Skip to main content
Premium Trial:

Request an Annual Quote

Genomic Health Says Future Lies in Next-Gen Sequencing


By Julia Karow

Genomic Health
has been exploring how next-generation sequencing can be harnessed for biomarker discovery in clinical cancer samples, and plans to make its research capabilities available to pharmaceutical partners, for example to develop companion diagnostics.

Eventually, the cancer MDx test provider's current diagnostics, which are based on real-time PCR, will be replaced by a sequencing-based "universal cancer assay," company executives said, that will reveal individual differences between cancers and how they respond to different therapies.

"In essence, we have taken our entire research group at Genomic Health and refocused it to become, if you will, a commercial cancer genome center," Randy Scott, Genomic Health's executive chairman, told In Sequence last week. "It's not just a pilot project at Genomic Health but it's actually pivotal to how we see the future of the development of genomic diagnostics."

Up until now, the maker of the pioneering Oncotype DX breast cancer diagnostic has been using RT-PCR as its main platform, which provides exquisite sensitivity, specificity, and dynamic range, and has been "a workhorse for the clinical laboratory," Scott said.

"But we do see the future as migrating to next-gen sequencing over time," he said. "Probably the first stage is that next-gen sequencing will move into all of our discovery programs, and over the course of the next several years, it will be brought to a level that will allow us to move into clinical development, and once it moves into clinical development, it will be on a path to be a commercial assay."

Long-term, he said, a sequencing-based universal cancer assay will be developed that will include both a tumor's mutation and expression profiles, which will be correlated with, for example, response to therapy.

Between the next five to 20 years, "we expect to see [sequencing-based] products beginning to come to market, and we absolutely believe that most of the products we have on the market today, or that are in late-stage development, will be replaced by next-generation sequencing platforms at some point in the future," Scott said. "And, of course, we would love to be the company that does that."

Genomic Health will not be the first company to run diagnostic tests on a next-generation sequencing platform, as several firms and non-profit labs have launched sequencing-based genetic tests for specific diseases, including Ambry Genetics, GeneDx, Correlagen Diagnostics, and Emory Genetics Laboratory (IS 5/25/2010).

But "for the most part," next-gen sequencing platforms are not yet ready for clinical applications, Scott said, since they have been designed for the research community. "That doesn't mean that they can't be adapted to the clinical laboratory," he added, and he expects that vendors will seek FDA approval for their sequencing platforms.

For next-gen sequencing to be ready for the clinic, "a lot of robust methods development has to take place," he said, including sample preparation, reproducible sequencing, improved sensitivity and specificity, and better bioinformatics tools.

Ultimately, said Scott, the greater value for Genomic Health will be in managing the information. The company currently employs between 50 and 60 staffers in IT, bioinformatics, and biostatistics, and according go Scott, "that will be one of the fastest-growing areas for the company for many years to come."

[ pagebreak ]

Platform Perusal

Genomic Health began looking into next-generation sequencing in 2006, initially by sending libraries to various platform providers in order to evaluate different sequencing technologies.

While the firm is still exploring different platforms today, within the last year it brought in-house the Illumina Genome Analyzer and expects to have the Illumina HiSeq 2000 in place in the near future.

In addition, Genomic Health has sufficient storage for "several scores of" terabytes worth of data, "and it's very clear we are going to need to continue to grow that over time," according to CSO Joffre Baker.

The advantage of next-gen sequencing, he said, is that it provides "vastly more information" that has "enabled us to look at many more mutations than we have looked at in the past, as well as many more transcripts than we have been able to look at in the past with RT-PCR."

While Genomic Health currently only has the Illumina sequencer in-house, Baker and Scott are quick to point out that the company is "technology-agnostic" and is still considering other sequencing platforms. "Different technologies and different platforms will be useful for different applications," Scott explained. "In some areas, you want long sequence reads at high quality, in others, you really want short reads, but high throughput that allows you to count, for example in transcriptome profiling." Also, the technology landscape keeps changing. "It's pretty clear that this field is moving, in terms of platforms, at just a blinding rate," Baker said.

So far, company researchers have focused on enabling sequencing to analyze RNA from formalin-fixed paraffin-embedded tumor tissues, both for whole-transcriptome profiling and targeted profiling. In addition to analyzing RNA, they are studying mutations in panels of genes known to be involved in major cancers the company is interested in.

Discovery of novel cancer mutations – the aim of many ongoing academic cancer genome projects -- will not be a focus for the company, however. Rather, it wants to harness results from these projects in its clinical studies. The plan is to "validate these [mutations] in a clinical setting and connect either mutations or expression profiles to actual clinical outcomes, especially around treatment selection," Scott said.

Earlier this year at the Advances in Genome Biology and Technology conference, scientists from Genomic Health and Illumina presented results from a preliminary study in which they extracted RNA from estrogen receptor-positive and negative FFPE breast tumors that were fixed seven to eight years earlier, depleted it of ribosomal RNA using a novel method, and sequenced the RNA on the Illumina GA.

According to the meeting abstract, they detected about 25,000 unique RefSeq gene transcripts, representing rare, intermediate, and abundant transcripts. Also, the relative abundance of estrogen receptor transcripts was correlated with results from qPCR. Fewer than 5 percent of the reads came from ribosomal RNA, and fewer than 5 percent from mitochondrial DNA. The study demonstrates, they said, "the feasibility of using RNA-seq with FFPE tumor specimens" and they "anticipate that RNA-seq will provide a new perspective about transcriptional changes that underlie the phenotypes of individual tumors."

Right now, Baker said, company researchers are engaged in a project that compares a couple of dozen normal and breast cancer samples, and "it's clear that in the near future, we will be doing discovery programs with next-generation sequencing."

To analyze the data, the company is using a blend of open-source bioinformatics tools, tools developed in-house, and programs provided by Illumina. In addition, it is currently evaluating commercial tools.

Analyzing the transcriptome, in particular, is challenging because other than for coding genes, the boundaries of transcripts are not yet well defined in the human genome, and novel splice sites are still being discovered. "The bioinformatics around these questions is far from optimized anywhere in the world," Baker said, adding that "even at the level of mapping, I don't think anyone would say that we have a completely satisfactory program."

Another challenge is to develop robust data processing and analysis tools to be able to compare, in the future, hundreds to thousands of patient samples from clinical trials. This will include handling and visualizing large amounts of data, and presenting them to clinicians. As an example, the researchers need to determine how to correct for differences in the amount and quality of RNA from different patients.

For its internal product development, Genomic Health focuses on breast, colon, and prostate cancer, and in all three areas, "we will be applying next-gen sequencing as a discovery platform," according to Scott. "We certainly, long term, believe that we will bring next-gen sequencing technology to bear on those three market places."

In addition to developing its own diagnostic tests, the company wants to make its technology platform and analytical tools available to pharmaceutical companies as a service, in particular to develop companion diagnostics for targeted drugs. Eventually, it hopes to become the commercial provider of such companion diagnostics as well, banking on the global sales and distribution network it has already in place for its Oncotype DX test.

Scott said the company has already talked to several undisclosed pharmaceutical companies about such partnerships at the recent annual meeting of the American Society of Clinical Oncology. Genomic Health has already had collaborations with Pfizer, Bristol-Myers Squibb and Imclone, and "we're now in a position where we want to begin to expand that program and pursue it built around the next-generation sequencing technology," he said.

So far, Genomic Health has not published results from next-gen sequencing projects in a peer-reviewed journal, but it is scheduled to present its platform at the "Beyond the Genome" meeting conference at Harvard Medical School in October.

The Scan

Study Links Genetic Risk for ADHD With Alzheimer's Disease

A higher polygenic risk score for attention-deficit/hyperactivity disorder is also linked to cognitive decline and Alzheimer's disease, a new study in Molecular Psychiatry finds.

Study Offers Insights Into Role of Structural Variants in Cancer

A new study in Nature using cell lines shows that structural variants can enable oncogene activation.

Computer Model Uses Genetics, Health Data to Predict Mental Disorders

A new model in JAMA Psychiatry finds combining genetic and health record data can predict a mental disorder diagnosis before one is made clinically.

Study Tracks Off-Target Gene Edits Linked to Epigenetic Features

Using machine learning, researchers characterize in BMC Genomics the potential off-target effects of 19 computed or experimentally determined epigenetic features during CRISPR-Cas9 editing.