Objective: Isoniazid (INH) remains a cornerstone drug in tuberculosis (TB) therapy, but its clinical response and hepatotoxicity vary widely among patients due to differences in metabolic pathways. This study aimed to (1) develop and validate a robust...
Objective: Isoniazid (INH) remains a cornerstone drug in tuberculosis (TB) therapy, but its clinical response and hepatotoxicity vary widely among patients due to differences in metabolic pathways. This study aimed to (1) develop and validate a robust LC–MS/MS method for simultaneous quantification of INH and its four major metabolites—acetylisoniazid (AcINH), isonicotinic acid (INA), hydrazine (HZ), and acetylhydrazine (AcHZ)—in human plasma; (2) characterize exposure distributions, variability, and metabolic ratios of these analytes in a large cohort of Korean TB patients.
Methods: A dual LC–MS/MS analytical strategy was developed to quantify INH and its metabolites in plasma. INH, AcINH, and INA were analyzed after protein precipitation, while HZ and AcHZ were derivatized with p-tolualdehyde to form stable hydrazones. The method was validated following FDA bioanalytical criteria. The validated assay was applied to 824 plasma samples from 776 Korean TB patients obtained from the multicenter cPMTb cohort. Concentrations, normalized to an equivalent 300 mg INH dose, were analyzed by time-after-dose windows (centered at 2 ± 0.5 h and 3 ± 0.5 h). Pathway-based molar ratios were calculated to represent acetylation and hydrolysis pathways.
Results: The LC–MS/MS method demonstrated excellent linearity for all five analytes over clinically relevant ranges, with accuracy and precision within regulatory limits and robust performance. Across 0–24 h, median (range) concentrations (µM) were: INH 10.6 (0.29–61.6), AcINH 13.8 (0.22–51.6), INA 7.5 (0.32–39.1), HZ 0.3 (0.02–1.46), AcHZ 4.8 (0.26–15.7). INH and AcINH dominated early after dosing, whereas INA and AcHZ showed more delayed and sustained presence and HZ remained low but measurable throughout the 24-hour period.
At approximately 2 h post-dose, the majority of rapid acetylators had INH concentrations below the conventional therapeutic range (3–6 µg/mL), whereas slow acetylators more often fell within or above this window; by 3 h, INH levels exceeding the suggested toxicity cut-off of 3.69 µg/mL were observed predominantly in slow acetylators. HZ concentrations were generally higher in slow acetylators, and AcHZ showed a phenotype-specific temporal pattern, being relatively lower in slow acetylators during the early post-dose period but relatively higher at later times, consistent with delayed accumulation.
Acetylation-linked indices (AcINH/INH and (AcINH+AcHZ)/(INH+HZ)) increased with greater NAT2 acetylation capacity, clearly differentiating slow, intermediate, and rapid phenotypes. The hydrolysis-linked INA/(INH+AcINH) ratio also increased from slow to rapid acetylators, supporting AcINH hydrolysis as a dominant contributor to INA formation. In contrast, hydrazine-focused ratios HZ/INH and (HZ+AcHZ)/(INH+AcINH) showed no significant differences between phenotypes, suggesting that hydrazine generation itself is less dependent on NAT2 activity. Notably, the AcHZ/AcINH ratio was substantially higher in slow acetylators, indicating reduced secondary acetylation of AcHZ to non-toxic diacetylhydrazine and a greater relative propensity for AcHZ accumulation in this group.
Conclusion: This thesis establishes a validated LC–MS/MS platform and provides the first large-scale dataset describing INH and its major metabolites in Korean TB patients. The concentration distributions and ratio profiles characterize real-world variability in INH metabolism. This work established a quantitative foundation for future pharmacogenetic and model-informed dosing studies.