Assessing the Reliability of Hysteroscopic Sampling Methods for Diagnosing Atypical Endometrial Hyperplasia

Background/Objectives: The diagnosis of atypical endometrial hyperplasia (AEH) is associated with a high rate of concurrent endometrial cancer (EC). This occurrence can be particularly challenging in premenopausal women wishing to become pregnant, as they may be subjected to conservative treatment. The type of endometrial sampling may affect this outcome. Currently, the recommended type of endometrial sampling is under hysteroscopic guidance. There is scant literature regarding the reliability of hysteroscopically guided biopsy (HSC-bio) and hysteroscopic endometrial resection (HSC-res) on this topic. We aimed to assess the underestimation rate of EC in AEH, according to different hysteroscopic sampling methods. The secondary outcome was to evaluate the procedure performance in pre- and postmenopausal women. Methods: We conducted a multi-institutional retrospective study that included 536 women diagnosed with AEH who underwent hysterectomy between 2015 and 2020. Patients were divided into two groups based on the initial diagnostic approach for AEH: HSC-bio and HSC-res. The comparison was performed using univariate and multivariate analyses. Results: 160/536 women (29.9%) showed EC at hysterectomy. Overall, the following rate of EC underestimation was found: HSC-bio = 32.1%, HSC-res = 24.2%, p = 0.07. After adjusting for baseline characteristics using logistic regression analysis, overall, there was no significant association of EC underestimation according to the type of sampling procedure. Interestingly, in premenopausal women, including 161 cases, the rate of EC underestimation in HSC-bio and HSC-res was 28.8% vs. 14.0%, respectively (p = 0.034). Conclusions: There were no significant differences in EC underestimation between the two hysteroscopic procedures in the entire cohort of women with AEH. Limited to the secondary objective, the significant findings in premenopausal women may be of particular clinical interest, as this population may undergo conservative treatment.