Advancements in pediatric drug development require innovative approaches that overcome challenges in assessing age-dependent drug disposition and toxic risks related to metabolism. Cytochromes P450 dominate drug metabolism yet roles for individual enzymes depend on genetics, disease states, co-medications, and ontogeny. Failure to account for those differences contributes to dosing challenges and possibly toxicity as reported for dextromethorphan, midazolam, and phenytoin. The NICHD Pediatric Formulation Initiative (PFI) workshops emphasized the need for better modeling to describe and predict how drug metabolism changes for children. Current pharmacokinetic models predict how drug clearance changes with age that affect the optimal dose, but those models are limited in two ways; (1) they require experimentally determined kinetic data for several enzymes that is not often available, and (2) they do not model formation of specific drug metabolites, which is important in predicting toxicity and drug interactions, regardless of whether clearance changes with age. We hypothesize that computational models of mixtures of P450 enzymes can predict how the ?metabolic fate?, i.e. the kinetics of drug metabolism and the resulting metabolite structures, of drugs changes with age. We propose building hierarchical mathematical models that at first predict the drug metabolites formed by metabolic enzymes (Aim 1), then the efficiency of formation for each metabolite (the kinetics) (Aim 2), and combine these models to predict metabolites formed by age-specific mixtures of P450s (Aim 3). This proposal makes significant steps toward achieving PFI goals. First, datasets created for this study will be made publicly available to foster model refinement and validation by the community. Second, simulation of metabolite profiles would yield tractable biomarkers and support studies on possible age-dependent drug-drug interactions, off- target biological activities, pro-drug activation, and formation of toxic species. Third, the models will indicate, for both new and existing drugs, when metabolic fate (consequently, toxicity and interactions) changes in pediatric patients, even when pharmacokinetics stay the same. Fourth, though not the primary aim of this study, successfully modeling metabolic efficiency could enable pharmacokinetic studies for predicting problematic pediatric drugs prior to carrying out any necessary experimental kinetic studies. Taken together, this proposal lays a strong foundation for developing models relevant that resolve challenges in optimizing drug dosages and minimizing toxicity risks for children.
The impact of age on cytochrome P450 metabolism makes drug disposition and toxic risks moving targets for pediatric patients, such that robust strategies to identify and assess ontogenetic (age-dependent) effects on drug metabolic fate (in vitro kinetics and metabolite structures) are clearly needed to better personalize drug development for pediatric dosing. As an extension of our prior work, we will build and test computational models and datasets for individual P450 isozymes that accurately predict drug metabolites and the rate (kinetics) at which they form. We will then combine these models for an effective, low cost approach that simulates hepatic P450 metabolism and takes into account age-dependent contributions from individual isozymes to assess the metabolic fate of drugs during child development.
| Khojasteh, S Cyrus; Miller, Grover P; Mitra, Kaushik et al. (2018) Biotransformation and bioactivation reactions - 2017 literature highlights *. Drug Metab Rev :1-35 |
| Dang, Na Le; Hughes, Tyler B; Miller, Grover P et al. (2018) Computationally Assessing the Bioactivation of Drugs by N-Dealkylation. Chem Res Toxicol 31:68-80 |
| Matlock, Matthew K; Dang, Na Le; Swamidass, S Joshua (2018) Learning a Local-Variable Model of Aromatic and Conjugated Systems. ACS Cent Sci 4:52-62 |
| Southerland, William M; Swamidass, S Joshua; Payne, Philip R O et al. (2018) The diversity and disparity in biomedical informatics (DDBI) workshop. Pac Symp Biocomput 23:614-617 |
| Barnette, Dustyn A; Davis, Mary A; Dang, Na L et al. (2018) Lamisil (terbinafine) toxicity: Determining pathways to bioactivation through computational and experimental approaches. Biochem Pharmacol 156:10-21 |
| Marsh, Jon N; Matlock, Matthew K; Kudose, Satoru et al. (2018) Deep Learning Global Glomerulosclerosis in Transplant Kidney Frozen Sections. IEEE Trans Med Imaging 37:2718-2728 |
| Hartman, Jessica H; Miller, Grover P; Meyer, Joel N (2017) Toxicological Implications of Mitochondrial Localization of CYP2E1. Toxicol Res (Camb) 6:273-289 |
| Dang, Na Le; Hughes, Tyler B; Miller, Grover P et al. (2017) Computational Approach to Structural Alerts: Furans, Phenols, Nitroaromatics, and Thiophenes. Chem Res Toxicol 30:1046-1059 |
| Hughes, Tyler B; Swamidass, S Joshua (2017) Deep Learning to Predict the Formation of Quinone Species in Drug Metabolism. Chem Res Toxicol 30:642-656 |
| Ruan, Shuxiang; Swamidass, S Joshua; Stormo, Gary D (2017) BEESEM: estimation of binding energy models using HT-SELEX data. Bioinformatics 33:2288-2295 |
Showing the most recent 10 out of 14 publications
