T. S. Samant,1 V. Lukacova,2 L. J. Lesko,1 S. Schmidt1; 1University of Florida, Orlando, FL, 2Simulations Plus, Inc., Lancaster, CA

BACKGROUND: PBPK modeling and simulation approaches have gained popularity for pediatric drug development as they account for changes in body composition and metabolic capacity from birth. Our objective was to study the effects of phase I enzymes and whole body ontogeny on the clearance (CL) of drugs.
METHODS: Specific substrates for a particular phase I metabolic pathway selected were CYP1A2: caffeine, theophylline; CYP3A4: midazolam; CYP2C9: warfarin; CYP2C19: omeprazole; CYP2D6: desipramine. In vitro Km and Vmax values for each drug-enzyme pair were scaled to in vivo values in adults using GastroPlusTM (Simulations Plus, Inc.). Once developed and externally qualified, adult PBPK models were expanded for application in children by accounting for pathway-specific enzyme ontogeny and other physiological changes from birth. Overall, CL was simulated for each substrate-enzyme combination from neonates to adults. Changes in simulated CL were overlaid with observed literature CL values for qualification of pediatric model.
RESULTS: PBPK model framework was successfully developed and qualified for five major phase I metabolic pathways using in vitro enzyme kinetics and literature clinical pharmacokinetics in children. Our model effectively predicted drug exposures for the enzyme-specific probe compounds as predicted AUC and CL were within two-fold error of the respective observations for adults and pediatrics. Slowest bodyweight normalized CL was in neonates, fastest at 2.5-5 years and lower adult values were reached by 10-15 years.
CONCLUSION: Our findings indicate that changes in CL from neonates to adults follow a non-monotonic function. These changes in ontogeny, captured better by PBPK modeling than by conventional allometry, can lead to effective dose estimation in pediatrics.