Updates on the Molecular Classification of Endometrial Carcinoma

The risk of recurrence and metastasis in patients with endometrial carcinoma varies widely, making evaluation of risk in individual patients critically important for planning effective adjuvant treatment. Tumor histotype and grade have been used, along with stage and lymphovascular space invasion (LVSI), but suboptimal reproducibility of histotype assignment (especially for high-grade tumors) and tumor grade have made risk assessment challenging. Similarly, single biomarkers such as L1CAM and beta-catenin have shown potential but none have emerged as well-validated and reproducible markers of risk in endometrial carcinoma. This changed with the finding of The Cancer Genome Atlas that there are 4 molecular subtypes of endometrial carcinoma, based on genomic architecture. Subsequently, a clinically applicable approach to endometrial carcinoma molecular subtype has emerged, based on sequencing of POLE to detect a limited number of pathogenic mutations, and immunostaining for mismatch repair (MMR) proteins and p53. This approach places each endometrial carcinoma into one of 4 molecular subtypes (POLEmut, MMRd, p53abn, and no specific molecular profile [NSMP]) that can be reproducibly diagnosed, show excellent concordance between the biopsy and hysterectomy specimen, are of prognostic significance, predict response to specific adjuvant therapies, and can be used to screen for underlying hereditary cancer susceptibility syndromes. Molecular subtype is now integrated into widely used endometrial carcinoma risk assessment guidelines and is a key component of the 2023 update to the International Federation of Gynecology and Obstetrics (FIGO) staging system for endometrial carcinoma. Recently, prognostic refinement of the 4 molecular groups introduces a possible fifth group showing absence of estrogen receptor (ER) expression and/or high histologic grade within the NSMP subtype. We will review the key features of endometrial carcinoma molecular subtypes, focusing on recent advances in molecular subtype diagnosis in practice.