Han Yujing, Huang Yingqin, Luo Deyi
Pelvic Diseases Center, West China Tianfu Hospital, Sichuan University, Chengdu, China.
Medicine (Baltimore). 2025 Jun 6;104(23):e42583. doi: 10.1097/MD.0000000000042583.
Uterine corpus endometrial carcinoma (UCEC) poses a significant to women's health. Accurate prediction of prognosis plays a crucial role in facilitating clinical decision-making processes. Therefore, this study aimed to develop a robust prognostic model based on gene expression profile. Gene expression profile of 546 UCEC samples of The Cancer Genome Atlas were retrieved. A multi-step strategy was employed to develop and validate a prognostic model predicting all-cause mortality rates. Receiver operating characteristic curve and decision curve analysis were performed to assess the predictive accuracy and net benefit of the model. Besides, model-associated immunological features were explored. The UCEC Prognostic Model (TUPM) performed well in identifying patients at high mortality risk. Patients with risk scores above the upper quartile had significantly decreased overall survival compared to patients with risk scores below the lower quartile (HR = 12.56, CI95: 4.629-34.09, P = 6.76E-7), indicating a prominent discriminability. The model accurately predicted patient survival from 1 to 5-year (area under the curve [AUC]1-year = 0.766, AUC2-year = 0.816, AUC3-year = 0.764, AUC4-year = 0.783, AUC5-year = 0.814) and provided excellent calibration. Meanwhile, The UCEC Prognostic Model encompassing transcriptome scores yielded a higher net clinical benefit than the baseline model that only included patient age and clinical stage. Furthermore, the prolonged survival in the low-risk group may be associated with increased infiltration of follicular T cells and regulatory T cells in the tumor microenvironment. We have developed a robust prognostic model for UCEC that may provide preliminary evidence for individualized management and treatment modality decision.
子宫体子宫内膜癌（UCEC）对女性健康构成重大威胁。准确预测预后对于促进临床决策过程起着至关重要的作用。因此，本研究旨在基于基因表达谱开发一种强大的预后模型。检索了癌症基因组图谱中546例UCEC样本的基因表达谱。采用多步骤策略来开发和验证预测全因死亡率的预后模型。进行了受试者工作特征曲线和决策曲线分析，以评估模型的预测准确性和净效益。此外，还探索了与模型相关的免疫特征。UCEC预后模型（TUPM）在识别高死亡风险患者方面表现良好。风险评分高于上四分位数的患者与风险评分低于下四分位数的患者相比，总生存期显著降低（HR = 12.56，CI95：4.629 - 34.09，P = 6.76E - 7），表明具有显著的辨别力。该模型准确预测了患者1至5年的生存期（曲线下面积[AUC]1年 = 0.766，AUC2年 = 0.816，AUC3年 = 0.764，AUC4年 = 0.783，AUC5年 = 0.814），并提供了良好的校准。同时，包含转录组评分的UCEC预后模型比仅包括患者年龄和临床分期的基线模型产生了更高的净临床效益。此外，低风险组生存期的延长可能与肿瘤微环境中滤泡性T细胞和调节性T细胞浸润增加有关。我们已经为UCEC开发了一种强大的预后模型，可为个体化管理和治疗方式决策提供初步证据。
