NEW YORK (GenomeWeb) – A new study from Cepheid and several international collaborators demonstrates how a new version of Cepheid's Xpert MTB/RIF assay, Xpert MTB/RIF Ultra, addresses multiple issues of the previous assay, improving tuberculosis detection and providing conclusive detection of rifampin susceptibility and resistance.
In the first peer-reviewed study on Xpert Ultra, published in mBio yesterday, the researchers tested the Ultra and Xpert's limits of detection, dynamic ranges, and RIF-R rpob mutation detection on Mycobacterium tuberculosis DNA, or sputum samples spiked with known numbers of M. tuberculosis H37Rv or Mycobacterium bovis BCG CFU.
The original Xpert MTB/RIF assay was endorsed by the World Health Organization in December 2010. Since then, 23 million Xpert tests have been purchased in 130 countries, leading to more than eight times the amount of testing for MDR TB worldwide.
While Xpert has successfully diagnosed TB in developing countries in the past seven years, the device has been noted to have a number of limitations.
Despite excellent sensitivity in tests of smear-positive sputum samples, Xpert is less sensitive when testing smear-negative sputum. According to the study, when Xpert was performed following negative smear microscopy results, its pooled sensitivity was 67 percent. In contrast, when Ultra was performed following negative smear microscopy results, its pooled sensitivity was 79 percent. However, when comparing the overall specificity in testing clinical sputum samples, both assays performed the same at 98.7 percent.
In an email, co-author Claudia Denkinger of the Foundation for Innovative New Diagnostics (FIND) explained that the Ultra assay distinguishes itself from its predecessor because of its sensitivity in paucibacillary samples. The mBio study reported that, for tuberculosis case detection among smear-negative culture-positive participants, the assay's sensitivity improved by 13 percent. Denkinger was also principal investigator on a larger version of this study that showed a 17 percent improvement. That data helped guide WHO in its decision earlier this year to recommend that the Ultra assay replace the original assay and be administered in countries to all adults and children with signs and symptoms of TB.
In the study published yesterday, the performance of Ultra on clinical sputum samples was assessed retrospectively with only 275 samples, based on an earlier multicenter diagnostic accuracy study done by FIND and the Tuberculosis Clinical Diagnostics Research Consortium to evaluate the new assay's performance in 10 sites and eight developing countries.
More than 1,500 patients at these locations with signs/symptoms of pulmonary TB were ultimately enrolled in the study. Researchers performed Xpert and Ultra on the same samples and examined their ability to detect TB, comparing them to research standard methods for TB detection. The prospective multicenter study is expected to be submitted for publication later this year.
"In certain populations, especially those with substantial paucibacillary diseases (through HIV co-infection in Sub-Saharan Africa or in children) the assay could substantially improve outcomes of patients through improved and earlier detection," Denkinger said.
In addition, Xpert's assay sensitivity has also been limited in certain types of extrapulmonary samples, which are known to contain lower levels of bacilli than pulmonary examples. Xpert has generally performed very well as rapid test for rifampicin resistance, with a pooled sensitivity and specificity of 94 percent and 98 percent respectively. However, the assay has a limited capacity to detect RIF-R-associated mutations in mixed samples. In some reports it even has a decreased capacity to detect rpoB C533G mutations responsible for certain cases of RIF-R.
In response to Xpert's drawbacks, Cepheid and collaborators developed the Ultra assay to help improve point-of-care detection of tuberculosis. The inclusion of two new PCR assays that target two different multicopy genes, conversion of the rpoB and IS6110 assays into fully nested PCRs, and use of a larger PCR tube that doubles the amount of sample DNA have all allowed for an almost tenfold increase in analytical sensitivity for M. tuberculosis H37Rv.
Testing Ultra's capacity to detect a wide variety of M. tuberculosis strains with varied numbers or no copies of IS6110, researchers found that it detected the disease's presence in all four replicates of each of the 25 DNA samples. Sensitivity more than doubled when detecting BCG, which contains only one IS6110 copy, suggesting that additional sensitivity gained by Ultra is expected to be true for different clinical M. tuberculosis strains containing both high and low numbers of iS6110 copies. The Ultra assay's enhanced sensitivity also allowed the researchers to identify more cases of smear-negative and culture-positive TB.
Because the manual steps required to perform Xpert and Ultra are identical and can be run in identical machines after a software upgrade, the researchers noted that "Ultra can be implemented with little additional training in sites that already use the current Xpert assay." The Ultra is CE-marked and ready to deployed to developing countries, as work is ongoing to assess extended stability of the assay to facilitate longer storage. Each Ultra cartridge will be available at the same concessional price of $9.99, the WHO noted in March.
The study's results suggest that, besides improving detection rates in paucibacillary cases, Ultra will lead to higher TB case detection in pediatric patients and patients with extrapulmonary TB. Denkinger also noted that FIND has supported a study on TB meningitis using the Ultra assay, which she believes will substantially benefit from the Ultra assay's sensitivity to achieve a rapid, accurate diagnosis. Data from that study is forthcoming.
That is not to say the Ultra assay does not come without risks or issues. The study noted that the increase in the assay's sensitivity is offset by a decrease in specificity. The drop in specificity could possibly predispose to false-positive results due to sample cross contamination, especially in laboratories that produce multiple samples from TB suspects and M. tuberculosis. Ultra may also be prone to detecting smaller number of non-replicating bacteria present, especially in patients with a recent history of TB treatment, reducing the assay's specificity and leading to further false positive results for TB.
"The implementation of the assay has to be carefully planned as it also has a slightly reduced specificity, which in certain populations can lead to overdiagnosis," Denkinger said.
Overall, the researchers believe that assays with increased sensitivity such as Ultra may "be instrumental in identifying the remaining cases of TB," and are "likely to prove valuable in further WHO goals to eradicate the disease."