Deciding the time is right to cash in on the potential of proteomics, the Framingham Heart Study last week announced an agreement with BG Medicine aimed at discovering biomarkers, including protein biomarkers, associated with heart disease.
The collaboration marks the first time that the landmark study is incorporating proteomics on any scale into its research, and is the first cooperative research and development agreement, or CRADA, between FHS and a commercial partner in the study's 60-year history.
The CRADA will cover other types of biomarkers such as metabolomic and lipidomic biomarkers, but most of the work will focus on proteomics, Pieter Muntendam, CEO of Waltham, Mass.-based BG Medicine, told ProteoMonitor this week. Both FHS and BG Medicine declined to disclose the financial terms of the agreement. Work on the CRADA is expected to start soon.
About a year and a half ago, FHS began making public on the web about 5 billion genetic data points for Framingham participants "to allow the scientific community to make discoveries about genetic mechanisms of disease," said Daniel Levy, director of FHS and the Center for Population Studies at the National Heart, Lung, and Blood Institute.
FHS is funded by NHLBI and conducted in collaboration with the Boston University School of Medicine and School of Public Health.
Moving to proteomics is a natural progression from the genomics work that has been done for FHS, Levy said, as the proteomics technology "is ready to allow us to leverage the many scientific resources that we've been creating here for the Framingham Heart Study. …
"We decided that a CRADA was the best method for enhancing the research that we do in Framingham in the area of biomarkers while maximizing the opportunities to fulfill that promise of bringing those discoveries into the marketplace," he said.
FHS first put out a solicitation for the CRADA in November 2006 for commercial and/or academic groups interested in doing biomarker discovery using FHS blood samples. Forty-two candidates responded to the solicitation, from which FHS whittled the list down to 27. It then asked the 27 candidates to submit detailed proposals describing their technologies and how they would be used in the CRADA.
Among the things FHS wanted to see was a track record of success with the technologies being proposed, and one reason BG Medicine was chosen was an earlier CRADA the firm had with the US Food and Drug Administration for a project to discover biomarkers linked to hepatoxicity, Levy said [See PM 02/22/07].
According to Muntendam, BG Medicine completed phase 1 of that CRADA in January and identified a biomarker signature and a difference in liver effects with two closely related agents, which Muntendam declined to name. BG Medicine and its pharmaceutical industry partners are now evaluating whether to continue with the project, he added.
"That project went very well, so they had a good report card," Levy said.
BG Medicine's technology is a mass spec-based method. While that is nothing special as virtually all protein biomarker work is done by mass spectrometry, BG Medicine's proprietary workflow allows it to detect and identify an unusually high number of proteins in small samples — 800 proteins in a 100-microliter sample, Muntendam said.
The technology, he added, is at the 1- to 10-nanogram-per-milliliter range, "where we can reliably detect proteins," and has a coefficient of variation of 12 percent. While research labs "are very good" at doing proteomics on small numbers of samples, Muntendam said that his firm can analyze hundreds of samples at once and get "consistent measurements across" them.
[ pagebreak ]
According to the firm's website, proprietary informatics tools developed by the company allow it to identify specific protein variants and generate detailed protein sequence information associated with disease or drug response.
In addition to mass spectrometry, BG Medicine uses microformat, multiplexed immunoassay technologies for the targeted measurement of low-abundance proteins.
More Than Just a Handful
For FHS, the CRADA is a chance to delve more deeply into what kinds of new information certain biomarkers can yield about heart disease. Since the study began in 1948, environmental factors such as obesity, a sedentary lifestyle, and cigarette smoking have been identified as important risk factors in the development of heart disease, yet the only blood-based biomarker of "risk that has resulted in important approaches to prevention" is cholesterol, Levy said.
FHS has identified about 20 candidate protein markers of heart disease, including fibrinogen, C-reactive protein, and von Willebrand factor, but "we need to add more biomarkers to our armamentarium, and hopefully some of them will be useful for diagnostics, [and] some of them will be perhaps useful therapeutic targets," Levy said.
"We wish to go beyond just a handful [of biomarkers] and increase those numbers substantially, but also wish to identify many through discovery types of approaches, and that's where proteomics comes in," he said.
The BG Medicine CRADA is one of four projects the FHS is conducting as part of a larger initiative called the Systems Approach to Biomarker Research in Cardiovascular Disease, or SABRe CVD, to identify and validate new biomarkers for heart disease.
SABRe CVD includes another project directed at validating about 180 candidate protein biomarkers by immunoassays that have been identified by researches inside and outside of NHLBI. The ultimate goal is to use the biomarkers to identify an individual's risk for developing heart disease, so that a clinician can tailor a therapy addressing that patient's specific needs, Levy said.
BG Medicine will not be part of the 180-marker project, however. A contract has not yet been awarded for that work, Levy said.
SABRe CVD includes another project targeting gene-expression profiling, and one looking at the role microRNAs play in heart disease.
For the five-year CRADA announced last week, BG Medicine will do all of the laboratory work. The study design, selection of cases and controls, and biosample repository will be handled by NHLBI and Boston University.
The project will be carried out in three studies: one looking at coronary atherosclerosis; another looking at metabolic syndrome; and one looking at new clinical cardiovascular events. In total, the three studies will include 1,700 case and control samples.
In total, BG Medicine, NHLBI, and Boston University will study about 1,000 blood-based biomarkers as part of the three studies. BG Medicine will validate any candidate biomarkers by measuring them in the blood samples of up to 7,500 Framingham participants, Levy said.
Single analytes that are shown to be biomarkers for heart disease will become part of the public domain, but BG Medicine will have the rights to develop multivariate markers into "products that [have to] make it through a regulatory pathway," such as in vitro diagnostics, Muntendam said.
In addition to the work it will be conducting in partnership with FHS, BG Medicine is involved with the High-Risk Plaque Initiative, a four-year, $30 million initiative started in 2007 to develop better methods and technologies to identify individuals at risk for heart disease. As part of that, BG Medicine last year began work with the Copenhagen General Population Study to use targeted assays to discover and validate protein biomarkers linked to myocardial infarction [See PM 03/06/08].
The company is now set to validate candidate protein biomarkers it has identified, Muntendam said.
Also as part of the HRP Initiative, BG Medicine has been conducting a BioImage Study to discover and validate biomarkers from blood that may be predictive of heart attacks and stroke. Those biomarkers are correlated with images of patients' hearts and cardiovascular systems.
In addition to providing information on ethnic variables that may play a role in heart disease and strokes, the BioImage Study will provide "an imaging read-out because … if you find a marker set that you think is plaque-originated that correlates with risk, [you want to know if] it really correlates the presence of plaque, and if so what kind of plaque," Muntendam said.
"So the bioimage will allow a biology-to-imaging comparison," he added.