Combined Molecular Subclass and Immune Phenotype Correlate to Atezolizumab Plus Radiation Therapy Response in Invasive Bladder Cancer: BPT-ART Phase 2 Study.

Abstract

Bladder preservation therapy in combination with atezolizumab and radiation therapy trial, which was a multicenter, open-label, single-arm phase 2 study, showed a promisingly high interim clinical complete response (cCR) rate of 84.4% (38/45). In the present study, we aimed to identify potential tissue biomarkers for achieving cCR using bladder preservation therapy in combination with atezolizumab and radiation therapy. We used tumor tissue samples of the bladder and blood samples collected from patients at baseline to analyze the tumor immune microenvironment at baseline using an integrated approach of immunophenotyping, genomic, and tumor-infiltrating lymphocyte (TIL) profiling. Immune phenotype analysis revealed that cCR rates of patients with the desert phenotype were as similarly high as patients with excluded/inflamed phenotypes (73.3% [11/15] vs 93.3% [14/15], P = .33) despite lower programmed death-ligand 1 expression levels in the desert phenotype. Genomic and TIL profiling then revealed that increased CD8+ and CD204+ TIL infiltration, high CD8:forkhead box protein P3 ratios in the stroma of the excluded/inflamed phenotypes, and gene alterations, such as CDK12, GNAS, NOTCH2, and AR1D1A, were associated with a high cCR rate (93.3%). Furthermore, the characteristics of these dual TILs, CD8-forkhead box protein P3 ratios, and gene alterations (especially FGFR3) bifurcated the desert phenotype into 2 subgroups with different cCR rates (100% [11/11] and 0% [0/4]). These potential subgroups, defined by combined molecular subclass and immune phenotype, could lead to the identification of good responders to atezolizumab plus radiation therapy for invasive bladder cancer. However, given the small cohort size and limited number of tumor samples, these findings should be viewed as hypothesis-generating and require further validation in larger studies.