Grant: Project Grant

Colorectal cancer (CRC) is the 2nd most deadly cancer in Canada, notably because many patients develop metastasis. Of those, over 50% develop liver metastasis (CRCLM), of which 70% ultimately die from CRCLM. There are two main subtypes of CRCLM, with one leading to worse outcomes. We recently found that the commonly used anti-angiogenesis therapy is harmful to the worse subtype. However, the subtype can only be determined after surgery using resected tissue, and all patients receive this expensive therapy, while ~ 20-40% are harmed. Tumor cells and clusters of tumor cells may detach from the lesion and enter the blood stream, and become circulating tumor cells (CTCs). CTCs are a window into the tumor, and are easily and repeatedly collected from a blood draw. However, CTCs are like a needle in a haystack, as many millions of white blood cells mask the handful CTCs typically found in a spoonful of blood. We developed a microfiltration method to purify both single CTCs and clusters from patients. Our method is much more sensitive than other methods, cheap, and easy-to-use. Here we propose to discover CTC biomarker for CRCLM from the first blood draw at diagnosis, and at each subsequent therapeutic step. We will (i) microfilter CTCs, (ii) develop and use new single cell imaging methods to capture morphological, molecular, and clustering features of CTCs associated with cancer, disease progression and immune response; and (iii) use artificial intelligence to identify biomarkers for CRCLM subtype and disease outcome. The data and analysis will identify CTC biomarkers, help guide therapy, benefit thousands of Canadians a year by salvaging them from harmful therapies, and save millions in unnecessary expenses. Time-course CTC analysis will advance our understanding of CRCLM, could lead to adaptive therapy based on changes in CTC biomarkers, and may identify new therapeutic targets (e.g. immunotherapies).