NEW YORK (GenomeWeb) – A group led by researchers from the Francis Crick Institute in London and the University of Leicester has developed a host gene signature that distinguishes patients with active tuberculosis from individuals with latent TB, viral, or bacterial infections.
The team has partnered with BioMérieux and French research institute Bioaster to develop a commercial diagnostic assay on BioMérieux's BioFire FilmArray platform based on the 20-gene biosignature.
Clinicians typically perform a whole blood test called an interferon-gamma release assay (IGRA) on patients they suspect might have tuberculosis. The assay diagnoses latent Mycobacterium tuberculosis infection by measuring the patient's immune reactivity to the pathogen. White blood cells in the body that have been infected with M. tuberculosis will release interferon-gamma when mixed with TB antigen.
While current IGRAs can detect if the patient has been exposed to TB, they can't reveal if a patient is currently healthy or predict who will progress to active TB disease. In addition, the assays can be expensive and require up to a day to determine if the patient has TB.
In a proof-of-concept study published earlier this week in Nature Communications, senior author and group leader at the Francis Crick Institute Anne O'Garra and her team used RNA sequencing to validate their previous microarray findings published in December 2010 in Nature that revealed a 393-transcript signature for active TB in blood collected from patients across London and Africa.
The researchers also evaluated TB gene signatures developed by other groups against TB and other lung infections that mimic TB, and found that these signatures unintentionally detected viral infections due to their inclusion of general interferon-inducible genes.
Afterward, the team developed and tested a modular TB signature in multiple TB cohorts and other diseases. Applying computational methods developed by O'Garra's team, they used the global genetic signatures to develop a reduced TB-specific gene signature comprising 20 key genes. The team ignored genes that signaled a more general response to infection, testing the panel against other gene sets with latent TB infections, influenza, and other viral and bacterial infections.
The team then evaluated the blood transcriptional signature in individuals who had been in contact with TB patients and went to develop TB, and in TB contacts who remained healthy. They used the new gene signature to analyze a new set of 53 patients in Leicester, UK, as well as 108 of the patients' close contacts over two years to see if the signature could predict who developed active TB.
"[The UK] is a low burden country, which has big pockets of TB, but not obviously on the same level as African countries," O'Garra explained. "But [it] has a lot of advantages, as we don't need to deal with reinfection issues."
The team noticed that the signature did not endure in the patient contacts who remained in good health, while six of the nine patients who eventually developed active TB demonstrated a sustained, consistent gene signature for the disease.
"This suggests that we will be able to see a sign of TB from this blood signature before individuals become sick and develop visible symptoms," O'Garra said. "It's important because the signature could tell you to treat people earlier, and that could then stop people from transmitting the disease to each other."
According to O'Garra, the new gene signature has both a clinical sensitivity and specificity of 100 percent, based on unpublished data. She noted that the team was pleasantly surprised to find out that the 20-gene panel did not detect viral infections or other similar conditions, such as lung cancer or pneumonia.
O'Garra emphasized that her team has "developed a novel approach that can determine active TB, separating it from … other viral infections, [and] that we can use it to detect a signature of early infection preceding any symptoms in context of TB patients."
O'Garra and her team are currently working with BioMérieux to improve the assay's automation and speed. BioMérieux Chief Medical Officer Mark Miller explained that the firm initially partnered with O'Garra's team in 2012 because it saw potential in developing a diagnostic tool based the group's original 393-transcript signature.
Back then, however, it "was clear that no company was going to create an assay based on 390-plus genes," Miller said. "So, we aimed to reduce the gene signature so that it was amenable to one of our diagnostic platforms."
BioMérieux acquired BioFire in 2014, along with its FilmArray platform that performs multiplex PCR for infectious disease diagnosis with panels in the range of 20 to 40 genes. According to Miller, BioMérieux plans to eventually place the TB gene signature assay on the FilmArray platform, which will potentially expedite time to results due to its automation and relative ease of use.
"This will take RNA sequencing — a difficult, laborious, manual method, where you need trained technologists and competence — to a simplified, diagnostic test," Miller noted. "However, we are not at the automated stage yet, but we are down to the stage where the process only needs a certain number of genes."
Miller explained that BioMérieux plans to commercialize the test if it can demonstrate in larger patient cohorts that the assay can both detect TB in blood and predict active TB before symptoms occur. "That would be the place we would have the biggest impact on TB globally," he said.
A test like this "would be positioned to diagnose active TB and differentiate from latent TB symptoms … as right now we don't have any tools to detect the disease at the active [stage] except for direct biopsies and direct cultures," Miller noted. "A blood-based test to direct those cases would be non-invasive and very useful."
While certain regions may have different needs for blood-based TB testing, Miller pointed out that BioMérieux has established a strong presence in countries, particularly low-income regions, that would directly benefit from the new assay.
"TB is unfortunately one of those global diseases that exists prevalently in places that people can't afford diagnostic testing and therapeutics in order to control, and in those countries that have the resources to test and control TB, the incidence is very low," Miller explained. "There has to be some careful deployment so that [the assay] can serve a population that needs it, and not just areas where people are able to afford it."
Depending on the region, BioMérieux envisions the assay being used in public health clinics, hospitals, nursing homes, and potentially refugee settlements and other crisis centers for other individuals.
According to Miller and O'Garra, the assay will potentially address at least a few unmet clinical needs. Besides diagnosing active TB, it may help monitor treatment response, which is currently very difficult and not possible in many patients. The assay could also eventually quickly determine whether a patient has developed multiple drug resistance to TB or other issues. , O'Garra also aims to use the test for early detection to help clinicians decide whether they need to treat patients proactively or leave them alone.
Miller also noted that the team will need to address whether one biosignature will serve all cases of TB, or if the group will need to develop additional signatures.
In addition, BioMérieux aims to search for other applications or diseases they could tackle using the same approach. Miller noted that published literature has suggested that researchers can explore the host's genome for gene signatures that may help distinguish bacterial versus viral infections and detect sepsis earlier in patients.
"These tests might give additional host information [where] we normally run up against a dead end when looking at the pathogen side," Miller explained. "If the BioFire platform can adequately measure, quantitate, and deliver highly complex signatures using [between] 20 to 40 genes, then there are many other uses that we can explore with gene signatures."