Abstract
PURPOSE: This study aimed to resolve inconsistencies in prothionamide (PTO) drug susceptibility testing (DST) results between solid and liquid media by validating the appropriate critical concentrations for PTO DST in solid media. It also aimed to assess the correlation between phenotypic resistance and genetic mutations in ethA and the inhA promoter. METHODS: A total of 100 multidrug-resistant tuberculosis (MDR-TB) isolates were tested for PTO susceptibility using the MGIT 960 system, Löwenstein-Jensen (L-J) medium, and Middlebrook 7H10 medium with varying PTO concentrations. Genotypic resistance was determined by PCR and sequencing of ethA and the inhA promoter. RESULTS: When MGIT 960 was used as the reference, 20 mg/L for L-J and 10 mg/L for 7H10 achieved the highest agreement (kappa = 0.865 and 0.827, respectively). Using Sanger sequencing as the reference, the highest kappa (0.594) was observed for MGIT 960 when defining resistance as inhA c-15t and/or loss-of-function mutations in ethA. At 20 mg/L, L-J medium provided a balanced sensitivity (71.4%) and specificity (84.6%), whereas 40 mg/L increased specificity but markedly reduced sensitivity. Six isolates showed suboptimal growth on 7H10, suggesting it may not support robust culture of MDR strains. Wild-type inhA and ethA profiles demonstrated high negative predictive value (NPV 90.2–100.0%), reliably excluding resistance. CONCLUSIONS: A critical concentration of 40 mg/L for PTO in L-J medium may underestimate resistance. A lower concentration of 20 mg/L offers improved detection but requires genotypic confirmation owing to lower PPV. The 7H10 medium may not be optimal for MDR-TB DST due to limited growth support. Integrating phenotypic and molecular methods is recommended for accurate resistance detection. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12866-025-04388-4.