ASCO 2017: Comprehensive molecular characterization and analysis of muscle-invasive urothelial carcinomas

Chicago, IL ( Dr. Lerner presented a continuation of a study published in 2014 on 131 tumors that had integrated molecular analysis.1  This original publication analyzed 131 tumors of advanced bladder urothelial carcinoma showing 32 significantly mutated genes, 9 not previously reported. This current presentation focused on the follow-up analysis of the entire cohort, composed of 412 tumors from the TCGA project in chemotherapy-naïve, muscle-invasive urothelial bladder cancer. 

Dr. Lerner initially showed the summary of new findings of this current study. This included 58 significantly mutated genes, 5 mutation signatures, HLA and neoantigen analysis, updated expression subtypes, integrated no-coding RNA analysis, multivariable analysis, and more.

In this study, tumors were analyzed for DNA copy number variants, somatic mutations (WES), DNA methylation, mRNA, non-coding RNA (lncRNA and miRNA) and protein expression, gene fusions, viral integration, pathway perturbation, clinical correlates, outcomes, and histopathology. 

All patients from 36 source sites had muscle invasive disease with 61% having a T3-T4 disease, 31% having N+ and 2.4% with M+ disease. 70% had a smoking history and none of the patients received prior chemotherapy or radiotherapy. Median follow-up was 17.5 months with 35% of the patients having a recurrence and 181 deaths (44%) had occured with 65% of them being cancer related. 13% of the specimens were variant histology while the rest were urothelial carcinoma.

There was a high overall somatic mutation rate (8.2/Mb), with 58 significantly mutated genes (SMG), 34 not identified in the previous initial study1, and 16 not identified as SMGs in pan cancer analysis.  The authors identified 5 mutation signatures including APOBEC-a and b (67%), ERCC2 (8%), C > T CpG dinucleotieds (20%), and a single ultra-mutated sample with a functional POLE mutation. APOBEC mutagenesis was associated with an extraordinary 89% 5 year survival (p = 0.0013). High mutation burden and neo-antigen load were also associated with improved outcome (p = 0.00014 and 0.00078, respectively). For each of the previously identified 5 mRNA subtypes, specific characterizations were found, for the Luminal papillary: 44% FGFR3 alterations, more CDKN2A deletions and low CIS signature. For the luminal infiltrated: low purity, high lymphocyte infiltrate, immune marker expression, highest p53 like and EMT signature. For the Luminal – high uroplakins (umbrella or intermediate cell origin with luminal differentiation?). For the Basal squamous – basal and bladder cancer stem cell markers, squamous differentiation markers, strongest immune expression including T cell markers. And finally for the neuronal subtype: high expression of neuronal differentiation and development genes, and loss of TP53 and RB1. When correlating these subtypes to survival outcomes, Luminal-papillary had the best outcome while the neuronal subtype had the worst (p=0.0014).  Clustering converged for mRNA, lncRNA and miRNA expression, and for inferred activity of gene sets associated with regulator expression. The authors also identified subsets with differential epithelial-mesenchymal transition scores, carcinoma-in-situ scores, and survival, with implications for distinct therapeutic potential. A future treatment paradigm for muscle invasive bladder cancer was therefore suggested by Dr. Lerner (see figure below).

In summary, the results provide unique insights into mechanisms of bladder cancer development, and identify novel subsets of MIBC that may benefit from differential treatment approaches. A high mutation rate mainly due to APOBEC mutagenesis was noted with 58 significantly mutated genes. mRNA expression clustering refined TCGA version 1 and identified a poor survival neuronal subtype, lacking small cell or neuroendocrine histology. Finally, basal/squamous and immune infiltrated subtypes may represent unique opportunities for checkpoint targeted therapy. 

Presented By: Seth P. Lerner, TCGA Network; Baylor College of Medicine, Houston, TX

Written By: Hanan Goldberg, MD, Urologic Oncology Fellow (SUO), University of Toronto, Princess Margaret Cancer Centre
Twitter: @GoldbergHanan

at the 2017 ASCO Annual Meeting - June 2 - 6, 2017 - Chicago, Illinois, USA

1. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature 2014; 507(7492): 315-22.