肿瘤治疗方案多决策节点的个体化多级预后预测标志识别策略

The main causes of recurrence and mortality after complete surgical resection for early and middle stage cancer patients are occult metastases and resistance to chemotherapy and targeted therapies. As there are differences in the basic drug resistance of cancer patients treated with a drug that are not related to the specific drug, it is necessary to correct for this confounder to identify specific predictive biomarkers for drugs. The aim of this project is to develop an analytical method for identifying multi-level prognostic/predictive biomarkers for multiple decision points of TNM in personalized treatment of lung adenocarcinoma (LUAD). (1) To estimate the robustness of the individualized prognostic (micrometastasis) biomarker (first level biomarker) for LUAD with early and middle stage, and identify the patients with micrometastasis for developing a predictive biomarker (second level biomarker) of platinum resistance. (2) Applying the first and second level biomarkers to LUAD cell lines, we select the cell lines which can represent the tissues with micrometastasis and resistance to platinum based chemotherapy. In the selected cell lines, we try to identify a specific predictive biomarker (third level biomarker) for EGFR inhibitors resistance after normalizing the difference in basic resistance among cancer cells (the median IC50 of a group of drugs), and validate it in tissue gene expression datasets. (3) To identify the multi-omics characters of patients who are resistant to platinum based chemotherapy and EGFR inhibitors, respectively, and find potential targeted genes to enhance their sensibility to corresponding drugs, which will be tested in the experiment. In addition, we would apply the method to lung squamous carcinoma for identifying their multi-level prognostic/predictive biomarkers in expand researches.

病理诊断为无远端转移的早中期肿瘤患者的微转移及其对化疗与靶向治疗的耐药是导致患者术后复发的主要原因。对接受药物治疗的患者，癌细胞存在与特定药物无关的基础耐药性差别，需排除其影响以识别药物特异的耐药标志。本课题拟发展识别个体化多级预后预测标志策略，针对肺腺癌按TNM分期的治疗方案的多决策节点：（1）评估我们已报道的早中期肺腺癌术后复发（微转移）风险标志（1级标志）的稳健性，再在由该标志识别的微转移患者中识别铂类耐药预测标志（2级标志）；（2）耦联应用1、2级标志筛选能代表微转移且对铂类耐药癌组织的肺腺癌细胞，标化各细胞的基础耐药水平（对已有抗癌药物的IC50中值）后，再识别EGFR抑制剂的特异性耐药标志（3级标志），并通过临床组织样本验证；（3）识别对铂类、EGFR抑制剂耐药癌细胞的高频多维组学特征，分别筛选逆转耐药的候选分子靶标并进行细胞系实验验证。最后，拟将该分析策略应用于肺鳞癌等亚型。

肺腺癌的发病率逐步上升，已成为肺癌最主要的亚型。病理诊断为无远端转移的早中期肿瘤患者的微转移及其对辅助药物治疗的耐药是导致患者术后复发的主要原因。本课题以肺腺癌为主要研究对象，发展识别肺腺癌的个体化多级预后预测标志的分析策略。. 首先，针对肺癌患者的治疗策略逐渐转向以组织学类型为指导的靶向治疗模式，而传统的病理学诊断技术鉴别低分化肺癌组织学亚型存在困难的问题，我们基于样本内基因表达水平的相对高低秩序关系（relative expression orderings, REOs）的定性特征，开发一个稳健的鉴别肺腺癌和鳞癌的个体化基因对标志（AGR2-KRT5）。该标志在收集的独立的冷冻样本数据以及临床诊断困难样本（如石蜡包埋、小活检等）中得到了有效验证。针对标志分类与临床组织学诊断不一致的样本，我们采用生存分析及亚型特异基因的差异表达分析，支持标志分类的准确性。与原始病理学诊断相比，基于REOs识别的标志能更准确地区分肺腺癌和肺鳞癌亚型，为肺癌患者个性化治疗提供可靠的依据。. 接下来，针对早期肺腺癌患者术后复发、转移的问题，我们开发肺腺癌个体化预后标志。在研究中我们发现，在基于不同参考基因组和基因注释处理的RNA测序数据中，基因表达值的差异可能导致转录组标志对患者亚型的预测产生变化。针对该问题，我们选择在不同基因注释版本里稳定的基因，基于该类基因的REOs，开发一个包含40对基因的预后标志，并在独立数据集中得到验证。该标志在临床收集的30例I期肺腺癌的石蜡包埋样本中得到了验证。基于REOs开发的定性预后标志，能够个体化识别出术后需接受辅助治疗的早期微转移患者。该类标志识别策略已成功应用于卵巢癌、结肠癌等多个癌型。.最后，针对（微）转移肺腺癌患者，我们基于基因组数据，识别一组包含106个突变编码序列的突变标志，能够识别出可接受免疫治疗的敏感患者，提高（微）转移患者的预后。. 综上所述，该项目已经识别了肺腺癌个体化多级预后预测标志，能够为患者临床个体化治疗方案的制定提供帮助。已发表SCI论文11篇。