Abstract
Molecular classification has emerged as a critical tool for guiding personalized treatment in endometrial cancer (EC) and atypical endometrial hyperplasia (AEH). This retrospective study aimed to assess the impact of molecular classification on fertility-sparing treatment outcomes in patients diagnosed with EC and AEH who underwent fertility preservation therapy between 2006 and 2021. Patients were categorized into four molecular subtypes using immunohistochemistry (IHC) and Sanger sequencing, based on the Proactive Molecular Risk Classifier for Endometrial Cancer (ProMisE): POLE-ultramutated, mismatch repair (MMR) deficient (MMRd), p53 abnormal (p53abn), and p53 wild-type (p53wt). All patients were evaluated for oncological prognosis and fertility outcomes, with a total of 103 patients included in the analysis. Recurrence rates exhibited significant differences among the molecular classifications, with the lowest recurrence rate observed in the p53wt subtype (19.7%), followed by MMRd (30.4%), POLE-ultramutated (66.7%), and p53abn (71.4%) subtypes. Multivariate Cox regression analysis indicated that the p53abn subtype was a significant risk factor for recurrence following conservation therapy when compared to the p53wt subtype. Additionally, there was a notable disparity in standard surgical treatment due to treatment failure, with operation rates of 7.5%, 19.2%, 66.7%, and 57.1% for the p53wt, MMRd, POLE-ultramutated, and p53abn subtypes, respectively. Regarding fertility outcomes, the p53wt group demonstrated the highest pregnancy rate after achieving a complete response compared to the other subtypes; however, no significant differences were observed in overall pregnancy outcomes. The ProMisE molecular classification holds significant prognostic value for patients with EC and AEH undergoing fertility-sparing treatment. Among the molecular subtypes, p53wt appears to be the most favorable for fertility-preserving interventions. This study provides essential insights into reproductive outcomes for this patient population.