Introduction/Background In the past few years, the molecular classification system, which categorize endometrial cancer (EC) into four risk classes, has become a crucial tool for determining the appropriate treatment following surgical staging. The primary objective of this study was to compare recurrence-free survival among the four molecular classes.

Methodology Starting from April 2019, we prospectively performed molecular characterization of all EC patients according to the WHO-endorsed algorithm. Molecular analysis included IHC for p53 and MMR proteins, microsatellite instability assay and Next Generation Sequencing for POLE exonuclease domain and TP53. ECs were classified into 4 molecular classes (POLEmut, MMR deficiency [MMRd], p53 abnormality [p53abn], and non-specific molecular profile [NSMP]). For the current analysis patients with stage IV disease, vaginal approach, no lymph node assessment, no molecular analysis, and follow-up < 6 months were excluded.

Comparison of clinicopathological characteristics between molecular classes was performed with univariate analysis. Survival analysis was performed using Kaplan-Meier curves to assess recurrence-free survival by molecular class.

Results In total, 239 patients were included: 26 (18.9%) POLEmut, 76 (31.8%) MMRd, 101 (42.3%) NSMP, 36 (15.1%) p53abn. Overall, 29 (12.1%) recurrences were observed, including 0/28 (0%) in the POLEmut, 4/76 (5.2%) in the MMRd, 12/101 (12.9%) in the NSMP, and 13/36 (36.1%) in the p53abn. The median time of recurrence was 12 months (IQR 8–17); while median follow-up of those without recurrence was 18 months (IQR 11–29). At survival analysis, we found a significant difference in recurrence-free survival among the four groups (p-value<0.001).

• Download figure
• Open in new tab
• Download powerpoint

Abstract #908 Figure 1

Tab. A) Comparison of clinical-pathological features by molecular classes; Tab. B) Survival analysis by molecular classes

Conclusion Our study prospectively confirms in a large population that molecular classification can predict the risk of recurrence in EC and underlines the protective role of POLEmut and negative predictive value of p53abn. These results highlight the importance of performing molecular evaluation in all patients with EC to improve risk stratification and guide clinical decision-making.

Disclosures The authors declare no conflict of interests