Abstract 3922: Proteomic profiling of muscle invasive bladder cancer

Abstract

Abstract Background: Only one quarter of patients with muscle invasive urothelial bladder cancer (MIBC) gain approximately 5% improvement in 5-year overall survival from cisplatin-based neoadjuvant chemotherapy (NAC). For patients with residual invasive cancer post radical cystectomy there is no standard of care and high mortality. Previous TCGA projects focused on pre-NAC MIBC genomic and transcriptomic alterations and identified 5 molecular subtypes with differential risk and response. For this study we hypothesized that comprehensive proteomic and phosphoproteomic profiling of MIBC prior to NAC will further define mechanisms responsible for chemotherapy resistance, and identify specific and actionable targeted therapies for patients with NAC-resistant tumors. Methods: OCT-embedded and flash frozen tissue samples from 143 eligible patients were processed and tested for proteomics quality control (QC). Samples containing >45% tumor content, less than 10% muscle content and >2,000 protein identifications in a single-shot quality control assay were selected for deep-scale proteomic and phosphoproteomic profiling. A final cohort of 60 samples (52 pre-treatment and 8 patient-matched post-treatment tumors) were multiplexed using tandem mass tags (TMT-11), fractionated by basic reverse phase chromatography and analyzed by liquid chromatography and mass spectrometry (LC-MS). Results: Over 12,000 proteins were identified in total, and 8,353 proteins identified in all samples, including 425 kinases and 77 targets of FDA approved cancer therapies. Principal component analysis identified two distinct resistant clusters and one sensitive cluster. The samples were clustered based on the 5 TCGA subtypes which resulted in the sub stratification of the resistant clusters into two resistance-enriched basal-squamous clusters (BS1 and BS2 respectively), one infiltrated-luminal cluster (L1), one sensitive luminal cluster (L2) and one intermediary cluster (L3) (with 86% samples resistant). UV response, Epithelial to Mesenchymal transition and Myogenesis were significantly elevated in both L1 and L3 (resistant) relative to L2 (sensitive). Similar, these pathways are significantly altered in L1 relative to L3. EGFR, CDK6, ITPKC and CSNK1 were elevated in the B4 subtype, all of which are potentially druggable targets. Matched samples and phosphoproteomics data will be presented. Conclusion: Surrounding non-tumor tissue can obfuscate true tumor signatures in MIBC, cohort selection through stringent pathologic QC allows proteomic profiling to identify tumor features that correlate with NAC resistance. This dataset provides proof of principle that actionable targets and subtype distinctions can be identified through discovery proteomics and further analysis in NAC treated cohorts are justified. Citation Format: Matthew V. Holt, Meggie N. Young, Karoline Kremers, Alexander Saltzman, Antrix Jain, Mei Leng, Hugo Villanueva, Lacey E. Dobrolecki, Beom-Jun Kim, Meenakshi Anurag, Matthew J. Ellis, Anna Malovannaya, Seth P. Lerner. Proteomic profiling of muscle invasive bladder cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3922.