Preliminary feasibility and safety of domestic probe-based confocal laser endomicroscopy in the diagnosis of bladder lesions: a prospective single-arm cohort study.

Background: White light cystoscopy (WLC), the first-line diagnostic modality for bladder lesions, has inherent limitations in detecting flat lesions and distinguishing low-grade malignancies. Imported probe-based confocal laser endomicroscopy (pCLE) systems address these gaps but are restricted by high costs and limited accessibility. This study aimed to assess the technical feasibility and safety (primary outcomes) of a domestic pCLE system for bladder lesion evaluation, while preliminarily comparing the clinical performance of two fluorescein sodium administration routes and three probe diameters (secondary outcomes) to inform future clinical research. Methods: A single-center, prospective single-arm cohort study was conducted at a tertiary hospital, enrolling 10 patients with suspected bladder masses or a history of bladder cancer undergoing cystoscopy. Feasibility was defined as the technical completion rate of pCLE imaging, and safety was evaluated using the Clavien-Dindo classification system for adverse events within 72 hours post-procedure. Patients with negative fluorescein sodium skin tests received either intravesical instillation or intravenous injection. Probes of 1.9, 2.6, and 3.2 mm were tested for imaging quality and maneuverability. Baseline clinical characteristics and histopathological results were collected for descriptive analysis; no inferential statistics for diagnostic accuracy were performed. Results: The technical completion rate was 100%, with no Clavien-Dindo grade ≥ I adverse events observed. Transient yellowish-green urine occurred in all patients who received intravenous fluorescein sodium, resolving spontaneously without intervention. The 2.6 mm probe exhibited the best balance of imaging quality and maneuverability, while the 3.2 mm probe had poor maneuverability in narrow anatomical sites. Intravenous injection shortened procedure time but caused mild surgical field interference in 50% of cases during transurethral resection, whereas intravesical instillation had no interference but a longer duration. Pathological results included eight malignant lesions (five high-grade and three low-grade urothelial carcinoma), one benign lesion, and one papillary urothelial neoplasm of low malignant potential (PUNLMP). Distinguishable cellular and tissue features were observed on pCLE imaging across different lesion types. Conclusions: This domestic pCLE system demonstrates favorable technical feasibility and safety in this pilot cohort. Due to the small sample size, no conclusions regarding diagnostic accuracy can be drawn. Larger, multicenter studies are urgently needed to validate these preliminary findings and evaluate the system's diagnostic efficacy in bladder cancer management.