Identification of biomarkers for early progression in muscle invasive bladder cancer (MIBC) using real-world data.

Abstract

e16568 Background: MIBC accounts for ~25% of all newly diagnosed bladder cancers and is associated with a high rate of recurrence despite radical treatment. The time to progression (TTP) can vary significantly between patients and can affect their prognosis. We leveraged a real world clinico-genomics database to classify patients as early vs late progressors, understand its impact on their outcomes and identify genetic drivers of early progression. Methods: This retrospective study was executed using the deeply curated ConcertAI Genome360 bladder cancer dataset. Only patients with confirmed MIBC status with reported tumor progression after initial treatment were considered. Based on the distribution of TTP from MIBC diagnosis of this patient cohort, we subdivided the patients into early vs late progressors. Overall survival (OS) and progression free survival (PFS) analyses were performed between these 2 cohorts. The underlying deleterious genomic mutations were identified in each cohort and differential genomic enrichment analysis was performed to identify statistically significant biomarkers correlated with early progression (p-value < = 0.05). Additionally, identified biomarkers were analyzed in the context of administered therapies. Results: The TTP exhibited a bimodal distribution with a break at around 500 days, so we chose this as the cut-off to define early (N = 203) vs late progressors (N = 51). As expected, there was a significant difference in the median PFS of early vs late progressors (0.62 vs 2.41 years (p < 1e-6)). The OS was also significantly different (1.98 vs 4.36 years (p = 6e-6)). We further identified mutations in 6 biomarkers which were enriched in early vs late progressors. These were CDK12 (p = 0.05), EGFR (p = 0.034), LRP1B (p = 0.05), MET (p = 0.045), PTEN (p = 0.026) and SMAD4 (p = 0.047). These patients were primarily treated with platinum-based therapies +/- gemcitabine/PD-L1 therapies in-line with MIBC treatment guidelines. The identified driver genes activate PI3K-AKT-MTOR, JAK-STAT3, b-catenin and ERK pathways which are known to interfere with the mechanism of action of platinum-based therapies, thus potentially forming the basis of early progression. Conclusions: Using our Genome360 bladder dataset, we have identified prognostic biomarkers for early progression in MIBC. This may provide clinicians with a tool to identify and modify treatment for a subset of MIBC patients who are at risk of early progression on standard of care therapies.