Molecular heterogeneity of endometrial cancer in the real-world: Biomarker patterns by tumor stage, histology, and molecular subtype

OBJECTIVE: To describe the prevalence and distribution of molecular biomarkers in endometrial cancer across tumor stage, histology, and molecular subtype using real-world data. METHODS: This retrospective cohort study used de-identified clinical and tumor sequencing data from the Oncology Research Information Exchange Network between 2006 and 2020. Patients diagnosed with endometrial cancer and next-generation sequencing results for ≥1 biomarker (POLE, MSI, TP53, PTEN, PIK3CA, ARID1A, TMB, ESR1, ERBB2 (HER2) mutation, and ERBB2 amplification) were included. Biomarker prevalence was summarized and stratified by histology, stage, and The Cancer Genome Atlas (TCGA)-defined molecular subtypes. RESULTS: The study cohort included 671 patients with a mean age of 62.4 years. Most patients (76%) had tumors with endometrioid histology. The most prevalent biomarkers included PTEN (65.9%), ARID1A (52.0%), and PIK3CA (39.2%). TP53 mutations (25.3%) and ERBB2 amplification (4.8%) were more prevalent in non-endometrioid histology (64.2% and 13.2%) and advanced-stage disease (41.4% and 11.2%) than in endometrioid or early-stage tumors. In contrast, POLE, PTEN, and TMB-high were more prevalent in early-stage disease and endometrioid histology. Across TCGA molecular subtypes, distinct biomarker patterns were observed: POLE-positive and MSI-H groups had high PTEN, ARID1A, and TMB-high prevalence, while TP53-mutated tumors showed the highest rates of ERBB2 amplification. CONCLUSIONS: The results of this real-world study demonstrate that endometrial cancer is a complex disease, characterized by substantial molecular heterogeneity and varying biomarker distributions across histology, stages, and molecular subtypes. This real-world study supports the integration of comprehensive molecular profiling into routine practice to refine prognostic stratification and guide biomarker-driven therapies.