Physiologically based pharmacokinetic modelling of levofloxacin as a tool for paediatric dose optimization.

AIMS: Levofloxacin (LFX) is critical for paediatric community-acquired pneumonia (CAP) and multidrug-resistant tuberculosis (MDR-TB). Current guideline doses may provide suboptimal exposure in younger children. This study aims to develop a paediatric physiologically based pharmacokinetic (PBPK) model to evaluate current guidelines and propose optimized oral regimens. METHODS: An adult PBPK model was developed and verified against 21 adult clinical studies. The model was scaled to virtual paediatric populations (0.5-16 years old) using age-dependent physiology and verified against paediatric data. Simulations assessed probability of target attainment (PTA ≥ 90%) for CAP (fAUCss,0-24/MIC ≥ 33.7) and MDR-TB (fAUCss,0-24/MIC ≥ 100). RESULTS: The adult model was verified (test dataset geometric fold error (GMFE): 1.04; mean relative deviation [MRD]: 1.33). The scaled paediatric model accurately predicted observed concentrations (86% within 90% prediction interval). For CAP, guideline doses (10 mg/kg QD) were inadequate in children >5 years (PTA: 46.7%-63%); optimized regimens (e.g., 14 mg/kg QD for 5-10 years) achieved PTA ≥ 90.4%. For MDR-TB, WHO-recommended doses (~15-20 mg/kg/day) caused severe underexposure in children <24 kg (PTA: 26%-63%). Optimized regimens (16-33 mg/kg/day) achieved PTA > 90%. CONCLUSIONS: Current paediatric LFX guidelines are suboptimal for CAP in children >5 years of age and critically inadequate for MDR-TB targets in children <24 kg. This PBPK model provides rational, age-specific oral dosing strategies to improve target attainment and support dose refinement for vulnerable paediatric populations.