Establishment of Breast Cancer Organoids: A Systematic Review and Meta‐Analysis

Background: Organoid culture has emerged as a promising model for studying normal and tumor tissues. Organoids may resemble their tissue of origin, allowing assessment of disease behavior and drug response. In breast cancer (BC), successful organoid establishment may depend on factors like time from tissue collection to culture, dissociation methods, and media composition, which may affect tumor growth and tissue similarity across culture passages. This systematic review and meta-analysis are aimed at assessing the success rate of BC organoid establishment and their similarity to parental tumors. Methods: Following PRISMA guidelines, we conducted an electronic search in PubMed, Embase, and Web of Science using terms related to "breast cancer" and "organoids". Heterogeneity was assessed by I2 (%), with forest plots for all studies and subgroups. A fixed-effects meta-analysis estimated the overall establishment rate with 95% confidence intervals. Results: We analyzed 59 full-text articles. Five culture media types and 37 added components were reported, most frequent EGF (80%), glutamine, FGF, R-spondin (66%, each), Y27632, and N-acetylcysteine (64%, each). Ten studies provided data on initial samples and established organoids, totaling 504 and 349, respectively, with an establishment rate of 71.34% (95% CI: 67.55-75.14), varying from 31.25% to 86.21%. By subtype, establishment rates were as follows: Luminal A (57.47%), Luminal B (72.51%), HER2+ (71.24%), and triple-negative (78.75). Concordance between receptor expression in tumors and organoids was observed, with kappa values of 0.692 for ER, 0.597 for PR, and 0.711 for HER2. However, positive receptor expression reverted to negative in 19%, 31%, and 28% of organoids for ER, PR, and HER2, respectively. Conclusion: Breast cancer organoids were successfully established from primary tumor samples, demonstrating good overall concordance in receptor expression with the original tumors. These findings support the feasibility of patient-derived organoids as translational models; however, continuous phenotypic characterization across passages is recommended.