Amgen, Brigham & Women’s reach milestone on population study.
By Kevin Davies
Sept. 5, 2008 | Sometime this month, Amgen’s Alex Parker should be throwing a party. That is when Parker and his small team at Amgen’s Cambridge, Massachusetts research facility aim to complete a landmark project—the genotyping of the 28,000 volunteers in the Women’s Genome Health Study, producing a veritable goldmine of 10 billion genotypes.
Parker joined Amgen in 2007 to help set up the company’s first high-throughput genetics lab. Amgen had entered a partnership with investigators at Brigham and Women’s Hospital, led by epidemiologist Julie Buring, for the Women’s Genome Health Study. (See sidebar: “Tracking the Long Term”)
The participants in this ongoing longitudinal study are mostly nurses and physicians recruited across the United States. To complement the 13 years of cumulative disease and exposure data on the participants, Parker’s task was to pull off “by far the largest genetic study ever done in a female cohort—and one of the largest genetic studies ever done, period—in under 18 months.”
“This really is a public/private partnership,” says Parker. “We’re working together to make this a really positive, beneficial experience for both parties.”
It is a testament to the new world of genome analysis that Parker did this with a team of just four associates in his lab—Sunita Badola, Lauren Young, Kim Tsui, and Elizabeth Robinson. On a recent tour, the spacious lab was eerily quiet save for the hum of robotic devices going about their business. Using Illumina bead stations, Parker’s group is characterizing some 360,000 SNPs throughout the thousands of genome samples.
So far, the study has focused on cardiovascular disease and cancer. The first paper was published earlier this year in the American Journal of Human Genetics. “They’re producing actionable data that we’re using in Amgen drug discovery projects, so it’s already bearing fruit,” says Parker.
By incorporating DNA data with the other elements of the study, Amgen now has the opportunity to study the genetics and risk factors for many other common diseases, including osteoporosis, diabetes, and Parkinson’s. For example, about 3,500 women in the study have reported a diagnosis of osteoporosis, and one third of those had fractures. Another 90 patients have Parkinson’s disease.
Amgen senior management has enjoyed a long-standing relationship with Harvard University/Brigham cardiologist Paul Ridker. The idea for whole-genome scanning had existed for some time. “Finally the cost came down to where it was doable,” says Amgen Executive Director Scott Patterson. “Senior VP, R&D, Joe Miletich believes in the idea of the industrial/academic/government partnership. And he’s always said, we could write a check and let somebody else do it.”
“Because the Women’s Genome Health Study is unique in its size and the length of time participants have been followed, it will allow scientists to focus on the genetic risk factors for diseases with extraordinary statistical power,” commented Miletich.
“This was really an opportunity for Amgen,” says Parker. “This was our excuse, if you will, to put this platform in place. We knew Amgen needed it. Amgen had a gap in its portfolio in terms of genetics, whereas we were strong in immunology, very strong in proteomics. This was the opportunity to put the genetics capability in place.”
Parker didn’t waste any time. By the time he and Ridker met in late 2007 to discuss some of the initial analyses, the Amgen group had already genotyped 7,000 individuals.
Slicing the Data
The first slice of data was published by Ridker, Buring, Parker, and colleagues last May. Although elevated levels of C-reactive protein (CRP) are known to predict risk of metabolic syndrome, diabetes, myocardial infarction, and stroke, Amgen decided to undertake a comprehensive analysis of the influence of genetic variation on CRP. Parker and the Brigham group conducted a genome-wide association study in more than 6,000 healthy women for more than 335,000 SNPs as potential determinants of plasma CRP concentration.
The team found seven loci significantly associated with plasma CRP levels. Two of those genes have been implicated in a hereditary form of diabetes, while others reside close to genes for the leptin receptor, apolipoprotein E, and CRP itself: in all, a very intriguing set of candidate genes involved in insulin metabolism, weight homeostasis, and atherothrombosis. The paper concludes that, “Common variation in several genes involved in metabolic and inflammatory regulation have significant effects on CRP levels, consistent with CRP’s identification as a useful biomarker of risk for incident vascular disease and diabetes.”
Parker says that genotyping the first few thousand individuals provides sufficient power to detect genetic effects on a whole range of plasma biomarkers, including cholesterol and CRP, which can in turn be linked to environmental factors such as age, body weight, smoking, and menopause.
In another study in press, the Amgen/Brigham team identified SNPs influencing the levels of LDL cholesterol and other cardiovascular risk factors. The results show levels are influenced by variants in genes including ApoE, ApoB, the LDL receptor, and a novel gene on chromosome 1. Importantly, the results were confirmed by a second genome scan in a different population.
“This is a beautiful illustration of the power of the technology that’s now available, the fact that we can do thousands if not tens of thousands of genome scans at a density that gives us this kind of ability to localize these signals. Now you’ve got two candidate genes,” says Parker.
Isn’t it frustrating, I ask Parker, to pinpoint a candidate gene, only to find it has no known function? “No, I love it,” counters Parker, “because that means we’ve found something new. It’s boring to find something that you already know about, right?”
Another glimpse at the early data involves the biomarker Lipoprotein A (LPA)—another component of cholesterol particles. Parker describes a rare variant near the LPA promoter. The small percentage of the population that carries the minor (TC) allele shows greatly elevated levels of LPA. “We’re in the process of finding the SNP that demarcates this high LPA population. This is obviously a dramatic effect for a single locus.”
Interestingly, there is tantalizing evidence that carriers of the LPA minor allele may be the population that benefits the most from aspirin therapy against heart attack. “Admittedly, it’s only pulling out 4 percent of the population, and when we find the “real SNP” as opposed to the marker SNP, maybe it’s only 3 percent of the population. Nevertheless, these people have a very different benefit-risk picture.”
In his previous work on rheumatoid arthritis, Parker saw a major difference in patient response to various targeted therapies. “There’s exactly the same opportunity to try to understand why some people are just biologically different than others, and some drugs are appropriate for some patients and not for others,” says Parker. If anything, the situation is a little clearer in oncology, but he expects rapid progress in many disease areas. “I think some of it may emerge from this study and studies like it that we’re doing.”
Parker hints that other examples are already revealing new potential targets and impacting Amgen’s drug discovery programs. “It shushes the naysayers who think that this isn’t a good thing for industry to be investing in. We’re getting benefits as drug developers out of this already.” Moreover, the genotyping platform is already being applied to at least one fairly large clinical trial, with several more in the queue.
Once Parker and his team have finished celebrating the data collection, the real fun will begin: analyzing the data, “so that ultimately patients will reap the full scientific benefits of the data discovered during this project.”
Ridker, P. et al. Am J Hum Genet. 2008: 82, 1185-1192. Chasman, D. et al. Circulation (in press).
Tracking the Long Term
Brigham and Women’s Hospital epidemiologist Julie Buring launched the Women’s Health Study clinical trial back in the early 1990s. The initial goal was to determine whether low-dose aspirin could have a prophylactic effect in preventing cardiovascular disease, particularly heart attacks in women (as was known in males). Participants, who had to be at least 45 years old and in good health, were randomized either to aspirin or placebo (as well as to vitamin E or placebo).
That ten-year study showed only a modest benefit to daily aspirin in relation to major cardiovascular events, although there was greater protective benefit to the risk of ischemic strokes than heart attack. (There was also an increased risk of gastrointestinal bleeding.)
In October 2006, Buring and Paul Ridker forged a collaboration with Amgen and the National Heart, Lung, and Blood Institute to genotype almost 28,000 study participants (those from the original study of 40,000 who provided blood samples). Buring performs the extensive biochemical characterization of the trial participants. Her colleague Dan Chasman heads up the statistical analysis efforts.
Importantly for the Amgen work, three-quarters of the 28,000 women gave a blood sample at baseline. DNA was prepared and plasma biomarkers were measured in the entire population. “This is a very, very large data set—all generated in the same laboratory under the same conditions, so highly comparable,” says Parker.
Among the study participants, 4,500 women have developed hypertension, 2,500 have osteoporosis, 1,200 have developed diabetes, 1,000 have had a serious cardiovascular event, and 900 have developed breast cancer.
This article appeared in Bio-IT World Magazine.
Subscriptions are free for qualifying individuals. Apply Today.