Dr. Xinning Yang reviewed how modeling approaches, particularly PBPK modeling, have increasingly been used during drug development to address unresolved DDI questions involving CYP enzymes and transporters. Using recent regulatory case studies at US FDA, he illustrated both the strengths and limitations of model-informed drug development and emphasized the growing role of endogenous biomarkers in reducing uncertainty. Dr. Houda Hachad presented real-world experience implementing pharmacogenomics- and DDI-based clinical decision support systems. Drawing on data from large healthcare networks, she demonstrated how integration of pharmacogenomics, DDIs, and pharmacist-led medication management can improve prescribing decisions and support personalized medicine. A recurring theme throughout the session was that scientific knowledge alone is insufficient; successful implementation requires effective decision-support tools embedded within clinical workflows.

The second session, “The EMA PBPK Model Qualification Pathway,” was inspired by discussions at the 2025 workshop regarding the European Medicines Agency’s landmark qualification opinion for PBPK platforms in the context of CYP-mediated DDI prediction. Dr. Efthymios Manolis (EMA) described the rationale and regulatory framework underlying platform qualification and explained how qualification differs from drug-specific model evaluation. Dr. Karen Rowland Yeo (Certara) reviewed the scientific journey leading to qualification of the Simcyp platform for selected CYP-mediated DDI applications and discussed the use of a large qualification matrix and Bayesian meta-analysis to characterize predictive uncertainty. Perspectives from regulators, industry, and model developers were provided by Susan Cole (MHRA), Loeckie De Zwart (Johnson & Johnson), Rolf Burghaus (Bayer/Open Systems Pharmacology), and Yuching Yang. The panel discussion highlighted broad support for the qualification initiative while also identifying important future challenges, including expansion of qualification to enzyme induction, transporters, UGT-mediated interactions, additional software platforms, and broader patient populations.

Participants agreed that the qualification framework is a significant step toward greater regulatory consistency and transparency in model-informed drug development. It also supports more efficient decision-making by reducing the need to reassess established platform components that have already been reviewed and approved.

A remaining challenge is to clearly differentiate between formal regulatory qualification—the highest standard—and qualifications granted internally within individual pharmaceutical companies or consortia. Participants also noted that, despite differences in terminology, the U.S. FDA’s Model Master File (MMF) approach and the EMA’s qualification pathway are broadly comparable.

Session III, “Complex DDIs,” focused on clinical scenarios where traditional DDI paradigms may fall short. In the keynote lecture, Prof. Eric Chan (National University of Singapore) described the evolution of DDI prediction from static approaches to dynamic PBPK models and emerging artificial intelligence (AI)-enabled frameworks. Using examples involving Paxlovid, rivaroxaban, and atorvastatin, he demonstrated how mechanistic modeling can be integrated with machine learning to support individualized DDI risk assessment. Dr. Zhu Zhou (University of the Pacific) discussed the challenges of managing DDIs in elderly patients, a population frequently excluded from clinical trials despite carrying the highest burden of polypharmacy. Using cannabinoids as a case study, she illustrated how PBPK models can be adapted to characterize age-related physiological changes and guide safer medication use in older adults. Dr. Lisa Almond (Certara) presented approaches for managing complex DDIs in global health settings, while Dr. Nada Abla Geiser (Medicines for Malaria Venture) reviewed the use of PBPK modeling in antimalarial drug development. Both highlighted the unique challenges posed by coexisting diseases such as HIV and tuberculosis, ethnic variability, and vulnerable populations including children and pregnant women. Collectively, the session emphasized that DDI management increasingly requires integrated approaches accounting for patient characteristics, disease states, pharmacogenetics, and comedications.

The final scientific session addressed “In Vitro Assays for Non-CYP Induction”. As UGTs are the predominant non-CYP drug metabolizing enzymes, advancing DDI prediction caused by UGT induction is an emerging challenge in DDI science. Dr. Diane Ramsden opened the session by demonstrating the limited dynamic range for UGT and CYP2C induction with conventional human hepatocyte 2D models and highlighted the utility of long-term, all-human co-culture systems which preserve an “inducible reserve”. This enables the generation of robust UGTs EC50 and Emax values. Additionally, she discussed reaction phenotyping for accurate UGT contribution estimation by leveraging siRNA-mediated knockdown methodologies. Dr. Xinning Yang shifted the focus to the clinical evaluation of UGT induction, noting that UGTs are frequently co-induced alongside CYP3A but to a lower extent. He proposed a decision tree that leverages a drug’s clinical CYP3A induction data alongside in vitro data to assess potential UGT induction risk and also illustrated how PBPK modeling can be successfully deployed to extrapolate the effect of induction on the UGT1A1 substrate cabotegravir. Dr. Mattia Berton concluded the session by evaluating the current state of PBPK extrapolation for UGT-mediated induction DDIs. He emphasized that the accuracy of these extrapolations heavily relies on reliable induction parameters (EC50 and Emax) and UGT degradation rates (Kdeg), where data is sparse. Participants agreed that although substantial progress has been made, important scientific and regulatory gaps remain before non-CYP induction can be evaluated with the same confidence currently applied to CYP-mediated interactions.

The workshop concluded with a poster session and general discussion. A major theme across multiple sessions of this year’s workshop was the increasing convergence of pharmacogenomics, PBPK modeling, global health, real-world evidence, and AI-based decision support in both drug development and clinical practice. Participants also revisited the implications of the EMA PBPK platform qualification initiative and discussed future opportunities to extend qualification principles beyond CYP-mediated DDIs. The discussions highlighted a growing need for collaboration among regulators, software developers, industry scientists, clinicians, and academic researchers to establish best practices for emerging methodologies.

The 2026 workshop marked the first full year under the leadership of the renewed organizing faculty following the retirement of founding organizers Dr. Robert Hermann and Ms. Karen Grave-Hermann in 2025. The faculty team, consisting of Prof. Uwe Fuhr, Prof. Amin Rostami-Hodjegan, Dr. Ping Zhao, Dr. Isabelle Ragueneau-Majlessi, and Dr. Wafaâ Jabrane, continued the workshop’s tradition of fostering open scientific dialogue and interdisciplinary collaboration. As DDI science evolves toward increasingly personalized and model-informed approaches, the Marbach DDI Workshop represents leading international forum for advancing both the science and practice of clinical DDI management.

The Faculty

Uwe Fuhr, MD
Dr. Wafaâ Jabrane
Isabelle Ragueneau-Majlessi, MD
Amin Rostami-Hodjegan, PhD, FCP
Ping Zhao, PhD

… read less