Many patient-derived tumor models have emerged recently. However, their potential to guide personalized drug selection remains unclear. Here, we report patient-derived tumor-like cell clusters (PTCs) for non-small cell lung cancer (NSCLC), capable of conducting 100–5,000 drug tests within 10 days. We have established 283 PTC models with an 81% success rate. PTCs contain primary tumor epithelium self-assembled with endogenous stromal and immune cells and show a high degree of similarity to the original tumors in phenotypic and genotypic features. Utilizing standardized culture and drug-response assessment protocols, PTC drug-testing assays reveal 89% overall consistency in prospectively predicting clinical outcomes, with 98.1% accuracy distinguishing complete/partial response from progressive disease. Notably, PTCs enable accurate prediction of clinical outcomes for patients undergoing anti-PD1 therapy by combining cell viability and IFN-γ value assessments. These findings suggest that PTCs could serve as a valuable preclinical model for personalized medicine and basic research in NSCLC.

最近出现了许多源自患者的肿瘤模型。然而，它们指导个性化药物选择的潜力仍不清楚。在这里，我们报告了来自患者的非小细胞肺癌 (NSCLC) 肿瘤样细胞簇 (PTC)，能够在 10 天内进行 100-5,000 次药物测试。我们已经建立了 283 个 PTC 模型，成功率为 81%。 PTC包含由内源性基质细胞和免疫细胞自组装的原发肿瘤上皮，在表型和基因型特征上与原始肿瘤具有高度相似性。利用标准化培养和药物反应评估方案，PTC 药物测试分析显示，前瞻性预测临床结果的总体一致性为 89%，区分完全/部分反应与进展性疾病的准确度为 98.1%。值得注意的是，PTC 通过结合细胞活力和 IFN-γ 值评估，能够准确预测接受抗 PD1 治疗的患者的临床结果。这些发现表明 PTC 可以作为 NSCLC 个性化医疗和基础研究的有价值的临床前模型。