The effect of metformin on the pharmacokinetics of rifampicin, isoniazid, and pyrazinamide in adults with tuberculosis

Metformin is under investigation as adjunctive host-directed therapy for tuberculosis (TB), which might interact with first-line TB treatment. We used population pharmacokinetic modeling to assess whether metformin alters first-line TB-drug pharmacokinetics in HIV/TB-coinfected adults without diabetes. Rifampicin, isoniazid, and pyrazinamide pharmacokinetics were investigated in participants from a randomized clinical trial of adjunctive metformin (500 mg twice daily to week 12) in adults starting HIV-associated TB treatment. Antiretroviral therapy (ART)-naïve participants initiated dolutegravir-based ART within 8 weeks. Intensive and semi-intensive sampling was conducted at week 5; concentration–time data were analyzed using non-linear mixed-effects modeling. Data from 78 individuals (43 receiving metformin, median weight 60.8 kg, 62.8% male, 79.5% on ART) were analyzed. Rifampicin and pyrazinamide were described by one-compartment models with linear elimination; typical clearances were 15.8 L/h (95% CI: 13.5–18.7) and 3.88 L/h (95% CI: 3.54–4.08), respectively. Isoniazid followed a two-compartment model with a mixture model for acetylator status; clearance was 10.5 L/h (95% CI: 9.44–11.9) in slow acetylators and 28.8 L/h (95% CI: 25.6–31.0) in fast/intermediate acetylators. Metformin reduced isoniazid bioavailability by 15.6% (95% CI: 4.45–26.4%, P < 0.009) and rifampicin bioavailability by 24.0% (95% CI: 7.83–36.3%, P < 0.007), decreasing the area under the curve from 0 to 24 h from 18.2 to 15.6 mg·h/L and from 35.9 to 27.1 mg·h/L, respectively. No significant effect on pyrazinamide was detected. We found that non-diabetic patients on metformin had lower isoniazid and rifampicin bioavailability. Lowered rifampicin exposure might be clinically relevant; simulations suggest that this could be offset by a single 150 mg rifampicin dose.