Isoniazid resistance-conferring mutations are associated with highly variable phenotypic resistance.

ABSTRACT

Background: High-dose isoniazid is recommended in the 9-12 months short-course regimen for multidrug-resistant tuberculosis with inhA mutation. However, there is insufficient evidence to support the assumption of genotypic-phenotypic concordance. This study aimed to identify the genetic mutations associated with high-level phenotypic isoniazid resistance. Methods: Clinical isolates from patients with drug-resistant tuberculosis were profiled by whole-genome sequencing and subjected to minimum inhibitory concentration (MIC) testing using MGIT based-method. MICs were performed in concentration ranges based on the mutation present isolates with no isoniazid resistance-conferring mutations and H37Rv, 0.016-0.256 µg/ml; inhA, 0.256-4.0 µg/ml, katG 1.0-16.0 µg/ml; and inhA + katG, 4.0-64.0 µg/ml. Isolates demonstrating resistance at the upper limit of the concentration range were tested up to the maximum of 64.0 µg/ml. Bootstrap of the mean MICs was performed to increase the robustness of the estimates and an overlap index was used to compare the distributions of the MICs for each mutation profile. Results: A total of 52 clinical isolates were included in this analysis. Bootstrap MIC means for inhA, katG and inhA + katG were 33.64 (95% CI, 9.47, 56.90), 6.79 (4.45, 9.70) and 52.34 (42.750, 61.66) µg/ml, respectively. There was high overlap between inhA and inhA + katG mutations (eta = 0.45) but not with inhA and katG (eta = 0.19). Furthermore, katG showed poor overlap with inhA + katG mutations (eta = 0.09). Unexpectedly, 4/8 (50.0%) of all InhA mutants demonstrated high-level resistance, while 20/24 (83.3%) of katG mutants demonstrated moderate-level resistance. Conclusions: InhA mutations demonstrated unexpectedly high MICs and showed high overlap with inhA + katG. Contrary to the common belief that katG mutants are associated with high-level resistance, this mutation primarily showed moderate-level resistance.