Association of PD-1, LAG-3 and TIM-3 expression on intratumoral CD8 T-cells with response to atezolizumab in a Real-World-Evidence biomarker study for advanced urothelial carcinoma patients.
de Andrea CE., Abengozar-Muela M., Arranz JÁ., Climent MA., Puente J., Vizcay Á., Montero de la Fuente L., Jurado JM., Bonfill T., Santander C., Villa JC., Pujol E., Rosero AC., Gomez J., Fernández EM., Álvarez Fernández C., Ramirez I., Arnáiz P., López-Janeiro Á., Melero I., Sanmamed MF., Pérez-Gracia JL.
Blockade of the PD-1/PD-L1 pathway is part of the standard treatment for advanced urothelial cancer, but reliable predictive biomarkers have not been identified. Here, we analyze with Multiplexed Quantitative Immunofluorescence the pretreatment tumor microenvironment (TME) of urothelial cancer samples from patients treated with atezolizumab to identify correlations with treatment efficacy in a Real-World-Evidence (RWE) study. We assessed with Multiplexed Quantitative Immunofluorescence the expression of CD8, PD-1, TIM-3 and LAG-3 on T-cells in the different compartments of the TME (tumor, stroma and whole tissue) in pre-treatment tissue microarrays. We studied associations between the expression of the markers and clinical efficacy. One hundred-nine patients received atezolizumab, showing an overall response rate of 23.8%. Safety was comparable to previous studies with atezolizumab. Pre-treatment tumor samples were available from 45 patients. CD8+ T-cell density was significantly increased in the tumor compartment, but not in the stromal compartment, of patients experiencing complete and partial responses, as compared with patients presenting stable disease or progression. Similar results were observed for co-expression of CD8/PD-1, CD8/TIM-3 and CD8/PD-1/TIM-3/LAG-3. Our findings support the relevance of the density and spatial distribution of CD8+ T-cells and its co-receptors for the clinical efficacy of single-agent PD-L1 blockade in patients with advanced urothelial cancer. RWE studies are a valuable tool for identifying predictive biomarkers.