DEVELOPMENT OF A SUPER-SENSITIVE AUTOMATED PLATFORM FOR GENE EXPRESSION ANALYSIS AND ITS VALIDATION FOR BLADDER CANCER DIAGNOSIS

Background: Current diagnostic methods for bladder cancer (BC) are constrained by low sensitivity, their invasive nature, and susceptibility to subject-dependent variability. Consequently, there is a critical need for developing a rapid, non-invasive, and highly sensitive test capable of detecting BC. A new basis for diagnostic tests is being provided by the vast amount of data on gene expressions that are now becoming available through large-scale measurements of mRNA. These insights are valuable to better understand the disease mechanism and to identify biomarkers for precise diagnosis. In this study, a highly sensitive and automated platform KAIROS was developed for gene expression analysis and validated through measuring mRNA targets frequently overexpressed in BC. Objective: To develop a highly sensitive automated platform and validate its performance for the diagnosis and follow-up of bladder cancer. Design, setting, and participants: Urine samples were prospectively collected from patients with suspected BC at first diagnosis undergoing TURB (Transurethral Resection of Bladder) and follow-up patients undergoing cystoscopy. This study addressed the critical barriers of diagnostic methods by developing and implementing a super-sensitive automated platform KAIROS for high-throughput gene expression patterns in BC. This platform is designed to streamline the entire workflow, with robust integrated software and gene expression analysis via RT-qPCR to ensure data reliability and minimize human error. Technical validation of the method was also performed, and specificity was determined by analyzing participants without BC. Outcome measurements and statistical analysis: Test characteristics were calculated based on TURB and cystoscopy results after comparing with KAIROS results. Absolute quantitative real-time polymerase chain reaction (qRT-PCR) data was used for expression analysis, and statistical analysis was performed using GraphPad Prism software version 8.0.2 and Excel (version 2010). Non-parametric statistical tests (Mann-Whitney U test, descriptive tests) were applied, and a p-value < 0.05 was considered statistically significant. Data was presented as mean + SE, derived from Ct values (high Ct means low expression and low Ct means high expression) of samples run in duplicates. In addition, to evaluate the diagnostic performance of combined biomarkers together, 5 an algorithm was built into KAIROS software. The diagnostic performance of selected transcript markers (based on expression and detection) was also evaluated using receiver operating characteristics (ROC) curve analysis. The effectiveness of each biomarker was calculated with corresponding area under the curve (AUC) values. Results: In this study, 210 BC patients (mean age 73 yr, 77% male, 47% never smoked), 72 follow-up patients, and 137 controls were included. The promising candidates (CK2O, GLUT1, IGF2) had sensitivity (SN) of 93% (95% confidence interval [CI]: 0.91-0.95) and specificity (SP) of 86% (95% CI: 0.80-0.88) in BC patients. Whereas these targets showed 62% sensitivity (95% confidence interval [CI]: 0.49-0.73) and 92% specificity (95% CI: 0.65-0.99) in BC follow-up patients. Conclusion: This automated platform represents a promising tool for BC diagnosis: a non-invasive and rapid test with improved SN and SP compared with standard tests for BC. This urine-based assay demonstrates the potential and feasibility of urinary gene expression profiling (CK20, GLUT1, IGF2) as a cost-effective and clinically applicable approach. Super sensitivity for low-abundance transcripts also makes it an ideal choice for analyzing challenging liquid-biopsy markers, but further investigation in larger clinical studies is needed.