Endometrial carcinoma diagnosis in the molecular era: a review and new considerations for low- to middle-income countries

Abstract

The incidence of endometrial carcinoma (EC) has seen a steady increase globally over the years. According to the Cancer Association of South Africa (CANSA), EC is the fourth most common carcinoma in women. The publication of the International Federation of Gynecology and Obstetrics (FIGO) 2023 staging for the management of EC has incorporated molecular EC subtyping in addition to commonly used histopathological diagnosis. The four molecular subtypes of EC defined by The Cancer Genome Atlas (TCGA) research group have been renamed: 1) POLEmut (POLE ultramutated), 2) p53abn (p53-abnormal, copy number high), 3) MMRd (mismatch repair deficient, microsatellite instability hypermutated), and 4) NSMP (no specific molecular profile, copy number low). The molecular profiling of EC by TCGA used whole exome sequencing with determination of somatic copy number alterations (SCNA) on multiple platforms, which may prove costly in low- to middle-income countries (LMIC). The Proactive Molecular Risk Classifier of Endometrial Cancer (ProMisE) aims to validate TCGA results and investigate the correlation between the traditional clinicopathological prognostic features using a simple decision tree focused on immunohistochemistry and simple molecular testing. In a LMIC healthcare facility, the two molecular subtypes that can be tested using surrogates in immunohistochemical stains are p53 and MMRd, MLH1, PMS2, MSH2, and MSH6. There is no surrogate at present to test for POLE mutation; however, there is QPOLE – a low-cost quantitative multiplex polymerase chain reaction (PCR) that can detect the 11 most common and pathogenic POLE mutations. The use of a flow diagram in the daily clinical practice of patients with EC should be considered in LMIC. Future translational research in determining the cost-effectiveness of preoperative molecular pathology testing for treatment planning and risk stratification of patients is needed in resource-limited settings.