#AI reads Urine# Development and Validation of a Urinary Exosomal miRNA Diagnostic Panel for Early Detection of Esophageal Cancer
Published 29 March, 2026
This study addresses the challenge of early detection of esophageal squamous cell carcinoma (ESCC) by developing and validating a non-invasive diagnostic panel based on urinary exosomal miRNAs. Urine samples from ESCC patients and healthy controls were prospectively collected across five institutions. Using small RNA sequencing and machine learning with recursive feature elimination, a diagnostic panel containing 57 miRNAs was constructed. In the proof-of-concept cohort, the panel achieved an AUC of 0.90, while in the validation cohort, it demonstrated an AUC of 0.85 (sensitivity: 84%, specificity: 66%). It exhibited exceptionally high diagnostic accuracy for early-stage patients, with AUCs of 0.95 for Stage 0 and 0.90 for Stage I, and its performance was not significantly affected by factors such as sex, BMI, alcohol consumption, or smoking habits. Furthermore, the diagnostic score of the panel significantly decreased after treatment in early-stage cases, indicating its potential to reflect treatment efficacy and tumor burden. This study provides a convenient and non-invasive screening tool for ESCC; however, standardized sample processing protocols need to be established, specificity should be improved, and further validation in diverse populations is required to enhance its applicability.
Cancer Sci. 2026 Jan 19. doi: 10.1111/cas.70298.
Youhe Gao
Statement: During the preparation of this work the author(s) used Doubao / AI reading for summarizing the content. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article.
For earlier AI Reads Urine articles：