Department of Gastroenterology, the First Affiliated Hospital of Nanchang University, Nanchang, China.
Cancer Med. 2021 Mar;10(6):1925-1935. doi: 10.1002/cam4.3632. Epub 2021 Feb 27.
Current the surveillance and management are controversial for patients with IPMN. We aimed to develop an alternative nomogram to individualize IPMN prognosis and LNM.
Based on the data from SEER database of patients diagnosed with IPMN between 2004 and 2015, a nomogram predicting the survival and LNM of IPMN based on univariate and multivariate and Lasso regression analysis was performed, internally and externally validated, and measured by C-index, and decision curve analysis (DCA), and compared to the 7 TNM stage.
A total of 941 patients were included. Age, T stage examined nodes, tumor size, and pathology grade were identified as an independent factor for predicting LNM. The nomogram we established to predict LNM had a high predicting value with a C-index value of 0.735 and an AUC value of 0.753. Interestingly, including T1 stage, we found an inverse correlation was between age and LNM. In addition, nomogram for predicting CSS also performed better than TNM stage both in the internal validation group (1-year AUC:0.753 vs. 0.693, 3-year AUC: 0.801 vs. 0.731, 5-year AUC: 0.803 vs. 0.733) and external validation group (1-year AUC: 0.761 vs. 0.701, 3-year AUC: 0.772 vs. 0.713, 5-year AUC:0.811 vs. 0.735). DCA analysis showed the nomogram showed a greater benefit across the period of follow-up compared to 7 TNM stage.
A nomogram based on multivariate and Lasso regression analysis showed great clinical usability compared with current criteria. Also, for LNM of IPMN, younger age patients with IPMN should be attached more importance.
目前对于 IPMN 患者的监测和管理存在争议。我们旨在开发一种替代的列线图，以实现个体化的 IPMN 预后和 LNM 预测。
基于 2004 年至 2015 年 SEER 数据库中诊断为 IPMN 的患者数据，通过单变量和多变量以及 Lasso 回归分析，建立了一个预测 IPMN 生存和 LNM 的列线图，进行了内部和外部验证，并通过 C 指数、决策曲线分析（DCA）进行了测量，并与 7 个 TNM 分期进行了比较。
共纳入 941 例患者。年龄、检查的 T 分期、肿瘤大小和病理分级被确定为预测 LNM 的独立因素。我们建立的预测 LNM 的列线图具有较高的预测价值，C 指数值为 0.735，AUC 值为 0.753。有趣的是，包括 T1 期在内，我们发现年龄与 LNM 之间存在负相关。此外，预测 CSS 的列线图在内部验证组（1 年 AUC：0.753 比 0.693，3 年 AUC：0.801 比 0.731，5 年 AUC：0.803 比 0.733）和外部验证组（1 年 AUC：0.761 比 0.701，3 年 AUC：0.772 比 0.713，5 年 AUC：0.811 比 0.735）中均优于 7 个 TNM 分期。DCA 分析表明，与 7 个 TNM 分期相比，该列线图在整个随访期间具有更大的获益。
基于多变量和 Lasso 回归分析的列线图与当前标准相比具有更好的临床实用性。此外，对于 IPMN 的 LNM，应更加重视年龄较小的 IPMN 患者。
