Colorectal cancer (CRC) remains a significant public health burden as the third-most commonly diagnosed and second-deadliest malignancy. While CRC is potentially preventable through the removal of precursor lesions known as advanced adenomas (AAs), current screening modalities have limitations. The fecal immunochemical test (FIT) exhibits reduced sensitivity for early-stage CRC and AAs, primarily lowering mortality rather than incidence. Colonoscopy, while effective, is invasive, costly, and suffers from suboptimal adherence. Additionally, individuals with Lynch syndrome (LS), who carry a pathogenic germline variant in mismatch repair (MMR) genes, lack effective and reliable methods for early-stage detection of non-CRC tumors associated with the syndrome.
The rationale of this study, BEACON/2026, is to develop and validate two blood-based tests to address these clinical gaps. The first is a screening test sensitive to both CRC and AAs to complement existing screening options; the second is a multi-cancer early detection (MCED) test tailored to the LS spectrum. The central hypothesis is that a liquid biopsy approach combining cell-free microRNAs (cf-miRNAs, which are sensitive) and exosome-bound microRNAs (exo-miRNAs, which are specific) can effectively distinguish patients with AAs, CRC, and LS-associated cancers from controls.
This study follows the Early Detection Research Network (EDRN) recommendations for biomarker development, including biomarker discovery, prioritization, training, and independent validation (Phases I through III). The experimental design complies with PROBE recommendations for prospectively collected biospecimens that are blindly evaluated. The procedure involves the collection of peripheral blood to isolate RNA from plasma or serum for reverse transcription, quantitative polymerase chain reaction (RT-qPCR) analysis and machine-learning modeling . This non-invasive approach aims to provide a cost-effective, patient-friendly alternative to improve screening participation and disease monitoring
