555P - Association between colon cancer molecular subtypes obtained by expression profiling and by microRNA (miR) profiling

Date 29 September 2014
Event ESMO 2014
Session Poster Display session
Topics Colon Cancer
Translational Research
Presenter Beatriz Perez-Villamil
Citation Annals of Oncology (2014) 25 (suppl_4): iv167-iv209. 10.1093/annonc/mdu333
Authors B. Perez-Villamil1, M. Paz-Cabezas1, A. Pascual-Montano2, M. Fuentes3, J. Sastre1, E. Díaz-Rubio1
  • 1Medical Oncology, IdISSC Hospital Clinico San Carlos, 28040 - Madrid/ES
  • 2Functional Bioinformatics Group, National Center for Biotechnology-CSIC, Madrid/ES
  • 3Preventive Medicine, IdISSC hospital Clínico San Carlos, Madrid/ES

Abstract

Aim

To analyze correlation between tumor subtypes obtained by expression profiling and by miR profiling in colon cancer and to identify miR targets in the colon tumors. Novel tumor subtypes have been identified in colon cancer with different clinical behavior. A gene signature associated with high-stroma recognized poor outcome patients (1). A similar classification associated high-stroma-subtype to an EMT-subtype and to a diferent response to cemotherapy treatment (2,3).

Methods

mRNA and miR expression profiling were analyzed in 88 colorectal tumors (24 Dukes A, 26 B, 19 C, 19 D) using gene expression and miR microarrays. Hierarchical clustering was used to uncover tumor subtypes TALASSO software (http://talasso.cnb.csic.es) was used to find miR targets and their correlation wth mRNA expression. Pathways analysis was carried out using GeneCodis software (http://genecodis.cnb.csic.es).

Results

There was a clear association of the tumor subtyes obtained by expression profiling wih the tumor subtypes obtained by miR profiling (p < 0.001). Stroma (p < 0.001) and BRAF mutations (p = 0.024) were also associated to the miR pofiling classificaton. 788 genes showed a significant (p < 0.05) interaction with specific miRs. Pathways jointly regulated by mRNAs and miRs showed differences between tumor subtypes. Relevant KEGG and Panther pathways identified were related to ECM-receptor interaction, focal adhesion, cytokine-cytokine receptor interaction, integrin sigaling and inflammation mediated by chemokine and cytokine signaling pahways as well as WNT and TGFbeta signaling pathways.

Conclusions

1. There is a clear correlation of tumor classification obtained by expression profiling and by miR profiling 2. Stromal genes and BRAF mutations are asociated to both classifications 3. Pathways related to extracellular matrix and inflamation are implicated References 1-Perez-Villamil et al. BMC Cancer 2012 12:260 2- Schlicker A et al. BMC Medical Genomics 2012 5:66 3- Roepman P et al.Int J Cancer 2013 134:552.

Disclosure

All authors have declared no conflicts of interest.