Affymetrix last week provided a fresh update on a targeted SNP genotyping panel for drug metabolism that has languished in R&D for several years, as well as a general overview of how it intends to use the underlying technology for the assay, which it acquired through its purchase of ParAllele Bioscience two years ago.
The company also addressed its strategy for the whole-genome genotyping market, where it plans to emphasize the use of its 1.8 million-marker SNP 6.0 array for replicating genome-wide association studies rather than move existing customers into new custom genotyping applications or launch a higher-density genotyping array.
Doug Farrell, Affy’s head of investor relations, told investors at Leerink Swan’s Life Sciences Tools Roundtable Conference in New York last week that the firm has renewed its focus on bringing the molecular inversion probe, or MIP, assay it acquired with ParAllele to the clinical market as well as making it available to researchers that would like to move from large-scale association studies to more focused genotyping.
Farrell said in particular that development of the MIP assay was tabled in 2006 as the company underwent a shuffle in its administrative team and focused on the roll-out of its 500,000-plus-feature SNP 5.0 array and SNP 6.0, which was launched in June. Now with the 6.0 on the market, the company is turning its attention back to ParAllele’s technology with implications for the academic and clinical markets.
“Obviously the [technology] was back-burnered last year as we focused on some of our whole-genome genotyping products, but MIP is a very powerful product for looking at subsets of SNPs,” Farrell said during the conference of Affy’s lower-multiplex genotyping capabilities.
John Blume, vice president of RNA products at Affy, told investors at the conference that the MIP assay has “long legs” and that it is already being used by some in the pharmaceutical industry on a research-use-only basis “in some clinical development trials.”
While “one could imagine taking it all the way through to the clinic,” he said, “I think that probably we need to work on improving the all-around usability of the assay — it’s a complicated assay — in terms of making it more robust and easier to manufacture” before marketing it for clinical use.
Blume also said that the company is seeing interest from pharma in its drug-metabolism, or D-MET, assay panel, which is based on the MIP platform. The product is a ParAllele legacy that tests around 160 genes that are involved in drug metabolism and transport pathways. Though the D-MET panel has been in the hands of some customers since Affy bought ParAllele in 2005, Affy has yet to offer it as a standard product to the general research market.
In February, Affy CEO Steve Fodor said that the firm expected to make D-MET available as a catalog product, as well as to recognize revenue from the $120 million ParAllele buy, by the end of this year. (see BAN 2/20/2007).
Affy officials did not update this timeline during the Leerink Swan conference.
Last week, Blume described D-MET as a “targeted panel that we have for drug-metabolizing enzymes and transporters [that] has come out of a collaboration with Eli Lilly.” He said that Lilly is currently using D-MET in clinical trials and that other pharmas have shown interest in the panel. According to Blume, D-MET is a “downstream [tool] that is looking at must-have content that is not addressed by whole-genome tools.”
The U133 of the Genotyping World?
While the impending roll-out of the D-MET panel indicates that Affy’s ParAllele buy could eventually pay off, the company has said in no uncertain terms that its new flagship product — at least for the DNA analysis side of its business — is the SNP 6.0 Array. As Kevin King, president of life sciences, pointed out during a second-quarter earnings call last month, Affy’s “improved top line growth was driven by the successful launch of the 6.0 Array, which has been an important catalyst for our DNA analysis business” (see BAN 7/31/2007).
Farrell pointed out last week that the 6.0, which launched June 1, was only on the market for about four weeks of the second quarter, “yet still represented about 35 percent of the DNA samples run during the quarter, so the uptake of it was fantastic.”
He also said that the firm expects more customers to transition to the 6.0 over the next few quarters, which will in turn continue to drive growth in Affy’s DNA analysis business. “Typically with a new product launch like this over a period of one or two quarters the vast majority of product users will transition to the new one,” he said.
“So I think that as more users transition over to 6.0, and about 70 percent of customers will transition at some point, that will give a chance to boost the overall margins in that space,” he said.
Blume said that Affy was able to ensure quick adoption of the 6.0 through its strong links to the research community, and that the firm sees the tool as being Affy’s flagship product for some time. He said that in particular, it is the copy number variant content on the array that has spurred interest from Affy’s loyal base in the academic community. Moreover he likened the 6.0 to Affy’s Human Genome U133 Plus 2.0 Array, which is widely used in the gene expression market.
“Affy’s close ties to the research community ... have led us to early on put on a lot of copy number content on 6.0s, and I think we timed that very well,” he said.
“We have reached the point of diminishing returns on coverage.”
“The feedback from researchers right now has been that this is going to be the 133 of the genotyping world,” said Blume. “It’s going to be a tool of choice for doing pure genotyping studies and its going to be a discovery tool for copy number as well. The power of being able to do both is going to be tremendous,” he added.
Indeed, Affy appears convinced of the power of the 6.0 and it is encouraging users to run replication studies on the 6.0 rather than move to custom projects. The company also recommends that users adjust the statistical power of their studies to use the 6.0, rather than wait for a higher-density array from Affymetrix.
“The worst thing in the world would be to run [association] studies and miss things ─ you found one gene instead of 10,” Blume said. He said that replication studies could help avoid these scenarios “because once you do downstream, 90 percent of the answer is lost to you. You’ll never see it again. Because the arrays are so dense the need for fine-mapping is going away.”
He also said that a higher-density chip is not in the cards, and that studies should be adjusted to take advantage of the SNP 6.0. “We could do a 2-million-SNP chip, but I think we have reached the point where the coverage is so good that the dominant factor is actually getting people to run appropriately powered studies,” he said. “So it’s not running 2 million SNPs on a thousand [samples]; more important will be running a million SNPs on two thousand. We have reached the point of diminishing returns on coverage,” he added.
Blume said that Affy’s customer base is also content with the 6.0 and that demand for a 2-million-feature chip is not strong. “When I have gone to people and said, ‘Hey, on an experimental basis, how would you like for me to put all 5 million addressable SNPs on a four-chip screening set?’ Surprisingly the answer has been ‘No, you’ve already got all the coverage I need’,” he said.
Blume also said that Affy is currently spending its R&D dollars on supporting the 6.0 array for use by more pharma and biotech customers who traditionally demand a more automated format than those in the research community.
“Our R&D investments at the moment are not around trying to do 2 million [features] and lower the feature size. They are around trying to put the content on pegs and onto plates and having it work out of saliva samples and having it work out of [formalin-fixed paraffin embedded] samples,” said Blume. “That’s really where I think there is going to be the greatest gain in terms of making the technology accessible to more people.”