1519O - Availability of tumour gene expression data facilitates clinical decision-making for patients with advanced cancers

Date 08 October 2016
Event ESMO 2016 Congress
Session Basic science and translational research
Topics Translational Research
Basic Principles in the Management and Treatment (of cancer)
Presenter Janessa Laskin
Citation Annals of Oncology (2016) 27 (6): 526-544. 10.1093/annonc/mdw392
Authors J. Laskin1, C. Ho1, Y. Shen2, M. Jones2, K.A. Gelmon1, H. Lim1, D.J. Renouf1, S. Yip3, A. Tinker1, K. Khoo4, C. Lohrisch1, S. Chia1, B. Deol1, K. Schrader5, Y. Ma2, R. Moore2, A. Mungall2, S. Jones2, M. Marra2
  • 1Medical Oncology, British Columbia Cancer Agency, V5Z 4E6 - Vancouver/CA
  • 2Canada's Michael Smith Genome Sciences Centre, British Columbia Cancer Agency, Vancouver/CA
  • 3Pathology, BC Cancer Agency, Vancouver General Hospital University of British Columbia, Vancouver/CA
  • 4Medical Oncology, BC Cancer Agency, Kelowna/CA
  • 5Hereditary Cancer Program, British Columbia Cancer Agency, Vancouver/CA

Abstract

Background

Whole genome analyses have the potential to identify the full landscape of genomic abnormalities within cancers, and can therefore be used to provide rationales for cancer treatments. The Personalized OncoGenomics study integrates DNA and RNA sequencing information into oncology practice; this analysis evaluates how physicians use the data for clinical decisions and the role of RNA data in identifying actionable targets.

Methods

Pts with advanced tumors and a survival > 6 m were eligible. Each had a tumour biopsy and blood sample which underwent comprehensive DNA (80X) and RNA sequencing followed by bioinformatic analysis including assembly, annotation, and mining of the data to identify somatic aberrations, gene expression changes or other putative cancer “drivers” or actionable targets. Gene expression from tumour RNAseq was compared to the TCGA and Illumina body map. Results were discussed in a multidisciplinary Clinical Genomics Tumour board and characterized as informative, actionable or neither. Clinical actionability can be based on abnormalities of DNA, RNA or both.

Results

Between July 2012 and April 2016, 217 pts had complete sequencing data available. Of these, 165 pts had clinically actionable, and 52 had no actionable pathways identified. Of the165 actionable patients: 34 pending progression, 60 received no POG directed therapy and 71 had POG informed treatment. 60 with no POG directed therapy: 24 poor PS/deceased, 16 clinical trial or off-label treatment not available, 20 POG data was not utilized. 71 treated using POG data: 13 clinical trial, 29 off label treatment and 29 treatment within guidelines of disease site. 40% of the treatment decisions were based on the RNA information; 45% on a combination of DNA and RNA; and 15% based on DNA alone.

Conclusions

Data from DNA abnormalities alone corresponds to the rate noted in panel associated drug matching trials. The availability of RNA expression information greatly increased our ability to identify clinically actionable targets. With the support of the multidisciplinary tumour board meetings and a tiered data system, oncologists had sufficient confidence in the results to seek clinical trials and off-label therapies based on genomic data in the majority of pts.

Clinical trial identification

NCT02155621

Legal entity responsible for the study

BC Cancer Agency

Funding

BC Cancer Foundation

Disclosure

J. Laskin: Academic talk honoraria: Az, Roche, BI Research funds to institution: Lilly, BI. C. Ho: Honoraria AZ, Bayer, BMS, BI, Pfizer, Lilly, Roche Research grants BI, Genzyme. Travel grant BI. H. Lim: Honoraria/Consulting/Research Funding: Eli Lilly, Leo, Bayer, Ipsen, Amgen. D.J. Renouf: Honoraria – Celgene. S. Chia: Personal honorarium Hoffmann LaRoche, Novartis, Genomic Health and Amgen. My institution has received funding for clinical trials from Hoffmann LaRoche, Novartis, Genomic Health, AstraZeneca, Genetech, and Amgen. All other authors have declared no conflicts of interest.