Proteomics diagnostics firm Somalogic reported data this week supporting the utility of its blood-based 13-protein panel for the detection of malignant mesothelioma.
The study, published this week in PLoS One, found that the test, which relies on Somalogic's aptamer-based Somamer technology, could distinguish between mesothelioma cases and asbestos-exposed controls with an area under the curve of 95. The researchers evaluated the test in serum from 117 mesothelioma cases and 142 controls.
The company is now planning additional validation studies and has begun "early discussions with some partners" about licensing opportunities for the test, Rachel Ostroff, clinical research director at SomaLogic, told ProteoMonitor. She did not name any potential partners specifically, but the company has partnered in the past with Quest Diagnostics on tests for diseases including lung and pancreatic cancer.
Mesothelioma is typically caused by long-term exposure to asbestos. The disease is relatively rare, with around 3,000 new cases in the US annually. However, Ostroff noted, with more than 20 million people in the US at risk due to asbestos exposure, and millions more globally, a screening test to aid in early diagnosis of the condition could find a sizable market.
Mesothelioma "develops decades after asbestos exposure, but once it develops it is very aggressive," she said. "Most people are dead within a year or two, and [existing] interventions – surgery or chemotherapy or both – are shown to be more effective if the disease is detected earlier when the disease burden is lower."
Given that, an effective early detection test could prove valuable, and, Ostroff said, Somalogic is currently working to put together longitudinal cohorts to better study the panel's suitability for this purpose.
"Longitudinal samples taken prior to disease presentation will be our next phase of validation," Ostroff said. "The value is really in the longitudinal samples leading up to diagnosis. Can you see the biomarkers changing prior to clinical presentation? That's really the holy grail."
Finding such cohorts, however, is a challenge due to the rareness of the disease, Ostroff noted. "There are very few cohorts in the world where those types of samples have been collected and well-documented over the years."
She cited as one possibility samples collected by the University of Western Australia as part of a study on the effects of asbestos exposure due to mining. Somalogic might also be able to obtain samples through its relationship with Harvey Pass, director of the division of thoracic surgery and thoracic oncology at New York University's Langone Medical Center and leader of the PLoS One study.
In an interview with ProteoMonitor last year following the release of initial results from the study, Pass observed that he and Somalogic could potentially access a 9,000-individual asbestos-exposed cohort from the Carotene and Retinol Efficacy Trial – an 18,314-subject trial studying the effects of beta-carotene and retinyl palmitate on patients at high risk of lung cancer – through their participation in the Early Detection Research Network (PM 4/8/2011).
"Blood [from these 9,000 subjects] was measured on a yearly basis over ten years, and 49 mesotheliomas developed from that cohort," Pass said at the time. "It's a perfect uniform cohort for the next step of validation for this sort of [test] and I think the [US Food & Drug Administration] will look very favorably if the performance in that cohort is comparable to what we've seen."
Ostroff said this week that Somalogic continues to collaborate with Pass and that he could potentially help the company move forward with its validation work. Somalogic aims to do the longitudinal studies in 2013, she said, adding that it would be ideal if the company could access 50 to 100 patients with longitudinal samples leading up to diagnosis.
Because mesothelioma is rare even in asbestos-exposed populations, an effective screening test will require high specificity to avoid passing on too many false positives for further clinical procedures like CT scans, Ostroff said, noting that the high 90 percent range would be a good target for such a diagnostic.
In the recently published study, the 13-protein panel detected mesothelioma cases with a specificity of 89 percent and sensitivity of 90 percent. When the test's decision thresholds were changed to raise the specificity to 98 percent, the test had a sensitivity of 60 percent.
Sensitivity also varied with the cancer's pathological stage, with 77 percent sensitivity for stage I tumors, 93 percent for stage II, 96 percent for stage III, and 96 percent of stage IV.
"Our markers and disease algorithm reflect disease burden, which means that the signal gets stronger as the cancer progresses," Ostroff said. The panel, though, is "still quite good at detecting early disease, with a sensitivity of 88 percent for stages I and II [combined], so we think our test will be useful for early detection," she added.
Other researchers and industry players are exploring other approaches to mesothelioma diagnosis, with interest focusing on microRNA in particular. For instance, in May a group from the Concord Hospital Asbestos Diseases Research Institute in Sydney, Australia, reported that levels of the microRNA miR-625-3p were four times higher in mesothelioma patients than in healthy controls, and could distinguish the disease with 82 percent accuracy (GSN 5/10/2012).
And in 2010 researchers led by NYU's Pass reported on miR-29c* as a mesothelioma marker (GSN 2/18/2010). Rosetta Genomics also offers an miRNA-based test – its miRview Meso product – for distinguishing mesothelioma from lung cancer.
In addition to an early detection test, Somalogic has "studies underway [using the panel] to monitor disease burden in diagnosed patients undergoing treatment," Ostroff said, but hasn't yet finalized its analysis of the data from this work.
Somalogic's Somamer technology uses a modified form of aptamers – single-stranded nucleic acids capable of binding target molecules – to detect proteins. The molecules represent an alternative to antibodies for protein discovery and identification, offering a high level of specificity that allows for multiplex arrays significantly larger than are possible using traditional ELISA platforms.
The company's Somascan platform currently has Somamers to more than 1,100 proteins, with a 50-femtomolar median sensitivity and average coefficients of variation of 5 percent, Ostroff said. According to Somalogic's director of communications Fintan Steele, the company plans to expand the platform with Somamers to several hundred additional proteins in the near future and could hit 2,000 by the end of the year.