A Study on Endometrial Polyps Recurrence Post-Hysteroscopic Resection: Identification of Influencing Factors and...

This study aimed to explore influencing factors and develop a predictive model of endometrial polyps (EP) recurrence after hysteroscopic resection.

This retrospective study included 180 patients who underwent hysteroscopic resection for EP between January 2021 to December 2023. The patients were divided into a modeling group (n = 135) and a validation group (n = 45) in a 3:1 ratio. The patients in the modeling group were further divided into a recurrence group (n = 35) and a non-recurrence group (n = 100) based on whether their polyps recurred. General information on patients was compared between the two groups. Univariate and multiple logistic regression analyses were conducted to identify factors influencing EP recurrence post-hysteroscopic resection. A predictive model was developed, and the receiver operating characteristic (ROC) curve analysis was performed to determine the clinical utility of the model.

Comparison of baseline characteristics between the modeling and validation groups showed no statistically significant differences (p > 0.05). However, 35 patients in the modeling group had recurrence, while 12 patients experienced recurrence in the validation group. Binary logistics regression analysis revealed matrix metalloproteinase-9 (MMP-9)/tissue inhibitor of metalloproteinase-1 (TIMP-1), hypoxia-inducible factor-1α (HIF-1α) and platelet-derived growth factor (PDGF) as independent predictors for polyp recurrence (p < 0.05). Furthermore, a model formula, p = eZ/1 + eZ, was developed. The slope of the calibration curve of this model in both groups were straight lines close to 1, indicating that the model's predicted recurrence risk strongly agreed with the actual risk. ROC analysis demonstrated that the area under the curve in the modeling group was 0.902, with standard error of 0.028 (95% confidence interval (CI): 0.885-0.954). The model yielded the Youden value of 0.79, with a sensitivity of 82.96% and a specificity of 95.66%. Moreover, the area under the curve in the validation group was 0.871, with a standard error of 0.040 (95% CI: 0.859-0.920). However, the model showed the Youden value of 0.59, with a sensitivity of 79.29% and a specificity of 79.96%. The Decision Curve Analysis (DCA) demonstrated significant clinical advantages of the model.

This study identified the influencing factors of EP recurrence and successfully constructed a predictive model based on these factors. After validation, the model demonstrates significant clinical utility.