BRAF V600E Mutant Colorectal Cancer Subtypes Based on Gene Expression.
- Barras D Swiss Institute of Bioinformatics, Bioinformatics Core Facility, Lausanne, Switzerland. david.barras@unil.ch Mauro.delorenzi@unil.ch sabine.tejpar@uzleuven.be.
- Missiaglia E Swiss Institute of Bioinformatics, Bioinformatics Core Facility, Lausanne, Switzerland.
- Wirapati P Swiss Institute of Bioinformatics, Bioinformatics Core Facility, Lausanne, Switzerland.
- Sieber OM Systems Biology and Personalized Medicine Division, Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia.
- Jorissen RN Systems Biology and Personalized Medicine Division, Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia.
- Love C Systems Biology and Personalized Medicine Division, Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia.
- Molloy PL CSIRO Preventative Health Flagship, Animal, Food & Health Sciences Division, North Ryde, NSW, Australia.
- Jones IT Department of Surgery, University of Melbourne and Colorectal Surgery Unit, Royal Melbourne Hospital, Parkville, Victoria, Australia.
- McLaughlin S Department of Colorectal Surgery, Western Hospital, Footscray, Victoria, Australia.
- Gibbs P Systems Biology and Personalized Medicine Division, Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia.
- Guinney J Sage Bionetworks, Fred Hutchinson Cancer Research Center, Seattle, Washington.
- Simon IM Agendia NV, Amsterdam, the Netherlands.
- Roth AD Oncology, Hôpitaux Universitaires de Genève, Genève, Switzerland.
- Bosman FT Department of Pathology, Lausanne University, Lausanne, Switzerland.
- Tejpar S Laboratory of Molecular Digestive Oncology, Department of Oncology, Katholieke Universiteit Leuven. david.barras@unil.ch Mauro.delorenzi@unil.ch sabine.tejpar@uzleuven.be.
- Delorenzi M Swiss Institute of Bioinformatics, Bioinformatics Core Facility, Lausanne, Switzerland. david.barras@unil.ch Mauro.delorenzi@unil.ch sabine.tejpar@uzleuven.be.
- 2016-06-30
Published in:
- Clinical cancer research : an official journal of the American Association for Cancer Research. - 2017
Amino Acid Substitution
Biomarkers, Tumor
Cluster Analysis
Codon
Cohort Studies
Colorectal Neoplasms
Computational Biology
DNA Methylation
Gene Expression
Gene Expression Profiling
Gene Expression Regulation, Neoplastic
Humans
Kaplan-Meier Estimate
Models, Biological
Mutation
Neoplasm Metastasis
Neoplasm Staging
Prognosis
Proteomics
Proto-Oncogene Proteins B-raf
Signal Transduction
Workflow
English
PURPOSE
Mutation of BRAF at the valine 600 residue occurs in approximately 10% of colorectal cancers, a group with particularly poor prognosis. The response of BRAF mutant colorectal cancer to recent targeted strategies such as anti-BRAF or combinations with MEK and EGFR inhibitors remains limited and highly heterogeneous within BRAF V600E cohorts. There is clearly an unmet need in understanding the biology of BRAF V600E colorectal cancers and potential subgroups within this population.
EXPERIMENTAL DESIGN
In the biggest yet reported cohort of 218 BRAF V600E with gene expression data, we performed unsupervised clustering using non-negative matrix factorization to identify gene expression-based subgroups and characterized pathway activation.
RESULTS
We found strong support for a split into two distinct groups, called BM1 and BM2. These subtypes are independent of MSI status, PI3K mutation, gender, and sidedness. Pathway analyses revealed that BM1 is characterized by KRAS/AKT pathway activation, mTOR/4EBP deregulation, and EMT whereas BM2 displays important deregulation of the cell cycle. Proteomics data validated these observations as BM1 is characterized by high phosphorylation levels of AKT and 4EBP1, and BM2 patients display high CDK1 and low cyclin D1 levels. We provide a global assessment of gene expression motifs that differentiate BRAF V600E subtypes from other colorectal cancers.
CONCLUSIONS
We suggest that BRAF mutant patients should not be considered as having a unique biology and provide an in depth characterization of heterogeneous motifs that may be exploited for drug targeting. Clin Cancer Res; 23(1); 104-15. ©2016 AACR.
Mutation of BRAF at the valine 600 residue occurs in approximately 10% of colorectal cancers, a group with particularly poor prognosis. The response of BRAF mutant colorectal cancer to recent targeted strategies such as anti-BRAF or combinations with MEK and EGFR inhibitors remains limited and highly heterogeneous within BRAF V600E cohorts. There is clearly an unmet need in understanding the biology of BRAF V600E colorectal cancers and potential subgroups within this population.
EXPERIMENTAL DESIGN
In the biggest yet reported cohort of 218 BRAF V600E with gene expression data, we performed unsupervised clustering using non-negative matrix factorization to identify gene expression-based subgroups and characterized pathway activation.
RESULTS
We found strong support for a split into two distinct groups, called BM1 and BM2. These subtypes are independent of MSI status, PI3K mutation, gender, and sidedness. Pathway analyses revealed that BM1 is characterized by KRAS/AKT pathway activation, mTOR/4EBP deregulation, and EMT whereas BM2 displays important deregulation of the cell cycle. Proteomics data validated these observations as BM1 is characterized by high phosphorylation levels of AKT and 4EBP1, and BM2 patients display high CDK1 and low cyclin D1 levels. We provide a global assessment of gene expression motifs that differentiate BRAF V600E subtypes from other colorectal cancers.
CONCLUSIONS
We suggest that BRAF mutant patients should not be considered as having a unique biology and provide an in depth characterization of heterogeneous motifs that may be exploited for drug targeting. Clin Cancer Res; 23(1); 104-15. ©2016 AACR.
- Language
-
- English
- Open access status
- bronze
- Identifiers
-
- DOI 10.1158/1078-0432.CCR-16-0140
- PMID 27354468
- Persistent URL
- https://sonar.ch/global/documents/137802
Statistics
Document views: 9
File downloads:
- fulltext.pdf: 0