Han Tianxiao, Cheng Sida, Wang Xun, Qi QingYi, Chen Jinchuan, Wang Wenxiang, Zhou Jian, Li Yun, Chen Kezhong, Li Hao, Yang Fan
Department of Thoracic Surgery, Thoracic Oncology Institute, Peking University People's Hospital, Beijing, China.
Department of Radiology, Peking University People's Hospital, Beijing, China.
Thorac Cancer. 2025 May;16(9):e70073. doi: 10.1111/1759-7714.70073.
Metastatic lymph nodes (mLNs) exhibit different responses to neoadjuvant immunotherapy compared to the primary tumor (PT) in non-small cell lung cancer (NSCLC). Evaluating mLNs' response is crucial for predicting treatment efficacy and prognosis; however, such assessments are currently insufficient.
We enrolled 101 NSCLC patients with mLNs who underwent neoadjuvant chemoimmunotherapy followed by surgery. Survival outcomes and radiological and metabolic changes were analyzed across different lymph node pathological response groups, and a least absolute shrinkage and selection operator (LASSO) logistic regression model was developed to predict mLNs' response. RNA sequencing was performed to characterize the immune microenvironment of lymph nodes with different pathological responses.
Residual tumors in mLNs were significantly associated with worse recurrence-free survival (p = 0.003) and a trend toward worse overall survival (p = 0.087). Combining the pathological responses of mLNs and PTs improved prognostic stratification. Neither radiological size changes (AUC: 0.621) nor the SUVmax reduction rate (AUC: 0.645) were effective in distinguishing mLNs response. A model combining radiological and metabolic parameters demonstrated fair prediction efficacy (AUC: 0.85). In separate analyses of N1 and N2 nodes, radiological and metabolic changes of N1 mLNs partly reflected their pathologic response (AUC: 0.734; 0.816), unlike in N2 mLNs. RNA sequencing revealed that immune infiltration in responding lymph nodes differed from non-responding ones, with higher CD8+ T cells, NK T cells, B cells, and dendritic cells in the former.
The pathological response of mLNs provides additional prognostic information, but current tools are ineffective for detecting residual tumors. A model integrating radiological and metabolic parameters may offer better prediction.
在非小细胞肺癌（NSCLC）中，与原发性肿瘤（PT）相比，转移性淋巴结（mLN）对新辅助免疫治疗表现出不同的反应。评估mLN的反应对于预测治疗效果和预后至关重要；然而，目前此类评估尚不充分。
我们纳入了101例伴有mLN的NSCLC患者，这些患者接受了新辅助化疗免疫治疗后进行手术。分析了不同淋巴结病理反应组的生存结果以及放射学和代谢变化，并建立了最小绝对收缩和选择算子（LASSO）逻辑回归模型来预测mLN的反应。进行RNA测序以表征不同病理反应的淋巴结的免疫微环境。
mLN中的残留肿瘤与无复发生存期较差显著相关（p = 0.003），且总生存期有变差的趋势（p = 0.087）。结合mLN和PT的病理反应可改善预后分层。放射学大小变化（AUC：0.621）和SUVmax降低率（AUC：0.645）均不能有效区分mLN的反应。结合放射学和代谢参数的模型显示出较好的预测效能（AUC：0.85）。在对N1和N2淋巴结的单独分析中，N1 mLN的放射学和代谢变化部分反映了其病理反应（AUC：0.734；0.816），与N2 mLN不同。RNA测序显示，反应性淋巴结中的免疫浸润与无反应性淋巴结不同，前者的CD8 + T细胞、自然杀伤T细胞、B细胞和树突状细胞含量更高。
mLN的病理反应提供了额外的预后信息，但目前的工具在检测残留肿瘤方面无效。整合放射学和代谢参数的模型可能提供更好的预测。
