Evaluating local neurotoxicity potential of extractables and leachables: Comparison of in vivo and in vitro data and need for new approach methods.

Animal testing remains the main regulatory approach in preclinical drug development, and may be used to analyze safety concerns (qualification) of potential leachables from container closure systems for high risk routes of administration such as intrathecal or epidural products. These studies are technically challenging, costly and potentially hurtful to animals, while providing limited scientific progress; replacement by new approach methodologies is therefore highly desirable. This study reviewed existing literature on in vivo and in vitro E&L local neurotoxicity, focusing on intrathecal studies, and assessed whether parallels between both methods could support the development of in vitro test batteries. Publicly available data repositories were searched for neurotoxicity data for 865 PQRI- and ELSIE-listed potential leachables. Data analysis revealed limited true neurotoxicity endpoints in vivo and limited availability of NOAECs for most potential leachables. However, 21 compounds (2.4%) were tested both intrathecally and in vitro. Of these, 16 showed similar results, while five led to discordant evaluations. For non-intrathecal in vivo administration routes, 108 potential leachables (12.5%) were tested in parallel; 91 showed similar and 17 discordant results. Deeper analysis revealed that these discordant results may be caused by the presence of the blood-brain barrier and cell-type mismatches. Overall, the findings identified overlaps and discrepancies between new approach methodologies and animal models in identifying neurotoxicity, information that may help in developing cell-based methods.