PD-L1 expression across urothelial carcinoma subtypes: A multicenter comparison of 22C3 and SP142 for clinical decision-making in the immunotherapy era

BACKGROUND: Urothelial carcinoma with histological subtypes (UC-s) presents diverse biological behaviours and therapeutic responses. PD-L1 immunoexpression has become a key biomarker for guiding immune checkpoint inhibitor therapy in advanced UC. However, the role of PD-L1 across UC-s remains limited characterized, with discrepancies reported between assays. OBJECTIVE: This multicenter retrospective study aimed to assess PD-L1 expression patterns in 169 UC-s cases using two assays (22C3 and SP142), with analysis of tumor cells (TC), immune cells (IC), combined positive score (CPS), interobserver variability, and to assess their diagnostic relevance. METHODS: Samples were obtained from five Spanish hospitals. Whole-slide immunohistochemistry was performed using both PD-L1 assays. Positivity thresholds were CPS ≥ 10 (22C3) and IC ≥ 5 % (SP142). Histological subtypes were subclassified and scoring was performed by four trained pathologists. RESULTS: PD-L1 expression varied widely among UC-s subtypes. Squamous and sarcomatoid subtypes showed the highest PD-L1 expression with both assays. A greater number of UC-s were positive with SP142, especially in IC, while 22C3 showed higher TC expression. Discordant cases highlighted the influence of assay and cellular compartment on PD-L1 status. Interobserver agreement was high. Notably, 27 cases were SP142-positive but 22C3-negative for IC, suggesting potential therapeutic implications. CONCLUSION: The 22C3 and SP142 assays show analytical concordance for overall PD-L1 status but significant variability in IC scoring. Given the variability in PD-L1 expression across UC-s, precise characterization of both TC and IC staining per assay is crucial. These findings support incorporating UC-s subtype and assay-specific PD-L1 profiles into routine diagnostics to optimize immunotherapy decisions.