Meta-analysis of the clinicopathologic features of endometrial cancer molecular staging

The 2013 TCGA identified four molecular subgroups of endometrial cancer; however, the data results for most of the pathological features were varied and of low value for clinical application. Therefore, a meta-analysis of articles related to the clinicopathological features of molecular typing was performed to observe how the prevalence of the four subgroups varied across different pathological features and whether they were associated with certain specific pathological features and to understand how molecular typing may influence current pathological assessments.

PubMed, Embase, Web of Science, CNKI, Wanfang, and VIP were searched from the time of library construction until May 2024, and the following data were extracted: histological type, FIGO grade, FIGO stage, LVSI, depth of muscularis propria infiltration, and lymph node status of each TCGA group. Two reviewers used the Cochrane Diagnostic Research Scale assessment, and the data were analyzed using Review Manager 5.4.1 and Stata 14.0.

Fourteen diagnostic research papers were included in this study, with a total of 4,776 patients with endometrial cancer. Non-estrogen-related endometrial carcinoma (NEEC) vs. estrogen-related endometrial carcinoma (EEC) was low in polymerase epsilon (POLE) (OR = 0.49), microsatellite instability (MSI) (OR = 0.45), and copy number low (CNL) (OR = 0.11), while it was high in CNH (OR = 26.76). G3 EEC vs. G1-2 EEC POLE (OR = 1.98), MSI (OR = 1.74), and CNH (OR = 5.57) were high, whereas it was low in CNL (OR = 0.23), low in FIGO II-IV vs. FIGO I in POLE (OR = 0.39) and CNH (OR = 0.64), and high in FIGO II-IV vs. FIGO I in CNH (OR = 3.05). There was no difference in MSI prevalence in FIGO II-IV vs. FIGO I. POLE (OR = 0.64) and CNL (OR = 0.75) were low in myometrial invasion depths ≥50% and lower in myometrial invasion depths <50%, and CNL (OR) was higher in CNH (OR) than in myometrial invasion depths <50%. There was no difference in MSI between different myometrial invasion depths. MSI (OR = 1.69) and CNH (OR = 2.12) were higher in lymphatic vascular infiltration (LVSI) vs. no LVSI; CNL (OR = 0.39) was lower in LVSI than in no LVSI. There was no difference in POLE in the presence or absence of LVSI. Lymph node metastasis with and without lymph node metastasis in POLE (OR = 0.25) and CNL (OR = 0.31) were lower, and CNH (OR = 3.06) was higher in lymph node metastasis than in no lymph node metastasis. There was no difference in MSI in the presence or absence of lymph node metastasis.

POLE patients predominated in pathological features of early-stage endometrial cancer and had better prognosis. MSI patients were more likely to be found in EEC and G3 EEC as well as LVSI. Nearly half of G3 EEC as well as LVSI were present in MSI patients, and CNH patients were more likely to be found to have pathological features of advanced endometrial cancer and poor prognosis, providing evidence that CNH is a high-risk cancer. Patients with CNL were more likely to be found to have pathological features of early-stage endometrial cancer and good prognosis, and CNL was present in large numbers in both early-stage and late-stage endometrial cancers. CNL does not yet have a precise prognostic value.

https://www.crd.york.ac.uk/PROSPERO/, identifier CRD42024563661.