癌症是全球死亡率最高的疾病之一。癌症负担持续增加，不仅影响患者的健康相关生活质量，还造成全球经济影响加大。癌症的复杂性和异质性给研究和临床实践带来了重大挑战，导致抗肿瘤药物临床试验的失败率增加。部分原因是临床前模型在忠实再现人类肿瘤以作为药物有效性的可靠指标方面仍然存在重要局限性。到目前为止，研究和开发策略采用昂贵的动物模型（包括所谓的“人源化小鼠”），这不仅引起伦理问题，而且经常无法准确预测抗癌药物的反应，因为它们不能忠实地复制人类生理学以及患者的肿瘤微环境。另一方面，传统的二维（2D）细胞培养物无法充分再现肿瘤的结构组织和体内发现的细胞异质性。开发更准确的癌症模型的必要性日益增加，这日益强调了三维 (3D) 体外细胞培养（例如癌症来源的球体和类器官）的重要性，作为弥补 2D 和动物模型之间差距的有希望的替代方案。在这篇综述中，我们简要概述了 3D 体外细胞培养作为临床前模型，能够在微调的微环境中正确再现体内肿瘤的组织组织、生物组成和复杂性。尽管存在局限性，这些模型共同增强了我们对癌症机制的理解，并可能在临床测试之前提供更可靠的药物疗效评估，从而改善癌症患者的治疗结果。©作者。

Cancer represents one of the diseases with the highest mortality rate worldwide. The burden of cancer continues to increase, not only affecting the health-related quality of life of patients but also causing an elevated global financial impact. The complexity and heterogeneity of cancer pose significant challenges in research and clinical practice, contributing to increase the failure rate of clinical trials for antitumoral drugs. This is partially due to the fact that preclinical models still present important limitations in faithfully recapitulating human tumors to serve as reliable indicators of drug effectiveness. Up to now, research and development strategies employ expensive animal models (including the so-called "humanized mice") that not only raise ethical concerns, but also frequently fail to accurately predict responses to anticancer drugs because they do not faithfully replicate human physiology as well as the patient's tumor microenvironment. On the other side, traditional two-dimensional (2D) cell cultures fail to adequately reproduce the structural organization of tumor and the cellular heterogeneity found in vivo. The growing necessity to develop more accurate cancer models has increasingly emphasized the importance of three-dimensional (3D) in vitro cell cultures, such as cancer-derived spheroids and organoids, as promising alternatives to bridge the gap between 2D and animal models. In this review, we provide a brief overview focusing on 3D in vitro cell cultures as preclinical models capable of properly reproducing the tissue organization, biological composition, and complexity of in vivo tumors in a fine-tuned microenvironment. Despite their limitations, these models collectively enhance our understanding of the mechanisms underlying cancer and may offer the potential for a more reliable assessment of drug efficacy before clinical testing and, consequently, improve therapeutic outcomes for cancer patients.© The author(s).