Pan-oesophageal and spatially resolved genetic analysis reveals instability signature for earlier diagnosis of squamous cell cancer

Early detection of oesophageal squamous cell carcinoma (OSCC) significantly improves patients survival. Nonetheless, the availability of non-endoscopic, effective, and minimally invasive diagnostic approaches is limited. Here, we evaluated the utility of minimally invasive pan-oesophageal sampling using a capsule-sponge and aneuploidy detection using shallow whole genome sequencing (sWGS) for early detection of OSCC and precancerous intraepithelial neoplasia (IEN). In the prospective arm, 200 participants underwent the capsule-sponge procedure, and we performed sWGS from 178 successfully collected specimens (89%). We combined newly developed genome-wide copy-number alteration (GW-CNA) score and copy number alterations (CNAs) at chromosomal arm level to measure global and local aneuploidy, respectively. Logistic regression model identified GW-CNA and CNAs on chromosomal arms 2q, 3q, 9p and 11q as key diagnostic predictors differentiating OSCC and IEN from healthy controls (AUC of 0.920 (95% CI: 0.907-0.933), accuracy: 0.888, sensitivity: 0.896, specificity: 0.887). The model outperformed histology-based diagnosis using H&E staining and p53 immunohistological assessment. Finally, the analysis of microdissected samples derived from retrospective endoscopic en-bloc resections, and spanning the entire pathological continuum of OSCC demonstrated stepwise increase in GW-CNA and CNAs of 2q, 3q, 9p and 11q, validating their biological significance. This study demonstrates the high potential of combined pan-oesophageal sampling and sWGS aneuploidy analysis for early detection of OSCC and as a potential path to improved patients outcomes.