the fluidic transport trends inside a solid tumor predictable solely from the medical scans through assessment of the packing fraction.

Our central hypothesis is: an integration of numerical computations with theoretical modeling can cover a diverse range of tumor microenvironments, rendering greater usability for the fluid mechanics tools.

The resulting in silico framework will generate percolation data over a wide parameter space of tumor geometry features, e.g., the packing fraction inside the stroma and the local curvatures of blood vessels in the tumor vasculature.

The work is structured around the following specific aims: (a) Aim 1 will numerically simulate multiphase transport inside the tumor vasculature, considering realistic blood vessel shapes and electrohydrodynamic effects on the mean transport, (b) Aim 2 will import from Aim 1 the plasma dynamics information at the endothelial openings and use the data as initial conditions to develop an integrative numerical-theoretical model for intratumoral transport through the extracellular matrix.

The theoretical setup will invoke a convection-diffusion approach, where the boundary conditions on the local concentrations at the tumor inlet and near the necrotic core (deep into the tumor) will be fed from the numerical simulations.

Finally, the in silico perfusion predictions will be compared against physical experiments performed in simplified bio-inspired microfluidic systems and with cell culture-derived realistic tumor spheroids embedded in a fluidic environment.