We reported a case of a 31-year-old female with therapy-resistant, refractory, and metastatic luminal B breast cancer. Using organoids derived from the patient's malignant pleural effusion for drug screening, we found that the combination of gemcitabine and cisplatin was the most sensitive, considering both IC50 and AUC values. In clinical practice, it was observed that the patient responded well to the selected treatment regimen, resulting in a significant reduction of pleural effusions, a marked decrease in tumor markers (e.g., CA125), and improved performance status (PS 2→1). The organoid model enabled the rational use of the patient's metabolic waste. It replicates the complexity of human tumors and facilitates extensive screening of beneficial drugs for patient diseases, particularly those with advanced tumors showing heterogeneity and rapid disease progression. This method swiftly identifies the optimal therapeutic drug regimen, minimizing the risk of drug resistance and trial costs, thereby providing maximum patient benefits.