Document Type



Doctor of Philosophy



First Adviser

Liu, Yaling


Cancer is a significant health risk to people living in developed and developing countries, which continues to prove difficult to treat. Common treatment options of cancers include surgical removal, radiation, and chemotherapies, which are often used in combination to improve the likelihood of successful treatment. Such combinatory approaches towards treatment are often taken because each approach is not targeted enough to function perfectly on its own. Being able to delivery therapeutic loads in a more targeted manner to sites of cancer has the capability of improving therapeutic efficiency and improving patient responses. The development of improved therapeutic delivery vehicles and screening systems can help serve the goal of improved targeted therapeutic delivery. The use of microfluidic devices for the study of therapeutic delivery has become popular over the past few decades because of the many benefits that they offer. Specifically, microfluidic devices only require small volumes of therapeutics for testing, which is often ideal because of limited drug supply during screening. Additionally, the high degree of control over channel geometries, ease of fabrication and low cost make microfluidic therapeutic testing devices well suited for higher throughput screening when run together in parallel. The ability to generate shear flow within the microfluidic channels also offers a means of more closely mimicking vascular physiology and conditions that would be experienced during drug delivery in the human body. Lastly, the use of microfluidic in therapeutic testing enables micro-scale data on characteristics such as binding, uptake, cellular permeability and others to be easily collected due to the transparent nature of the devices and ability to facilitate cell cultures. As such, the focus of this dissertation is mainly based around the establishment of microfluidic systems capable of mimicking cancerous environments and testing of various therapeutic vehicles and delivery methods targeted for cancer. In brief, the dissertation demonstrates a few methods of establishing cancerous environments within microfluidic systems of increasing complexity, and how screening of various nanoparticle vehicles and therapeutics is performed.First, a single layer microfluidic device is developed to facilitate the growth of cancer monolayers for the screening of solid lipid nanoparticle drug delivery performance. The device is designed to assist in identifying an optimal ratio of antibody to polymer chains exposed on the surface of the nanoparticles. Improved targeting of nanoparticles to cancer cells is achieved by increasing target specific binding through addition of cancer antibody while reducing non-specific binding through addition of polymer chains on the nanoparticles surface. Conditions for optimal targeting specifically to cancer cells were identified for nanoparticles with 37% of their surface area occupied by polyethelyene glycol (PEG). The cancer cell targeting efficiency for the 37% coated nanoparticles was determined to be a maximum of 81% when a cancer specific antibody was used in conjunction on the nanoparticles surface.Next, to improve the physiological relevance of the microfluidic screening system, a bi-layer setup was fabricated. The nature of the bi-layer device is designed to facilitate the co-culture of cancer and endothelial cells (ECs) in different compartments while still permitting signaling and chemical interactions to occur between the two cell types. The presence of ECs in the device is designed to mimic a blood vessel, as therapeutic delivery within the body relies heavily on the circulatory system from drug transport. As such, understanding the mechanics of therapeutic delivery from mimicked vasculature to cancer is an important consideration. Conditions in the bi-layer system influencing therapeutic transport include endothelial permeability, therapeutic size, system flow rate, and treatment time. Improved therapeutic delivery was achieved using smaller molecules, slower system flow rates, and when the EC monolayer was highly permeabilized. Increased treatment times, resulted in less and less therapeutic transport from the mimicked vessel to the cancer environment as the EC monolayer regained confluency. It was shown that the bi-layer microfluidic system functions to screen therapeutic delivery to a mimicked cancer environment under more physiologically relevant conditions.The next progression with the system was to test nanoparticle delivery and transport from the mimicked vessel to the cancer environment. This was accomplished utilizing the same bi-layer microfluidic setup in conjunction with a range of nanoparticle shapes that were utilized to identify characteristics that facilitate the greatest degree of therapeutic delivery. Specifically, spherical, short rod and long rod/worm-like nanoparticles were tested for their ability to transport therapeutic loads to the cancer environment over the course of 5 day treatments. Optimal nanoparticle shapes for each flow rate varied based on treatment time. Overall, nanoparticle drug delivery should be varied based on the degree of EC permeability which changes with time as the cancer environment is treated.Lastly, to improve the physiological relevance of cancer environments being used, a method for establishing and growing tumor spheroids within the microfluidic devices in an expedited fashion was developed. The ability to perform therapeutic and nanoparticle carrier screening on tumor spheroids as opposed to cancer monolayers provides feedback on efficiency and performance which more closely mimics outcomes observed in animal and clinical testing. In addition, the ability to form tumor spheroids in an expedited manner allows the screening process to be completed in a shorter period of time and with fewer initial cells. The use of convective driven nutrient flow is utilized to achieve such expedited cancer growth in a microfluidic system which also has the potential to facilitate therapeutic screening. The system has been shown to function with adherent and non-adherent cell types where 1.5 to 4.5 times faster growth can be achieved. The ability to cut tumor culturing times from 1 week to 3 days and reducing required cell counts from thousands to tens of cells has the potential to save lives in clinical settings when using patient derived samples.