The potential of a novel pharmaceutical compound to cause an adverse event (AE) in humans is currently assessed in animals. Testing for potential safety concerns in the central nervous system is carried out in a group of tests in rats or mice (named the Functional Observational Battery, FOB, or Irwin Test, IT). A typical FOB/IT study will use 24 rats or mice per compound, and up to ten compounds will be tested to deliver a single candidate drug. The ability of these animal tests to predict the most common side effects seen in man has never been analysed.
In this study, five pharmaceutical companies collaborated to provide data on 141 candidate compounds under development. The NC3Rs, acting as an honest broker, anonymised and analysed the data to assess the predictive nature of the FOB/Irwin test batteries. By comparing the occurrence of AEs in the rodent and human data for each compound, the ability of the animal model to predict pain, fatigue/somnolence, dizziness and nausea was assessed. The results of this analysis demonstrated that, within this dataset, the FOB/Irwin tests do not predict these common side effects.
This study demonstrates that animal models do not predict AEs for which there may not be a direct equivalent (nausea, dizziness). It was perhaps more surprising that the tests did not predict when a drug would cause fatigue in the human, as the animal tests included measures of locomotor activity and other relevant outcome measures. Within the confines of the dataset (as described in Mead et al 2016), this study demonstrates that current models do not adequately predict many CNS safety liabilities. The application of human tissue- and cell-based models may improve the predictive ability in safety pharmacology, and further work is required to identify the most human-relevant preclinical tests.
Mead AN, Amouzadeh HR, Chapman K et al. (2016) Assessing the predictive value of the rodent neurofunctional assessment for commonly reported adverse events in phase I clinical trials; Regul Toxicol Pharmacol 80:348-357 doi: 10.1016/j.yrtph.2016.05.002