Originally published Feb. 7.
Genomic Health last week said that it has used next-generation sequencing to survey the human transcriptome and identify an undisclosed number of candidate biomarkers linked to breast cancer recurrence.
Through the hypothesis-generating exercise, enabled by Illumina's next-generation sequencing platform, Genomic Health identified "hundreds of coding and non-coding transcripts that are differentially expressed in tumor versus normal, non-cancerous breast tissue, including a subset of genes statistically associated with the recurrence of breast cancer." Additionally, the company found "new candidate biomarkers linked to breast cancer recurrence from regions outside of the previously identified protein-coding sections of the genome."
Genomic Health presented its findings at the 12th Annual Advances in Genome Biology and Technology meeting in Marco Island, Fla., last week.
According to Joffre Baker, Genomic Health's chief scientific officer, the study demonstrated the ability of next-generation sequencing to accelerate the biomarker discovery process using formalin-fixed tissue. "Our long-term goal for this and future studies is to validate the findings with large, well-designed clinical trials in an effort to connect either gene mutations or expression profiles to clinical outcomes and provide additional personalized information to physicians and patients beyond that already available through the use of Oncotype DX," Baker said in a statement.
In the study, Genomic Health compared gene expression profiles between 12 normal and 12 tumor formalin-fixed breast specimens — all of which were between 10 and 13 years old. On average, each sample yielded 19 million base pairs of DNA, 85 percent of which came from unique sites in the genome, the company reported.
Company researchers identified differences in the expression of more than 3,500 coding genes when comparing tumor and normal specimens. Next they investigated the relationship between these genes and the risk of breast cancer recurrence in published studies involving 3,000 patients with a record of clinical outcome. "This analysis showed that sets of transcripts over-expressed in tumors compared with normal tissues produced only a modest enrichment for prognostic significance, but further evaluation of these transcripts by gene set analysis produced a set that is highly enriched for prognostic genes," Genomic Health said in a statement.
The investigators also developed an algorithm that allows them assess transcripts from non-coding regions of the genome. "Through this process more than 1,000 differentially expressed non-coding sequences were identified," Genomic Health stated.
In order to home in on specific biomarkers linked to breast cancer recurrence, researchers designed reverse transcriptase-PCR assays to survey several non-coding transcripts using breast cancer specimens from 136 patients. "Several of these non-coding RNAs proved to be associated with breast cancer recurrence risk," Genomic Health reported.
Genomic Health's next steps will be to validate these preliminary associations. "We expect that for most validation studies we will continue to do well designed studies on archival tissue with prospectively designed endpoints (similar to those used to validate Oncotype DX and KRAS," a company spokesperson told PGx Reporter. "The specifics of study size will depend on the clinical population and the purpose of the test."
Last year, Genomic Health announced plans to apply whole-genome sequencing to develop a "universal cancer assay." Genomic Health officials have noted, however, that it will be several years before its sequencing efforts are translated into a commercial product (PGx Reporter 09/29/10).