From Prediction to Decision Making: PBPK and QSP as Regulatory-Grade NAMs

New approach methodologies (NAMs) are experimental and computational tools that generate human-relevant mechanistic data for drug development and regulatory decisions. This review argues that PBPK and QSP models are computational NAMs that convert data from experimental NAMs (organoids, organ-on-chip, proteomics) into quantitative predictions of human exposure, efficacy, and safety. It presents PBPK case studies (a generic topical diclofenac bioequivalence approval, a moxidectin lactation-safety reassessment, and tafenoquine’s COVID-19 repurposing) and QSP case studies across CD3 T-cell engagers (ISB 2001, mosunetuzumab, epcoritamab, teclistamab, elranatamab), showing how mechanistic modeling shaped FDA/EMA decisions on dosing, labeling, and first-in-human trial design — including the FDA’s June 2026 draft guidance endorsing QSP for MABEL determination and the ICH M15 framework for model credibility.