Concordance between initial GeneXpert MTB/RIF results and follow-up diagnostic testing in patients treated for multidrug-resistant tuberculosis in Lusaka, Zambia.

INTRODUCTION: diagnosis of multi-drug-resistant (MDR) tuberculosis (TB) has been revolutionized by the introduction of the GeneXpert MTB/RIF test, which is much faster and easier to use than traditional methods. Concerns have been raised about false positive results for rifampicin susceptibility, especially in cases of samples with low bacillary loads and settings with a low burden of MDR-TB. The objective of this study was to assess the concordance between initial GeneXpert MTB/RIF results and subsequent diagnostic testing for rifampicin susceptibility in patients with suspected MDR-TB in Lusaka, Zambia.

METHODS: this was a retrospective cohort study of patients who started MDR-TB treatment in Lusaka province, Zambia, from 2019 to 2021. Initial GeneXpert MTB/RIF bacillary load and rifampicin susceptibility, as well as results for any confirmatory testing on rifampicin resistance (including repeat GeneXpert MTB/RIF, phenotypic drug susceptibility testing, and line probe assay results) and patient outcomes were collected from patient files.

RESULTS: the study included 342 patients. Initiation of MDR-TB treatment was based on rifampicin resistance detected by GeneXpert MTB/RIF in all patients. Twenty-nine percent of patients had a very low bacillary load in the initial sample. Additional diagnostics for rifampicin resistance were available in 56% of patients, and revealed rifampicin susceptibility in 28% of tested patients. Rifampicin-susceptible TB was found more often if the bacillary load in the initial sample was very low (78% versus 18%; p=0.0001).

CONCLUSION: false positive results for rifampicin susceptibility on GeneXpert MTB/RIF appear to be common in Lusaka, Zambia, especially in patients with very low bacillary loads on the initial test. Repeat testing to confirm MDR-TB prior to initiation of treatment can be considered in this group.