The National Cancer Institute's Early Detection Research Network plans this summer to launch a series of validation trials for lung, colon, and prostate cancer biomarkers that will compare the effectiveness of a number of industry and academic platforms.
The trials will evaluate biomarker-based diagnostic panels from firms including Somalogic, Laboratory Corporation of America, Exact Sciences, Gen-Probe, and Beckman Coulter; and from researchers at institutions including the MD Anderson Cancer Center and Johns Hopkins University.
Protein-based tests will be most heavily represented in the lung cancer study, which will include a 12-protein panel from Somalogic and is tentatively slated to include a six-protein panel from Celera. The lung cancer study will also examine a four-gene DNA methylation panel developed by Johns Hopkins researcher David Sidransky.
Celera was a confirmed participant in the study, but, as Sudhir Srivastava, head of the EDRN's Cancer Biomarkers Research Group, told ProteoMonitor last week, "the future participation of Celera in our EDRN project is currently in limbo" due to Celera's pending acquisition by Quest Diagnostics (PM 04/15/2011).
Quest announced the proposed $671 million acquisition on March 18 (GWDN 03/18/2011). Last week the deal cleared US antitrust review (GWDN 04/13/2011), and this week Celera said it had reached settlements regarding several putative class action lawsuits filed in Delaware and California related to the acquisition (GWDN 04/18/2011).
If the acquisition goes through and Celera participates in the EDRN lung cancer trial as originally planned, Quest – which has rights to Somalogic's lung cancer diagnostic – would have a stake in two of the panels included in the study.
This, noted Harvey Pass, director of the division of thoracic surgery and thoracic oncology at New York University's Langone Medical Center and head of the lung cancer trial, raises the question of how Quest might use results from the study to either select one of the two tests for commercialization or explore combining the best performing markers from each into a single panel.
"That's the whole purpose of this study," he told ProteoMonitor. "We agree that [the best test] is not going to be just one platform or one profile, and that if you integrate the best platforms together you'll get the best sensitivity and specificity.
"It doesn't make sense to have two [lung cancer diagnostics], so maybe [combining them] is the way to go," he said. "Maybe this is a sort of blinded validation to see, if one company is going to control them both, how they could be integrated."
Quest didn't respond to a request for comment on the EDRN trials. Mark Messenbaugh, Somalogic's director of corporate strategy, declined to comment on Pass's speculation, telling ProteoMonitor in an e-mail that the company "look[s] forward to continuing a great working strategic partnership with Quest."
The study will run the three platforms against roughly 300 lung cancer cases and controls collected by Pass and his NYU colleague William Rom. The cohort consists of patients with undiagnosed pulmonary nodules of less than 3 cm in diameter.
According to Pass, the Somalogic, Celera, and Johns Hopkins tests were selected for participation in the trial in part because they have used EDRN sample sets for discovery and pre-validation work.
"These [tests] have been with us through the EDRN process, so these are mature, ready to go," he said.
In a study published in the December 2010 edition of PLoS One, Somalogic's panel distinguished cases of non-small cell lung cancer from controls with 89 percent sensitivity and 83 percent specificity (PM 12/10/2010). At the American Association for Cancer Research's annual meeting this month, Celera presented data showing its test distinguished between benign and malignant lesions in a cohort of 80 cases and controls with a sensitivity and specificity of 83 percent.
If Celera doesn't participate in the study, the EDRN plans to replace its test with another proteomics-based diagnostic, Pass said, although he declined to say what tests are under consideration.
"These other candidates still need some pre-validation in our opinion, but if we can get that done, [we will] put them in the trial to go with Somalogic and [Johns Hopkins'] Sidransky," he said. "It's very dependent on what [Quest and Celera] are going to do."
The researchers hope to get the trial running within the next two months and have preliminary data to present at the EDRN Scientific Workshop in September, Pass said. Data analysis for the trials will be done by the network's Data Managing and Coordinating Center which, he said, will "do all the unblinding and get the individual data on the individual markers in the individual profiles.
"Then they'll be able to say [to the participants], 'Hey, your marker from this platform goes well with this marker from this platform and creates a superprofile'," he added. "Then the companies can do what they want with it."
Srivastava said that a key "philosophy" of the EDRN "is that if we have different platforms and we have the same sample set, we can find out if all three platforms give some common biomarkers [that] distinguish between cancer and non-cancer,".
The network also plans this summer to launch similar trials for colon and prostate cancer. The prostate cancer study will compare gene-fusion assays from Gen-Probe and Beckman Coulter. The colon cancer study will examine vimentin methylated DNA assays from LabCorp and Exact Sciences, as well as a protein biomarker assay from MD Anderson researcher Robert Bresalier.
Bresalier's test measures serum levels of an aberrantly glycosylated form of haptoglobin, which he identified as a potential colon cancer biomarker during proteomics research looking for ligands of the cancer and heart disease protein biomarker galectin-3. Since then the marker has "been through multiple steps of increasing stringency that have shown it a promising marker," he told ProteoMonitor.
The EDRN study will be "a prospective randomized trial in patients that are undergoing a colonoscopy for either screening or surveillance of colon cancer," he said. "We'll be collecting those sera and they'll be analyzed in a blinded fashion."
The assays will be run by an independent biomarker validation laboratory operated by the EDRN, and the data will be analyzed by the DMCC. According to Dean Brenner, professor of internal medicine at the University of Michigan and the trial will examine roughly 6000 stool samples, serum samples, and urine samples and take roughly four to five years to complete.
Like Pass and Srivastava, Bresalier said he expects the results will show that a combination of markers from different panels produces the best diagnostic.
"I think it's pretty unlikely that a single marker is going to be what we're looking for," he said. "It's going to have to be a combination of the markers."
As for whether LabCorp and Exact would be amenable to combining panels if results suggested it could improve their assays, he suggested that would be a matter "for future discussions."
Srivastava noted that the EDRN operates on a five-year funding cycle and hopes to complete the lung, prostate, and colon cancer trials within five years.
"Sometimes these trials have their own life and can be delayed," he said. "But our goal has always been five years."
The organization has a number of similar studies in progress, Srivastava added, but he said that they aren't yet far enough along to discuss.
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.