A nomogram to predict the survival of stage IIIA-N2 non-small cell lung cancer after surgery
A nomogram to predict the survival of stage IIIA-N2 non-small cell lung cancer after surgery
复制标题
DOI：
10.1016/j.jtcvs.2017.11.098
复制
发表时间：
2018-04-01
影响因子：
6
通讯作者：
Xu, Lin
中科院分区：
文献类型：
--
作者：
Mao, Qixing;Xia, Wenjie;Xu, Lin
目的：IIIA-N2期非小细胞肺癌(NSCLC)患者的术后生存率具有高度的异质性。在这里，我们的目的是识别与术后生存相关的变量，并开发一种生存预测工具。方法：对2004年1月至2009年12月的监测、流行病学和最终结果数据库进行回顾。采用后向逐步回归法筛选出显著变量。采用多变量COX回归方法构建诺模图。通过协调度指数和校准曲线对模型的性能进行了评价。该模型通过江苏省肿瘤医院肺癌中心的独立队列进行了验证。结果：共有1809名接受手术的IIIA-N2期非小细胞肺癌患者进入训练队列。采用多因素逐步回归分析，年龄、性别、分级、组织学、肿瘤大小、内脏胸膜侵犯、阳性淋巴结、淋巴结检查和手术类型(肺叶切除与全肺切除)是影响预后的重要因素。从训练队列中开发了一个诺模图，并使用一个独立的中国队列进行了验证。模型的一致性指数训练队列为0.673(95%可信区间，0.654~0.692)，验证队列为0.664(95%可信区间，0.614~0.714)。校正图显示诺模图预测存活和观察存活之间的最佳一致性。生存分析显示不同亚组间预后评分有显著差异。结论：该图提供了IIIA-N2期非小细胞肺癌患者术后的个体化生存预测，有助于患者和临床医生的生存咨询，临床试验设计和随访，以及术后策略的制定。
Objective: Postoperative survival of patients with stage IIIA-N2 non-small cell lung cancer (NSCLC) is highly heterogeneous. Here, we aimed to identify variables associated with postoperative survival and develop a tool for survival prediction.Methods: A retrospective review was performed in the Surveillance, Epidemiology, and End Results database from January 2004 to December 2009. Significant variables were selected by use of the backward stepwise method. The nomogram was constructed with multivariable Cox regression. The model's performance was evaluated by concordance index and calibration curve. The model was validated via an independent cohort from the Jiangsu Cancer Hospital Lung Cancer Center.Results: A total of 1809 patients with stage IIIA-N2 NSCLC who underwent surgery were included in the training cohort. Age, sex, grade, histology, tumor size, visceral pleural invasion, positive lymph nodes, lymph nodes examined, and surgery type (lobectomy vs pneumonectomy) were identified as significant prognostic variables using backward stepwise method. A nomogram was developed from the training cohort and validated using an independent Chinese cohort. The concordance index of the model was 0.673 (95% confidence interval, 0.654-0.692) in training cohort and 0.664 in validation cohort (95% confidence interval, 0.614-0.714). The calibration plot showed optimal consistency between nomogram predicted survival and observed survival. Survival analyses demonstrated significant differences between different subgroups stratified by prognostic scores.Conclusions: This nomogram provided the individual survival prediction for patients with stage IIIA-N2 NSCLC after surgery, which might benefit survival counseling for patients and clinicians, clinical trial design and follow-up, as well as postoperative strategy-making.