Nonparametric Population Pharmacokinetic Modeling of Isoniazid in Colombian Patients with Tuberculosis

Background:Isoniazid (INH) is a first-line antituberculosis (TB) agent with a pharmacokinetic profile characterized by high interindividual variation; however, population pharmacokinetic studies in patients with TB are scarce. The aim was to develop a population model for INH in Colombian patients with TB suitable for predicting drug exposure and assessing the probability of target attainment of pharmacodynamic goals.Methods:Ten hospitalized adult patients with TB undergoing INH treatment were recruited. After an 8-hour fasting, subjects took 300 mg of INH, and 10 samples were taken from 0 to 12 hours. INH was quantified by high-performance liquid chromatography-UV, and data were analyzed with the Pmetrics R package software. A Monte Carlo simulation with the model parameters was run to determine the probability of target attainment for optimal efficacy.Results:The best model included 2 compartments, first-order absorption (K^a), delayed absorption (T^lag), and linear clearance (CL). Median T^lag was 0.25 hours, 5.54 hour^-1 for K^a, for CL, for the volume of the central compartment (V^c), 1.04 L/h for intercompartmental clearance (Q), and 788 L for the volume of the peripheral compartment (V^p). CL and V^c were allometrically scaled on basis of the normalized body weight.Conclusions:The Monte Carlo simulation indicated that 300 mg of INH per day is appropriate for Mycobacterium tuberculosis strains with minimal inhibitory concentration (MIC) up to 0.03 mg/L (target: area under the concentration-time curve/MIC >597); however, to cover strains with MIC up to 0.125 mg/L (80% of clinical isolates), a dose of 900 mg per day would be required.