Abstract

Pyrazinamide and isoniazid are first-line drugs for tuberculous meningitis (TBM), but limited information is available on their plasma pharmacokinetics, and particularly cerebrospinal fluid (CSF) penetration, in patients with TBM. Any potential effect of co-administration with high-dose rifampicin, also being evaluated in trials for TBM, is unknown. Understanding this is important for dose optimisation. We characterized pyrazinamide and isoniazid plasma and CSF pharmacokinetics among adults enrolled in a phase 2 clinical trial of intensified antibiotic therapy for HIV-associated TBM. Participants were randomized to receive either standard TBM treatment (including rifampicin 10 mg/kg) or high-dose rifampicin (35 mg/kg) plus linezolid, with or without aspirin. Plasma and lumbar CSF samples were collected on days 3 and 28 after study enrollment, and drug concentrations were measured using liquid chromatography-tandem mass spectrometry. Data were analysed using nonlinear mixed-effects modeling. Forty-nine participants provided 414 plasma and 44 CSF concentrations. Pyrazinamide CSF concentrations equilibrated with plasma with a half-life of 0.66 h and a pseudo-partition coefficient of 1.05. Isoniazid concentrations equilibrated with a half-life of 3.87 h and a pseudo-partition coefficient of 1.04. Pyrazinamide clearance increased by 30% from day 3 to day 28. NAT2 phenotype determined multi-modal isoniazid clearance. High-dose rifampicin did not affect pyrazinamide or isoniazid plasma pharmacokinetics or CSF penetration. Both drugs achieved exposure in CSF similar to plasma, supporting their crucial role in TBM treatment. Plasma pharmacokinetics of pyrazinamide and isoniazid in TBM were consistent with previously reported values in pulmonary tuberculosis, even when co-administered with high-dose rifampicin.