News

Molecular characterisation of urothelial bladder carcinoma

29 Jan 2014
Molecular characterisation of urothelial bladder carcinoma

by ecancer reporter Clare Sansom

Cancers of the bladder are almost always carcinomas involving the epithelial lining of the organ, known as the urothelium.

Urothelial bladder carcinoma is a relatively common malignancy that is diagnosed significantly more often in men than in women and that accounts for about 150,000 deaths each year worldwide.

Early stage bladder cancer in which the carcinoma is confined to the urothelium is relatively easily treated, but few treatment options are available when the cancer is more advanced.

Several studies have identified genes that carry recurrent somatic mutations or copy number alterations in bladder carcinomas, and most of these genes have already been linked to at least one other tumour type.

A large group of researchers led by John Weinstein of the University of Texas MD Anderson Cancer Center, Houston, Texas, USA and working as part of the Cancer Genome Atlas project has now conducted a comprehensive molecular and genetic study of 131 urothelial bladder carcinomas.

All tumours involved in the study had invaded the muscular wall of the bladder and had not been exposed to chemotherapy.

The researchers performed a wide range of analyses including gene sequencing and copy number analysis, and analyses of RNA expression, protein expression and phosphorylation, alternative splicing and pathway perturbation.

These molecular features of the tumours were correlated with their histological and clinical characteristics.

In 118 cases, the tumour samples were compared to samples of peripheral blood and in 23 cases they were compared to adjacent normal bladder tissue.

Whole-exome sequencing of tumour and matched normal samples revealed significant levels of somatic mutation in a total of 32 different genes, many of which had not been linked to this tumour type before.

Many of these genes were known to be involved in biochemical processes linked to cancer such as kinase signalling, cell cycle and chromatin regulation; however, nine had not previously been reported as being significantly mutated in any cancer studied within the Cancer Genome Atlas Project.

Seven tumours with mutations in the gene RXRA (retinoid X nuclear receptor alpha) also showed increased expression of genes involved in lipid metabolism and differentiation into adipose tissue.

Other genes mutated in a significant number of these tumours that had not previously been implicated in cancer through the Cancer Gene Atlas project included a nucleotide excision repair gene, ERCC2, and a transcription factor involved in response to oxidative stress, NFE2L2.

Three tumours featured translocations of DNA on chromosome 4 leading to the expression of a fusion protein comprising the amino-terminal end of the fibroblast growth factor receptor 3 (FGFR3) and the carboxy-terminal end of the protein encoded by ERCC2; this was predicted to constitutively activate the FGFR3 kinase.

Clustering the samples based on genomic aberrations grouped them into three classes and clustering based on RNA expression grouped them into four.

One of the expression signatures obtained based on RNA expression (“cluster I”) was enriched in tumours with papillary morphology and mutation or over-expression of FGFR3; another (“cluster III”) was similar to a pattern found in basal-like breast cancers and squamous cell tumours of the lung, head and neck.

In contrast, both cluster I and cluster II shared some characteristic expression patterns with the luminal A type of breast cancer.

The papillary and basal/squamous-like tumour subtypes could also be distinguished using microRNA sequencing and analysis of protein expression patterns. 

Viral DNA or transcripts could be identified in a small proportion of the samples, indicating that viral infection might occasionally play a role in the development of this tumour.

An integrated analysis of all the molecular data suggested that three pathways were commonly dysregulated in urolthelial bladder carcinoma: cell cycle regulation; kinase and phosphatidylinositol-3-OH kinase (PI(3)K) signalling; and chromatin remodelling.

In fact, genes involved in chromatin regulation or remodelling were identified as mutated more frequently in this sample of bladder carcinomas than in any other common cancer studied through the Cancer Genome Atlas project.

Mutations in genes that are the targets of actual or potential anti-cancer drugs were identified in 69% of the tumours, and these included several members of the epidermal growth factor receptor (EGFR) family. 

Taken together, these results suggest that a majority of urothelial bladder carcinomas may respond to one or more targeted agents, and drugs targeting the chromatin remodelling pathway might be particularly beneficial.

This would greatly increase the currently limited options for treating invasive or metastatic bladder tumours.


Reference

The Cancer Genome Atlas Research Network (2014). Comprehensive molecular characterization of urothelial bladder carcinoma. Nature, published online ahead of print 29 January 2014. doi:10.1038/nature12965