Advancing chemical mixture toxicity assessment using advanced in vitro NAMs: current status and future perspectives

Chemical exposure in real-world settings occurs predominantly as mixtures, whereas toxicological assessment has long been shaped by single-substance paradigms. As interest in more human-relevant and mechanistically informative testing strategies has grown, new approach methodologies (NAMs) have been increasingly incorporated into mixture toxicity research. However, the role of advanced in vitro NAMs in chemical mixture toxicity assessment has not yet been examined in a structured manner. This review therefore evaluated how advanced in vitro NAMs are currently being applied in mixture toxicity assessment and considered their potential value for regulatory translation. A PubMed-based literature search covering studies published between 2021 and 2025 identified 353 in vitro NAM-based studies related to chemical mixture toxicity, of which 70 were classified as advanced in vitro NAM-based studies and included for in-depth analysis. Among the advanced in vitro NAM categories, route-of-exposure-relevant models were the most frequently applied, followed by Omics, 3D models, stem cell-based models, and HTS/HCI approaches. The reviewed studies addressed diverse mixtures, including air pollutants, PFAS, phthalates, pesticides/herbicides, consumer product-related mixtures, and other environmentally relevant chemical combinations. Across these studies, advanced in vitro NAMs were used not only to assess overall mixture effects, but also to compare observed and predicted responses, distinguish mixture effects from those of individual chemicals, identify effect-driving components, and interpret non-additive responses under biologically relevant conditions. The reviewed evidence indicates that advanced in vitro NAMs are being used not simply as alternative test systems, but as tools to address key challenges intrinsic to mixture toxicity assessment, including defining relevant mixture entities, identifying toxicity drivers, interpreting non-additivity, and supporting prioritization under combinatorial complexity. These findings suggest that the future value of advanced in vitro NAMs in mixture assessment will depend not only on continued technological development, but also on their integration into question-driven and fit-for-purpose testing strategies. From this perspective, advanced in vitro NAMs may contribute most effectively when used to generate decision-relevant evidence aligned with specific mixture assessment needs and regulatory contexts.