BACKGROUND: Pharmacokinetic variability arising from drug-drug interactions and pharmacogenetics may influence the effectiveness of treatment regimens for TB. The Improving Treatment Success Trial compared the WHO-recommended standard treatment in TB patients with an experimental regimen substituting ethambutol with moxifloxacin (MFX) in Durban, South Africa.METHODS: Non-linear mixed-effects modelling was used to investigate the population pharmacokinetics of rifampicin (RIF), isoniazid (INH) and pyrazinamide (PZA). A total of 25 single-nucleotide polymorphisms, including pregnane-X-receptor, were selected for analysis.RESULTS: TB drug concentrations were available in a subset of 101 patients: 58 in the MFX arm and 43 in the control arm. Baseline characteristics were well-balanced between study arms: median age and weight were respectively 36 years and 57.7 kg; 75.2% of the patients were living with HIV. Although weight-based drug dosing was the same in the two arms, we found that RIF exposure was increased by 19.3%, INH decreased by 19% and PZA decreased by 19.2% when administered as part of the MFX-containing regimen. Genetic variation in pregnane-X-receptor (rs2472677) was associated with a 25.3% reduction in RIF exposure.CONCLUSION: Optimised weight-based TB treatment dosing is essential when RIF, INH and PZA are co-administered with fluoroquinolones. The reduction in RIF exposure associated with pharmacogenetic variation is worrying.