Rebecca Leyla Huber 1, Oliver Scherf-Clavel 1
1 Ludwig-Maximilians-Universität München (Munich, Germany)
Introduction and Objectives: Major depressive disorder affects an estimated 5.7% of the global adult population and represents a substantial contributor to the global burden of disease [1], [2]. Although pharmacotherapy is a central component in the treatment of moderate to severe depression [3-5], vulnerable populations such as pregnant women, older adults, and pediatric patients are frequently underrepresented or systematically excluded from randomized controlled trials [6]. This underrepresentation is typically due to ethical constraints or the limited availability of patients who meet specific inclusion criteria. Physiologically based pharmacokinetic (PBPK) modeling has emerged as a valuable tool in this context, providing a mechanistic in silico framework for investigating drug exposure, variability, and drug–drug interactions (DDIs) in clinical scenarios that are difficult or unethical to study empirically. Although numerous PBPK models have been developed for several antidepressants, no comprehensive overview of modeled compounds, modeling purposes, and software platforms has been available. The objectives of this scoping review were therefore to (1) identify antidepressants for which PBPK models have been developed, (2) analyze the specific purposes and applications of these models, and (3) document the software tools used in their implementation, in order to map the current research landscape and identify knowledge gaps.
Methods: A scoping review was conducted in accordance with the PRISMA-ScR guidelines. Seven electronic databases (PubMed, Embase via Ovid, Cochrane Library, Web of Science, PubPharm, ClinicalTrials.gov, and the Open Systems Pharmacology based-publications-and-content via GitHub) were systematically searched for studies published from 1977 until July 2025 in English or German. Eligible studies were original research articles describing the development, adaptation, or new application of a PBPK model for an antidepressant. Whole-body, minimal, and semi-physiological PBPK models were included, provided they were relevant to human pharmacokinetics. Models developed exclusively for other indications than depression, and studies in which antidepressants were used solely as reference or methodological compounds were excluded unless clinically relevant outcomes for the treatment of depression were reported. Screening and study selection were performed using Rayyan® [7]. Data extracted included first author, year of publication, antidepressant(s) modeled, modeling objectives, and the software platforms used for model development.
Results: Of the 2,823 identified records, 60 studies met the inclusion criteria after screening and eligibility assessment. The distribution of PBPK models across antidepressants was highly heterogeneous. Paroxetine, fluvoxamine, and venlafaxine were the most extensively modeled compounds (each n = 10 models), whereas several antidepressants, including mirtazapine, trimipramine, and milnacipran, were represented by only one or no PBPK model.
The most common primary modeling purposes were simulation of DDIs (n = 18 models), pharmacokinetics during pregnancy (n = 14 models), and drug–(drug–)gene interactions (n = 13 models). In contrast, only two models specifically addressed ethnic populations, and no PBPK model was identified for pediatric patients treated for depression. Other clinically relevant contexts, such as toxicity, lactation and brain exposure, have been studied only moderately. Substance-specific analyses revealed clustering of research efforts around narrow application areas. For instance, several amitriptyline models primarily examined cardiac pharmacokinetics and toxicity, while sertraline and paroxetine models frequently addressed pregnancy-related scenarios and pharmacogenetic interactions.
Regarding software platforms, the Simcyp® Simulator was used in 46% of identified models, followed by PK-Sim®/MoBi® and GastroPlus® (each 17%), and R®/RStudio® (8%). In 12% of cases, alternative platforms were used or not explicitly reported. Additionally, 96 excluded studies incorporated antidepressants as perpetrator or victim drugs in PBPK-based DDI simulations without focusing on the antidepressant itself, underscoring their frequent use in CYP-mediated interaction research.
Conclusions: This scoping review demonstrates that while PBPK models are available for several antidepressants, their distribution and application are uneven. Modeling efforts are concentrated on a limited number of compounds and primarily focus on DDIs, pregnancy, and pharmacogenetics. Clinically important areas, including pediatric use and specific ethnic groups, remain underrepresented. Expanding PBPK modeling to additional antidepressants and underserved populations represents a key priority for future research. Furthermore, systematic evaluations of model quality and predictive performance are warranted to enhance clinical translation and regulatory applicability.
References:
[1] De Luca B, et al. Efficacy and tolerability of antidepressants in individuals suffering from physical conditions and depressive disorders: network meta-analysis. Br J Psychiatry. 2025 Aug;227(2):553-566. doi: 10.1192/bjp.2025.18. [2] Rizea-Savu S et al. The Effect of Food on the Single-Dose Bioavailability and Tolerability of the Highest Marketed Strength of Duloxetine. Clin Pharmacol Drug Dev. 2020 Oct;9(7):797-804. doi: 10.1002/cpdd.759.
Reference: PAGE 34 (2026) Abstr 12178 [www.page-meeting.org/?abstract=12178]
Poster: Drug/Disease Modelling - Absorption & PBPK