Susan Leggett Builds Tumors on Chips

A tumor is more than just cancer cells; it is part of a complex ecosystem that includes multiple cell types and a surrounding physical environment known as the matrix. Tumors can sense and modify their surroundings, which, in turn, shape the tumors’ development and have a significant impact on the disease’s progression. The mechanical environment, including local fluid flows in the tissue, plays an important role in drug delivery and can contribute to metastases – when part of the tumor separates and implants in another part of the body.

Just as recent advances in manufacturing are revolutionizing tumor model creation, bioengineering professor Susan Leggett is using similar techniques to construct larger, customizable fluidic devices that serve as artificial environments for the models to inhabit. 3D printing allows these tumor model housings to be efficiently and reproducibly created, and techniques from computer vision and data science allow complex tumor-environment interactions to be captured and characterized.

“We call our manufactured models ‘organ chips’ or ‘tumor chips’ because they integrate various components – such as cells, matrix, fluids, and tiny channels – into a single system while having connections to external monitors and controls like an electronic chip,” Leggett said. “We study the impacts of biophysical cues like matrix stiffness and fluid flow, and biochemical cues like nutrient gradients. We can even observe how tumors grow and spread in these chips like they do in the body. And we’re optimizing these procedures so that any group with a commercial 3D printer can replicate them.”

Leggett’s group uses an engineering mindset at two levels: developing the technology to manufacture and deploy organ chips, and studying tumor mechanics in addition to the biology and chemistry.

“We use imaging to identify and track cells as they interact with each other and their environment,” Leggett said. “But we also take an engineering approach, using machine learning to analyze and classify these cells into different groups based on their behavior. This allows us to develop quantitative models grounded in physics, helping us uncover fundamental principles that govern cell behavior and tumor dynamics.”

When a cell is tracked, its shape, its movement, its interactions with other cells, the proteins it expresses, and many other features are recorded. Leggett’s group uses dimensionality reduction and machine learning techniques to quickly simplify and process the gigantic data sets that result.

“We don’t just look at one thing at a time: we can analyze these cell behaviors all at once with more sophisticated computer vision and machine learning,” Leggett said. “But we do need to put a lot of careful thought into how we collect our data and design our algorithms to efficiently leverage computing resources. This is another area where an engineering approach comes in handy.”

Susan Leggett is an assistant professor of bioengineering the Department of Bioengineering at Illinois Grainger Engineering. She is an associate member of the Cancer Center at Illinois, and she is affiliated with the Carl R. Woese Institute for Genomic Biology at Illinois.