Integrated bioinformatics and experimental validation identify FOXQ1, GRIN2D, and SCNN1B as novel biomarkers for distinguishing high-risk sessile serrated lesions from hyperplastic polyps.

Serrated polyps, key precursors in up to 30% of colorectal cancer (CRC) cases, present substantial diagnostic challenges due to morphological similarities, especially between sessile serrated adenomas/polyps or sessile serrated lesions (SSLs) and hyperplastic polyps (HPs). Current histopathological methods often fail to reliably differentiate these lesions, with diagnostic accuracy rates as low as 70–85%. This underscores the urgent need for molecular markers to facilitate early detection and surveillance of CRC. We integrated RNA sequencing datasets (GSE76987, GSE46513, ERP109626) using meta-analysis to identify differentially expressed genes (DEGs) distinguishing SSL from HPs. Candidate genes (FOXQ1, GRIN2D, SCNN1B) were validated using quantitative real-time PCR (qPCR) on 30 SSLs, HPs, adjacent normal tissues, and 15 CRC samples. FOXQ1 and GRIN2D were significantly upregulated in SSLs and CRC samples compared to HPs and normal tissues (P < 0.0001), while SCNN1B was consistently downregulated (P = 0.0045). These gene expression patterns were consistent across both bioinformatics and wet-lab analyses. This integrative approach identified FOXQ1, GRIN2D, and SCNN1B as promising molecular biomarkers for distinguishing high-risk SSLs from HPs. These genes hold potential for enhancing diagnostic accuracy and early CRC detection. Further functional studies are warranted to explore their roles in serrated neoplasia and clinical application in CRC risk assessment.