NEW YORK – A team led by researchers at the Broad Institute and Harvard University has developed a protein-based test for detection of active tuberculosis.
Described in a paper published this week in Science Translational Medicine, the test is based on the single-molecular array (Simoa) immunoassay technology developed by Harvard professor and study author David Walt and commercialized by Quanterix, which Walt cofounded.
In addition to presenting an initial version of the tuberculosis assay, the STM study points toward an application of a point-of-care Simoa technology that Walt is working on.
The Simoa technology uses femtoliter-sized reaction chambers in which single molecules can be isolated and detected, allowing for digital immunoassays offering roughly a thousandfold better sensitivity than a conventional ELISA.
Quanterix has commercialized the technology as a research instrument, but in the STM paper, Walt and his coauthors raise the possibility of implementing Simoa as a point-of-care device, which could expand the technology's reach and range of applications.
Walt said that his laboratory is currently working "on developing the technology that would allow us to port the Simoa technology onto something that is either sort of a benchtop device that could be used to go from sample to answer or even potentially a handheld kind of system."
The challenge, he said, was packaging the technology, including sample prep and data readout, into a system that does not require either a number of manual steps or complex instrumentation.
"The existing Quanterix instruments have a sophisticated imager, very sophisticated robotics that do all the sample handling and processing and washing steps," Walt said. "It's not right now in a format that could be converted into a point-of-care type system."
He said that his lab has various pieces of a point-of-care Simoa workflow working.
"We have the ability to do certain steps of the process," he said, "We don't, at this point, have a fully integrated system … because we are really trying to work on optimizing and simplifying each of the different components of the process."
Walt is a member of Quanterix's board of directors but said that his lab's point-of-care Simoa work is entirely independent of the company, though the company would have rights to the technology through its exclusive license covering Simoa.
He suggested that a point-of-care device would likely be of more interest to a diagnostic company than it would Quanterix, given that the latter firm hasn't entered the diagnostic space. It has indicated that it is interested in doing so, however. Last year Quanterix ended its license with BioMérieux covering commercialization of Simoa for in vitro diagnostic purposes, a move that was widely seen as the company taking back control of its IVD program.
In an email, Kevin Hrusovsky, CEO and president of Quanterix said that the company was "working to capitalize on a broad range of markets," including point of care, though he did not comment on the company's interest in the Walt lab's point-of-care work specifically. He also cited the recent launch of the company's benchtop SP-X and SR-X instruments as "two early examples of our efforts to miniaturize our platforms."
Walt suggested that the tuberculosis work presented in the STM study, though done with a conventional Quanterix instrument, was an example of an application that would benefit from the combination of high sensitivity and portability a point-of-care Simoa device could provide.
Improving tuberculosis control in resource-constrained areas requires a simple and accessible assay for detecting the disease and distinguishing between active and latent cases. Existing sputum-based tests have low sensitivity, particularly in children and adults infected with HIV. Nucleic acid-based tests require more sophisticated laboratories than are typically available in the areas in question, as do culture-based tests.
Given these issues, Walt and his colleagues set out to determine if a panel of host response proteins could effectively identify patients with active tuberculosis.
To develop the test they collected 387 plasma samples, 199 from patients with active tuberculosis and 188 from patients with conditions that are not tuberculosis but that are similarly characterized by a sustained cough. Using a Luminex system, the researchers measured 47 host proteins in these samples, then used machine learning to develop a four protein panel ( IL-6, IL-8, IL-18, and VEGF) that was able to distinguish between the two patient groups with sensitivity of 86 percent and specificity of 65 percent, a level of performance that the authors noted approaches the target of 90 percent sensitivity and 70 percent specificity set by the World Health Organization as the minimum for a rule-out tuberculosis triage test.
They then developed Simoa assays to the four proteins and validated the panel in an independent cohort of 317 samples representing patients from Vietnam, South Africa, and Peru, finding that the assay performed with 80 percent sensitivity and 65 percent specificity. Adding an antibody to the tuberculosis antigen Ag85B boosted the performance of the Simoa assay to a sensitivity of 86 percent and specificity of 69 percent.
While the test's performance does not quite meet the WHO benchmark, Walt said that he was encouraged that it proved effective in samples from several different geographies, suggesting that it could be broadly applicable.
He added that including information in the test algorithm on where a patient was from and their HIV status could further boost its performance.
Walt said that he and his colleagues planned to do additional discovery work to identify proteins that could improve the test's accuracy, noting that he hoped to use the Quanterix instrument for this discovery work, as opposed to starting with the Luminex platform as the researchers did initially.
He said he also planned to continue his work on simplifying the Simoa assay in order to package it in a system that could be used in resource-constrained areas, including, potentially, as a point-of-care device.