CEA Cell Adhesion Molecule 6 (CEACAM6), Collagen Type I Alpha 2 Chain (COL1A2), Galectin 4 (LGALS4) and Tetraspanin 8 (TSPAN8) mRNAs as blood biomarkers for colorectal cancer

Colorectal cancer (CRC) develops over several years, thus, early diagnosis and prevention are fundamental to reduce CRC burden. A new non-invasive blood test based on 4 mRNAs encompassing adhesion molecules such as CEACAM6, as well as a member of the collagen I superfamily, namely COL1A2, combined with LGALS4 and TSPAN8, was recently discovered and referred to as CELTIC panel [1]. The expression of the CELTiC panel was measured by quantitative PCR. The panel of putative biomarkers was subsequently evaluated in 101 subjects resulted positive to fecal immunochemical screening test (FIT) highlighting the potential to distinguish colonoscopy negative-FIT positive patients (NFIT, n=36), patients with low risk lesions (LR, n=36) and with high risk lesions or CRC (HR/CRC, n=92) [2]. In this study 174 healthy FIT negative subjects (FITN) were analysed [3] and compared to previous groups, evaluating also the influence of age and sex. Within the FITN group, CEACAM6 and COL1A2 display significantly lower expression in female than in male, confirming this sex difference of CEACAM6, a differentiation marker in normal colonocytes, also in the older groups (60–70 y.o.) and supporting current data on the importance of gender- and age-specific reference intervals for the early diagnosis of CRC [4,5]. The four genes showed significantly lower expression in FITN than in HR/CRC. Interestingly, TSPAN8 and COL1A2 were significantly lower expressed in FITN also than in NFIT and LR patients. TSPAN8, an integral membrane protein which upregulation promotes metastasis [6], was confirmed by logistic model as able to discriminate FITN from NFIT, LR and HR/CRC. Interestingly, also LGALS4, involved in cell-cell interaction and studied in CRC patients [7,8], was able to differentiate FITN from NFIT (false FIT positive). TSPAN8 and galectins have been described also in exosomes [9] suggesting the origin of the mRNAs of the CELTiC panel. Finally, the CELTIC panel showed good sensitivity (84-90%), specificity (76-81%) and AUC (0.87-0.89) to discriminate FITN from the other groups. The CELTiC panel was confirmed as a useful tool to diagnose CRC highlighting the importance of sex and age. A multicenter cross-sectional study will enroll 800 FIT positive subject at the Universities of Amsterdam and Bologna to better estimates the performances of the CELTiC panel and to further validate its robustness and usefulness in the early diagnosis of CRC and to exclude false FIT positive subjects.