Efficacy of Anti-PD(L)1 Treatment in Patients with Metastatic Urothelial Cancer Based on mRNA- and Protein- Based PD-L1 Determination: Results from the Multicentric, Retrospective Fosmic Trial

Background Immunotherapy (IO) with antibodies against PD1 or PD-L1 has been approved for 1st and 2nd line treatment of metastasized urothelial bladder cancer (mUC). Here we evaluated efficacy in a retrospective Real-World-Phase IV trial based on molecular subtypes and PD-L1 status. Methods Cancer specific survival from start of IO treatment to death was retrospectively analyzed in a multicenter cohort of 65 patients having received palliative immune checkpoint inhibition for mUC in 1st and 2nd line setting. Intrinsic molecular subtypes were assessed by CK5, CK20, GATA3, FOXA1 and CD44 immunohistochemistry and RT-qPCR of KRT5 and KRT20. PD-L1 status was determined using the 28-8 Dako-assay. Stromal tumor infiltrating lymphocytes were assessed on HE slides (Salgado et al, 2015). Survival analyses were performed by applying clinically relevant cut-offs and using Kaplan-Meier analysis and logistic regression for cancer specific survival (CSS). Results Altogether, 43% of tumors presented with basal differentiation, 57% were luminal. As expected, basal tumors exhibited significantly higher levels of PD-1 and PD-L1 gene expression, higher amounts of sTILs and PD-L1+ immune and tumor cells. Hierarchical clustering of all immunological biomarkers revealed three cluster: “Inflamed high” (n = 13, 20%), “Inflamed low” (n = 22, 33.8%), and “Uninflamed” (n = 30, 46.2%). There were no significant survival differences for any of these groups (CSS). However, patients with good performance status (ECOG 0) and low PD-L1 expression showed significantly better CSS compared to those with ECOG0 / PD-L1 high and ECOG1 or ECOG2, respectively (30.3 months vs. 16.33, 12.07 and 15.13 months; p = 0.0047). Conclusions Immunological determinants of mUC show comparable distributions and correlations with subtypes of localized curatively treated MIBC patients. Interestingly, data indicate improved CSS after PD-1/PD-L1 treatment in patients with good performance and low PD-L1 expression, which warrants validation in larger cohorts. Clinical trial identification DRKS00016925. Legal entity responsible for the study STRATIFYER Molecular Pathology GmbH. Funding Janssen-Cilag GmbH. Disclosure R. Wirtz: Leadership role: CEO & CSO, STRATIFYER Molecular Pathology GmbH. All other authors have declared no conflicts of interest.