BackgroundThe Xpert MTB/RIF (Xpert) assay has been widely used to diagnose suspected active tuberculosis (TB) and rifampicin-resistant TB cases. Despite its excellent performance record, false-positive Xpert rifampicin (RIF) resistance results are obtained for specimens with extremely low bacterial loads. ObjectiveWe aimed to study the feasibility of repeat Xpert testing as a strategy for reducing the odds of obtaining false-positive results when testing paucibacillary TB patients. MethodsWe enrolled previously tested TB patients with very low initial bacterial loads from May 2016 to February 2022 for Xpert retesting. A total of 251 TB patients were retested using the Xpert assay. ResultsRIF resistance was noted in 65 (25.9 %) patients when tested by Xpert at initial diagnosis. Only 107 (42.6 %) of 251 patients tested positive for MTB when retested via Xpert. The majority (98.6 %) of RIF-susceptible cases were still susceptible to RIF when retested. Initial Xpert testing yielded 35 positive results for MTB in the RIF-resistant group, of whom 25 (71.4 %) still exhibited RIF resistance when retested. All culture-positive MTB isolates in the RIF-susceptible group were also RIF-susceptible by phenotypic DST. In the RIF-resistant group, 10 of 14 culture-positive MTB isolates exhibited RIF resistance, of which 4 isolates were deemed RIF-susceptible by phenotypic DST. The proportion of double mutations within the MTB rpoB RRDR sequence, as detected by hybridization of Xpert D and E probes, was significantly higher in the RIF-susceptible group than in the RIF-susceptible group. ConclusionsOur results demonstrated that initial RIF-susceptible results were more accurate than RIF-resistant results. Additionally, patients with double mutations that delayed probe D/E hybridization were more likely to have false-positive Xpert results. Our findings emphasize that repeat Xpert MTB/RIF testing is necessary for TB patients with extremely low bacterial loads who are at high risk for RIF-resistant TB.
Read full abstract