Somalogic and HUNT Biosciences said this week that they are collaborating on a 2,000-patient validation study of protein biomarkers for predicting cardiovascular events including stroke, heart attacks, and heart failure.
The study will use samples provided by HUNT, which is the commercial arm of Norway's HUNT (Helseundersøkelsen i Nord-Trøndelag) Research Center. The partners will use Somalogic's aptamer-based SOMAscan proteomic platform to evaluate a panel of CVD protein markers identified in a previous 1,000-subject study of high-risk CVD patients performed by Somalogic in collaboration with the University of California, San Francisco.
Somalogic aims through the study to validate its marker panel for use with high-risk CVD patients as well as to explore the panel's usefulness in predicting CVD events in low- and normal-risk subjects, Steve Williams, the company's chief medical officer, told ProteoMonitor.
"The first [1,000-patient] cohort is to completely mimic what we've already done. 'Does it work in an independent cohort?'" he said. "And the second set is to ask, 'If we use those same markers in people who have not had a previous event, does it still work in predicting events?'"
"The study we did [with USCF] was done in people who have already had a cardiovascular event," he said. By current standards, all such patients are identified as high risk and subjected to intensive monitoring. However, Williams noted, in fact less than half of this population will suffer another CVD event.
"If you do the Framingham Risk Score [currently the standard measure of CVD risk], all of these [patients] look like very high risk, but actually most of them are never going to have another event," he said. "So if you really want to manage your medical resources, you would like to know who you really need to intensively monitor and for whom standard of care is perfectly acceptable."
For patients who have not suffered a CVD event the Framingham Risk Score is useful for identifying potentially high-risk patients, but, Williams said, it might be possible to improve on it or add to it with a protein biomarker test.
"Framingham is validated for primary prevention, and it does work, but it's far from perfect," he said.
Williams suggested that with the combined data from the initial UCSF collaboration and the upcoming HUNT work, Somalogic could have enough to launch a laboratory-developed test for CVD. He noted, however, that because Somalogic doesn't commercialize its own tests, "that is a decision we would make in collaboration with a commercial partner."
"We're talking to a few [potential partners] right now, but we haven't got anything to announce about who that might be," he said.
Williams also raised the possibility of taking a test through the US Food & Drug Administration, though he noted that such a submission would likely require additional studies.
"This is an area people have been working in for a long time, and people have gotten rather skeptical about new biomarkers," he said. "So I think the weight of evidence you would need before people got excited is quite high."
Were it to bring a test to market, Somalogic would face competition from several other protein biomarker firms – most notably BG Medicine, which in December filed a 510(k) submission with the FDA for its CardioScore cardiac risk test.
In February the company released data from a 6,600-patient study of the test – which uses measurements of seven plasma proteins to identify at-risk patients – in which it predicted patients who would go on to suffer near-term major cardiovascular events with more than double the sensitivity of conventional methods (PM 2/17/2012).
Single-molecule detection firm Singulex is also a player in the CVD protein biomarker space. In March the company announced that a study in press at the journal Clinical Chemistry suggested that its high sensitivity cardiac troponin-I assay could help identify increased risk of CVD in healthy asymptomatic subjects eight to 15 years in the future.
The company has also developed what, in an interview with ProteoMonitor last month, Singulex CEO Philippe Goix called a "comprehensive cardiovascular menu" consisting of a panel of protein biomarkers as well as lipid and metabolite analysis. The company currently runs these assays out of its CLIA laboratory and, Goix said, hopes to submit them for FDA approval by 2014.
Beyond the CVD collaboration, Williams said there might be additional opportunities for Somalogic and HUNT to work together. HUNT was launched in 2007 to offer an interface with industry players interested in using data and samples collected through the HUNT Health Study – a population-based study following more than 100,000 participants in Norway. The study has collected samples including blood, urine, DNA, and immortalized cells as well as detailed lifestyle and health exam information from these participants at three time points spanning more than 25 years.
The HUNT samples "are particularly suited to conditions that are reasonably common," such as CVD and more common cancers like lung cancer, for which Somalogic also has an ongoing protein biomarker program, Williams said. "We will probably do other things with them. I do think it's a useful set."
HUNT CEO Per Foss told ProteoMonitor that the company is quite interested in using its samples for protein biomarker projects like that with Somalogic. Thus far, he noted, the study has been used primarily for genomics work, but, he said, at present "I believe more in the proteomics approaches than the genomics."
"If you look at all the work done with GWAS studies, not much has come out of them," Foss said. "So I'm much more optimistic when it comes to" proteomic approaches.
The HUNT collection contains validated samples for a variety of disease endpoints including cardiovascular disease, cancers, and neurodegenerative diseases, Foss said, adding that the study is currently putting together a cohort from its patients that could be useful for Parkinson's disease research and has plans to put together a 1,000-patient Alzheimer's cohort as well.
The CVD study is slated to last two years, Foss said, with IP rights to any markers identified through the collaboration belonging to Somalogic.
The company will use its Somascan platform for the work, Williams said. That platform can currently measure levels of 1,001 proteins, but, he said, the next version – which the researchers hope to use for the HUNT collaboration – will measure 1,150.
While Somalogic plans to run the samples against the full Somascan platform, it identified in its discovery work a panel of 10 proteins that offer good CVD risk prediction in high-risk patients, Williams said. The company might, however, include additional markers to provide information on subspaces underlying the overarching risk levels, such as inflammation, obesity, and kidney function.
"We know we can get risk prediction with 10 proteins," Williams said. "The question we're answering now is, 'Should we include more markers because it might be useful for clinicians to know these other things as well?'"
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.