ENDOESTRO score: Where is the dark side? Evaluation of estrogen receptor expression cutoff in endometrial cancer molecular classification, an ambispective study

Background: Estrogen receptor (ER) expression has prognostic value in endometrial cancer (EC), also in the context of molecular classification. However, a standardized cutoff to define ER positivity is lacking. Objective: To identify the ER expression cutoff that best correlates with survival outcomes overall and within each molecular subgroup. Study design: We conducted an ambispective study comprising a retrospective phase (Feb 2017–Feb 2022) and a prospective phase (Mar 2022–Jan 2024), including patients with EC who underwent surgery at our institution. In the retrospective cohort, molecular subgroups were defined by immunohistochemistry (IHC): 1) MMRpIHC: MMR-proficient and wild-type p53 (p53wt); 2) p53abnIHC: MMRp and abnormal p53; 3) MMRdIHC: MMR-deficient. In the prospective phase, classification followed current IHC plus NGS standards: POLE-ultramutated (POLEmut), No specific molecular profile (NSMP), p53abn, and MMRd. ER expression was stratified into three groups: ≥ 10 %, < 10 %, and < 1 %. Results: Among 2084 patients, ER < 10 % was the strongest cutoff for prognosis based on disease-free survival (DFS). Patients with ER < 10 % and < 1 % shared similar unfavorable clinicopathological and molecular features, while ER ≥ 10 % was associated with more favorable profiles. ER < 10 % was an independent adverse prognostic factor in uni- and multivariate analyses for DFS and overall survival, particularly in MMRpIHC and p53abnIHC groups. Prospective data confirmed < 10 % as the most reliable cutoff. Conclusion: ER < 10 % expression is a robust prognostic indicator, associated with worse outcomes and may serve as a clinically meaningful cutoff in endometrial cancer risk stratification. Further prospective validation is warranted.