2026年世界肺癌大会（WCLC 2026）于9月12—15日在韩国首尔举行，本届大会以“Science Without Boundaries: Uniting the World Against Thoracic Cancer”为主题，汇聚全球胸部肿瘤领域专家学者，共探肺癌诊疗前沿进展。
编者按：2026年世界肺癌大会（WCLC 2026）于9月12—15日在韩国首尔举行，本届大会以“Science Without Boundaries: Uniting the World Against Thoracic Cancer”为主题，汇聚全球胸部肿瘤领域专家学者，共探肺癌诊疗前沿进展。
9月15日，在大会“Highlights of the Day”环节中，新加坡国家癌症中心Aaron C. Tan教授对本次大会流行病学、危险因素、预防与烟草控制及病理与生物标志物领域口头报告亮点进行了系统总结。《肿瘤瞭望》特邀Aaron C. Tan教授接受专访，围绕肺癌高危人群再定义、筛查与预防策略、戒烟整合挑战，以及人工智能（AI）在肺癌病理中的应用前景展开探讨。
《肿瘤瞭望》：您在Highlights of the Day环节中，分享了关于流行病学、危险因素、预防与烟草控制及病理学与生物标志物方向的口头报告亮点内容，首先请您为我们介绍下相关亮点内容及临床意义？
Aaron C. Tan教授：在“Highlights of the Day”环节中，我汇总了OA07和OA11两个专场的9项重要研究进展。OA07聚焦肺癌预防新前沿，涵盖空气污染、戒烟及实施科学；OA11聚焦循环和AI驱动生物标志物。
整体来看，这些研究围绕三条主线展开：其一，空气污染等新型风险因素与肺癌及胸膜间皮瘤风险、预后之间的关系；其二，女性肺癌流行病学转变及戒烟实施策略与药物干预；其三，循环肿瘤DNA（ctDNA）多组学、循环蛋白及AI病理模型在新辅助治疗反应与毒性预测中的应用。上述研究为重新定义高危人群、优化筛查与预防路径，以及将循环生物标志物和AI工具引入临床决策提供了新兴证据。正如我在报告中所强调，对流行病学和生物学认识的不断演变，将指导肺癌预防和筛查策略。
ONCOLOGY FRONTIER：In today’s Highlights of the Day session, you shared key oral abstract highlights in epidemiology, risk factors, prevention and tobacco control, and pathology and biomarkers. Could you first introduce these highlights and their clinical significance?
Prof. Aaron C. Tan: I presented the Highlights of the Day session to summarize nine important abstracts across two oral abstract sessions: OA07 on new horizons in lung cancer prevention, smoking cessation, air pollution and implementation science, and OA11 on circulating and AI-driven biomarkers in lung cancer. These covered epidemiology, risk factors, prevention and tobacco control, and pathology/biomarkers. Importantly, many explored epidemiological shifts, circulating biomarkers for risk stratification and molecular residual disease, and AI/pathology-based approaches for predicting response or toxicity to neoadjuvant therapy. So these studies provide emerging evidence for new risk factors and new technologies that we can potentially apply in the clinic.
《肿瘤瞭望》：本次大会中，空气污染、女性肺癌流行病学转变等研究为重新定义肺癌高危人群提供了哪些新证据？这些发现可能如何影响未来筛查策略与预防政策？
Aaron C. Tan教授：首先，UK Biobank空气污染暴露与胸膜间皮瘤风险的前瞻性分析研究（摘要号：OA07.01）纳入392,286名参与者，中位随访14.4年，结果显示细颗粒物（PM2.5）与胸膜间皮瘤风险显著相关：Q4 vs Q1的风险比（hazard ratio，HR）为1.55（95%CI 1.16 - 2.07），p=0.003；每四分位HR为1.12（95% CI 1.03 - 1.23），p=0.012；每1-SD HR为1.10（95% CI 1.00 - 1.22），p=0.048。PM2.5是最强且最一致的关联。蛋白质组学发现96种蛋白与所有污染物相关，例如WFDC2；肿瘤细胞分析提示免疫相关通路参与其中。该研究的局限性包括未测量石棉暴露、病例数有限和未建模竞争死亡。
第二，国际肺癌联盟（ILCCo）环境空气污染暴露与肺癌预后汇总与比较分析研究（摘要号：OA07.02）纳入6,261例患者，显示PM2.5和二氧化氮NO2）显著增加肺癌患者死亡风险；PM2.5关系总体呈非线性，且在非吸烟者中效应更大，非吸烟者5年总生存（OS）汇总Kaplan-Meier曲线差异具有统计学意义（p=0.048）。
此外，中东和北非（MENA）地区女性肺癌流行病学转变：基于GLOBOCAN的纵向研究（2012 - 2022）（摘要号：OA07.04）发现，MENA地区17国女性肺癌年龄标准化发病率（ASIR）相对增加46.6%，显著高于全球的19.1%；研究者讨论指出，这一趋势不能仅用吸烟解释，其他因素包括二手烟、水烟、家庭固体燃料、氡和环境空气污染。
这些证据提示，未来筛查和预防策略除继续强调戒烟外，还需关注空气污染等非吸烟相关风险因素，并探索针对新风险人群的筛查路径。但必须强调，戒烟仍是降低肺癌风险最重要手段。
ONCOLOGY FRONTIER: Based on the relevant research advances and discussions at this congress, what new evidence do these studies provide for redefining high-risk populations for lung cancer? How might this influence future lung cancer screening strategies and prevention policies?
Prof. Aaron C. Tan: There were abstracts that looked at air pollution—both its role in lung cancer prognosis [OA07.02] and its potential role in mesothelioma risk [OA07.01]. There was also an abstract on shifting epidemiological trends of lung cancer in women [OA07.04]. These suggest that in the future, screening and prevention strategies may focus on alternative methods apart from smoking cessation, which still remains the most important method to reduce lung cancer risk. But there may be other risk factors we need to control.
《肿瘤瞭望》：人工智能AI在病理学领域的应用正从学术探索走向临床实践。请您结合相关研究进展及实践经验，谈谈您认为AI在肺癌病理学领域最具临床价值的应用场景有哪些？
Aaron C. Tan教授：目前AI与机器学习模型仍需更多数据支持，尤其前瞻性研究。但新兴数据提示，其在数字病理中潜力较大。
首先，T-PRISM——苏木精-伊红（H&E）衍生T细胞状态深度学习预测非小细胞肺癌（NSCLC）新辅助免疫治疗反应研究（摘要号：OA11.03）利用H&E全切片图像（WSI）衍生T细胞状态，通过深度学习预测NSCLC新辅助免疫治疗反应。当前生物标志物如程序性死亡配体1（PD-L1）和肿瘤突变负荷（TMB）预测不完整，且依赖检测和采样。T-PRISM在GDPH OOF队列（n=267）中受试者工作特征曲线下面积（AUROC）为0.661，优于ABMIL的0.606；外部队列（n=452）中AUROC为0.613，优于ABMIL的0.569；与PD-L1 TPS相比也显示数值优势。
其次，可解释机器学习模型预测NSCLC新辅助化免疫治疗高级别不良事件的多中心研究（摘要号：OA11.04）构建可解释机器学习模型，预测NSCLC新辅助化免疫治疗高级别治疗相关不良事件（TRAEs）。该研究纳入1,096例患者，45.3%发生≥3级TRAEs；最终模型基于十个常规变量，验证集AUROC为0.728（95% CI 0.644 - 0.808），准确率67.8%。该模型旨在支持风险适应监测，而非强制化疗剂量减少。这些工具有望在未来辅助治疗前识别可能获益或高风险患者，但仍需进一步前瞻性验证。
ONCOLOGY FRONTIER: Based on relevant research advances and your practical experience, what do you consider the most clinically valuable application scenarios of AI in lung cancer pathology?
Prof. Aaron C. Tan: At the moment, we are still waiting for further data on the potential role of AI and many machine-learning models. However, much of the emerging data suggests a potential role, particularly in digital pathology. There was an abstract [OA11.03] that looked at the use of machine-learning models from H&E-stained whole-slide images to potentially predict whether patients will respond to neoadjuvant immunotherapy. There was also a related abstract [OA11.04] using an explainable machine-learning model to predict high-grade adverse events in neoadjuvant chemo-immunotherapy. So this is a potential role for AI; however, we await further studies, in particular future prospective studies.
《肿瘤瞭望》：您认为将戒烟干预整合到肺癌筛查路径中，在临床实践中面临的主要挑战是什么？
Aaron C. Tan教授：本次大会中，一项将戒烟支持嵌入肺癌筛查实施策略的国际专家调查研究（摘要号：OA07.03）纳入23个国家/地区的181名专家，评估16项实施策略，识别出可行且有效的“Go-zones”，包括：在肺癌筛查（LCS）临床工作流中添加戒烟支持提醒或提示；改善吸烟史记录及其在LCS服务中的使用；与合格LCS参与者共同设计或修改戒烟患者教育材料。
其次，RISP试验——金雀花碱用于老年人戒烟研究（摘要号：OA07.05）评估金雀花碱（cytisine）在老年人戒烟中的效果。结果显示，金雀花碱组戒烟率为50.1%，显著高于单纯咨询组的8.1%，调整比值比（OR）为14.53（95% CI 11.67 - 18.08）；在≥65岁人群中，金雀花碱组戒烟率为49.1%，单纯咨询组为8.2%，调整OR为12.65（95% CI 8.78 - 18.24）；未发现新的安全信号。但当前欧洲产品标签通常不推荐65岁以上成人使用cytisine，需特别关注肾功能、心血管疾病、精神病史和多重用药。
总体而言，戒烟整合的主要挑战仍在于不同干预措施的可行性和有效性，以及如何在广泛LCS人群中落地实施。
ONCOLOGY FRONTIER: What do you see as the main challenges in integrating smoking cessation into lung cancer screening pathways in clinical practice?
Prof. Aaron C. Tan: There were a couple of abstracts that explored the potential role of different implementation strategies for smoking cessation in lung cancer screening studies [OA07.03], as well as the potential role of other pharmacological interventions such as cytisine [OA07.05]. The main issue with smoking cessation for participants in lung cancer screening is really the feasibility and effectiveness of different interventions. There is a wide range of potential interventions ranging from pharmacological to counseling. So it is really the implementation and feasibility of implementing these across a wide population of participants undergoing lung cancer screening that has led to the difficulty in implementation.
《肿瘤瞭望》：请您分享下本届WCLC您的参会感受及让您个人印象最深刻的研究或方向？
Aaron C. Tan教授：WCLC为全球科学和临床社区提供了交流最新数据的平台。令我印象最深的是大量新化合物和新疗法正在开发，许多早期试验数据在肺癌全程管理和不同分子亚型中非常有前景，正在改变未来治疗选择。
与此同时，新辅助治疗虽然具有早期持久控制、扩增抗肿瘤免疫、降低微转移复发风险、提供预康复窗口、实现手术降级和指导辅助治疗等潜在优势，但仍面临手术延迟、新辅助治疗进展、治疗相关不良事件、检查点敏感性生物标志物缺乏、哪些患者可从额外辅助治疗获益不明确，以及I-II期肺癌领域的临床试验需求等挑战。因此，增强患者选择仍是改善新辅助治疗疗效和耐受性的关键。总体而言，肺癌研究正处于令人振奋的发展阶段。
ONCOLOGY FRONTIER: Could you share your overall impressions of this year’s WCLC and the study or direction that impressed you most personally?
Prof. Aaron C. Tan: I think this WCLC has been an amazing opportunity for the entire scientific and clinical community around lung cancer to come together to discuss the latest data and research being conducted. What impressed me most is really the large number of new compounds and new therapies being developed. Many of the early-phase trial data are very promising across the continuum of lung cancer care and across different molecular subtypes of lung cancer. It is really changing the way we may potentially practice in terms of treatment options for our patients. So I think it is a really exciting time to be involved in lung cancer research.
专家简介
Aaron Tan，MBBS，PhD，FRACP
助理教授；新加坡国家癌症中心（NCCS）肿瘤内科医生。
Aaron C. Tan 现任NCCS肿瘤内科医生，在悉尼皇家北岸医院（Royal North Shore Hospital）完成肿瘤内科专科培训，并在悉尼大学（The University of Sydney）获癌症药物发现方向博士学位。此后，他在NCCS完成药物开发/早期临床试验及肺癌转化研究方向的专科研修。其临床与研究方向包括早期和晚期临床试验、肺癌转化研究及亚洲区域学术研究，目前持续积极参与上述领域工作。