子宫内膜癌预后影响因素的研究进展：从传统病理到分子整合

Research Progress in Prognostic Factors of Endometrial Cancer: From Traditional Pathology to Molecular Integration

摘要: 子宫内膜癌是全球女性第六大常见恶性肿瘤，其发病率呈持续上升趋势。尽管多数患者预后良好，但仍有相当比例的患者面临复发和死亡风险。准确识别预后相关因素对于个体化治疗决策至关重要。本文系统综述了子宫内膜癌预后影响因素的研究进展，从传统临床病理因素、分子分型、生物标志物、免疫微环境、影像学评估及新型预测模型等多个维度进行阐述。传统因素如年龄、FIGO分期、组织学类型、肿瘤分级、肌层浸润深度、淋巴血管间隙浸润及淋巴结转移等已被证实与预后显著相关。分子分型(POLE突变型、MMRd型、NSMP型、p53异常型)的临床应用极大地提升了对患者预后判断的精准度。此外，HE4、L1CAM、ER/PR等免疫组化标志物及外周血炎症指标在风险分层中也显示出重要价值。高龄作为独立因果性预后因素，其对不良结局的影响持续增加至80岁。新修订的FIGO 2023分期系统整合了分子特征与病理参数，展现出更优的预后预测效能。基于多因素的列线图和机器学习模型为临床决策提供了新的工具。未来，整合分子特征与临床病理参数的个体化预后评估体系将进一步完善，推动子宫内膜癌精准治疗的发展。

Abstract: Endometrial cancer is the sixth most common malignant tumor among women worldwide, with a persistently increasing incidence. Although most patients have a favorable prognosis, a considerable proportion remain at risk of recurrence and mortality. Accurate identification of prognostic factors is crucial for individualized treatment decision-making. This article systematically reviews the research progress in prognostic factors for endometrial cancer, covering multiple dimensions including traditional clinicopathological factors, molecular classification, biomarkers, immune microenvironment, radiological evaluation, and novel predictive models. Traditional factors such as age, FIGO stage, histological type, tumor grade, depth of myometrial invasion, lymphovascular space invasion, and lymph node metastasis have been confirmed to be significantly associated with prognosis. The clinical application of molecular classification (POLE-mutated, MMR-deficient, NSMP, p53-abnormal subtypes) has greatly improved the accuracy of prognostic stratification. In addition, immunohistochemical markers including HE4, L1CAM, ER/PR, and peripheral blood inflammatory indicators also demonstrate important value in risk stratification. Advanced age, as an independent causal prognostic factor, shows a progressively increasing impact on adverse outcomes up to 80 years old. The newly revised FIGO 2023 staging system integrates molecular features and pathological parameters, exhibiting superior prognostic performance. Multivariatebased nomograms and machine learning models provide new tools for clinical decision-making. In the future, individualized prognostic evaluation systems integrating molecular characteristics and clinicopathological parameters will be further refined, promoting the development of precision therapy for endometrial cancer.