Abstract
Polyvalent antivenom potency tested for every batch and at four different stages of production is a regulatory requirement to be followed by the industry. Antivenom manufacturers have been following the gold standard test methodology for estimating median effective dose in mice for ages. Here, we report a non-animal New Approach Methodology that aligns with 3Rs in animal testing agenda, leveraging in vitro human stem cell technology for recreating microphysiological system complemented with process automation, Artificial Intelligence and Machine Learning digital workers’ effective utilization in the assay system to measure Naja naja snake antivenom potency. In vitro neutralization performed on specially configured human Microphysiological System acquired phenotype data sets at 20X magnification were analyzed against benchmark panels in the trained prediction model while the Artificial Intelligence predicted median effective dose value of venom-antivenom mixture was 3.9μL, showing a potency value of 2.04 mg recorded as the readout. This method adopted in the antivenom producer’s workflow will reduce reliance on mice-based testing and showcases potential for acceptance of robust alternative strategy to traditionally practiced protocols.