Yang Dong, Liu Zifeng, Xie Jingwei
Oncology Department, Affiliated Hospital of Jining Medical University, Jining, 272000, Shandong, People's Republic of China.
Oncology Department, Jining No. 1 People's Hospital, #6Jiankang Road, Jining, 272029, Shandong, People's Republic of China.
Sci Rep. 2025 Apr 30;15(1):15155. doi: 10.1038/s41598-025-00289-8.
Upper esophageal cancer (ESCA) is a distinct subtype of ESCA that accounts for < 10% of ESCA cases. However, its unique clinical characteristics remain unclear, and without specialized prognostic model. We aimed to clarify its unique clinical characteristics and develop a specialized prognostic model. Data for a total of 1371 upper ESCA cases and 15,434 cases of ESCA at other segments were collected from the Surveillance, Epidemiology, and End Results (SEER) database. Compared with that of patients with ESCA at other segments, a greater proportion of patients with upper ESCA were older and female; had an abnormal marital status; had tumors at the T4 stage, N0 stage, and M0 stage; and had squamous cell carcinoma and differentiation grade II. Moreover, the prognosis of upper ESCA was significantly poorer, and the constituent ratio stratified by the above characteristics from 2004 to 2015 showed no significant changes of average annual percent change (AAPC). Cox regression analysis was used to identify independent prognostic factors. Age, sex, marital status, histologic type, grade, and T, N and M stage were included in the development of the nomogram. The C-indexes of the training cohort and validation cohort were 0.64 (95% CI 0.62-0.66) and 0.62 (95% CI 0.58-0.64), respectively. The area under the curve (AUC), calibration curve, and decision curve analysis (DCA) results confirmed the good performance of the upper ESCA model. The C-index, integrated discrimination improvement (IDI), net reclassification improvement (NRI), time-dependent AUC, and DCA and survival analysis results confirmed that the upper ESCA model performed better than the TNM model in predicting the prognosis of upper ESCA. Finally, compared with the total ESCA model, which is based on a total of 16,805 ESCA cases, the upper ESCA model showed better performance in predicting the prognosis of upper ESCA. In conclusion, we outlined the unique clinical characteristics of upper ESCA and developed a specialized prognostic model that exhibited better performance in predicting the prognosis of upper ESCA than did the TNM model and total ESCA model.
食管上段癌（ESCA）是ESCA的一种独特亚型，占ESCA病例的不到10%。然而，其独特的临床特征仍不明确，且缺乏专门的预后模型。我们旨在阐明其独特的临床特征并开发一种专门的预后模型。从监测、流行病学和最终结果（SEER）数据库中收集了总共1371例食管上段癌病例和15434例其他节段的ESCA病例的数据。与其他节段ESCA患者相比，食管上段癌患者中年龄较大和女性的比例更高；婚姻状况异常；肿瘤处于T4期、N0期和M0期；且为鳞状细胞癌和分化二级。此外，食管上段癌的预后明显较差，2004年至2015年按上述特征分层的构成比显示年均变化百分比（AAPC）无显著变化。采用Cox回归分析确定独立预后因素。年龄、性别、婚姻状况、组织学类型、分级以及T、N和M分期被纳入列线图的构建。训练队列和验证队列的C指数分别为0.64（95%CI 0.62 - 0.66）和0.62（95%CI 0.58 - 0.64）。曲线下面积（AUC）、校准曲线和决策曲线分析（DCA）结果证实了食管上段癌模型的良好性能。C指数、综合判别改善（IDI）、净重新分类改善（NRI）、时间依赖性AUC以及DCA和生存分析结果证实，食管上段癌模型在预测食管上段癌预后方面比TNM模型表现更好。最后，与基于总共16805例ESCA病例的总ESCA模型相比，食管上段癌模型在预测食管上段癌预后方面表现更好。总之，我们概述了食管上段癌的独特临床特征，并开发了一种专门的预后模型，该模型在预测食管上段癌预后方面比TNM模型和总ESCA模型表现更好。
