NEW YORK (GenomeWeb) – SomaLogic is investigating the use of its SomaScan platform to identify markers that could be useful in the diagnosis of active and latent tuberculosis.
In a paper published last month in the Journal of Clinical Microbiology, the company and its collaborators identified a series of proteins that could be used to identify patients with latent infections. It is now working on a separate study looking at whether these markers could help predict patients with latent disease who are likely to progress to an active TB infection, Urs Ochsner, principal scientist at SomaLogic and leader of the company's infectious disease work, told GenomeWeb this week.
The company is also pursuing a test for diagnosing active disease using funding from the Bill & Melinda Gates Foundation. It received an initial grant from the foundation in 2012 and another grant in 2014. Ochsner said that SomaLogic had in that work identified a protein signature based on host response markers that hits certain desired sensitivity and specificity targets. The company is currently putting together a manuscript detailing those findings, he added.
With regard to latent infection, SomaLogic hopes to improve over existing tests, which suffer from relatively poor accuracy, Ochsner said. In particular, he said, they are susceptible to false positives, especially in low-prevalence populations.
Currently, there are three main tests in use. There is the skin test, in which TB antigen is injected under a patient's skin and the patient is then monitored for swelling around the area of injection which would indicate an immune response, suggesting that the patient has previous exposure to the disease. Then there are Qiagen's QuantiFERON-TB Gold and Oxford Immunotec's T-SPOT.TB tests, both of which measure levels of interferon gamma (IFN-γ) released by patient T cells in response to exposure with TB antigens.
In the JCM study, the researchers used samples originally collected for use in the TB Epidemiologic Studies Consortium, a multicenter study funded by the Centers for Disease Control and Prevention that aimed to compare the performance of the three existing TB tests in patients at high risk of latent infection. The SomaLogic study looked at 13 subjects that had tested positive on all three tests and 26 subjects that had tested negative on all three.
Running patient supernatant samples on SomaLogic's SomaScan assay, a high-throughput platform that measures protein levels using the company's SOMAmer aptamer reagents, the researchers compared levels of roughly 4,300 proteins between the two sample sets, identifying several human proteins that appeared useful for distinguishing between patients with and without TB.
One of these was IFN-γ, which provided a confirmation of sorts of the quality of the SomaScan assay, Ochsner said, noting that they saw excellent correlation between IFN-γ levels as measured by the existing ELISA-based tests and the SomaScan platform.
This finding was particularly reassuring given that, as the authors noted, the SomaScan platform has been developed mainly for measuring proteins in serum or plasma, not the supernatant samples used in the JCM study.
The researchers also observed differential expression of ten other proteins, three of which, interleukin-2 (IL-2), monocyte chemotactic protein 2 (MCP2), and interferon gamma inducible protein-10 (IP-10), appeared the strongest markers, Ochsner said.
Beyond using these proteins to identify patients with latent disease, SomaLogic is exploring whether they could be useful in predicting patients with latent disease who are likely to progress to active disease. In the US, patients with latent disease are typically treated, and so the question of who is likely to progress is not particularly pressing. However, in developing countries, such a test could be used to help target treatment, Ochsner said.
"In some [regions] the rate of latent TB is as high as 80 percent, and there are not programs for treating everybody," he said. "So instead of treating everyone, the goal could be to identify those people who are at risk of developing active TB, which is only about 10 percent of the pool of people with latent TB."
Such a test could also be useful for selecting cohorts for clinical trials for TB vaccines, Ochsner said. One challenge in putting together such trials is enrolling enough patients who go on to develop active disease.
Because people with latent TB may convert to active infection at a rate of only around 1 percent a year, vaccine makers may need to enroll tens of thousands of subjects in their studies, Ochsner said. "So if you could enrich those trials with people at high risk using such a test, that would save a lot of money."
In June, an international consortium funded in part by the Gates Foundation published a study in The Lancet describing an RNA-based real-time PCR test for predicting progression to active TB. Ochsner noted, though, that that assay had relatively low accuracy.
Indeed, in a commentary published last month in The Lancet, researchers from the London School of Hygiene & Tropical Medicine and the Amsterdam Institute for Global Health and Development observed that the RNA test's performance would lead to a high number of false positives, though, in their reply, the original researchers noted that they had seen improved specificity in more recent cohorts.
In any case, SomaLogic's initial data suggests that proteins could provide better performance than RNA, Ochsner said, though he noted that this finding has yet to be validated. The company plans to publish on these initial findings sometime next year, he said, adding that it is also collaborating with researchers in South Africa on longer term testing of the identified markers for predicting progression to active disease.