Discordant rifampicin resistance results in repeat Xpert MTB/RIF testing: Risk factors and diagnostic challenges...

The Xpert MTB/RIF (Xpert) assay can detect Mycobacterium tuberculosis complex and rifampicin resistance (RIF-R) with high sensitivity. However, it can produce false RIF-R results, particularly in samples with low and very low bacillary loads, leading to inappropriate treatment.

This study investigated discrepancies in RIF-R results detected by Xpert. Between July 2022 and September 2023, a total of 1531 Xpert-detected RIF-R TB patients were enrolled from 10 programmatic management of drug-resistant TB sites in Bangladesh. Fresh sputum samples were collected for re-testing with Xpert, phenotypic drug susceptibility testing, line probe assay, and targeted next-generation sequencing to confirm rifampicin susceptibility status.

Upon repeat Xpert testing, out of 1531 initially diagnosed RIF-R cases, 593 (38.7%) were RIF-R, 327 (21.4%) were RIF-S, 21 (1.4%) were RIF-indeterminate, and 590 (38.5%) were MTB not detected. The RIF-S (false RIF-R) results were most common in cases with low (20.0%) and very low (27.3%) bacterial burden samples. In RIF-S cases, phenotypic drug susceptibility testing, line probe assay, and targeted next-generation sequencing showed concordance rates of 98.0%, 97.4%, and 96.0%, respectively, compared to repeat Xpert results.

Newly diagnosed patients and those with low or very low bacillary burden were found to be more likely to report false RIF-R. The diagnostic algorithm requires review for cases diagnosed with RIF-R, particularly with low bacterial burden or no previous TB history. RIF-R status should be confirmed with repeat testing or additional DST methods, and the newer Xpert Ultra, which is more specific for determining RIF-R status, should be considered.