The potential of a physiologically-based pharmacokinetic (PBPK) model to predict oral amoxicillin bioavailability, by considering the physiological changes after "Roux-en-Y gastric bypass" (RYGB) surgery in bariatric patients, was evaluated. A middle-out approach for parameter estimations was undertaken using in vitro, in situ, and in vivo data. The observed versus predicted plasma concentrations and the model sensitivity of the simulated parameters of AUC0-inf and Cmax of amoxicillin (AMX) were used to confirm the reliability of the estimation. The model considers that a drug-transporter (Transp) in the initial segments of the normal intestine plays a significant role in the AMX absorption. A lower fraction absorbed (Fabs) was observed in RYGB patients (54.43% for suspension and 45.21% for tablets) compared to healthy subjects (77.48% capsule). Furthermore, the tablet formulation presented a lower dissolved fraction (Fd) and Fabs compared to the suspension formulation of AMX in RYGB patients (91.70% and 45.21% versus 99.92% and 54.43%, respectively). The AUC0-inf and Cmax were sensitive to changes in Rtintestine, PeffAMX, and Transp for both healthy and RYGB models. Additionally, AUC0-inf and Cmax were also sensitive to changes in the tlag parameter for tablet formulation in RYGB patients. The PBPK model showed a reduction in AMX bioavailability as a consequence of reduced intestinal length after RYGB surgery. Additionally, the difference in the predicted Fd and Fabs between suspension and tablet suggests that liquid formulations are preferable in postbariatric patients.
Keywords: absorption; amoxicillin; gastric bypass; physiologically based pharmacokinetic model.