How Deeplex® Myc-TB fills precision gaps in rapid triage tests for rifampicin resistance detection
Avoiding misdiagnosis: How Deeplex® Myc-TB helped resolve false-positive rifampicin resistance in a French TB case
Rapid detection of rifampicin (RIF) resistance is the cornerstone of modern tuberculosis (TB) control, enabling early rifampicin-resistant (RR)/multidrug resistant (MDR)-TB diagnosis, appropriate treatment, and interruption of transmission. The Xpert® MTB/RIF (Ultra) assay, has considerably accelerated the detection of rifampicin-resistant tuberculosis. However, two recent publications, a French clinical case report and a large programmatic study from Bangladesh, remind us that such rapid genotypic tests may have some inherent limitations, and that more accurate molecular detection by the rapid targeted next-generation sequencing (tNGS) assay Deeplex® Myc-TB (GenoScreen, Lille), recommended by the WHO, can have a crucial impact in securing treatment decisions.
In a case study published in the journal Mycobacteria, a team of microbiologists and infectious disease specialists from the Hospices Civils de Lyon (France) describe the management of a young patient hospitalized for smear-positive pulmonary tuberculosis. On day 4, the Xpert® MTB/RIF Ultra assay, followed by a second rapid test (MDR/MTB ELITe MGB®), both detected rifampicin resistance, indicating the need for an intensive MDR-TB treatment regimen. Indeed, at the time of diagnosis (end of 2024), the 6-month short-course treatment with bedaquiline, pretomanid, linezolid and moxifloxacin was not available in France.
For confirming this indication, the team chose to rapidly apply targeted next-generation sequencing on a patient’s sputum sample using Deeplex® Myc-TB, capable of directly identifying resistance mutations to 15 anti-tuberculosis drugs in a single test. The result, available within just a few days, identified the P454L mutation in the rpoB gene, which is located just outside the rifampicin resistance-determining region (RRDR) normally targeted by Xpert assays. This mutation is classified by the WHO as of uncertain significance (group 3, no clear evidence of association with resistance). No resistance mutation was detected in any of the targets of the 14 other anti-TB drugs.
This information proved decisive: in consultation between clinicians and microbiologists, the patient was treated as a drug-susceptible tuberculosis case, with a bridging strategy including levofloxacin while awaiting phenotypic confirmation. Eighteen days later, whole genome sequencing performed after culturing confirmed the P454L mutation in the rpoB gene in isolation and the absence of resistance mutation in any other gene targets. By day 31, phenotypic drug susceptibility testing using MGIT and determination of the minimal inhibitory concentration (0.125 mg/L) using broth microdilution confirmed full susceptibility to rifampicin. Treatment was simplified by discontinuating levofloxacin, and the patient was cured within six months, with no relapse observed after three months’ follow-up.
This case of false-positive rifampicin resistance detected by indirect, real-time PCR tests highlights the clinical value of culture-free direct sequencing with the Deeplex® Myc-TB assay: by precisely identifying the nature and gene position of the mutation, the use of this approach prevented an unnecessary prolonged and toxic MDR-TB treatment, while ensuring a multidisciplinary evidence-based decision.
Prevalence of false-positive rifampicin-resistant Xpert MTB/RIF results in a large programmatically representative population in Bangladesh, detected by Deeplex Myc-TB
The second study published in the Journal of Global Antimicrobial Resistance by a team from the International Centre for Diarrhoeal Disease Research in Dakha (Bangladesh) documents the high prevalence of false-positive rifampicin resistance in patients with samples with low or very low bacillary burdens tested by Xpert MTB/RIF. The study included 1531 patients identified as rifampicin-resistant by Xpert MTB/RIF across 10 programmatic drug-resistant TB management sites in the country. Fresh sputum samples were collected for re-testing with Xpert, phenotypic drug susceptibility testing, a line probe assay, and tNGS with Deeplex® Myc-TB to confirm rifampicin susceptibility status.
Of the 1531 cases initially classified as rifampicin resistant, only 38.7% remained so on repeat Xpert testing, while 21.4% tested susceptible and 38.5% no longer showed detectable M. tuberculosis. The discrepancies were markedly more frequent in paucibacillary samples, reaching discordance in “very low” (94.6%) and “low” (62.3%) bacterial load samples.
Among the cases re-classified as rifampicin susceptible, targeted sequencing showed a 96% concordance with the repeat Xpert results, comparable to that of the reference phenotypic drug susceptibility test (98%) and the line probe assay (97.4%). Even further importantly, when compared against a composite reference standard combining several methods, Deeplex® Myc-TB achieved the best performance of all the techniques tested, with 99.3% agreement and a kappa coefficient of 0.91, indicating near-perfect agreement, higher than that of the phenotypic drug susceptibility test (kappa = 0.74) and far higher than the line probe assay (kappa = 0.43). Deeplex® Myc-TB also detected some rare mutations not covered by the line probe assay, providing a direct explanation for some discordances across methods.
The authors concluded that that field studies are required to determine the effectiveness of the new Xpert MTB/RIF Ultra assay to overcome this limitation of the former version and that confirmation with additional WHO-endorsed rapid molecular tests, such as Deeplex® Myc-TB is strongly recommended.
The added value of Deeplex® Myc-TB
In summary, these two studies, one at individual patient level, the other at the scale of a national program, show that, despite their undisputed value for rapid initial detection and triage, indirect, probe-based PCR tests are exposed to some inherent limitations and risk of false-positive resistance.
Targeted next-generation sequencing with Deeplex® Myc-TB can overcome these limitations through:
- Accurate molecular resolution: by using direct sequencing and comprehensively covering the known catalogue of resistance mutations in the rpoB gene (and 14 other resistance genes), Deeplex® Myc-TB can precisely identify and distinguish them from mutations not associated with resistance or of uncertain significance.
- High concordance with reference methods: both studies showed perfect or near-perfect agreement of Deeplex® Myc-TB results compared to phenotypic testing or molecular testing (WGS or a line probe assay), separately or considered in combination as a composite reference standard.
- Turnaround time compatible with clinical decision-making, with direct impact on patient management: In the French setting, where Deeplex® Myc-TB was used prospectively, the result was available within a few days, allowing the treatment to be adjusted (well) before the obtaining of WGS or phenotypic drug susceptibility testing results. A needless prolonged intensive treatment for multidrug-resistant tuberculosis was avoided, representing a tangible benefit for the patient.
Sources
- Genestet C, Eberl I, Joannard B, Dumitrescu O, Hodille E. “False detection of rifampicin resistance using Xpert MTB/RIF Ultra due to a P454L mutation on rpoB in Mycobacterium tuberculosis: a case report.” Mycobacteria, 2026;2:9.
- Uddin MKM, Nasrin R, Ferdous J, et al. “Discordant rifampicin resistance results in repeat Xpert MTB/RIF testing: Risk factors and diagnostic challenges among drug-resistant tuberculosis patients in Bangladesh.” Journal of Global Antimicrobial Resistance, 2026;50:364-370.