Advancing oncology drug development: Innovative approaches to enhance success rates while reducing animal testing

Drug development remains a high-risk endeavour, particularly in oncology, where failure rates exceed 90 %. This review examines emerging tools and strategies designed to enhance preclinical success rates, aligning with the 3Rs principle: Reduction, Refinement, and Replacement of animal testing. Traditional 2D in vitro screening remains fundamental in early anticancer drug development due to its cost-effectiveness and reproducibility. However, 3D in vitro culture systems, including patient-derived organoids, better recapitulate tumour structure, providing more accurate predictions of clinical response. Additionally, Organ-on-a-chip platforms further enhance physiological relevance and complement conventional animal toxicology models. Despite their promise, these technologies face challenges in standardisation, validation, and regulatory acceptance. Artificial intelligence is also emerging as a transformative tool in oncology drug discovery and development. However, its widespread adoption is currently constrained by limited access to high-quality datasets, concerns around data security, privacy, and underdeveloped computational infrastructure. For in vivo studies, patient-derived xenograft (PDX) models remain the gold standard, offering robust and translationally relevant platforms for efficacy testing. Hybrid models, such as PDX-derived organoids and PDX-derived cell cultures, provide complementary systems that integrate in vitro and in vivo insights. While these innovations offer long-term potential to reduce animal use, more innovative experimental designs and methods, such as the Single Mouse Trial and the Hollow Fibre Assay, may reduce animal numbers in the short term without compromising data quality. Together, these advances contribute to a more ethical, efficient, and predictive framework for the development of preclinical anticancer drugs.