Potential markers of response and resistance to programmed cell death-1 blockade in first-line therapy of cisplatin-inilegible advanced urothelial cancer.

Abstract

449 Background: Immune checkpoint inhibitors (ICIs) have demonstrated clinical benefit in advanced urothelial cancer (UC) patients. An alert was issued about a decrease in survival for bladder cancer patients with PD-L1-low status receiving immunotherapy versus chemotherapy as first-line therapy. We studied a small series of UC patients treated with first-line PD-1 checkpoint inhibition in order to analyze their characteristics and patterns of response to therapy. Methods: Eleven UC samples were obtained from patients before undergoing therapy with ICIs. Patients were classified according to their benefit from therapy in responders (n=5) and non-responders (n=6). Genomic and immunohistochemistry analyses were performed. Results: Both luminal and basal UC subtypes showed benefit from ICIs. Tumors from non-responders showed increased mutations in chromatin remodelling genes and the amplification of 3q26-28 region. Transcriptome analyses showed that tumors from responders displayed a significant enrichment of genes associated with interferon γ and α response, TNFα via NFκB, genes upregulated by MYC or E2F1, genes involved in G2/M checkpoint and epithelial-mesenchymal transition compared to non-responders. Specific immune cell subsets were present in the tumor microenvironment of tumors from responders and non-responders. Immunohistochemistry showed that none of the immune cell markers analyzed individually was sufficient to discriminate between responders and non-responders. However, the increase in FOXP-3, PD-L1, PD-1, CD8, β2 microglobulin and CD68 and the decrease in CD4 and CD163 cells identified UC patients that responded to ICIs. Conclusions: Our findings confirm that the evaluation of pre-treatment UC tumor samples provides valuable information that could influence treatment decisions. Despite the clinical benefit of PD-1/PD-L1 inhibition in UC, only a fraction of patients benefit from therapy. Our data suggest that responders and non-responders display diverse genomic and transcriptome changes as well as specific immune cell subsets in the tumor microenvironment that can be identified by conventional IHC staining.