Skip to content
DraconiticResearch
Search Sources API
226,742 studies31 sourcesUpdated 36 min ago
Peer-reviewedCC BYFrontiers in Drug Safety and Regulation

From signal to mechanism: registry pharmacology and the future of translational pharmacovigilance

  1. 1chool of Pharmacy, School of Health Sciences, College of Medicine and Health, University of Birmingham, Edgbaston, United Kingdom

Abstract

This perspective describes Registry Pharmacology, a translational framework that integrates spontaneous adverse drug reaction (ADR) reporting with prescribing data, physicochemical descriptors and polypharmacology, to move beyond signal detection towards mechanistic interpretation of real world drug safety. The framework rests on three pillars. First, exposure normalisation standardises ADR reporting frequency to prescriptions using national datasets, contextualising spontaneous ADR reporting relative to prescribing volume, without estimating comparative risks among exposed patients. Second, clinically achievable pharmacology filtering restricts off-target interactions to those occurring at or below clinically relevant Cmax, giving mechanistic insights. Third, physicochemical integration provides interpretive context for tissue specific ADR patterns. Across studies spanning oncology to respiratory medicine, the framework has consistently generated pharmacologically coherent, mechanistically plausible hypotheses for ADR divergence. Interpretable machine learning yielded ROC AUC values up to 0.94 for ADR-category classification, although discrimination alone does not establish reliable signal detection or prospective clinical utility. Acknowledged limitations include the inherent constraints of spontaneous reporting systems e.g., underreporting, notoriety bias, confounding by indication, incomplete pharmacology databases, largely unaddressed polypharmacy and drug drug interactions, and the hypothesis-generating rather than causal inference character of the methodology. Registry Pharmacology may offer a translational framework for converting pharmacovigilance data into pharmacologically interpretable intelligence. As data infrastructure, regulatory need for mechanistic evidence and interpretable AI capabilities matures, this framework may contribute to supporting a transition from reactive epidemiological surveillance to prospective, mechanistically informed drug safety science.

Related

  1. Ensemble signal detection reveals target-specific temporal risk of hypokalemia after immune checkpoint inhibition: a translational pharmacology framework
    Shengwang Chen, Ziyi Dai, Zihan Guo +3
    Frontiers in pharmacologyPeer-reviewedCC BYFull textPharmacology & toxicologyEurope PMCCrossref
  2. Safety signals of tocilizumab in pediatric patients: A real-world pharmacovigilance analysis based on the FAERS database with network pharmacology insights.
    Gang Wu, Xueqing Ma, Bin Zhu +1
    Human vaccines & immunotherapeuticsPeer-reviewedCC BY-NCFull textPharmacology & toxicologyEurope PMC
  3. Cannabinoids in oncology: pharmacology, clinical applications, and translational perspectives
    Emma Safran, Howard Safran, Stanley Kim
    Frontiers in pharmacologyPeer-reviewedCC BYFull textPharmacology & toxicologyEurope PMC
  4. Guidance on the interpretation of pharmacovigilance signal detection data and analyses
    Charles Khouri, Virginie Bres, Marion Allouchery +15
    TherapiesPeer-reviewedCC BYFull textPharmacology & toxicologyHAL
  5. Integrating pharmacovigilance and machine learning: bleeding signal detection and risk prediction for oral anticoagulants in cancer patients
    Meina Lv, Xiaochun Zheng, Hang Xie +1
    Research and Practice in Thrombosis and HaemostasisPeer-reviewedCC BYFull textPharmacology & toxicologyDOAJ
BibTeX
@article{cheng2026from,
  title = {{From signal to mechanism: registry pharmacology and the future of translational pharmacovigilance}},
  author = {Vicky Cheng and Alan M. Jones},
  journal = {Frontiers in Drug Safety and Regulation},
  year = {2026},
  month = nov,
  doi = {10.3389/fdsfr.2026.1992258},
  url = {https://doi.org/10.3389/fdsfr.2026.1992258},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - From signal to mechanism: registry pharmacology and the future of translational pharmacovigilance
AU  - Vicky Cheng
AU  - Alan M. Jones
PY  - 2026
DA  - 2026/11/16
JO  - Frontiers in Drug Safety and Regulation
DO  - 10.3389/fdsfr.2026.1992258
UR  - https://doi.org/10.3389/fdsfr.2026.1992258
AB  - This perspective describes Registry Pharmacology, a translational framework that integrates spontaneous adverse drug reaction (ADR) reporting with prescribing data, physicochemical descriptors and polypharmacology, to move beyond signal detection towards mechanistic interpretation of real world drug safety. The framework rests on three pillars. First, exposure normalisation standardises ADR reporting frequency to prescriptions using national datasets, contextualising spontaneous ADR reporting relative to prescribing volume, without estimating comparative risks among exposed patients. Second, clinically achievable pharmacology filtering restricts off-target interactions to those occurring at or below clinically relevant Cmax, giving mechanistic insights. Third, physicochemical integration provides interpretive context for tissue specific ADR patterns. Across studies spanning oncology to respiratory medicine, the framework has consistently generated pharmacologically coherent, mechanistically plausible hypotheses for ADR divergence. Interpretable machine learning yielded ROC AUC values up to 0.94 for ADR-category classification, although discrimination alone does not establish reliable signal detection or prospective clinical utility. Acknowledged limitations include the inherent constraints of spontaneous reporting systems e.g., underreporting, notoriety bias, confounding by indication, incomplete pharmacology databases, largely unaddressed polypharmacy and drug drug interactions, and the hypothesis-generating rather than causal inference character of the methodology. Registry Pharmacology may offer a translational framework for converting pharmacovigilance data into pharmacologically interpretable intelligence. As data infrastructure, regulatory need for mechanistic evidence and interpretable AI capabilities matures, this framework may contribute to supporting a transition from reactive epidemiological surveillance to prospective, mechanistically informed drug safety science.
SN  - 2674-0869
ER  -