From drug response to resistance evolution: Leveraging 3D cancer models for functional precision oncology.

Therapy resistance remains a major cause of relapse and cancer-related mortality, arising from dynamic interactions between tumor-intrinsic programs and microenvironmental constraints. Genetic and epigenetic alterations, transcriptional rewiring, metabolic adaptation, extracellular matrix signalling, hypoxia and limited drug penetration shape resistant phenotypes and therapeutic failure. Accurately modelling these processes is essential for identifying actionable resistance mechanisms and guide patient treatment. Conventional 2D cultures provide valuable mechanistic insights and scalable drug screening platforms, but fail to recapitulate the spatial organization, diffusion gradients and heterogeneous drug exposure characteristic of solid tumors. In contrast, advanced 3D preclinical models capture key aspects of tumor architecture and microenvironmental complexity that influence therapy response and resistance evolution. This review provides a comparative analysis of current 3D cancer models and proposes a conceptual framework linking mechanisms of therapeutic adaptation to the experimental platforms that most faithfully recapitulate them, with particular emphasis on patient-derived organoids (PDOs). By retaining clinically relevant features of the tumor of origin, PDOs enable the investigation of resistance evolution in both treatment-naïve and clinically treated tumors, as well as modelling experimentally acquired resistance through controlled therapeutic selection. These features enable mechanistic studies with high translational relevance, while supporting the development of precision oncology strategies. Finally, we propose practical recommendations for experimental design, model selection and reporting to improve reproducibility and facilitate cross-study comparisons. Collectively, advanced 3D cancer models represent a promising translational framework for uncovering mechanisms of therapy resistance and identifying actionable vulnerabilities that may guide more effective and personalized treatment strategies.