高发人群食管鳞癌癌前病变进展的风险预测与分级研究

Early detection and treatment of Esophageal Squamous Cell Carcinoma (ESCC) has been widely accepted as an appropriate strategy for disease control. However, the currently adopted endoscopic surveillance, which uses fixed interval according to the pathological diagnoses of precancerous lesions, would lead to the under-diagnosis of fast progressors and the over-diagnosis of slow progressors or non-progressors. Use a risk prediction model to distinguish high-risk individuals would be an efficient strategy for large scale ESCC screening and endoscopic surveillance programs. This risk stratification plays a key role in increasing the cost-effectiveness of endoscopic surveillance on both the population and individual level, and requires high-quality data from large-scale prospective population-based cohort studies to insure the robustness of performance. .In this study, we aim to establish risk stratification tools for ESCC precancerous lesion progression as well as the respective surveillance strategies using our prospective esophageal cancer cohorts (25,000 endoscopies, 1000 patients with precancerous lesions, up to 10 years’ follow-up) at a high-prevalence region in China. Systematic detection of potential biomarkers for ESCC precancerous lesion progression will be adopted. All potential contributing factors collected (including environmental and behavior factors, biomarkers and their interaction terms) will be evaluated and filtered according to the endoscopic surveillance and longitudinal follow-up data, using statistical approach established in our cross-sectional ESCC prediction model building. To insure the robustness and generalization of our models, validation processes using external independent cohorts will be performed. .Results from this study will shed light to a tailored ESCC precancerous lesion surveillance strategy, and promote the precise risk management and prevention of ESCC in China at both population and individual level.

内镜早筛是食管癌防治重要策略。目前，对筛查检出的癌前病变仅按照病理级别给予统一随访建议，未能实现癌前病变患者临床处置与管理策略的精准化制定。基于大规模人群数据，前瞻性构建具有良好外推能力的“预测模型”，对患者进展为癌的风险进行定量评估，是减少漏诊及过度检查、提高筛检及早期临床干预成本-效果的重要途径之一。依托本课题组在河南安阳高发区建立的食管癌专病队列（2.5万人筛查，1000例癌前病变，最长10年随访），本研究拟系统挖掘环境、行为暴露数据并对相关分子标志物进行深度检测与筛选，联合复查与纵向随访获得的发病数据，前瞻性构建“食管癌前病变进展风险预测模型”，并在人群队列内部及基于医院构建的门诊患者队列中进行模型验证与修订。最终建立科学可靠、简单易行的食管癌前病变进展风险分级标准与临床处置策略。本研究的成功开展将明确推进我国食管鳞癌防治工作的精准化与个体化进程，具有重要的公共卫生及临床应用价值。

内镜早筛是食管癌防治重要策略。目前，对筛查检出的癌前病变仅按照病理级别给予统一随访建议，未能实现癌前病变患者临床处置与管理策略的精准化制定。基于大规模人群数据，前瞻性构建具有良好外推能力的“预测模型”，对患者进展为癌的风险进行定量评估，是减少漏诊及过度检查、提高筛检及早期临床干预成本-效果的重要途径之一。.依托本课题组在河南安阳高发区建立的食管癌专病队列（2.5万人筛查，1000例癌前病变，最长10年随访），本研究报告了人群研究平台的基线入组情况及工作绩效；利用大规模人群筛查数据评估了碘染色内镜技术对于食管早癌及癌前病变诊断的灵敏度；利用随访数据评估了内镜筛查食管癌的效果；系统描述了血脂指标在高发区人群中的分布，探索了血脂与食管癌风险之间的关联；最终构建和验证了食管病变进展预测模型，建立了食管癌前病变进展风险分级方案。.本研究提出利用医保数据建立有效的肿瘤发病监测系统，评估显示其报告肿瘤发病的绩效显著优于主动随访，且工作量更小；依托人群研究平台评估碘染色内镜对于食管早癌及癌前病变检出的灵敏度，发现碘染对于早期病变存在较高的漏诊可能；证实食管碘染内镜筛查可有效降低人群食管癌发病及死亡率；首次检验食管癌高发区人群血脂异常分布的性别差异，发现食管癌家族史对于血脂与食管癌关联的效应修饰作用；按计划成功构建和验证了食管病变进展预测模型，建立了食管癌前病变进展风险分级方案，在业内首次提出碘不染色特征对于食管病变监测的重要作用。 .本研究基于课题组在食管癌高发区开展的大规模人群筛查工作，利用宝贵的人群流调数据和生物样品资源，系统探索并回答了筛查是否有效，如何筛查更高效以及筛查检出的癌前病变如何进行内镜监测等一系列瓶颈问题，将对我国食管癌筛查策略产生深远影响。.