By Monica Heger
Diagnostic company Asuragen has entered the next-generation sequencing space with targeted cancer panels for the Ion Torrent PGM and Illumina Genome Analyzer. The panels have been optimized to work on DNA extracted from formalin-fixed paraffin-embedded tissue as well as tissue extracted by fine needle aspiration.
In a webinar in late December, Gary Latham, the company's director of diagnostic research and technology development, said that Asuragen has developed three panels: a 52-gene panel on the GA for discovery purposes, and smaller 16-gene and five-gene panels on the PGM. It is also offering Ion's AmpliSeq Cancer Panel.
The internally developed panels — AsuraSeq-1000, AsuraSeq-35, and AsuraSeq-8 — target 1,000, 35, and eight amplicons, from 52, 16, and five known cancer genes, respectively, while Ion's AmpliSeq Cancer Panel targets 190 amplicons from 46 genes. Each of the panels will be sequenced to a median 1,000-fold coverage, and Asuragen will report variants detected with a frequency as low as 5 percent.
Currently the company is offering the panels for research purposes. Officials said the firm may look to eventually validate the sequencing instruments and panels in its CLIA-certified laboratory, but did not provide a timeline for that process.
Carole Berry, senior vice president and general manager at Asuragen, told Clinical Sequencing News that the company will market the panels to pharmaceutical companies, biotech companies, and academic groups doing translational medicine studies. The company did not disclose costs of the panels, and said that prices vary depending on a customer's requirements.
The panels mark Asuragen's first entry into the next-gen sequencing space. The molecular diagnostics company has historically developed companion diagnostics and offered pharmacogenomics services such as gene expression profiling, genotyping, and mutation testing using PCR and array technology.
In the next-gen arena, Asuragen is currently equipped with a GA, a PGM, and RainDance's RDT 1000 system for target enrichment. Latham said that the company is "platform agnostic" and "as we see business opportunities expand, we'll grow."
"We're certainly aware of what's happening with the next-gen and the next-next-gen [systems]. We're paying attention to the MiSeq … and as others come on the market we'll take those into consideration," he said.
Latham told CSN that the company decided to get into the next-gen space because it felt that the technology was at a "tipping point" for "being applicable in oncology," and that sequencing from FFPE tissue, in particular, was an "unmet need" in the field.
Sequencing DNA extracted from FFPE tissue is challenging due to crosslinking that occurs because of the fixing process and because the DNA is highly fragmented. Because most clinical samples are stored and banked using FFPE, however, there is a huge interest in using these specimens for clinical next-gen sequencing research, particularly in the oncology field.
Asuragen chose to use PCR for sample enrichment with the Ion Torrent panels, as opposed to a capture-based approach, because of its broad experience with PCR technology and also because PCR is "well-vetted in diagnostics," Latham said during the webinar. For its Illumina panel, it is using RainDance's microdroplet PCR technology.
Development and Validation
One of the main tasks in developing the panels was to determine next-gen sequencing's sensitivity for mutation detection from FFPE tissue. Latham said that there was some concern that the background mutation level for FFPE tissue would be substantially higher than from fresh frozen tissue, making it difficult to distinguish rare variants from noise.
He said that the company carried out a lot of internal work to design the primers and the PCR products, most of which are proprietary. It's "important to make sure that we maintain PCR characteristics that are consistent with highly fragmented samples."
When the company compared a cohort of 39 colorectal FFPE specimens to 27 cancer cell lines on the GA, it found "no fundamental difference in the background incidence of detection," Latham said.
Additionally, he said, to validate each panel, the company compared sample replicates, cell line mixtures, well-characterized cancer cell lines, DNA extracted from FFPE tissues of both known and unknown genotypes, and fine needle aspiration-extracted DNA of both known and unknown genotypes.
Additionally, it compared the panels against Sanger sequencing, a Luminex bead array, real-time PCR, and each other.
One problem with comparing next-gen to Sanger sequencing, said Latham, is that Sanger sequencing can only detect variants down to about a 10 percent to 20 percent frequency, but, to be relevant for cancer, a next-gen panel should be able to reliably detect low-level rare variants, present at 5 percent frequency, or even lower.
This is where having multiple technologies helps, Latham said. Comparing the 52-gene Illumina panel to Ion's AmpliSeq Cancer Panel yielded concordant results in the sections where the panels overlapped, Latham said, despite the fact that two use different sample-prep technologies and different sequencing chemistries.
Being able to cross-validate the different panels helped the company determine the cut-off point for detecting variants. While company researchers have detected variants as low as around 1 percent frequency, Latham said that the firm reports to customers "confidently" on variants down to 5 percent frequency, and at "lower confidence" on variants present at less than 5 percent.
Additionally, the company tested each panel and platform to determine minimal sample requirements. It requires 10 nanograms of DNA for its AsuraSeq-8 panel and the Ion AmpliSeq Cancer Panel, 50 nanograms for its AsuraSeq-35, and between 100 and 250 nanograms for AsuraSeq-1000.
Latham said he envisions the panels being used for a whole spectrum of clinical applications including screening, confirmation, and diagnosis. "The larger panel is definitely of value when looking at samples you're trying to get a broad profile on … and making sure you're not missing anything," he said.
Increasingly, pharmaceutical companies will want to start using next-gen sequencing to profile patients earlier in the drug development process to make sure they have "a really comprehensive portrait of the mutations that are prevalent" and also to identify any "trends with specific mutations that may not have been previously well-documented."
Following a broad screen, that information can be condensed into more focused panels, he said, for instance to guide patients into the correct clinical trial or to stratify responders from non-responders.
Latham hopes that the four different panels will help it compete with other companies and universities now offering targeted sequencing-based cancer panels.
"There's rarely a one size fits all," he said. The different panels "give us more latitude for how we can work with customers." Additionally, added Berry, the company's history in working with FFPE tissue will also help it compete in the market.
Berry said that Foundation Medicine would likely be Asuragen's biggest competitor for the cancer assays. Foundation is developing a targeted sequencing cancer test that will sequence nearly 200 genes, which it is aiming to launch in the second quarter of 2012 in a CLIA-certified laboratory (CSN 6/8/2011).
Foundation has already signed pharmaceutical companies Novartis and Cellgene as collaborators (CSN 4/13/2011 and 5/17/2011).
Some universities are also now offering targeted sequencing cancer panels, such as Washington University's School of Medicine, which in November began offering a 28-gene cancer panel on the HiSeq, and Baylor College's Genetics Laboratory, which will soon offer sequencing services using Ion Torrent's AmpliSeq Cancer Panel.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.