Do longer duration nonclinical toxicology studies provide predictive clinical safety value? The IQ consortium longer duration nonclinical to clinical translational database

Thomas M Monticello,David M Potter,Qihong Huang,Timothy K Hart,Dana Shuey,Sean Troth,James M Vergis,Nardos Tassew,Peter Glascott
DOI: https://doi.org/10.1016/j.taap.2024.117087
2024-09-05
Abstract:The IQ Consortium's DruSafe Leadership Group previously reported results of a nonclinical to clinical translational database for First-In-Human (FIH)-enabling animal toxicology studies. We have completed an additional translational database populated with longer duration (>1 month) animal toxicology studies and longer duration (Phase 2 and beyond) clinical trials. The blinded database was composed of 127 molecules. Animal and clinical data were categorized by organ system and animal model (e.g. rodent, dog). The 2 × 2 contingency table (true positive, false positive, true negative, false negative) was used for statistical analysis and both the positive predictive value (PPV) and negative predictive value (NPV) were determined. As also reported in the FIH database, the NPV was the strongest predictive performance measure at 96 %. The PPV was lower than the FIH database with the rodent at 29 %, dog at 21 % and NHP at 20 %. No new additional target organs were observed in 62 % of the entries. A new target organ was identified in 38 % of the entries, with the majority in a rodent (26 %) and fewer in the dog (8 %) or NHP (12 %). However, new target organ data resulted in only a PPV of 13 %, suggesting that current ICH requirements for longer duration animal general toxicology studies should be re-evaluated and better aligned with the 3Rs. A newer paradigm could include an appropriately justified single animal model for longer duration studies, in addition to utilizing New Approach Methods (NAMs) that would provide translational safety data, but additional research is needed.
What problem does this paper attempt to address?