Authors
Pranay Wal, Jyotsana Dwivedi, Kanika Pandey, Sudarshana Borah, Krishana Kumar Sharma, Abida Khan, Hanish Singh Jayasingh Chellammal, Amin Gasmi
Published in
Critical reviews in toxicology. Pages 1-27. Nov 17, 2026. Epub Nov 17, 2026.
Abstract
NAMs are revolutionizing safety assessment and bridging the substantial disconnect between traditional animal toxicology studies and human exposure within the pharmaceutical industry. AI predictive toxicology uses graph neural networks, deep learning, explainable AI, PBPK, and IVIVE modeling for human-dose predictions, and micro-physiological and organ-on-a-chip systems that replicate specific physiological states. The development of NAMs has been applied to FDA initiatives, such as the Predictive Toxicology Roadmap, ISTAND MIDD, and CiPA, to demonstrate how integrated NAM systems optimized at early drug development stages provide valuable information on mechanistic toxicities related to humans and accelerate drug development. AI-QSAR models, PBPK simulations, and organ-on-a-chip technologies are increasingly integrated with discovery and early-stage preclinical workflows in different sectors. Standardization and reproducibility across platforms, transparency of AI models, multiomics integration, and global regulator convergence are required to improve the power of AI. This review is the first to provide a unified mechanistic synthesis of all three pillars of NAMs, AI-based predictive toxicology, PBPK-IVIVE modeling, and microphysiological systems (MPS), while earlier reviews covered them independently. We suggest a cyclical NAM ecosystem in which AI predictions seek translation to human pharmacokinetics, with iterative mechanistic validation via MPS, solving current challenges of translation from animals to humans. This review is distinct from others in that it presents NAM technologies as an integrated, human-relevant system rather than disparate technologies, and thus offers regulators a roadmap.
PMID:
43043626
Bibliographic data and abstract were imported from PubMed on 18 Nov 2026.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 5
- Comments 0