Available at: https://digitalcommons.calpoly.edu/theses/3436
Date of Award
12-2026
Degree Name
MS in Engineering - Biochemical Engineering
Department/Program
Biomedical Engineering
College
College of Engineering
Advisor
Chris Heylman
Advisor Department
Biomedical Engineering
Advisor College
College of Engineering
Abstract
Preclinical models of colorectal cancer rarely reproduce the tumor vasculature or the transport environment that governs drug delivery, despite the vasculature being simultaneously the route of delivery and a therapeutic target. Interstitial flow is a primary cue for tumor-driven vessel formation, yet in most vascularized tumor-on-a-chip platforms it is imposed as a fixed operating condition and inferred from device geometry rather than measured within the tissue chamber. This thesis develops a three-chamber microfluidic colorectal tumor-on-a-chip in which the directed interstitial-flow field is measured, spatially uniform, and tunable. Two endothelial chambers flank a central tumor chamber, and flow is driven by hydrostatic head across three flow configurations. The flow field was characterized by two fluorescent microscopy imaging techniques: micro-particle image velocimetry and FITC-dextran wavefront velocimetry. Each technique has a matched finite-element simulation whose porous-domain permeability is supplied from empirical derivation rather than assumption. Interstitial velocity scaled linearly with applied head, giving a predictable relationship for input pressure to flow velocity. Filling the chambers with hydrogel decreased flow coefficient of variation from as high as 0.8 to below 0.18 and produced pressure-driven interstitial velocities of 0.4–1.9 µm/s, within the physiological range for normal tissue. This placed transport in the advection-dominated regime associated with vasculogenesis (Peclet number 13–39) across all 1D conditions and at the upper two 2D driving pressures. Simulated and measured velocity profiles agreed closely in shape, and the remaining magnitude offsets were consistent with known biases of each imaging technique. A proof-of-concept co-culture sustained, and in most devices, increased the area covered by endothelial fluorescence. Together these results establish interstitial flow as a measured tunable experimental variable and provide a characterized foundation for a physiologically representative vascularized colorectal cancer model.