Skip to main content
Premium Trial:

Request an Annual Quote

New Cepheid TB Assay Expected to Benefit Patients With Hard-to-Detect Illness


NEW YORK (GenomeWeb) – A Mycobacterium tuberculosis assay launched by Cepheid last week is likely to increase the number of pediatric, extrapulmonary, and HIV co-infected TB cases detected. The new test, which runs on Cepheid's GeneXpert platform, has increased sensitivity compared to the firm's original MTB/RIF test, leading the World Health Organization to draft new guidelines for its use.

The MTB/RIF Ultra assay has been in development by Cepheid and collaborators at Rutgers New Jersey Medical School and the Foundation for Innovative New Diagnostics for a few years, and has been anxiously anticipated by the TB community.

Cepheid's standard MTB/RIF assay, which was endorsed by WHO in 2010, is widely used around the world and offered at concessional pricing in certain countries. As of the end of 2016, a total of 6,659 GeneXpert instruments and more than 23 million Xpert MTB/RIF cartridges were procured in the public sector in 130 of the 145 eligible countries, with 6.9 million cartridges procured in 2016 alone, according to WHO.

Yet the test does not always meet the needs of patients who have low levels of bacteria, or paucibacillary TB.

For example, in pediatric populations, a group that is challenging to test due to low bacterial load and an inability to produce sputum samples, the standard MTB/RIF test sensitivity has been shown in a meta-analysis to vary across studies and specimen types.

The original test has also been shown to perform poorly in patients with TB infection in body compartments other than the lungs, also due to low levels of bacteria. Patients with HIV co-infection are more likely to have extrapulmonary TB, and they can represent a significant proportion of infected people in certain countries.

Furthermore, a 2015 study in Swaziland — a country where more than 80 percent of people with TB also have HIV co-infection and 26 percent of the adult population is HIV infected — showed that the standard MTB/RIF test failed to detect a circulating strain that was infecting up to 30 percent of patients there because the culprit strain carried a mutation in the rpoB gene that was not detected by Cepheid's test.

However, the latter problem may now be avoided with the new test because MTB/RIF Ultra has two additional targets.

"We added two conserved multicopy targets that are unique to tuberculosis," said David Persing, Cepheid's chief medical officer, in an interview. These targets, called IS6110 and IS1081, are present in the bacteria in approximately five and 35 copies per cell, as opposed to the single copy per cell of rpoB. The additional copies available for detection thus enables the assay's higher sensitivity.

Initial data on the Ultra test, reported by FIND at the Conference on Retroviruses and Opportunistic Infections earlier this year, showed a 17 percent increase in detection sensitivity in HIV-positive patients whose infections were missed by microscopy, compared to the standard MTB/RIF test.

Another feature enabling the test is an adaptation of the GeneXpert platform to perform automated melt curve analysis. Persing noted that this change did not require any modification of the platform's hardware. "The existing GeneXperts on the ground doing the current test can run [the Ultra] because it is the software that does the magic," he said.

Specifically, the software was updated to be able to read melt peaks automatically within certain temperature windows, so that it is now able to call out individual mutations occurring throughout the amplified region. "It's sort of the next best thing to having a sequence analysis," Persing said.

This melt technology can also now be used in any test Cepheid is developing. For example, another TB assay in the firm's pipeline, called XDR-TB, is for extensively drug-resistant tuberculosis. It is pioneering the firm's 10-color testing abilities, but with the new melt feature Cepheid can now detect up to five targets in each color channel, or a total of up to 50 unique targets.

The Ultra test is also potentially faster than the standard test. It splits detection of the rifampin resistance-conferring mutations from the TB detection conferred by the multicopy genes. "Even though both are TB-specific, we can build in something called early assay termination," Persing said. Because the multicopy targets are more sensitive that the rpoB target, if a sample is negative for those two targets, the test is negative, he said. Therefore, the TB detection element of the assay only takes about one hour, Persing said, which in turn creates more capacity for testing.

A representative at the company also confirned that the Ultra MTB/RIF test will have the same cost as the standard MTB/RIF test.

Karin Weyer, WHO's coordinator of laboratories, diagnostics, and drug resistance, noted in an interview that the Ultra test's level of sensitivity can potentially lead to false positives. The organization had "extensive discussions" on the relative harms and benefits of possible overtreatment, as documented in the WHO technical expert group report describing a non-inferiority analysis of Xpert MTB/RIF Ultra compared to Xpert MTB/RIF.

The Ultra test adds a new category of detection, Weyer explained. The standard MTB/RIF test grades positivity into four categories, from high to low, but the new test has a fifth category at the low end, called a "trace call," she said.

"The additional sensitivity is almost exclusively linked to the trace call [category], because it picks up low numbers of MTB," Weyer said, which can include detection of dead bacilli and ones circulating in patients that recently completed treatment and are cured of disease.

The WHO experts determined the risk of false positives to be less than five percent, Weyer said, so the WHO guidelines now being drafted will attempt to guide countries on in which patients a trace call can be considered a true positive.

As such, a "more nuanced diagnostic algorithm" will likely recommend the test for all patients with signs or symptoms of TB, cautioning that a positive result in an HIV-negative patient or in one without symptoms might be a false positive and recommending additional testing be done.

Overall, Weyer said she thinks this test is going to be an improvement over the current MTB/RIF, "particularly in population groups where we have trouble with diagnosis of TB because of low number of bacteria." The benefits of rapid diagnosis will likely be most pronounced in high-burden HIV settings, but pulmonary TB in children is also "seriously underdiagnosed" globally, she said, and obtaining a TB diagnosis can lead to treatment and cure.

Weyer further noted that Cepheid's Omni system will work with the MTB/RIF Ultra cartridge, allowing diagnosis of TB and rifampin resistance to be done almost at the point of care. "Its a one-module, robust machine that works with a battery and can be used in a doctor's consulting room or areas without electrical supply," she said. "When combined with the new Ultra test ... it would bring the diagnosis much closer to the patient, which is exactly what we want."

Cepheid is actively developing the Omni system, which was delayed in 2016, and plans to release it in the second half of 2017. "It's in final stages of engineering and validation right now," Persing said.