Genomic Health Discovers New Gene Networks Associated with Breast Cancer
Genomic Health, Inc announced the presentation of the first-of-its-kind clinical outcomes study for biomarker discovery applying next generation sequencing (NGS) for whole transcriptome profiling of archival formalin-fixed paraffin embedded (FFPE) tumor specimens. Results presented at the 2011 CTRC-AACR San Antonio Breast Cancer Symposium demonstrate that a proprietary RNA-Seq method developed at the company is capable of identifying validated genes that previously had been discovered by RT-PCR, the technology used to develop the Oncotype DX® breast cancer test.
RNA-Seq refers to the use of high-throughput sequencing to identify which RNAs are expressed in a given sample and generate a quantitative gene expression profile by counting the number of RNA copies that match each gene or region of the genome. The technology allows Genomic Health scientists to assay the entire genome simultaneously to discover regions of the genome that are turned on or off in disease. From these changes, researchers are focused on predicting disease outcomes. In addition to re-confirming the original 21 Oncotype DX genes originally identified by RT-PCR, this study also revealed more than 1,800 new biological relationships associated with breast cancer recurrence.
"This study marks an important milestone in the clinical application of next generation sequencing by establishing the biological and technical validity of a new proprietary method that accommodates very low amounts of RNA from fixed paraffin-embedded tumor tissue, and therefore is practical to use broadly for biomarker discovery and clinical development," said Joffre Baker, Ph.D, chief scientific officer, Genomic Health. "These findings position Genomic Health to lead in the delivery of next generation cancer diagnostics."
Based on these data, Genomic Health is accelerating its efforts to provide a comprehensive genomic platform for clinical research and development combining both whole transcriptome profiling and mutation analysis next year. Accordingly, the company is accelerating its investment in information technology infrastructure to support the flow and storage of vast amounts of patient genomic data.
"Next generation sequencing allows researchers to process and analyze large volumes of data in a fraction of the time currently required by conventional gene expression profiling technologies," said Jose Esteban, MD, Providence Saint Joseph Medical Center, Burbank, Calif., a co-investigator of the study. "New methodology used in this study applies NGS-based whole transcriptome profiling to very low amounts of RNA from FFPE tissue that had been archived for up to 12 years, to provide vast amounts of data, with data quality sufficient for broad biomarker discovery from regions outside the previously identified protein-coding sections of the genome."
Genomic Health scientists carried out whole transcriptome RNA-Seq on FFPE tumor RNA from a cohort of 136 breast cancer patients with tumor tissue obtained at the time of diagnosis and established clinical outcomes for disease recurrence. These tumors were originally analyzed by RT-PCR in the biomarker discovery phase of the development of the 21-gene Oncotype DX assay. Sequencing was carried out using the Illumina® HiSeq 2000 instrument and generated over 40 terrabytes of data for analysis. The data analysis methods used in this biomarker discovery study were the result of a two year effort to optimize methods for RNA-Seq data normalization and other statistical parameters.
RNA-Seq results were successfully generated for all patients using RNA inputs of just 100 nanograms. Bar-coded RNAs were sequenced with two samples per flow cell lane, yielding 43 million reads per each duplexed sample, 69 percent of which uniquely map to the human genome.
Standardized hazard ratios and p-values for the 21 Oncotype DX genes determined by RNA-Seq were comparable to those originally obtained using RT-PCR. Additionally, whole transcriptome RNA-Seq revealed more than 1,800 new coding and non-coding RNAs associated with breast cancer recurrence risk, many of which belong to previously unrecognized gene networks. Additional research will be performed to determine whether these new genes can be observed in other cohorts and whether they add value beyond the 21 Oncotype DX genes.