S548 Identification of Novel mRNA Biomarkers With Improved Diagnostic Clinical Performance From a Large Multicenter Study

Introduction: Colorectal cancer (CRC) is a major global health concern, ranking as the third most diagnosed cancer and the second leading cause of cancer-related deaths. Early detection is crucial, but the invasiveness and resource demands of colonoscopy limit its widespread use. Non-invasive alternatives like fecal immunochemical tests (FIT) and multitarget DNA assays exist but have varying accuracy, particularly in detecting advanced precancerous lesions (APL) and early-stage CRC. Methods: Messenger RNA (mRNA) stool-based biomarkers represent a promising approach for the diagnosis of colorectal cancer and advanced precancerous lesions. But it is unclear which mRNA biomarkers have the most clinical utility. This study aims to address this issue by performing the largest screen to date of mRNA biomarkers using individual PCR. We leverage bioinformatic analyses of publicly available RNA-seq datasets to identify candidate genes with differential expression in CRC versus normal tissues whose pattern of expression make them prime candidates for a stool-based screening test. A comprehensive screening process then evaluated the top 143 computationally predicted biomarkers on clinical stool samples across 2 independent patient cohorts (Cohort 1 N = 90, Cohort 2 N = 131). Results: The results demonstrate that several mRNA biomarkers exhibit strong predictive power, with genes such as OLR1, TGFBI, CXCL8, and MMP7 achieving high area under the curve (AUC) values for CRC detection. Additionally, the combination of top-performing biomarkers using machine learning approaches further improved diagnostic accuracy, yielding an AUC of 0.986 for CRC and 0.77 for APL detection. The sensitivity for CRC detection was 97.9% the APL sensitivity was 54.5% at a specificity of 86%. To enable a direct comparison, fecal immunochemical tests (FIT) were run on the same samples and the specificity was adjusted to 86%. FIT’s sensitivity of CRC detection was 85.7%, the sensitivity of detection for APLs was 13.6%. Conclusion: This is the largest screen, in terms of genes screened, to date identifying mRNA biomarkers with clinical utility for CRC and APL detection. Several novel biomarkers were identified with promising clinical performance. When combined into a panel and directly compared with FIT the mRNA panel had substantially improved performance over using FIT alone.