Table_8_A new tumorgraft panel to accelerate precision medicine in prostate cancer.xlsx

BackgroundDespite the significant advances in the management of advanced prostate cancer (PCa), metastatic PCa is currently considered incurable. For further investigations in precision treatment, the development of preclinical models representing the complex prostate tumor heterogeneity are mandatory. Accordingly, we aimed to establish a resource of patient-derived xenograft (PDX) models that exemplify each phase of this multistage disease for accurate and rapid evaluation of candidate therapies.
MethodsFresh tumor samples along with normal corresponding tissues were obtained directly from patients at surgery. To ensure that the established models reproduce the main features of patient’s tumor, both PDX tumors at multiple passages and patient’s primary tumors, were processed for histological characteristics. STR profile analyses were also performed to confirm patient identity. Finally, the responses of the PDX models to androgen deprivation, PARP inhibitors and chemotherapy were also evaluated.
ResultsIn this study, we described the development and characterization of 5 new PDX models of PCa. Within this collection, hormone-naïve, androgen-sensitive and castration-resistant (CRPC) primary tumors as well as prostate carcinoma with neuroendocrine differentiation (CRPC-NE) were represented. Interestingly, the comprehensive genomic characterization of the models identified recurrent cancer driver alterations in androgen signaling, DNA repair and PI3K, among others. Results were supported by expression patterns highlighting new potential targets among gene drivers and the metabolic pathway. In addition, in vivo results showed heterogeneity of response to androgen deprivation and chemotherapy, like the responses of patients to these treatments. Importantly, the neuroendocrine model has been shown to be responsive to PARP inhibitor.
ConclusionWe have developed a biobank of 5 PDX models from hormone-naïve, androgen-sensitive to CRPC primary tumors and CRPC-NE. Increased copy-number alterations and accumulation of mutations within cancer driver genes as well as the metabolism shift are consistent with the increased resistance mechanisms to treatment. The pharmacological characterization suggested that the CRPC-NE could benefit from the PARP inhibitor treatment. Given the difficulties in developing such models, this relevant panel of PDX models of PCa will provide the scientific community with an additional resource for the further development of PDAC research.

背景 尽管晚期前列腺癌（prostate cancer, PCa）的诊疗已取得长足进步，但转移性前列腺癌目前仍被视为不可治愈的疾病。为推进精准治疗领域的深入研究，构建能够模拟前列腺肿瘤复杂异质性的临床前模型是必要前提。据此，本研究旨在建立一套患者来源异种移植（patient-derived xenograft, PDX）模型资源，以覆盖该多阶段疾病的全部病程阶段，从而实现对候选治疗方案的精准、快速评估。

方法 本研究直接从手术患者体内获取新鲜肿瘤样本及配对正常组织。为确保所构建的模型能够重现患者肿瘤的核心特征，我们对多代传代的PDX肿瘤及患者原发肿瘤均进行了组织学特征分析。此外，还通过短串联重复序列（short tandem repeat, STR）分型分析确认了患者身份。最终，本研究还评估了PDX模型对雄激素剥夺治疗、PARP抑制剂及化疗的响应情况。

结果 本研究成功构建并鉴定了5株全新的前列腺癌PDX模型。该模型库涵盖了激素初治、雄激素敏感及去势抵抗性前列腺癌（castration-resistant prostate cancer, CRPC）原发肿瘤，以及伴神经内分泌分化的去势抵抗性前列腺癌（CRPC-NE）。值得注意的是，对这些模型的全基因组特征分析显示，其存在雄激素信号通路、DNA损伤修复通路及PI3K通路等常见的癌症驱动基因变异。转录组表达谱分析进一步验证了上述结果，并筛选出一批潜在的基因驱动靶点及代谢通路相关靶点。此外，体内实验结果显示，PDX模型对雄激素剥夺治疗及化疗的响应存在异质性，这与患者接受此类治疗后的响应模式一致。尤为重要的是，该神经内分泌分化型PDX模型被证实对PARP抑制剂具有响应性。

结论 本研究成功构建了包含5株PDX模型的生物样本库，覆盖了从激素初治、雄激素敏感型到去势抵抗性原发前列腺癌及CRPC-NE的各个病程阶段。模型中存在的拷贝数变异增加、癌症驱动基因突变累积以及代谢重编程现象，与临床治疗耐药机制的增强相一致。药理学特征分析结果显示，CRPC-NE模型可从PARP抑制剂治疗中获益。鉴于此类模型构建难度较大，这套具有代表性的前列腺癌PDX模型将为科研界提供额外的研究资源，以推动胰腺导管腺癌（PDAC）相关研究的进一步发展。