Incorporating Ontogeny in Physiologically Based Pharmacokinetic Modeling to Improve Pediatric Drug Development: What We Know About Developmental Changes in Membrane Transporters

Abstract

Developmental changes in the biological processes involved in the disposition of drugs, such as membrane transporter expression and activity, may alter the drug exposure and clearance in pediatric patients. Physiologically based pharmacokinetic (PBPK) models take these age-dependent changes into account and may be used to predict drug exposure in children. As a result, this mechanistic-based tool has increasingly been applied to improve pediatric drug development. Under the Prescription Drug User Fee Act VI, the US Food and Drug Administration has committed to facilitate the advancement of PBPK modeling in the drug application review process. Yet, significant knowledge gaps on developmental biology still exist, which must be addressed to increase the confidence of prediction. Recently, more data on ontogeny of transporters have emerged and supplied a missing piece of the puzzle. This article highlights the recent findings on the ontogeny of transporters specifically in the intestine, liver, and kidney. It also provides a case study that illustrates the utility of incorporating this information in predicting drug exposure in children using a PBPK approach. Collaborative work has greatly improved the understanding of the interplay between developmental physiology and drug disposition. Such efforts will continue to be needed to address the remaining knowledge gaps to enhance the application of PBPK modeling in drug development for children.

Keywords: PBPK; children; model-informed drug development; ontogeny; pediatric; physiologically based pharmacokinetic modeling; transporters.

Figure 1.

Summary of the human membrane transporters in the intestine, liver, and kidneys that are mentioned in this review. Transporters with only mRNA or limited data are depicted in brown circles, whereas those that have both gene expression and protein abundance data are depicted in green circles. (Adapted/modified from Brouwer et al and Chu et al.) Noteworthy, the localization of OATP2B1 remains questionable. Future investigation is needed to ascertain if its localization is subject to developmental changes. BCRP, breast cancer resistance protein; BSEP, bile salt export pump; GLUT, glucose transporter; MATE, multidrug and toxin extrusion; MCT, monocarboxylate transporter; MRP, multidrug resistance-associated protein; NTCP, sodium/taurocholate cotransporting polypeptide; OAT, organic anion transporter; OATP, organic anion transporting polypeptide; OCT, organic cation transporter; PEPT1, peptide transporter 1; P-gp, P-glycoprotein; URAT1, uric acid transporter 1.

Figure 2.

Workflow of the pediatric physiologically based pharmacokinetic model (PBPK) establishment to simulate drug exposure in children. An adult PBPK model was first established and verified by comparing the output from the simulations to that in observed data.After ensuring that the adult model was robust, the pediatric model was generated by scaling the anatomical and physiological parameters using default age-dependent algorithms and incorporating ontogeny information for the transporters that are pertinent to this study. The pediatric PBPK model was verified, once again, by comparing the output from the simulation input with observed data from literature. Predictive performance of allometry and PBPK in estimating the clearance in children was also evaluated. CL indicates clearance; ped, pediatric; vs, versus.

Figure 3.

The elimination pathway of tazobactam. After intravenous administration, approximately 80% of tazobactam would be cleared renally by glomerular filtration and active secretion via OAT1 and OAT3. The majority of the rest of tazobactam would undergo hydrolysis to form the inactive metabolite, tazo-M1, which, similar to the parent drug, will be eliminated renally. A small amount (<1%) of tazobactam would undergo biliary excretion. CL indicates clearance; CL_R, renal clearance; GFR, glomerular filtration rate; OAT, organic anion transporter.

Figure 4.

Simulation of tazobactam exposure in adult population (a) and 3 pediatric cohorts (b-d) using PBPK (PK-Sim v7.3, Bayer, St. Louis, Missouri). Following a 500-mg 60-minute infusion in the virtual adult population, the predicted C_max, AUC, and CL were between 1.02- and 1.2-fold of the observed data. Three pediatric cohorts were generated: 0–3 months old (b), 3 months to 2 years old (c), and 2–7 years old (d). By taking into account the physiological and anatomical changes during development, and the ontogeny of transporters that are pertinent to the disposition of tazobactam, the tazobactam exposure was predicted reasonably well with C_max, AUC, and CL were all within 1.5-fold of observed data. Allometric scaling approach resulted in CL estimation that were comparable to that predicted using PBPK. However, for the youngest age group, 0–3 months old (b), the PBPK model performed slightly better, as allometry slightly overpredicted the CL (1.2-fold vs 1.8-fold of observed CL). AUC indicates area under the concentration-time curve; CL, clearance; C_max, peak plasma concentration; PBPK, physiologically based pharmacokinetics.