Login to Access Video or Poster Abstract: MP88-05
Sources of Funding: none

Introduction

Genome wide profiling studies across cancers have been key to the increased understanding of tumour heterogeneity and recent efforts from large muscle invasive bladder cancer (MIBC) cohorts have led to their classification into molecular subtypes displaying distinct genomic and transcriptomic features. Herein, we performed a comprehensive in silico immune transcriptomic profiling using publicly available datasets to identify immune gene expression patterns associated with molecular subtypes of MIBC.

Methods

We utilized the publicly available global transcriptomic sequencing (RNA-Seq) data from 412 MIBC cases, with the corresponding clinical information downloaded from The Cancer Genome Atlas (TCGA) data Portal. Cases were divided into discovery (n=122) and validation (n=245) cohorts for downstream analysis and were divided into four clusters based on their genomic profiles. To investigate the presence of subtype associated immune signatures we assembled a defined set of 828 immune related genes, consisting of genes involved primarily in Type I and II interferon pathways in addition to other immune response and immune cell phenotype genes. All downstream data analysis was performed in R Bioconductor statistical environment. A one-way ANOVA was utilized to determine significantly differentially expressed genes with a Benjamini and Hochberg correction for false-discovery rate (FDR) correction of q<0.05.

Results

In the 122 case discovery cohort, we identified a total of 452 genes differentially expressed among the four clusters with an FDR q<0.05. The performance of these differentially expressed genes to accurately distinguish the four TCGA clusters was evaluated by unsupervised clustering of both genes and samples. The 64 top 20% of ranked genes were able to distinguish the four clusters in an unsupervised analysis of both the discovery and validation cohorts. The most enriched biological processes in the 452 gene list were response to IFN-γ, antigen processing and presentation, cytokine mediated signalling, cell proliferation, NK cell and macrophage activation and B cell mediated immunity The top five overrepresented pathways included, JAK/STAT signalling pathway, Toll receptor signalling pathway, interleukin signalling pathway, and T cell activation. Kaplan Meier survival analysis revealed that in combination, higher expression of three genes, SA100A7, S100A8 and SERPINB2 significantly associated with decreased survival only between clusters I and III in both discovery and validation cohorts.

Conclusions

Recent evolving findings from completed immunotherapy based clinical trials have emphasized the value of pre-existing tumour immune state that potentially determines response to treatment and survival. Our analyses reveal a grouping of immune gene expression patterns using both supervised and unsupervised clustering approaches. Given that specific genetic alterations associate with these molecular subtypes it seems that anti-tumour immune responses could be partly driven by oncogenic drivers. The findings provide further insights into the association between genomic subtypes and immune activation in MIBC and may open novel opportunities for their exploitation towards precise treatment with immunotherapy.

Funding

none