Discovery of circulating RNA biomarkers for pancreatic cancer detection and monitoring

About This Grant

Significance to VA: Pancreatic cancer affects nearly 1,000 Veterans annually, and most patients are diagnosed with advanced stage disease that is rapidly fatal. Early detection greatly improves the chances of long-term survival, but no effective screening methods currently exist. A non-invasive blood test (liquid biopsy) for pancreatic cancer could enable early detection, but past efforts have failed to outperform standard of care imaging. Plasma circulating cell-free RNA (cfRNA) shows tremendous potential as a liquid biopsy analyte, yet its disease specificity remains unclear, and technical challenges have impeded its clinical translation. Our novel cfRNA sequencing protocol, validated in pilot studies, supports our hypothesis that tumor tissue- derived cfRNAs can be utilized for cancer detection and monitoring. Innovation and Impact: Our methodology overcomes several technical barriers, indicating that cfRNA profiling can be highly sensitive and reproducible. The project will use high-resolution spatial transcriptomics technology to map cfRNAs to specific tumor cells and determine whether cfRNA profiling can infer disease features without a tissue biopsy. Early detection of pancreatic cancer would enable patients to undergo potentially curative surgery, return to the workforce, and maintain their quality of life. If this project is successful, our cfRNA approach could be rapidly applied to other cancer types that lack effective screening or monitoring. Specific Aims: Aim 1) Map RNA expression from pancreatic tumor tissue to the bloodstream. We hypothesize that distinct cfRNAs are directly produced from pancreatic tumor tissues and can be used for disease inference. The objective is to identify whether cfRNAs can predict disease characteristics and obviate the need for tissue biopsies. Aim 2) Identify circulating RNAs for pancreatic cancer detection and monitoring. We hypothesize that cfRNAs can be used for disease detection and estimation of disease burden in patients with established cancer. The objective is to determine whether cfRNA profiling is non-inferior to cross-sectional imaging and tumor markers for patients with established disease. Methodology: Aim 1 is a retrospective cohort study of Veterans (N=40) that will evaluate matched tissue and blood specimens using high-resolution spatial transcriptomics and plasma cell-free RNA sequencing. Aim 2 is a prospective case-control study of Veterans (N=225) with pancreatic cancer, benign conditions, and other cancers that will collect clinical information and blood specimens for molecular profiling over a 2-year period. Computational modeling of molecular profiling data will leverage modern AI/machine learning (ML) methods. Path to Translation/Implementation: The next step would be a clinical trial of cfRNA diagnostics in patients who are at high risk for developing pancreatic cancer, including patients with hereditary cancer mutations and/or a strong family history of pancreatic cancer. Training: The PI aspires to lead a translational research lab that pioneers molecular diagnostics to improve cancer care. This project incorporates training in spatial transcriptomics, AI/ML, and clinical biomarker development, coupled with mentorship toward research independence.