Endometrial cancer (EC) is the most common gynaecologic malignancy in developed countries, and its diagnostic and prognostic framework has evolved substantially following the introduction of the 2023 FIGO staging system, which integrates molecular classification with clinicopathologic features. Both histopathologic features, such as lymphovascular space invasion (LVSI) and molecular subtype, including POLE mutation status, mismatch-repair deficiency, and p53-abnormal phenotype, are incorporated into the updated staging system, highlighting the importance of tumour biology in risk stratification. Accordingly, the value and contribution of MRI to patient management must extend beyond macroscopic assessment to support a more biologically driven approach. This narrative review synthesizes recent advances in MRI for EC, highlighting developments that improve diagnostic accuracy and align imaging with the molecular paradigm. Multiparametric MRI remains the reference standard for local staging, while emerging quantitative diffusion techniques provide microstructural biomarkers associated with tumor aggressiveness and prognostic features. The consistency of nodal staging has been enhanced by Node-RADS, a structured reporting system that integrates nodal morphology and configuration, with the goal of improving reproducibility and diagnostic performance over size-based assessment alone. Radiomics and artificial intelligence (AI) represent the most transformative frontier, enabling MRI to infer biological behaviours previously accessible only via histopathologic assessment. Radiomics and deep-learning models have demonstrated high accuracy in predicting LVSI, DMI, nodal metastasis, and molecular subtypes, offering non-invasive biomarkers aligned with FIGO 2023 prognostic categories. Together, these advances position MRI as a quantitatively enriched, biologically relevant tool that supports precision oncology in endometrial cancer.
Gennarini et al. (2026) studied this question.