Guide on Using Quantitative Systems Pharmacology Models for Trial Dosing
FDA has released a draft guidance outlining how sponsors can use quantitative systems pharmacology (QSP) models to determine starting doses for first-in-human (FIH) clinical trials, particularly for higher-risk therapies where traditional animal-based approaches may be less predictive. The document focuses on using QSP models to estimate the minimum anticipated biological effect level (MABEL), a dose expected to produce a minimal biological effect in humans. FDA says the approach combines mathematical modeling, disease biology, pharmacology, and multiple sources of preclinical data to improve dose selection and potentially reduce reliance on animal toxicology studies.
According to the agency, QSP modeling may be particularly useful for products such as immunostimulatory monoclonal antibodies and other therapies with human-specific targets that are difficult to evaluate using traditional animal models. The guidance notes that regulatory submissions increasingly include QSP analyses to support decisions ranging from investigational new drug applications to marketing applications.
The document provides recommendations on model development, validation, uncertainty analysis, and regulatory interactions. FDA advises sponsors to incorporate all available data — including in vitro, in vivo, ex vivo, and in silico evidence — and to engage with the agency early through pre-IND and Model-Informed Drug Development meetings when planning to use QSP-based approaches for dose selection.