术后淋巴结阴性肺浸润性黏液腺癌患者的预后预测模型

Construction of a Prognostic Prediction Model of Patients with Pathologic N0 in Resected Invasive Mucinous Adenocarcinoma of the Lung

背景与目的肺浸润性黏液腺癌(invasive mucinous adenocarcinoma of the lung,IMA)是肺腺癌中一种少见且特殊的类型,该类肿瘤的特点往往是少有淋巴结转移,因此对于该类肿瘤的预后评估依靠现有的肿瘤原发灶-淋巴结-转移(tumor-node-metastasis,TNM)分期存在困难。本研究的目的是构建列线图来预测术后淋巴结阴性的IMA患者的预后。方法根据纳入标准和排除标准,回顾性分析2012年7月至2017年5月宁波大学附属李惠利医院(训练队列,n=78)和宁波市第二医院(验证队列,n=66)胸外科收治的术后病理为淋巴结阴性的IMA患者的资料,分析训练队列的临床病理特征的预后价值并建立预后预测模型,并对模型性能进行评价,最后将验证队列的数据代入进行外部验证。结果单因素分析显示肺炎型、较大的肿块、包含黏液和非黏液成分的混合型、较高的总分期是5年无进展生存期(progression-free survival,PFS)及总生存期(overall survival,OS)的影响因素。多因素分析进一步表明,影像学分型、肿块大小、黏液成分是5年PFS及OS的独立预后因素。5年PFS率和OS率分别为62.82%和75.64%,亚组的生存分析显示,肺炎型和包含黏液和非黏液成分的混合型IMA患者的5年PFS及OS分别明显低于孤立型和纯黏液型IMA患者。5年PFS和OS的Harrell’s C指数分别为0.815(95%CI:0.741-0.889)和0.767(95%CI:0.669-0.865),这两个模型的校准曲线及决策曲线分析(decision curve analysis,DCA)在两个队列中显示出良好的预测性能。结论本次基于临床病理特征构建的列线图在一定程度上可以作为IMA切除术后淋巴结阴性患者的一种有效预后预测工具。

Background and objective Invasive mucinous adenocarcinoma(IMA)was a rare and specific type of lung adenocarcinoma,which was often characterized by fewer lymphatic metastases.Therefore,it was difficult to evaluate the prognosis of these tumors based on the existing tumor-node-metastasis(TNM)staging.So,this study aimed to develop Nomograms to predict outcomes of patients with pathologic N0 in resected IMA.Methods According to the inclusion criteria and exclusion criteria,IMA patients with pathologic N0 in The Affiliated Lihuili Hospital of Ningbo University(training cohort,n=78)and Ningbo No.2 Hospital(validation cohort,n=66)were reviewed between July 2012 and May 2017.The prognostic value of the clinicopathological features in the training cohort was analyzed and prognostic prediction models were established,and the performances of models were evaluated.Finally,the validation cohort data was put in for external validation.Results Univariate analysis showed that pneumonic type,larger tumor size,mixed mucinous/non-mucinous component,and higher overall stage were significant influence factors of 5-year progression-free survival(PFS)and overall survival(OS).Multivariate analysis further indicated that type of imaging,tumor size,mucinous component were the independent prognostic factors for poor 5-year PFS and OS.Moreover,the 5-year PFS and OS rates were 62.82%and 75.64%,respectively.In subgroups,the survival analysis also showed that the pneumonic type and mixed mucinous/non-mucinous patients had significantly poorer 5-year PFS and OS compared with solitary type and pure mucinous patients,respectively.The C-index of Nomograms with 5-year PFS and OS were 0.815(95%CI:0.741-0.889)and 0.767(95%CI:0.669-0.865).The calibration curve and decision curve analysis(DCA)of both models showed good predictive performances in both cohorts.Conclusion The Nomograms based on clinicopathological characteristics in a certain extent,can be used as an effective prognostic tool for patients with pathologic N0 after IMA resection.