Urine Liquid Biopsies via Highly Integrated Digital PCR System for Accurate Detection of Bladder Cancer

Abstract Bladder cancer (BC) is a prevalent urological tumor with high recurrence rates, requiring long‐term monitoring. Although cystoscopy is the primary diagnostic method, its invasiveness and cost hinder routine screening and follow‐up. This study aimed to develop a novel diagnostic tool utilizing newly developed on‐chip heating dPCR platform, which features integrated and rapid temperature control capabilities, for non‐invasive BC detection. The dPCR platform is improved by integrating a multi‐color detection system, enabling precise quantification of mutant allelic fraction (MAF) of TERT promoter mutations with a limit of detection (LOD) of 0.29%. Diagnostic performance is enhanced by integrating the NRN1 methylation biomarker and employing machine learning to optimize biomarker weighting. Testing the model on urine samples from controls ( n = 35) and BC patients ( n = 41) yielded a sensitivity of 0.92, specificity of 0.94, and an AUC of 0.98, surpassing conventional cytology in sensitivity while maintaining comparable specificity. Furthermore, the model effectively differentiated between normal controls and different stages, achieving accuracies of 0.92, 0.71, and 0.79 for NC, stage I, and stage II+ respectively. These findings suggest the proposed dPCR assays could serve as a sensitive and non‐invasive approach for BC detection in clinical practice.