Results of an evaluation of the performance of Xpert® MTB/RIF Ultra (Ultra) are now available in the Lancet ID. Ultra is a second-generation, molecular tuberculosis (TB) test (Cepheid, Sunnyvale, USA) that has shown increased sensitivity over its predecessor. The assessment of Ultra’s performance was conducted in 2016 in a multi-centre study at ten sites in eight low- and middle-income countries, coordinated by FIND, and conducted by FIND and the NIAID-sponsored Tuberculosis Clinical Diagnostics Research Consortium.
TB was responsible for 1.7 million deaths in 2016 and caused TB disease in 10.4 million people. An estimated 600,000 TB patients were also resistant to rifampicin, a critical first-line TB drug. However, less than 70% of new TB cases and only 22% of drug-resistant cases were diagnosed or treated in the same year.1 These diagnostic gaps present a major challenge for controlling TB transmission and reaching the targets of the World Health Organization (WHO) End TB Strategy.
The first generation test, Xpert MTB/RIF (Xpert), introduced in 2010, offers rapid diagnosis of TB and rifampicin resistance. As of 2016, more than 23 million Xpert tests have been procured in 130 countries and multidrug-resistant TB (MDR-TB) detection has more than tripled.2,3 However, the sensitivity of Xpert and its detection of rifampicin resistance is imperfect. Ultra was designed to overcome these limitations.
The study found that Ultra sensitivity was up to 17% higher than Xpert. The greatest sensitivity gains were seen among patients that would have been missed by microscopy (smear-negative, culture-positive patients) and HIV-positive TB patients. The high sensitivity of Ultra could enable the diagnosis of TB at earlier stages of disease and the diagnosis of TB in patients with HIV who would otherwise be missed. The results are promising for other hard-to-diagnose populations such as children and those with extrapulmonary forms of TB, which were assessed in other studies.
As a result of its increased sensitivity, the use of Ultra also resulted in lower specificity (i.e. a higher rate of false positives) compared to Xpert. This may be due to the detection of non-viable TB bacteria that are present in people with a recent history of TB.
Earlier this year, WHO issued a recommendation that Ultra can be used as an alternative to the existing test for the diagnosis of TB and detection of rifampicin resistance in all settings.4 The recommendation noted the need for further discussion and study to understand the implications of the decreased specificity of Ultra as compared to Xpert.
The development and evaluation of Xpert® MTB/RIF Ultra (Ultra) were conducted with the support of UK aid from the UK government, the governments of Australia and the Netherlands, as well as the Bill & Melinda Gates Foundation.