Several articles published last week in the New England Journal of Medicine that questioned the utility of genome-wide association studies and direct-to-consumer genetic testing have prompted Affymetrix and Decode Genetics to defend the use of array-based GWAS and the resulting data.
While NEJM published a quartet of perspectives that offered a range of opinions on the value of GWAS and DTC testing, two articles that were critical of these methods attracted particular attention from financial analysts and the American news media, who concluded that the perspectives could negatively impact companies that provide tools for these studies.
In particular, a New York Times article that outlined the NEJM papers, entitled "Genes Show Limited Value in Predicting Disease," painted a bleak picture of GWAS, which it said have "turned out to explain surprisingly little of the genetic links to most diseases."
The article added that the publication of the papers "affects companies that offer personal genomic information and that had assumed they could inform customers of their genetic risk for common diseases, based on researchers’ discoveries."
In a research note published last week, Isaac Ro, an analyst at Leerink Swan, cited the Times article as evidence that the "sentiment in the research community continues to move in favor of [second-generation] sequencing-based approaches for future GWAS work, which favors [Illumina] and [Life Technologies division Applied Biosystems] while posing further pressure on [Affymetrix's] core franchise."
David Goldstein, director of the Center for Human Genome Variation at the Institute for Genome Sciences and Policy at Duke University, acknowledged in one of the NEJM articles that the associations discovered in GWAS "reflect real biological causation," but asked "do they matter?"
In Goldstein's opinion, the arrays used in GWAS, manufactured by Affymetrix and Illumina, contain common variants, but it is rarer human variation that might provide the most insight into various conditions. Goldstein recommended that researchers turn away from large-scale GWAS and instead focus on looking at rare variation in "thoughtfully selected" individuals.
"Even though genome-wide association studies have worked better and faster than expected, they have not explained as much of the genetic component of many diseases and conditions as was anticipated," Goldstein wrote. "We must therefore turn more sharply toward the study of rare variants."
In response to the questions raised by Goldstein's NEJM article and associated media coverage, an Affymetrix official agreed that studying rare variation is increasingly on the mind of researchers, but disagreed that studying rare variation goes hand in hand with a switch to smaller, sequencing-driven projects.
"One of the key messages that Goldstein conveyed in the editorial is that rarer variants will have a more profound effect than the common variants that have been examined in GWAS to date," Jay Kaufman, vice president of product marketing at Affy, told BioArray News this week.
"Does this mean that GWAS are of limited value, particularly relative to other approaches? Probably not," he said. "What it does imply is that now that common variation has been extensively studied it is likely time to turn the research focus to other types of variants."
According to Kaufman, second-generation sequencing will "prove to be a useful tool" for discovering putative common and rare variants. But from Affy's perspective, this "bolus of new SNPs that are entering and will continue to enter the public domain" will be useful for designing array products to enable "cost-effective and focused study of rare and structural variants as well as other valuable classes of mutations."
Once these chips become available, Kaufman said that Affy expects a "second, large wave" of GWAS to ensue. "We are very optimistic about this part of our business," he said. Also, Affy will continue to offer chips that feature copy number variation content, said Kaufman. "This is also emerging as an important component of understanding and characterizing the causes of disease," he said.
[ pagebreak ]
An Illumina spokesperson declined to comment specifically on the NEJM articles, but referred to past statements in which the San Diego firm had previously "discussed the synergies across sequencing and arrays."
Like Affy, Illumina is also ramping up for the second coming of GWAS. During the firm's most recent earnings call in February, CEO Jay Flatley said that the company's arrays stood to benefit from its involvement in the 1000 Genomes Project (see BAN 2/10/2009).
Launched in January 2008, 1000 Genomes is relying exclusively on second-generation sequencing technology in its goal to sequence at least 1,000 human genomes in order to create a new, more detailed map of the human genome.
However, as the 1000 Genomes Project generates content and other sources of sequencing data come into the public domain, "we’ll be able to take some of the additional discoveries of variants, of which there is predicted to be a very large number, [and] move those variants onto our chips," Flatley said at the time.
This, in turn, will enable researchers to "begin a whole other round of what we call rich genome-wide association studies. The expectation is that those next phases of studies will use significantly more samples because … to get statistical significance on variants that are at a lower percentage, you just need a larger sample size," he said. "I think that’s clearly going to be the trend, particularly as we get out toward the end of this year. I think that cycle will really begin to kick in a large impact in 2010."
Decode Strikes Back
While Goldstein's proposal to abandon large-scale GWAS in favor of second-gen sequencing perhaps caused the most discussion by users of Affy and Illumina SNP chips, a separate piece in NEJM had implications for companies that offer DTC genetic testing. Several of these firms — including Navigenics, 23andMe, and Decode Genetics — use arrays provided by Affy and Illumina to create genetic risk profiles for their clients for a number of common diseases.
Authored by Peter Kraft, an associate professor of epidemiology and biostatistics at the Harvard School of Public Health, and David Hunter, a professor in the departments of epidemiology and nutrition at the Harvard School of Public Health, the piece argued that "many" rather than "few" variants are responsible for genetic diseases.
In their editorial, Kraft and Hunter cautioned that it is "too early in the cycle of discovery for most tests that are based on newly discovered associations to provide stable estimates of genetic risk for many diseases." They concluded that, "although the major findings are highly unlikely to be false positives, the identified variants do not contribute more than a small fraction of the inherited predisposition."
Kraft and Hunter's editorial prompted a response from Reykjavik, Iceland-based Decode Genetics the next day. Decode has been offering a DTC genetic testing service based on the Illumina array platform called "DecodeMe" since November 2007.
"When we hear august voices argue that you shouldn’t have the right to look at your genome if you want to, or that we shouldn’t test for genetic risk factors until we know everything there is to know about the human genome, we feel obliged to disagree," wrote Decode spokesperson Edward Farmer on the company's website last week.
"In this week’s New England Journal of Medicine, we have heard again that it is 'too early' to measure genetic risk factors for common diseases. Why? In essence because in the coming years we are likely to discover many more genetic risk factors that will help to round out our understanding of all of the risk factors that exist," Farmer wrote.
Rather than wait for the availability of more information on genetic risk, Decode is trying to bring "genetic risk factors into clinical practice as swiftly as possible." He said the company needed to educate doctors and patients about what could be done with the advent of genetic risk information, but said that its customers should "set straight" those who would tell them to wait for genetic testing.
As to how long DTC users might have to wait, Kraft and Hunter wrote that "rapid progress being made through meta-analyses suggests that many more common variants conferring a risk of disease will be identified in the next several years, leading to increasing stability of individual risk estimates."
Once risk estimates are more stable, the "usefulness of genetic screening will need to be considered for each disease, and recommendations about potential interventions will need to be made for persons whose predicted risk exceeds some threshold," the authors wrote. They went on to speculate that, "although testing for inherited susceptibility on the basis of common risk alleles is premature for most diseases, the situation may be very different in just 2 or 3 years."