Leveraging multi-organ models for drug combination therapy safety

In this study, we developed a streamlined in vitro platform to assess the systemic toxicity of multidrug combinations within a clinically relevant timeframe, facilitating its application in personalized medicine. By incorporating cellular models that represent major toxicity-sensitive organs, such as the kidney, liver, and heart, we evaluated two previously optimized multidrug combinations (C2 and REMP). Our findings revealed distinct organ-specific toxicity profiles, with some drug-induced toxicities exacerbated upon combination. To evaluate the influence of model complexity, we compared responses between simple proliferative cell lines and more advanced models. Proliferative cell models, while useful for initial toxicity screenings, frequently failed to predict the severity of drug-induced toxicity. For instance, the C2 combination decreased cell viability by 50 % in patient-derived kidney organoids but only by 20 % in HEK293T cells. However, the C2 combination caused a 77 % viability reduction in differentiated hepatocyte spheroids, 32 % more than its effect on non-differentiated hepatocyte-like cells. C2 showed significant release of lactate dehydrogenase compared to 0.1 % DMSO. These findings underscore the critical need for systemic and biologically relevant drug safety assessments in the development of novel drug combinations, regardless of known single-drug safety. Despite some limitations, the platform accurately reproduced known single-drug toxicity profiles, confirming its translational potential. Overall, this biologically relevant approach enables efficient early-stage toxicity screening of drug combinations, supporting a safer development of personalized cancer therapy.