Microfluidic Isolation and Deep Learning-based Profiling of Subtypes of Circulating Tumor Cells during Epithelial Mesenchymal Transition

Title: Microfluidic Isolation and Deep Learning-based Profiling of Subtypes of Circulating Tumor Cells during Epithelial Mesenchymal Transition Project Summary/Abstract Circulating tumor cells (CTCs) are highly heterogeneous and contain many cellular subpopulations. It is known that specific CTC subpopulations, rather than the whole, are responsible for cancer metastases. Recently, epithelial mesenchymal transition (EMT) of adherent epithelial cells to a migratory mesenchymal state has been implicated in tumor metastasis. CTCs isolated from cancer patients exhibit dynamic changes in epithelial and mesenchymal composition, and serial CTC monitoring in those patients suggested an association of EMT in CTCs. Emerging microfluidic technologies have shown great promise for the complete capture of the CTCs population with high yield and enhanced purity. However, most existing devices simply isolate all CTCs in blood without resolving them into distinct subpopulations: isolation of CTC populations with specific EMT markers remains a significant challenge. Consequently, epithelial-mesenchymal plasticity of CTCs during cancer progression is largely unknown, preventing researchers from acquiring true insights into the metastatic potential of CTCs. In the R15 project, we propose to develop the HU microchip platform for isolation and profiling of CTCs in pre-clinical settings, with the focus on detecting and understanding features associated with CTC subtypes during the course (or intermediate states) of EMT. To do so, we will isolate CTCs and explore deep learningbased profiling platforms using both micro and nanometer-scale features found from microscope images obtained from our HU microfluidic devices. This is expected to achieve super-high accuracy for profiling CTC subtypes with varied cell plasticity (i.e., more “epithelial” or more “mesenchymal” type). Building on the success of this project, we expect to develop an integrated system comprising 1) a user-friendly HU microchip for highly efficient isolation of CTCs from blood samples, and 2) an AI-based image analysis platform that can identify CTC subtypes for monitoring cancer progression and support cancer diagnosis and treatment. CTCs with defined EMT stages generated from the NSG mouse model will used to provide critical ground-truth information for validation of the proposed AI model, which cannot be accomplished using other alterative models. This project is strongly integrated with an educational plan to expose undergraduates to advanced experiments in the areas of biomedical engineering and cancer research. It is our belief that involving all students in cutting-edge research will have a significant positive impact on attracting and retaining students in science and healthcare fields. In this grant, we will provide research opportunities for 6 undergraduate students and 1 graduate to learn advanced knowledge on cancer biology, obtain hands-on experience in microfabrication and biological assays. Texas Tech University (TTU) is located in Lubbock, a rural area in northwestern Texas and TTU has been classified as "Tier 1" status in 2016. Due to historical and geographical reasons, life sciences research is not a strength for TTU, and TTU is not a recipient of major NIH grants. This R15 grant will strengthen our research and training at TTU and help us to reach our goal to be one of the best bioengineering programs in Texas. Project Number: 1R15CA313429-01 | Fiscal Year: 2026 | NIH Institute/Center: National Cancer Institute (NCI) | Principal Investigator: Wei Li | Institution: TEXAS TECH UNIVERSITY, LUBBOCK, TX | Award Amount: $607,488 | Activity Code: R15 | Study Section: Special Emphasis Panel[ZRG1 MCST-N (85)] View on NIH RePORTER: https://reporter.nih.gov/project-details/11361375