MP39-08 A LISTERIOLYSIN EXPRESSING BCG WITH FAVOURABLE IMMUNOGENICITY AND PRECLINICAL TOXICITY AS A NOVEL TREATMENT FOR NON-MUSCLE INVASIVE BLADDER CANCER

INTRODUCTION AND OBJECTIVES: The molecular profile of bladder cancer has been suggested to play a critical role in prognosis and treatment response. A greater understanding of the somatic mutations that arise, and co-exist, in advanced tumors could aid in the development of targetable therapies for bladder cancer. METHODS: Fifty patients who underwent surgery for pT2+ urothelial carcinoma of the bladder with very high risk pathologic features and adequate tumor volume for DNA extraction were identified for whole exome sequencing of a set of 422 genes selected to represent the cancer genome. Demographic and clinical data were recorded. Known hotspot bladder cancer mutations (e.g. FGFR3, PIK3CA, KRAS, HRAS) were analyzed using standard Sanger sequencing. Whole genome sequencing using Solexa sequencing technology was completed. Bioinformatic analysis was subject to a rigorous analysis pipeline with multiple data filtering steps. Mutations were categorized by type, stratified against previously identified cancer loci in the COSMIC database, and evaluated in pathway analysis and co-mutation plots. RESULTS: Median age was 64 years and 67% of subjects were male. At cystectomy, 36% had positive nodes, 33% were T3, and 28% were T4. With a median follow-up of 1.4 years, 45% developed metastases and 24% died of bladder cancer, demonstrating the very high risk nature of the cohort. Sequencing of 422 genes and application of the multistep filtering algorithm revealed a core set of 212 mutations. When compared to the COSMIC database, higher mutation rates were seen in several genes including PIK3CA, NOTCH2, MSH2, APC and p53. Pathway analysis demonstrated highly mutated pathways including the PI3K/mTOR pathway, the MAPK/ERK pathway, the cell cycle regulators, and epigenetic regulators. Co-mutation analysis showed frequent co-occurrence of mutations in RB and P53 as well as NF1 and PIK3CA. CONCLUSIONS: Discovery phase analysis of the somatic mutations in 50 very high risk MIBC specimens revealed several interesting mutation patterns. These patterns may be useful for designing systemic therapy regimens for patients at very high risk of disease metastasis and death.