ESTRO 2024 - Abstract Book

S1063

Clinical - Gynaecology

ESTRO 2024

Material/Methods:

We retrospectively included patients who underwent endometrial cancer staging operations, including hysterectomy, at two institutions between 2010 and 2020. Only patients with the FIGO stage I, endometrioid histology, and grade 2 or 3 were included. Patients with less than 6 months of follow-up were excluded. Immunohistochemistry (IHC) results, which included estrogen receptor (ER), mismatch repair (MMR) proteins, and p53, were collected if analyzed. Adjuvant radiotherapy was administered at the discretion of radiation oncologists. The primary endpoint was any recurrence. Outcomes based on risk classes were compared by the log-rank test. Univariate Cox regression analyses were performed to identify potential prognostic factors. Recursive partitioning analysis (RPA) was conducted to classify patients based on significant factors, and Cox proportional hazard analysis was used for between-class comparisons.

Results:

A total of 298 women were included in the study, with a median follow-up of 80.1 months (interquartile range [IQR] 63.7-98.9 months). Among them, 157 women (52.7%) received adjuvant radiotherapy. ER status and molecular subtype were studied in 130 (43.6%) and 123 (41.3%) of the patients, respectively. Forty patients (13.4%) experienced disease recurrence, with 15 (5.0%) having locoregional recurrence and 32 (10.7%) having distant recurrence; 7 (2.3 %) had both locoregional and distant recurrence. The median time to recurrence was 15 months (IQR 8-32 months). Comparison of risk groups based on 2016 ESMO-ESGO-ESTRO risk classes showed statistical significance for disease recurrence and distant metastasis (log-rank test p<0.001), while locoregional recurrence was not different between risk classes (p=0.16). Age, tumor size, histologic grade, depth of myometrial invasion (MMI), lymphovascular space invasion (LVSI), ER status, progesterone receptor (PR) status, and radiotherapy treatment were statistically significant for disease recurrence in univariate analysis and were incorporated into RPA. Molecular profile was not statistically significant. When IHC results were blinded, the RPA model identified three classes based on age at surgery and MMI. Patients aged up to 70 years and MMI less than 1/2 had a favorable prognosis, with a 3-year recurrence-free rate of 96.0%. Patients over 71 years, regardless of MMI, had the worst clinical outcomes, with a 3-year recurrence-free rate of 62.5% (hazard ratio [HR] 9.41, 95% CI 4.06-21.84, p<0.001). For patients with a known ER profile, the RPA model identified three classes based on age at surgery and ER status. Patients aged up to 70 years with ER(+) had the best outcome. Patients aged up to 70 years with ER(-), and patients over 70 years, regardless of ER status, had similarly unfavorable outcomes, with 3-year recurrence-free rates of 54.5% (HR 7.16, 95% CI 2.63-19.55, p<0.001) and 46.2% (HR 9.46, 95% CI 3.65-24.55, p<0.001), respectively.