By Tony Fong
Using a comparatively new technology that has been largely untouched by proteomics researchers, Somalogic has developed an array-based method that company officials said could move protein biomarkers closer to use in the clinic.
Somalogic's technology, called Somamers, is now at the stage where the company is poised to launch diagnostics using the technology. In addition, the company in recent months has built up its management team, hiring Steven Williams as chief medical officer, Nebojsa Janjic as chief science officer, and Nicholas Saccomano as chief technology officer.
Based on aptamers as binding molecules, Somamers form the basis for a lung cancer diagnostic that Somalogic hopes to launch in 2011 as a laboratory-developed test. It also plans to seek US Food and Drug Administration for a launch possibly the year after.
The firm, based in Boulder, Colo., also is developing tests for mesothelioma and pancreatic cancer.
At a session on personalized medicine at the recent Association of Biomolecular Resource Facilities conference in Sacramento, Calif., Williams presented the Somamer technology as the potential impetus for greater FDA approval of protein and proteomic biomarkers.
Aptamers, discovered only about 20 years ago, are oligonucleic acid or peptide molecules that bind to other specific molecules. Somalogic's Somamer technology comprises chemically modified nucleic acid aptamers that are created using a process called systematic evolution of ligands by exponential enrichment, or SELEX. The process, developed by Larry Gold, now the company's CEO, selects RNA ligands against T4 DNA polymerase.
While Somamers are "undoubtedly aptamers," the additional chemical modification they undergo results in both high affinity and high specificity, which allows Somalogic to use its multiplex assay for biomarker discovery and then use the biomarkers for diagnostic purposes, Saccomano told ProteoMonitor this week.
The Somamers, he said, are third-generation aptamers. First-generation aptamers were unmodified single strands of nucleic acids, while second-generation aptamers were "chemically modified on the sugar backbone," he said.
Those aptamers were the first to show therapeutic promise, and Somamers build on that clinical utility, Saccomano said.
While Somamers share common attributes of DNA aptamers, such as enzymatic processes, they "allow entry into a whole … physical, chemical space that previous aptamers could not access," he said. "In so doing, they allow them to have new binding modalities to proteins, which are advantageous both from a therapeutic point of view as well as an affinity reagent."
The Somamers are formatted on an array for measuring large numbers of proteins at once. Typically, array-based technologies for protein discovery and identification have used antibodies. According to Saccomano, the Somamers offer several advantages over antibody technology. Foremost is that using the technology, Somalogic researchers have been able to look at the changes in the levels of proteins in up to 1,000 proteins simultaneously.
By comparison, antibody-based technology allows for the quantification of about 30 proteins maximum in one sample, Saccomano said. In a 1,000-protein experiment, that would translate to a 30-fold increase in sample volume using an antibody array compared to a Somamer array, while the expense would be four times greater, he said.
The multiplexing ability of Somamers is due to its specificity: "That's the specificity of binding to its protein target exclusively and its ability to avoid non-specific binding, or unproductive or misinformative binding events," Saccomano said.
In addition, because Somamers are synthetically developed and purified they are not prone to inconsistent performance as antibodies are because they are produced using biological systems.
"You will find folks that are using ELISAs [who] are quite concerned as to whether the cell line mutates or dies, or whether the group population has shifted in some way that their new antibodies are not the same chemical entities as their previous antibodies," Saccomano said.
[ pagebreak ]
A third advantage, he added, is the cost savings that can be achieved in the biomarker discovery-to-assay-development pipeline. That typically can cost as much as $250,000 and take between 18 and 24 months. Using its Somamers, Somalogic can do the same for about $5,000 and in as little as 20 weeks.
"What we do is we make 1015 molecules, we do SELEX, we find the clone, we characterize the clone, and we synthesize it to our specifications. And in a very short period of time after that, it's released as an analyte in an affinity reagent for biomarker discovery and diagnostic purposes," Saccomano said.
Along with antibody quality, another issue with antibody-based technologies is that for the majority of proteins, no antibody exists. Because aptamers are synthetic, though, more than 80 percent of the time, Somalogics will be able to manufacture a high-quality aptamer that can be used as an analyte-specific reagent, Saccoman said.
Williams said that if a client were to present Somalogics with 200 proteins, the company would be able to come up with Somamers against 160 proteins within three months.
"And you couldn't possibly do that with antibodies," he said.
As the company adds to its library of Somamers, the goal is to achieve 100 percent success in developing Somamers against all proteins, Saccomano said. So far, the firm has raised Somamers against approximately 1,000 proteins.
In 2008 researchers in Canada described a mass spec-based method called AptaBid for biomarker discovery using aptamers. That method selects DNA aptamers to biomarkers that are differentially expressed on cell surfaces in different states, and uses mass spectrometry to isolate and identify them.
While Williams said he was unfamiliar with that technology, he said that in general, mass spectrometry is based on fractionating the proteins in the front end. In this case, this would be done by the aptamers. Such a process, though, sacrifices breadth of coverage.
"If you use mass spec without affinity reagents, you can look at a large number of proteins but the sensitivity is very poor" — about 1,000-fold less than with Somamer or ELISA's, Williams said. "You can increase that sensitivity by using various manipulations of the cell before you do the mass spec, but you sacrifice the breadth, so you end up looking at a small fraction of the proteome.
"We don't think there's anything else [aside from the Somamer technology] that can do breadth and sensitivity," he said.
The main drawback of the Somamers is that they can be less sensitive than "a well-designed and well-engineered" sandwich ELISA assay run against a single analyte. According to Saccomano, though, the sensitivity on an ELISA is "no more than two- to four-fold."
A Focus on Lung Cancer
While the Somamer technology may have therapeutic use, he said that for now it is most applicable for diagnostics development.
Somalogic's lead product is a panel of 12 proteins still in development for the early detection of lung cancer, the leading cause of cancer deaths in the US. In 2009, an estimated 159,390 men and women in the US died from the disease and 219,440 new cases were diagnosed, according to the National Cancer Institute.
There are no generally accepted screening methods for lung cancer, however, though chest X-rays and CT scans are being investigated for their efficacy. It was that gap in medical technology as well as market opportunity that prompted Somalogic to direct is first commercial product to lung cancer, whereas other companies developing protein and proteomics-based diagnostics have focused on ovarian, breast, and prostate cancers.
[ pagebreak ]
According to Williams, about only one-third of patients with lung cancer are diagnosed in stage 1, when the cure rate is 90 percent. "So that's a great reason" for developing a lung cancer diagnostic, he said. "The medical need is the greatest and the diagnostic can make the greatest difference."
Beginning with 1,400 serum samples, Somalogic ran its 850-plex assay against 985 of the samples, with the balance used as a blinded set, to find proteins indicative of lung cancer.
Initially, Somalogic detected 61 candidate protein biomarkers, but only a handful of them in combination with each other showed any promise as biomarkers for the company's target population, stage 1 patients and people with poor lung function, such as long-time, heavy smokers, Williams said.
Somalogic used four different geographic centers as training sites. Accounting for different collection methods that may have affected the results, the company whittled the 61 potential biomarkers to 44, "which we liked better," Williams said.
It then tested those biomarkers in different combinations to determine which ones had the best performance as measured by area under the ROC curve in stage 1 disease and in poor lung function, and came up with its panel of 12 proteins.
At ABRF, Williams said the panel has an accuracy of better than 90 percent area under the ROC curve. This week, he said that "there are many alternatives" to the dozen biomarkers it now has, and in continuing work, Somalogic is investigating new markers and new combinations of them.
Using samples from the NCI's Early Detection Research Network, Somalogic expects to develop a final panel in August, at which time the company will start designing additional clinical studies in preparation of filing for FDA clearance, Williams said.
"Sometime next year" it will bring the test to market as a laboratory-developed test with its commercialization partner, Quest Diagnostics, he added.
The company is also "quite keen on getting into" the European market and "it's possible we might find the right partner," but there is no real strategy to bringing the test into that market currently, he said.
In addition to lung cancer, Somalogic is developing tests for pancreatic cancer and mesothelioma. The pancreatic cancer panel consists of 11 proteins, while the mesothelioma panel has "six or less" proteins.
In both instances the company plans to run the proteins against a smaller plex, rather than the 850-plex, assay "so that we're only measuring the proteins of interest," Williams said. However, because there were "many fewer" signaling proteins to start with in those two indications than with lung cancer, "our task may already be complete," and the company "may be pretty close to a final panel" for pancreatic cancer and mesothelioma, Williams said.
A target launch date for those two panels has not yet been determined, though. Quest has the option on partnering with Somalogic on any diagnostics developed from those panels.
While it does not offer services as a formal business, the firm has forged alliances with pharma, biotechs, and contract research organizations "who are interested in using the proteomic platform that we have in the context of various phases of clinical trials," Saccomano said.
"There is a lot more work to be done here, but people who have thought about biomarkers for a long time … recognize that finding validated biomarkers either for stratifying and understanding disease or for stratifying and designing clinical trials is a huge asset and all pharma I have seen are extremely interested … in using the platform for that," Saccomano said.