A Liquid Biopsy Assay For The Non-Invasive Early Detection of Advanced Adenomas and Colorectal Cancer

Colorectal cancer (CRC) is a significant global health concern, ranking third in diagnosis and second in mortality. Despite being potentially preventable, it remains a leading cause of cancer-related deaths. Traditional screening methods like fecal immunochemical testing (FIT) have shown benefits in reducing late-stage diagnoses but have not effectively prevented CRC incidence. This is because tests like FIT can effectively detect the cancers, but not the precursor lesions, called adenomas. On the other hand, endoscopy-first approaches offer higher sensitivity for such adenomas and, therefore, lower the risk of developing CRC but face challenges such as invasiveness, cost, and patient compliance.

Non-invasive tests are more appealing to patients than invasive tests and can increase participation rates. Biomarker studies have shown promise, but existing tests lack sensitivity for early-stage CRC and advanced adenomas (AAs). This is likely because they assume the same analyte can detect both CRC and AAs, which may not be accurate due to differences in analyte release and the biological changes that occur during the adenoma-carcinoma sequence.

This study proposes developing an innovative liquid biopsy test tailored for AAs and CRC to address this. An ideal screening test should be minimally invasive, highly sensitive, and cost-effective. This test would optimize patient compliance and resource allocation by detecting both conditions from a single blood draw. More specifically, circulating microRNA (miRNA) analysis shows promise: tests based on cell-free microRNA (cf-miRNA) have demonstrated high sensitivity, while those based on exosome-derived microRNA (exo-miRNA) offer high specificity. Therefore, combining both analytes in a single test could maximize sensitivity and specificity.

This study will develop a non-invasive blood test for AA and CRC in four phases:

1. Genome-wide profiling of cf-miRNA and exo-miRNA and selecting the best candidates for biomarker panels.
2. Utilizing machine learning to identify promising candidates and train algorithms for detecting AAs and CRC separately, based on results from quantitative polymerase chain reaction (qPCR) analysis.
3. Combining these algorithms to create detection signatures for both conditions.
4. Independently validating these signatures using diverse cohorts to ensure broad applicability and compare the effectiveness of the blood assay to standard care through retrospective and prospective studies.