Three-dimensional spheroid models in breast cancer: tumor microenvironment complexity, cancer stem cell-driven resistance, and translational model integration

Abstract

The limited clinical success of anticancer therapies highlights the need for preclinical models that more accurately recapitulate tumor complexity, therapeutic response, and patient-specific heterogeneity. Conventional two-dimensional (2D) culture systems fail to capture the structural organization, cellular diversity, and dynamic microenvironmental gradients of in vivo tumors, thereby limiting their translational predictive value. Three-dimensional (3D) spheroid models have emerged as widely adopted platforms that better mimic tumor architecture, microenvironmental gradients, and cell–cell interactions. In this review, we provide a comprehensive and critical evaluation of 3D spheroid systems in breast cancer, with particular emphasis on their role in enriching breast cancer stem cells (BCSCs), modeling therapeutic resistance, and improving translational relevance. We comparatively assess scaffold-free and scaffold-based approaches, including hanging drop systems, ultra-low attachment cultures, hydrogels, microfluidics, patient-derived organoids (PDOs), and bioprinting technologies, while clearly distinguishing multicellular tumor spheroids (MCTSs) from organoid-based models according to their structural complexity, self-organization hierarchy, patient-derived fidelity, and translational applicability. While 3D spheroid models recapitulate key features of the tumor microenvironment, including hypoxia, nutrient gradients, extracellular matrix interactions, and stemness-associated signaling, their utility remains constrained by experimental heterogeneity, incomplete vascular and immune integration, and limited cross-platform standardization. We further discuss conflicting evidence regarding predictive drug-response accuracy and examine the persistent translational gap between preclinical findings and patient outcomes. Importantly, we propose a tiered, fit-for-purpose framework for the strategic selection and integration of preclinical cancer models based on the biological question, cancer subtype, and translational objective. Emerging technologies, including patient-derived organoids, vascularized organ-on-chip systems, and AI-assisted analytical platforms, are evaluated as complementary strategies to overcome current biological, technical, and translational limitations. 3D spheroid models represent a critical intermediate platform for physiologically relevant cancer modeling; however, their translational impact depends on integration with more complex, standardized, and clinically validated systems. A multi-model, context-driven approach incorporating advanced bioengineering, immune-competent platforms, AI-assisted analytics, and patient-specific models will be essential for improving translational predictability and advancing precision oncology in breast cancer.