AUA 2017: Big Data in Prostate Cancer: Identifying Cancer Vulnerabilities through Personalized Network Identification

Boston, MA ( After opening remarks from the SBUR “Big Data in Prostate Cancer” session moderators Dr. Ganesh Raj and Dr. Christopher Evans, Dr. Stuart from UC-Santa Cruz delivered a high level presentation discussing how to identify cancer vulnerabilities through personalized network identification. At the heart of personalized medicine, Dr. Stuart notes, is that “all (many) roads lead to Rome”, highlighting that there may be one primary genetic alteration, but there are many ways to acquire this alteration (gene fusion, alternative splicing, etc). Using examples from the early TGCA work on gliobastoma multiforme, he notes that cellular pathways integrate seemingly disparate alterations in genetic mutations.

In the first part of his talk, Dr. Stuart focused on the ‘one for all’ mentality of identifying all cancer forms, which may be from alterations in microRNA, proteins, DNA copy number, DNA methylation or exome-mutations. Using their group’s PARADIGM integration system, they are able to incorporate multiple types of data, producing what Dr. Stuart describes as circuitry (or as he affectionately calls a ‘hairball’ of output). Using PARADIGM tumor maps they are able to identify clusters of signals within respective types of cancer tissue, for example with bladder cancer, the signals diverge into bladder-enriched, squamous and LUAD-enriched islands of signals. Based on the type of clustering signals, Dr. Stuart’s group has shown that this may affect clinical patient outcomes.

In the second half of his presentation, Dr. Stuart discussed the ‘all for one’ personalized medicine mentality of finding treatments specific to individuals. Using the example of pediatric malignancies, his group noted that one particular signal cluster (SARC) groups with neuroblastoma ALK fusion tumors. By using RNA-sequencing data, they can identify a set of genes that are significantly up- or down-regulated, subsequently matching the profile generated with a known cancer subtype to obtain robustness of transcriptome classification. Following this, they are able to link mutations to transcriptional changes with heat-diffusion on various mapping networks, which allows his group to infer active transcription factors. Specifically, his group has assessed this in mCRPC patients, developing a network-based selection of targets and target combinations for individual patients.

As Dr. Stuart concludes, the TCGA collection has produced the largest genomics compendium, enabling pan-cancer analysis of collections of adult genome signatures, revealing signature pathways with potential treatment targets.

Presented By: Joshua Stuart, University of California-Santa Cruz, Santa Cruz, CA, USA

Written By: Zachary Klaassen, MD, Urologic Oncology Fellow, University of Toronto, Princess Margaret Cancer Centre

Twitter: @zklaassen_md

at the 2017 AUA Annual Meeting - May 12 - 16, 2017 – Boston, Massachusetts, USA