目的：基于癌症基因组图谱（The Cancer Genome Atlas, TCGA）分子分型系统，探讨子宫内膜癌（endometrial carcinoma，EC）不同分子亚型的临床病理特征及预后差异，为个体化治疗提供理论依据。

方法：收集2019年12月至2023年3月在北京医院接受手术治疗并完成TCGA分子分型的146例子宫内膜癌患者，根据TCGA提出的分子分型的方法分为四种亚型，POLE突变型、微卫星不稳定型（Microsatellite instability-High mutation, MSI-H）、低拷贝数型（Copy-number Low，CNL）、高拷贝数型（Copy-number High，CNH）。收集整理相关临床资料，包括患者年龄、绝经状态、病理类型与分级、手术分期、淋巴血管间隙浸润（LVSI）、术前血清CA125水平、淋巴结转移情况，以及雌激素受体（ER）和孕激素受体（PR）等免疫组化标志物表达状态。应用SPSS 24.0统计分析软件分析四种TCGA分子分型的临床病理特征、免疫组化结果及预后情况。

结果：146例患者中，POLE突变型8例（5.5%）、MSI-H型29例（19.9%）、CNL型94例（64.4%）、CNH型15例（10.3%）。分子分型与患者年龄（p=0.012）、绝经状态（p=0.003）、ER/PR表达（p<0.001）显著相关。不同亚型在病理类型（p<0.001）、组织学分级（p<0.001）、FIGO分期（p=0.013）、淋巴结转移（p=0.005）及LVSI（p=0.030）方面差异显著，但术前CA125水平无统计学差异（p=0.587）。预后分析显示，POLE突变型、MSI-H型、CNL型和CNH型的2年无进展生存率（Progression-free surial ,PFS）分别为100%、100%、93.6%和73.3%，组间差异显著（p=0.006）。

结论：TCGA分子分型可有效区分子宫内膜癌的临床异质性：CNH型呈现显著高龄化趋势，POLE突变型平均年龄最小；CNH型聚集晚期表型，非子宫内膜样癌占比、高级别占比、晚期肿瘤占比高；CNH型具有显著的侵袭性生物学行为，淋巴脉管阳性率及淋巴结转移率均显著高于其他分型；CNH型预后最差（2年PFS 73.3%），而POLE突变型和MSI-H型预后良好（2年PFS均为100%）。该分型系统对预后评估和个体化治疗具有重要指导价值。

Objective: To Investigate the Clinicopathological Characteristics and Prognostic Differences Among Different Molecular Subtypes of Endometrial Cancer Based on The Cancer Genome Atlas (TCGA) Molecular Classification System: Providing a Theoretical Basis for Personalized Treatment.

Methods: A total of 146 EC patients who underwent surgical treatment and TCGA molecular typing in Beijing Hospital from December 2019 to March 2023 were collected.The endometrial cancer cases were classified into four molecular subtypes according to the TCGA classification system: POLE-mutated, MSI-H, CNL (copy-number low), and CNH (copy-number high). Comprehensive clinical data were collected for analysis, including patient age, menopausal status, pathological type and grade, surgical staging, lymphovascular space invasion (LVSI), preoperative serum CA125 levels, lymph node metastasis status, as well as the expression of immunohistochemical markers such as estrogen receptor (ER) and progesterone receptor (PR). The clinicopathologic features, immunohistochemistry, and prognosis of the four TCGA molecular types were analyzed retrospectively.

Result: Among 146 patients, molecular subtyping revealed the following distribution: POLE-mutated (8 cases, 5.5%), MSI-H (29 cases, 19.9%), CNL (94 cases, 64.4%), and CNH (15 cases, 10.3%). Significant associations were observed between molecular subtypes and age (p=0.012), menopausal status (p=0.003), as well as ER/PR expression levels (p<0.001). Distinct variations emerged across subtypes regarding histopathological type (p<0.001), histological grade (p<0.001), FIGO stage (p=0.013), lymph node metastasis (p=0.005), and lymphovascular space invasion (LVSI) (p=0.030), though no significant difference was detected in preoperative CA125 levels (p=0.587). Prognostic analysis demonstrated significantly divergent 2-year progression-free survival (PFS) rates: 100% for POLE-mutated, 100% for MSI-H, 93.6% for CNL, and 73.3% for CNH subtypes, with statistically significant intergroup differences (p=0.006).

Conclusions: The TCGA molecular classification effectively distinguishes the clinical heterogeneity of endometrial cancer: The CNH type shows a significant trend towards higher age, while the POLE-mutated type has the lowest average age; the CNH type is associated with advanced phenotypes, with higher proportions of non-endometrioid carcinomas, high-grade tumors, and advanced-stage tumors; the CNH type exhibits significantly aggressive biological behavior, with higher rates of lymphovascular space invasion and lymph node metastasis compared to other subtypes; the CNH type has the worst prognosis (2-year PFS of 73.3%), whereas the POLE-mutated and MSI-H types have favorable prognoses (both with a 2-year PFS of 100%). This classification system holds significant guiding value for prognostic evaluation and individualized treatment.