NEW YORK (GenomeWeb) – A team led by researchers from Sanofi, Hamilton Health Sciences, and McMaster University have identified a series of protein biomarkers that could help identify diabetes patients at increased risk of cardiovascular events and death.
Using serum samples taken from 8,401 participants in the Outcome Reduction with Initial Glargine Intervention (ORIGIN) trial, the researchers identified 10 markers linked to an increased risk of myocardial infarction, stroke, or death and an additional 5 markers that, when added to the original 10, further improved prediction of death.
They also identified a panel of nine markers that slightly improved their ability to predict progression to a composite outcome that, in addition to myocardial infarction, stroke, and death, also included heart failure hospitalization or revascularization.
Detailed in a paper published last month in the journal Circulation, the study used a modified version of Myriad RBM's DiscoveryMAP assay panel to measure levels of 237 proteins in patient blood.
Sanofi launched the seven-year, 12,500-participant ORIGIN trial to study its Lantus (insulin glargine) agent, comparing use of the drug in patients with impaired fasting glucose, impaired glucose tolerance, or early type 2 diabetes versus the standard of care for these patients. In addition to looking at the efficacy of the drug, the study investigated whether its use changed the relative risk for cardiovascular disease and cardiovascular mortality compared to standard of care, finding that it did not.
Upon completion of the trial in 2012, Sanofi and its collaborators at the HHSMU Population Health Research Institute, who led the trial, then turned their focus to investigating possible CVD biomarkers using samples from a subset of patients who had given permission for such additional research.
"When we were recruiting patients, we were able to get more than 8,000 who consented to have their blood stored for future analyses," said HHSMU researcher Hertzel Gerstein, principal investigator on the ORIGIN study and first author on the Circulation study. "And once [the study] was over, Sanofi was very interested, as were we, in supporting future studies to advance knowledge of cardiovascular disease in people with diabetes, and so they supported this study, which was designed to discover new biomarkers."
The size of the ORIGIN cohort provided the researchers with "a really rich dataset with … a lot of patients who had a lot of different health outcomes," Gerstein told GenomeWeb. Of the 8,401 patients studied, 1,405 progressed to myocardial infarction or stroke, 2,435 progressed to either myocardial infarction, stroke, heart failure hospitalization, or revascularization, and 1,340 died.
Each of the 15 identified markers contributed to improving prediction of patient health outcomes independent both of clinical measurements like sex and age and of the other markers. "So, any one of them alone is not as good as all of them together," Gerstein noted.
That said, three markers in particular — NT-proBNP, angiopoietin 2, and gluthathione S — stood out as being identified as predictive of all outcomes and across all models and, the authors noted, "may have the most pathophysiological and prognostic relevance for serious outcomes." Of these three markers, two, NT-proBNP and angiopoietin 2, have been previously identified as "possible mediators" of cardiovascular disease.
Gluthathione S, on the other hand, appeared to have a protective effect as it was inversely associated with the three measured outcomes.
While the other markers were less strongly predictive of outcomes, they, too, provide potential avenues for further investigation, Gerstein said. "We have 15 individual markers that are independently related to death [from CV events], many of which were not known to be related to death beforehand, and obviously it raises lots of questions."
He cited the example of apolipoprotein B, which was predictive of death and cardiovascular events independent of traditional clinical measures including LDL/HDL ratio.
"So, if you are a researcher interested in lipids, that is really interesting from a lipids perspective," he said. "I think this raises a lot of questions for research, and one could spend a career going through each one of these if one wanted to. It will speak to people who are actually already focused on some of these biomarkers in other contexts."
Gerstein noted that, while the researchers performed extensive internal validation of the markers, additional external validation is needed. He said that his role as deputy director of the Population Health Research Institute could help him and his colleagues access large sample sets with which to do such validation.
"We run large international studies and clinical trials all over the world and currently we have lots of samples from these study sets, some of which might be amenable to testing some of these hypotheses," he said.
In addition, the researchers would like to investigate how predictive the markers are for other health outcomes like renal disease, Gerstein said.
Currently, the researchers are working to add genetic analyses to their protein data, he said, noting that they recently secured funding to support this work.
"We're in the process of doing those analyses right now, so we hope to have some interesting findings," he said.
"If we know what a person's genetic signals are because of the genetic variants that are being assessed, and we know their biomarker levels, we can, for instance, do Mendelian randomization analyses to relate these things to outcomes down the line," Gerstein said, citing an example of the sort of questions he and his colleagues are pursuing.
"We can ask how the biomarker-gene-phenotype combination predicts future phenotypes," he said. "So it's much more powerful than just knowing the biomarkers alone."
The data and markers identified by the study are in the public domain, Gerstein said, though, he noted that certain of the proteins may already be under patent. For instance, Roche currently has a patented NT-proBNP on the market for aiding in the management and diagnosis of cardiovascular conditions including heart failure.