Future of Pharmacometrics

Abstract

Pharmacometrics represents a firmly established discipline and provides an essential component to model-informed drug discovery, development, and utilization. Grounded in basic principles of pharmacokinetics (PK) and pharmacodynamics (PD), modeling approaches can be readily extended to diverse data types, including dichotomous, ordered categorical, time-to-event, and continuous outcomes. Traditional PK/PD models of drug action utilize compartmental structures to integrate the time-course of drug exposure, pharmacological properties (capacity, sensitivity, and transduction of drug-target interactions), and (patho-) physiological turnover processes. Such semi-mechanistic models contain a minimal number of identifiable parameters to describe temporal profiles of macro-scale therapeutic and adverse drug effects. Coupled with nonlinear mixed effects modeling of relatively large clinical trials, a covariate analysis can be used to identify patient specific characteristics (e.g., genetic polymorphisms) that explain the inter-individual variability in model parameters and outcomes. This approach can be limited by specific study designs and is rarely sufficient for recapitulating multiple, complex genotype-phenotype relationships; however, a major opportunity for pharmacometrics is the extension of pharmacostatistical principles to systems pharmacology models. Significant insights have been realized from the recognition that both drugs and disease processes give rise to complex and dynamic clinical phenotypes by altering natural interconnected biochemical networks and support the emergence of systems pharmacology models of drug action. Multi-scale models that combine physiological PK/PD principles and signaling networks can serve as a platform for integrating genomic/proteomic factors that regulate drug effects and clinical outcomes. Models derived from combining genomic and proteomic databases with systems models and a statistical framework could be used to test confidence in early drug targets, project inter-individual variability and patient subpopulations likely to respond to new drugs or drug combinations, and to ultimately achieve precision medicine.