Introduction
South-East Asia accounts for about 45% of the global burden of tuberculosis, and the coalescing of social factors makes the population along the Thai-Myanmar border particularly vulnerable. While a six-months administration of rifampicin-isoniazid-pyrazinamide-ethambutol is a currently available first-line treatment for pulmonary tuberculosis, the impact of some demographic factors and HIV-status on pharmacokinetics and response is uncharacterized in the region.
Aims
Here, we evaluated the pharmacokinetics of rifampicin, isoniazid, pyrazinamide and ethambutol in a population from the Thai-Myanmar border to help identify and better understand the contributors to inter-individual variability in pharmacokinetics and drug response.
Methods
Sixty-one adults with pulmonary tuberculosis who were either HIV positive (n = 24) or negative were recruited for this study. All participants were administered daily doses of 10 mg/kg rifampicin, 5 mg/kg isoniazid, 25 mg/kg pyrazinamide and 20 mg/kg ethambutol for the first 2 months. This was then followed by daily doses of rifampicin and isoniazid for 4 months. Plasma drug levels were determined in dense samples collected over a 24-h period on Day-1 and Week-6, and sparse samples collected 6-h, post-dose, on 5 other occasions between the second and sixth month. Serial sputum samples were also collected at intervals to assess drug response. Population pharmacokinetic models describing the administered drugs were developed in Pumas (version 2.5) using a non-linear mixed-effect modelling technique. In addition, the impact of HIV-status and other covariates on the time-to-bacteriological-clearance were evaluated with Cox’s proportional-hazard models.
Findings
Two-compartment structural models with transit absorption-compartments adequately described the pharmacokinetics of isoniazid and ethambutol. The concentration versus time profiles of rifampicin and pyrazinamide were best captured by a transit absorption compartment model and a one-compartment structural model. Body weight, HIV status and fat-free mass explained some of the between-subject variabilities in the population. HIV status had no significant impact on the time to bacteriological clearance in these study subjects.