Pharmacometric modeling of opioid disposition in pregnancy: a systematic review of PopPK and PBPK approaches, model limitations, and future directions

Physiological changes during pregnancy substantially alter opioid pharmacokinetics, yet clinical pharmacokinetic data in pregnant populations remain limited. Pharmacometric modeling using population pharmacokinetic (PopPK) and physiologically based pharmacokinetic (PBPK) approaches offers a quantitative framework to characterize these changes and to estimate maternal and fetal drug exposure. Despite increasing application of these methods, the methodological quality, reproducibility, and clinical relevance of existing opioid models in pregnancy have not been systematically evaluated. This systematic review aimed to critically appraise published PopPK and PBPK models of opioid medications in pregnancy, identify recurring structural and methodological features, and outline priorities for improving model development and reporting. A structured PubMed search (database inception to September 2025) identified studies modeling opioid disposition in pregnant individuals using PopPK or PBPK frameworks. Data on model structure, parameterization, assumptions, evaluation strategies, and reported limitations were extracted and synthesized. Ten studies met inclusion criteria, with fentanyl having the most studies. PopPK analyses consistently identified pregnancy as a significant covariate associated with increased clearance. PBPK models showed substantial heterogeneity in physiological detail, ranging from simplified fetoplacental compartments to permeability-limited placenta representations and multi-compartment fetal systems. Reproducibility emerged as a concern, with at least one PBPK model unable to be independently replicated due to incomplete reporting. Across studies, common gaps included limited verification datasets, incomplete representation of gestational physiology, and reliance on single-time-point umbilical cord concentrations at delivery. Overall, pharmacometric modeling provides valuable mechanistic insight into opioid disposition during pregnancy, but its translational impact is constrained by inconsistent reporting practices, sparse empirical data, and limited incorporation of fetal and neonatal exposure. Future progress will be accelerated by standardized documentation, improved pregnancy-specific physiological data, integration of genetic and developmental variability, and closer linkage between pharmacokinetics and clinically relevant outcomes to support model-informed perinatal opioid therapy.