EBP50 and Merlin Expression Across the Spectrum of Serous Ovarian Tumors: Clinicopathological Associations and Exploratory Diagnostic Value

Background: Serous ovarian tumors comprise biologically distinct entities ranging from benign and borderline tumors to low-grade (LGSC) and high-grade serous carcinoma (HGSC). EBP50/NHERF1 and Merlin/NF2 regulate epithelial polarity, membrane–cytoskeletal organization, and signaling. We evaluated EBP50 and Merlin expression across the spectrum of serous ovarian tumors, explored their clinicopathological and diagnostic associations, and compared these findings with Ki67 and mitotic activity in serous carcinomas. Methods: EBP50 and Merlin immunoreactivity was evaluated in serous cystadenomas, serous borderline tumors (SBTs), LGSCs, and HGSCs. Ki67 and mitotic activity were assessed in malignant tumors. Exploratory discriminatory performance was evaluated by receiver operating characteristic analysis, and overall survival by Kaplan–Meier and Cox proportional hazards analyses. Results: EBP50 expression differed significantly among tumor groups (p < 0.001), with the lowest expression and most frequent complete loss in HGSC. EBP50 showed moderate discrimination between carcinomas and non-carcinoma tumors (AUC = 0.790) and lower expression was associated with more advanced disease, including higher FIGO stage. Merlin expression also differed among groups (p = 0.003) but showed limited discriminatory performance (AUC = 0.605). Ki67 and mitotic activity strongly distinguished LGSC from HGSC (AUC = 0.992 for both). Although Ki67 was associated with poorer overall survival in univariable analysis, none of the biomarkers independently predicted survival after adjustment for age and FIGO stage. Conclusions: EBP50 showed potentially useful differences in expression across serous ovarian tumor categories and may provide adjunctive information in selected diagnostic settings, whereas the diagnostic contribution of Merlin appeared limited. Ki67 and mitotic activity confirm the established proliferative distinction between LGSC and HGSC. Further validation of EBP50 and Merlin in independent cohorts is warranted.