A Clinicopathology and Simplified Molecular Marker-Based Risk Stratification Model for Predicting 3-Year Recurrence in Endometrial Cancer

基于临床病理学和简化分子标志物的子宫内膜癌3年复发预测风险分层模型

阅读:2

Abstract

BACKGROUND: The integration of clinicopathological parameters and molecular biomarkers holds significant prognostic value for endometrial carcinoma (EC) recurrence assessment. The purpose of this study is to combine clinical pathological parameters with simple molecular features to predict 3-year recurrence of EC. METHODS: Based on a retrospective cohort with 136 patients assessing recurrence risk within 3 years postoperatively, we developed PM-TERR (clinicopathology and simplified molecular markers-based EC tailored 3-year recurrence risk). All patients were randomly assigned to the training set (96 patients) and the test set (40 patients). The nomogram comprised 4 input variables (Stage, Grade, Pathology Type, and P53 status) for calculating a PM-TERR score. RESULTS: Kaplan-Meier analysis demonstrated a significant association between PM-TERR scores and disease-free survival (DFS) in both the training and test sets (both P < .001). The PM-TERR model outperformed the ESMO stratification in predicting DFS. The likelihood value of the PM-TERR model increased by 6.88 in the training set, and by 2.20 in the test set, both sets P < .05. Time-dependent ROC analysis further confirmed the PM-TERR enhanced predictive accuracy, with consistently higher AUC values across all time points than ESMO in both cohorts. CONCLUSION: In summary, the PM-TERR model integrates stage, grade, histology, and p53 status to predict 3-year recurrence risk. This retrospective, single-center study demonstrates its potential utility; however, future prospective, multi-center validation is required to confirm its robustness and clinical applicability before broader implementation.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。