Oops, you're using an old version of your browser so some of the features on this page may not be displaying properly.

MINIMAL Requirements: Google Chrome 24+Mozilla Firefox 20+Internet Explorer 11Opera 15–18Apple Safari 7SeaMonkey 2.15-2.23

Poster display session

3490 - Assessing functional Androgen Receptor (AR) pathway activity using a computational model

Date

11 Sep 2017

Session

Poster display session

Topics

Translational Research

Presenters

Anne van Brussel

Citation

Annals of Oncology (2017) 28 (suppl_5): v573-v594. 10.1093/annonc/mdx390

Authors

A. van Brussel1, M.A. Inda2, E. Den Biezen2, D. van Strijp2, J. Wrobel2, H. van Ooijen2, R. Hoffmann3, W. Verhaegh2, A. van de Stolpe2

Author affiliations

  • 1 Precision & Decentralized Diagnostics, Philips Research, 5656 AE - Eindhoven/NL
  • 2 Precision & Decentralized Diagnostics, Philips Research, Eindhoven/NL
  • 3 Oncology Solutions, Philips Research, Eindhoven/NL
More

Resources

Abstract 3490

Background

Cellular signal transduction research identified 10-15 signaling pathways responsible for driving tumor growth. Defining pathway activity in tumor tissue is necessary to optimize targeted therapy choice. Verhaegh et al (Cancer Research 2014) used a Bayesian network approach to model transcriptional programs of signaling pathways. These pathway models use mRNA expression levels of validated direct pathway target genes to infer a probability of pathway activity in individual patient samples. Here, initial results of the AR model are presented.

Methods

28 bona fide AR target genes were selected and a Bayesian network model for the AR pathway was built and calibrated. The model uses target genes mRNA levels (Affymetrix HG-U133Plus2.0 array) as input to infer probability of AR pathway activity. Evaluation was done using multiple public datasets from clinical studies. The model was also adapted for qPCR data as input, using a subset of most informative target genes.

Results

Biological validation on androgen stimulated LNCaP cultures showed expected AR activity (GSE7868), which was inhibited by the anti-androgen bicalutamide (GSE7708). In cell line xenograft models (GSE21887, GSE33316, GSE966), AR was active in the presence of androgen and inactive in castrated mice. In prostate hyperplasia and 90% of primary prostate cancer (PCa) samples (GSE17951, GSE28403, GSE32982, GSE3325, GSE45016) AR was active; in contrast, AR was inactive in 30-50% of castration resistant or metastatic samples. AR was active in primary PCa samples, but not in samples taken 3 days after surgical castration (GSE32982). In other cancer types AR was mostly inactive, except for a subset of Her2 subtype Breast Cancer (BCa), Luminal BCa (EM-TAB-365, GSE12276, GSE17097, GSE21653), and meningioma samples (GSE16581, GSE9438). Translation to qPCR-RNA measurement as input was successful, underscoring the portability of our approach to other measurement platforms

Conclusions

Our biologically validated computational AR model enables assessing functional AR pathway activity in individual patient tissue samples, based on mRNA microarray or qPCR input from respectively FF or FFPE material. Other pathway models and clinical validation studies are in progress.

Clinical trial identification

Legal entity responsible for the study

Philips Research

Funding

None

Disclosure

A. van Brussel, M.A. Inda, E. Den Biezen, D. van Strijp, J. Wrobel, H. van Ooijen, R. Hoffmann, W. Verhaegh, A. van de Stolpe: Employee of Philips Research.

This site uses cookies. Some of these cookies are essential, while others help us improve your experience by providing insights into how the site is being used.

For more detailed information on the cookies we use, please check our Privacy Policy.

Customise settings
  • Necessary cookies enable core functionality. The website cannot function properly without these cookies, and you can only disable them by changing your browser preferences.