NEW YORK (GenomeWeb) – Supported by a four-year, $2 million Grand Challenge Award from the American Heart Association, Cedars-Sinai Medical Center researcher Jennifer Van Eyk is leading an effort to discover and validate protein biomarkers for cardiovascular disease.
The project, which also incorporates genomics research led by University of Pennsylvania researcher Daniel Rader, will use a variety of proteomic platforms in both the discovery and validation stages to evaluate patient samples drawn from four separate well-characterized cohorts.
The ultimate goal, Van Eyk told GenomeWeb, is to identify a panel or panels of markers that can offer relatively near-term predictions of cardiac risk.
Currently, she noted, the field relies on tools like the Framingham Risk Score, which combines information like smoking, diabetes status, and family history to provide an assessment of a patient's 10-year risk of a cardiovascular event.
"What we are trying to do is bring that risk stratification closer to an event," she said. "To bring that long-term risk [assessment] closer to something that is really actionable."
To identify the markers, Van Eyk and her colleagues are undertaking a discovery effort looking at roughly 1,000 samples from three patient cohorts — the Framingham Heart Study, which consists primarily of subjects of European descent; the Jackson Heart Study, which includes mainly African-American subjects; and the Pakistan Risk of Myocardial Infarction Study (PROMIS), which looked at a South Asian population.
In addition to spanning different ethnic groups, the study is enriched for women, Van Eyk said, noting that cardiovascular disease has traditionally been understudied in women compared to men.
The discovery stage will be followed by validation work in roughly another 1,000 samples, these taken from the National Institute of Health's Atherosclerosis Risk in Communities (ARIC) study, which has followed roughly 16,000 subjects in four US communities beginning in 1987.
The project will also incorporate Rader's genomic research on cardiovascular risk markers, Van Eyk said.
"He has done some amazing work in genomics and GWAS where they have found some polymorphisms that really link closely to atherosclerosis," she said. "So he is a real success story of actually linking SNPs and variants to a disease phenotype."
On the proteomic side, the effort represents what Van Eyk called "something of a new paradigm in terms of how to balance discovery and validation." Traditionally, protein biomarker discovery is done in relatively small cohorts with significantly larger cohorts used for validating the markers that emerge from the discovery process.
In the Grand Challenge work, however, the researchers are using roughly equivalent numbers of samples for each stage, with the hope being that by doing such extensive discovery work, they will be able to identify a relatively narrow, high-quality group of markers for validation.
Improvements in mass spec throughput have made such large-scale discovery experiments feasible, Van Eyk said, citing the example of a biomarker discovery project she did several years ago in chronic kidney disease.
"We did 64 individuals, each fractionated into 22 fractions, and then we analyzed each of those fractions on an Orbitrap," she said. "It was like six months of running nothing but those samples."
By contrast, she said she expects running the roughly 1,000 cardiovascular discovery samples to take around a year and a half.
Van Eyk and her team will primarily use two platforms for discovery — Swath-style data-independent acquisition mass spectrometry and Somalogic's SomaScan platform, which uses SomaLogic's Slow Off-rate Modified Aptamer, or Somamer, affinity reagents to simultaneously quantify 1,129 protein analytes.
Essentially short strands of DNA, these Somamers bind to protein targets in samples of interest and can then be quantified via microarrays, with the quantity of a given Somamer corresponding to the quantity of its target protein.
Initially, Van Eyk said, she had planned to use only Swath mass spec for discovery. However, the American Heart Association asked that she and Rader combine their proteomic efforts and genomic efforts into a single project, and Rader's project had included use of the SomaScan platform.
Her lab's mass spec approach proved to be highly complementary to the SomaScan panel, Van Eyk said, noting that they consistently quantify around 1,500 proteins via their Swath-MS assays — and only around 20 percent of these 1,500 proteins overlap with the 1,129 proteins covered by the SomaScan platform.
"So we decided to use both technologies," she said.
Additionally, the researchers will be using Quanterix's Simoa platform for high-sensitivity measurement of cardiac troponin-I and, possibly, other markers where extremely high sensitivity is needed. The Simoa platform relies on arrays of reaction chambers designed to isolate single molecules, enabling each well to serve effectively as an independent assay for a single molecule. According to the company, it can measure proteins at concentrations 1,000 times lower than the best current immunoassays.
Van Eyk said she anticipates the researchers will move between 30 and 35 proteins identified in the discovery stage on to validation. Validation, she said, will be done primarily on mass spec using SISCAPA-MRM assays, which will be developed for the project by SISCAPA Assay Technologies. The validation stage will also include a number of established cardiovascular markers for which SAT already has existing SISCAPA assays. According to SAT CEO Leigh Anderson, the company has around 20 to 25 assays in its catalog that could be relevant to the Grand Challenge project.
Invented by Anderson, SISCAPA combines antibody-based peptide enrichment with mass spectrometry, increasing the sensitivity of mass spec instruments, which by themselves are often not sensitive enough to detect low-abundance proteins in complex samples.
By providing samples enriched for the peptides of interest, the technique also reduces the length of the LC gradients required prior to mass spec analysis, thereby boosting assay throughput, a key consideration in large-scale validation experiments like that being undertaken by Van Eyk and her colleagues.
She noted, as well, that the SISCAPA technology is appealing because it is amenable to clinical implementation. Indeed, SISCAPA assays have already been used in the clinic — for instance, Mayo Medical Laboratories offers mass spec-based serum thyroglobulin testing using SISCAPA technology.
And, this week, Anderson told GenomeWeb that the company was in the process of establishing a partnership that would allow the approach to be "easily and rapidly transportable into a clinical mass spectrometry platform."
Van Eyk said SISCAPA-MRM would most likely be the method of choice for the large majority of analytes moved on to validation. However, she noted, in some cases, other approaches might be necessary.
For instance, the researchers will expand their use of the Quanterix ELISA platforms if there are proteins that require extremely high-sensitivity assays. Likewise, if the SomaScan platform detected certain proteins that were best enriched using Somalogic's Somamers as opposed to antibodies, they could develop aptamer-MRM assays, using the aptamers for upfront enrichment.
SISCAPA assay development will be handled by SAT, with Van Eyk's lab responsible for developing assays to analytes that appear more amenable to different approaches.
In the end, the researchers hope to have in hand assays that improve upon current methods of gauging cardiovascular risk.
"It would be really nice if we could say you are at high cardiovascular risk for something that is going to happen to you in the next year or two years or next five years — to know whether or not an individual should be more aggressively tracked, or put on statins earlier, or moved on to routine stress tests and closer monitoring," Van Eyk said.
Cardiovascular risk has been a relatively active area of research in proteomics, with firms including BG Medicine and Aviir both pursuing similar protein biomarker tests for gauging patient risk.
These efforts have met with mixed success. Aviir went bankrupt last year, after which former employees including CEO Douglas Harrington launched a new firm, Global Discovery Biosciences, which is developing a new cardiovascular risk test named the Myocardial Assessment Profile. Aviir's test, the MIRISK test, which uses protein measurements to predict an individual's five-year heart attack risk was purchased by Cleveland HeartLab.
BG Medicine, meanwhile, was developing a proteomic test, CardioScore, for identifying individuals at high risk for near-term cardiovascular events. However, the company has in recent years shelved those development plans and narrowed its focus to its galectin-3 testing business.