Static or microfluidic? A comparative pilot study of an in vivo-in vitro validated human iPSC-derived blood-brain barrier model for neurotoxicity hazard assessment.

Blood-brain barrier (BBB) permeation determines human brain exposure to chemicals and is therefore a prerequisite for neurotoxicity. However, the lack of toxicokinetic BBB models remains a critical uncertainty in regulatory in vitro risk assessment, which was highlighted in the new EU roadmap towards phasing out animal testing. Human induced pluripotent stem cell (hiPSC)-derived BBB models combined with clinical positron emission tomography (PET) data provide translational benchmarks for evaluating whether in vitro permeability data predict human brain exposure. Microphysiological systems (MPS) may better emulate physiological flow than static transwell® systems. Nonetheless, it remains unclear whether additional complexity introduced by flow improves fitness-for-purpose, reproducibility, and regulatory utility of BBB new approach methodologies (NAMs). Therefore, we directly compared our static hiPSC-derived BBB model, which was previously validated with human clinical PET brain-penetration data, with its microfluidic counterpart to determine whether flow translates into measurable added value for human brain exposure assessment. The comparative pilot study demonstrated that microfluidic implementation of the static model was technically feasible and maintained key BBB functionality in terms of barrier integrity, efflux capacity, and permeability. However, the MPS displayed reduced in vivo-like barrier characteristics and lower brain-penetration predictive performance than the static counterpart within the tested compound set, which is relevant for regulatory neurotoxicity assessment. In addition, the static model offered advantages in standardization, robustness, and throughput that are important for future regulatory implementation. Finally, this study identified technical limitations of the selected microfluidic transwell®-in-chip implementation that may guide the future development of standardized, commercially available BBB systems.