Sun Yat-Sen University Cancer Center, 651 Dongfeng Road East, Guangzhou, 510060, People's Republic of China.
State Key Laboratory of Oncology in South China, 651 Dongfeng Road East, Guangzhou, 510060, People's Republic of China.
Eur Radiol. 2023 May;33(5):3682-3692. doi: 10.1007/s00330-023-09431-4. Epub 2023 Feb 3.
This study focused on developing and validating a nomogram to predict the overall survival (OS) of patients with nasopharyngeal carcinoma (NPC) without distant metastasis based on their clinical characteristics, serum biomarkers, and presence of nasopharyngeal (NP) necrosis.
This study included 9298 patients with NPC. Patients from January 2009 to December 2014 were randomly categorized into the training cohort and validation cohort A. Validation cohort B, whose data were collected from January 2015 to December 2017, was also included. OS was the primary endpoint of this study. Cox regression analysis was used to detect independent risk variables. Decision curve analysis, calibration curve, time-dependent receiver operating characteristic (ROC) curve, and concordance index (C-index) were used to evaluate the performance of the nomogram model.
A total of 267 patients developed NP necrosis after the first routine radiotherapy. After radiotherapy, patients with NP necrosis had significantly lower OS than other patients in all three cohorts (p < 0.001). Eleven factors, including NP necrosis, were involved in the nomogram, which had favorable discrimination and calibration with a C-index of 0.768 in the training cohort, 0.749 in validation cohort A, and 0.739 in validation cohort B. The nomogram exhibited a significantly larger area under the ROC curve for predicting OS than the TNM stage and Epstein-Barr virus (EBV) DNA (p < 0.001).
Compared with the TNM system and EBV DNA, we established a nomogram model with an accurate prognostic prediction for patients with NPC, which might help with patient management in NPC.
• This study included 9298 patients with NPC, and 11 factors were involved in the final model. • The nomogram had a significantly higher C-index and area under the ROC curve than the TNM stage and EBV DNA. • We established the first nomogram model for NPC involving the occurrence of NP necrosis, which was valuable for providing individual counseling and clinical assessments.
本研究旨在建立并验证一个列线图，用于预测无远处转移的鼻咽癌（NPC）患者的总生存期（OS），该列线图基于患者的临床特征、血清生物标志物和鼻咽（NP）坏死情况。
本研究纳入了 9298 例 NPC 患者。2009 年 1 月至 2014 年 12 月的患者被随机分为训练队列和验证队列 A。此外，还纳入了 2015 年 1 月至 2017 年 12 月期间的数据作为验证队列 B。OS 是本研究的主要终点。采用 Cox 回归分析检测独立的风险变量。决策曲线分析、校准曲线、时间依赖性接受者操作特征（ROC）曲线和一致性指数（C-index）用于评估列线图模型的性能。
共有 267 例患者在首次常规放疗后发生 NP 坏死。放疗后，所有三个队列中 NP 坏死患者的 OS 明显低于其他患者（p<0.001）。11 个因素（包括 NP 坏死）被纳入列线图，在训练队列中，该列线图具有良好的区分度和校准度，C-index 为 0.768，在验证队列 A 中为 0.749，在验证队列 B 中为 0.739。与 TNM 分期和 EBV DNA 相比，该列线图预测 OS 的 ROC 曲线下面积显著更大（p<0.001）。
与 TNM 系统和 EBV DNA 相比，我们建立了一个针对 NPC 患者具有准确预后预测能力的列线图模型，这可能有助于 NPC 患者的管理。
本研究纳入了 9298 例 NPC 患者，最终模型涉及 11 个因素。
列线图的 C-index 和 ROC 曲线下面积均显著高于 TNM 分期和 EBV DNA。
我们建立了第一个涉及 NP 坏死发生的 NPC 列线图模型，该模型对于提供个体咨询和临床评估具有重要价值。
