Comparative prognostic performance of the FIGO 2023 and FIGO 2009 staging systems integrated with molecular classification in endometrial cancer: a retrospective single-centre cohort study
Related Experiment Videos
Wei Xia1, Bingjie Liu1, Rui Kong1
1Department of Obstetrics and Gynaecology, Beijing Friendship Hospital, Affiliated to Capital Medical University, Beijing, China.
Background:
The 2023 revision of the International Federation of Gynecology and Obstetrics (FIGO) staging system incorporates molecular classification, aggressive histology, and substantial lymphovascular space invasion (LVSI) as key determinants. Its prognostic performance compared with FIGO 2009 requires validation in real-world cohorts.
Methods:
A retrospective review was undertaken of 712 consecutive patients with endometrial carcinoma treated between January 2019 and April 2025. Cases were restaged according to FIGO 2009 and FIGO 2023 integrated with molecular classification (FIGO 2023 m). Progression-free survival (PFS) and overall survival (OS) were analysed using Kaplan-Meier methods. Prognostic discrimination was assessed using Harrell's concordance index (C-index), time-dependent area under the curve, information criteria, and continuous net reclassification index (NRI).
Results:
Stage migration occurred in 26.4% of patients, predominantly due to upstaging driven by p53 abnormality, aggressive histology, and substantial LVSI. Molecular subtypes included. POLE-ultramutated (7.2%), microsatellite instability hypermutated (25.8%), no specific molecular profile (copy-number low) (51.3%), and p53 abnormal (copy-number high) (15.7%). Five-year PFS varied substantially across molecular groups. The C-index improved from 0.712 to 0.785 for PFS, and from 0.725 to 0.811 for OS. The continuous NRI for PFS was 0.241.
Conclusions:
FIGO 2023 m demonstrated superior prognostic discrimination compared with FIGO 2009, supporting incorporation of molecular subtypes into routine endometrial cancer staging and risk assessment. Clinically, this integrated framework may help identify patients at risk of under- or over-treatment, particularly in those with disease previously considered as early stage, although prospective studies are needed before treatment algorithms can be changed definitively.