Abstract:

Objective: To elucidate the mutation profiles of fluoroquinolones (FQs) resistance genes in Mycobacterium tuberculosis (MTB) and to analyze the correlation between specific genetic mutations and the minimum inhibitory concentrations (MICs) of FQs. Methods: Positive strains isolated and cultured from rifampicin-resistant tuberculosis (RR-TB) patients, who were admitted to tuberculosis prevention and control institutions and designated hospitals in Beijing between 2016 and 2021, were selected for drug sensitivity testing using the microplate method. The MICs of levofloxacin (Lfx) and moxifloxacin (Mfx) for these RR-MTB strains were determined and summarized. Concurrently, first-generation sequencing was performed. The relationship between the mutation characteristics of the gyrA and gyrB genes, which were associated with FQs resistance, and the MICs of FQs was analyzed. Phenotypic drug sensitivity testing (pDST) results were used as a reference standard to evaluate the performance of genotypic drug sensitivity testing (gDST) in detecting FQs resistance. Additionally, the reasons for discrepancies between phenotypic and genotypic resistance in RR-MTB strains were investigated. Results: Among the 303 RR-MTB strains analyzed, the pDST resistance rate to FQs was 27.7% (84/303), and the detection rate of gyrA gene mutations was 25.1% (76/303). No gyrB gene mutations were detected. Using pDST results as the reference standard, the sensitivity and specificity of gDST for detecting FQ resistance in RR-MTB were 84.5% (71/84; 95%CI: 74.6%-91.2%) and 97.7% (214/219; 95%CI: 94.5%-99.1%), respectively. There were 25 strains with inconsistent results between the two methods, resulting in an inconsistency rate of 8.3% (25/303). The MIC of FQs-resistant strains was predominantly 2 μg/ml, with the most common mutation site located at codon 94 (53.9%, 41/76). The gyrA gene mutations at codons 88 and 94 were completely consistent with pDST resistance, while the consistency rates of pDST resistance at codons 90 and 91 were 95.8% (23/24) and 3/5, respectively. The mutation at position 88 was linked to resistance to Lfx in pDST and high-level resistance to Mfx. The mutation at codon 90, predominantly an alanine-to-valine substitution (92.3%, 24/26), resulted in MICs at the critical concentrations for Lfx (1 μg/ml) and Mfx (0.25 μg/ml). The aspartic acid mutation at codon 94 was associated with high-level resistance to Lfx and Mfx when mutated to asparagine (1/1). Additionally, this aspartic acid mutation, when altered to tyrosine, was linked to resistance to Lfx (1/1) and high-level resistance to Mfx (1/1). Conclusion: The primary mechanism of FQs resistance in RR-MTB in Beijing is attributed to mutations in the gyrA gene. Different gyrA mutations are indicative of varying levels of FQs resistance. The concordance between pDST and gDST for detecting FQs resistance in RR-MTB is high. Consequently, gDST can be implemented early to identify FQs resistance in RR-TB patients, facilitating the clinical development of effective treatment plans.

Key words: Mycobacterium tuberculosis, Fluoroquinolones, Microbial sensitivity tests, DNA mutational analysis, Drug resistance, bacterial