Beyond Inert Fillers: A Systematic Review of Machine Learning Applications in Pharmaceutical Excipient and Formulation Design
Scientia Pharmaceutica · 14.11.2026 · 94 (4) · 111
DOI: 10.3390/scipharm94040111 ↗Kurz gefasst
Systematische Übersicht in der Zeitschrift Scientia Pharmaceutica, online erschienen am 14.11.2026. Verfasst von Maria Kannavou und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Background
Pharmaceutical excipients are increasingly recognized as active determinants of formulation performance, manufacturability, stability, and safety. Artificial intelligence (AI) and machine learning (ML) are being explored to support excipient-related decisions.
Methods
This systematic review was conducted and reported according to PRISMA 2020. Scopus, Web of Science and PubMed were searched in October 2026 for English-language, peer-reviewed original research published between January 2020 and July 2026 that applied trained ML models to excipient-related formulation problems.
Results
Of 463 unique records, 116 full-text reports were assessed and 60 studies were included; several share datasets or platforms and therefore represent fewer independent evidence streams. The studies were organized into five application domains: drug–excipient compatibility; dosage-form design and optimization; advanced manufacturing and drug-delivery systems; biopharmaceutical formulation and stability; and excipient safety and biological activity. Most studies relied on retrospective internal validation; only 12 reported experimental, prospective or external validation that fully met our criterion.
Conclusions
AI/ML may help narrow the experimental search space and support excipient decisions, but evidence of reduced time, material use or cost is rarely reported. Implementation is limited by heterogeneous datasets, class imbalance, incomplete reporting, restricted applicability domains, limited external validation, and interpretability. Prospective benchmarking, interoperable databases, and regulatory-grade validation are needed.
Zitieren
APA
Kannavou, M., & Davrazos, G. (2026). Beyond Inert Fillers: A Systematic Review of Machine Learning Applications in Pharmaceutical Excipient and Formulation Design. Scientia Pharmaceutica, 94(4), 111. https://doi.org/10.3390/scipharm94040111
Vancouver
Kannavou M, Davrazos G. Beyond Inert Fillers: A Systematic Review of Machine Learning Applications in Pharmaceutical Excipient and Formulation Design. Scientia Pharmaceutica. 2026;94(4):111. doi:10.3390/scipharm94040111
BibTeX
@article{Kannavou2026,
author = {Kannavou, Maria and Davrazos, Gregory},
title = {Beyond Inert Fillers: A Systematic Review of Machine Learning Applications in Pharmaceutical Excipient and Formulation Design},
journal = {Scientia Pharmaceutica},
year = {2026},
volume = {94},
number = {4},
pages = {111},
doi = {10.3390/scipharm94040111}
}
EUVIMED