Photodynamic diagnosis-assisted ureteroscopy for upper tract urothelial carcinoma achieves improved diagnostic performance using quantitative color image analysis

Upper tract urothelial carcinoma (UTUC) remains challenging to diagnose accurately by ureteroscopy, particularly when lesions are flat or inconspicuous under white-light (WL) imaging. We conducted an observational study to evaluate the diagnostic performance of 5-aminolevulinic acid (5-ALA)-based photodynamic diagnosis-assisted ureteroscopy (PDD-URS) and to investigate whether quantitative color image analysis could improve diagnostic accuracy. Thirty-six consecutive patients with suspected UTUC were included in this observational study, and a total of 126 biopsy specimens were analyzed. Color image analysis was subsequently performed using PDD images obtained during ureteroscopy to calculate the red-to-blue (R/B) ratio. Diagnostic performance was assessed using sensitivity, specificity, and receiver operating characteristic (ROC) analysis. PDD-URS demonstrated substantially higher sensitivity than WL-URS (94.4% vs. 63.9%) but lower specificity (68.9% vs. 98.7%), and detected 11 of 13 flat urothelial carcinoma lesions that were missed under WL imaging. Color image analysis was performed on 97 PDD images corresponding to the biopsy locations by calculating the R/B ratio. The median R/B ratio was significantly higher in urothelial carcinoma than in non-malignant lesions (1.47 vs. 0.84, P < 0.001), yielding an area under the ROC curve of 0.95. Using an R/B cutoff value of 1.3, sensitivity and specificity were 92.6% and 88.6%, respectively. These findings suggest that PDD-URS improves the detection of UTUC, particularly flat lesions, while quantitative R/B ratio analysis may reduce false-positive interpretations and provide an objective adjunctive tool for enhancing diagnostic accuracy.