Nov. 13, 2007
| ‘When we start to think about the biology of disease, our goal is not just new targets anymore but biomarkers also,” said Andrew Plump, executive director of Merck’s Cardiovascular Disease franchise. Plump was speaking last month at the Personalized Medicine meeting held at George Washington University’s Richard B. and Lynne V. Cheney Cardiovascular Institute, in Washington D.C.
The meeting was focused on “The Genomic Revolution in Cardiac Care,” and highlighted how biomarker research is now not just reaching across a wider range of conditions, but also using a broader mix of tools. Even more intriguing, omic tools are increasingly being used together to uncover valuable biomarkers to guide those now “must have” proof of concept studies bridging pre-clinical and clinical trials.
Not long ago that was deemed too expensive, too complicated, and unnecessary.
Finding biomarkers is particularly challenging in fields such as atherosclerosis, where biopsies are not a routine part of patient monitoring. It’s hard to figure out what’s happening to all that nasty arterial plaque the medications are supposed to whisk away.
Merck has a leg up because it has been a pioneer in omic analysis, dating back to a high profile acquisition of subsidiary Rosetta Inpharmatics in 2001. “We bought Rosetta in part for their real capacity in system analysis,” Plump said.
In trying to uncover blood-based biomarkers of atherosclerosis progression, Merck researchers are examining SNPs, histology, proteins, and RNA. They are not just looking for genetic causes of cardiovascular disease, but for markers of “inflammatory instability” as Plump said. “Right now, there are no tools for decision making in phase 2,” said Plump. “In the future, imaging tools will emerge, but in the meantime we are looking for markers in plaque.”
Not all pharmaceutical companies have the internal omics capability to find such markers by themselves, however, and that’s created opportunity for a new wave of technology vendors and service providers with specialized expertise or novel tools.
Digilab Biovision falls in the latter category. Offering a novel “peptidomics” service, the company’s tools uncover peptide and small-protein based biomarkers. Because they are signaling molecules, peptides are believed to be a more specific read out than proteins or RNA, says Hans-Dieter Zucht, chief technology officer. And what have we learned about analyzing omics so far? “The sequence does not speak to you,” warns Zucht. Experiments must be appropriately designed, and researchers need to get enough samples. In some cases, Zucht says, his company and their collaborators are aiming for as many as 10,000 samples to power a single study.
After sifting through all that data, everyone hopes to find just one golden marker in the end, but that’s not always possible. “Especially when you are looking at markers for a syndrome, such as inflammation,” Zucht says, “you’re probably going to need more than one marker.”
Proteome Sciences’ business development director Ian Pike adds another bit of hard-earned wisdom. “You have to find the markers and identify the proteins,” he says. Researchers working in proteomics have slowly shifted away from using random sets, or constellations, of unknown markers, to wanting to know exactly what proteins make up the marker. Proteome Sciences has developed novel tags and other methods that make it easier to quantify proteins and measure them in complex samples.
Few of these tool providers can parade big pharma customers who have used their tools, the field is too secretive. But Jack Reynolds, formerly a senior vice president in R&D at Pfizer, speaks generally about Genstruct, which has a platform that is both novel and specialized. Genstruct generates and analyzes data from metabonomics, proteomics, protein phosphorylation, and gene expression.
“People didn’t get that much out of genomics originally,” Reynolds says. “Because you have to create a baseline platform and generate knowledge from it, like Genstruct has done.” Reynolds says this platform, which Genstruct CEO Keith Elliston says has been used in about 30 partnerships so far, helped Pfizer become one of the first companies to submit to the FDA under the voluntary genomics data submission guidance.
“In one area of adverse events we submitted over a million data points,” Reynolds says. “People would ask how can you get something out of that much data, but that’s exactly why you need something like the Genstruct platform.”
Clearly, tangible progress has been made, but pioneers like Merck, Digilab Biovision, Genstruct, and Proteome Sciences are all still working to gain further understanding of the how these data types interplay and what types of conclusions can, and cannot, be drawn from them.
Subscribe to Bio-IT World magazine.