Predictive models for biological aggressiveness of pancreatobiliary cancers by integrative molecular analysis of their precursor lesions

胰胆管癌是一类生物侵袭性肿瘤，预后较差。近年来，已经定义了分子亚群，并开始研究PB癌的基因组和转录组特征。由于已知的易感因素数量有限，早期发现似乎是改善预后的最有效方法，而初级预防措施的相关性有限。因此，全面表征PB前体的遗传景观，一方面对于揭示其向癌症的分子进化，另一方面对于确定疾病发展早期生物学行为的预测因素具有重要意义。在这个项目中，我们的目标是通过分离已知的和假定的胰腺和胆管癌的前驱病变，并描述它们从发展到侵袭的特征，来破译胰腺癌和胆管癌异质生物学行为的分子基础。利用已建立的和新开发的基于人体组织的方法，胰腺癌和胆管癌将被用作发展肿瘤发生概念的模型，这将允许更好的早期发现和靶向治疗方法。为此，临床和形态学特征良好的人体组织群将用于从不同的正常组织区室和前体病变（侵袭前病变）中产生纯细胞群，然后对其进行基因组和转录组学表征。从不同患者的前体病变中获得的结果将相互比较，并与伴随的侵袭性肿瘤及其淋巴结和血源性转移进行配对，以确定癌症前体的风险概况，并使用生物信息学方法预测给定肿瘤的生物学行为。来自人体组织的类器官被用于基因组编辑和基因敲除实验，以评估预测的驱动事件的功能意义。最后，采用组织和液体为基础的方法，在回顾性和前瞻性系列中确定已识别的相关分子标记物在体内的价值。

Pancreatobiliary (PB) cancers are a group of biologically aggressive neoplasms with a poor prognosis. In recent years, molecular subgroups have been defined and work has begun to characterize the genomic and transcriptomic landscape of PB carcinomas. Due to the limited number of known predisposing factors, early detection seems to be the most effective approach to improve prognosis, whereas primary prevention measures show limited relevance. Therefore, a comprehensive characterization of the genetic landscape of PB precursors is of great importance in order to uncover their molecular evolution to cancer on the one hand and to identify predictors of biological behavior in the early stages of disease development on the other hand. In this project, our goal is to decipher the molecular basis of the heterogeneous biological behavior of pancreatic and biliary duct cancer by isolating their known and putative precursor lesions and characterizing them from their development to invasion. Using established and newly developed methods based on human tissue, pancreatic and bile duct cancer will be used as a model to develop a concept of tumorigenesis that will allow better early detection and targeted therapeutic approaches. For this purpose, clinically and morphologically well characterized human tissue collectives will be used to generate pure cell populations from different normal tissue compartments and precursor lesions (preinvasive lesions), which will then be subjected to genomic and transcriptomic characterization. The results obtained from precursor lesions from different patients will be compared with each other and as matched pairs with the accompanying invasive tumors and their lymph nodes and blood-borne metastases to identify risk profiles of cancer precursors that predict the biological behavior of a given neoplasm using a bioinformatics approach. Organoids from human tissue are used for genomediting and gene knock-down experiments to assess the functional significance of the predicted driver events. Finally, tissue and fluid based approaches are used to determine the value of identified relevant molecular markers in vivo in retrospective and prospective series.