Evaluation of pre-operative endometrial biopsy accuracy and etiological distribution in abnormal uterine bleeding using the PALM-COEIN classification.

Objective: Abnormal uterine bleeding (AUB) is one of the leading causes of gynecological consultations and indications for hysterectomy. The aim of this study is to determine the distribution of aetiological causes in patients undergoing hysterectomy due to AUB according to the FIGO PALM-COEIN system and to evaluate the diagnostic concordance between pre-operative endometrial biopsy (EB) results and final hysterectomy pathology. Methods: This retrospective, descriptive and analytical study included 296 patients who underwent hysterectomy due to AUB at Bandirma Onyedi Eylül University Hospital between 2020 and 2025. The patients' demographic data, clinical characteristics, and histopathological results were examined from hospital records. The causes of AUB were categorized according to the PALM-COEIN system (Polyps, Adenomyosis, Leiomyoma, Malignancy; Coagulopathy, Ovulatory, Endometrial, Iatrogenic, Unclassifiable). The concordance of pre-operative EB results with final pathology was analyzed using sensitivity, specificity, positive/negative predictive values, the Kappa coefficient, and ROC analysis AUC. Results: The mean age of the patients was 48.2 ± 6.9 years. According to the PALM-COEIN classification, structural causes (PALM) were the most common etiology, with leiomyomas (35.5% isolated, 59.8% total) forming the pre-dominant group. The rate of post-menopausal bleeding was determined to be 28.0%. Regarding concordance between pre-operative EB and final pathology, EB demonstrated high specificity but limited sensitivity, particularly for focal lesions. For AEH+ (atypical endometrial hyperplasia and above), sensitivity was 22.2%, specificity was 99.6%, and AUC was 0.609; for endometrial polyps, sensitivity was 30.6%, specificity was 83.0%, and AUC was 0.568. Overall agreement between biopsy and final pathology (Kappa) was low to moderate, and ROC analysis indicated limited discriminatory ability across subgroups. Conclusion: Among patients undergoing hysterectomy for AUB, structural pathologies, particularly leiomyomas, constitute the primary reason for hysterectomy. Although pre-operative EB is successful in ruling out pathologies, it may be insufficient in detecting focal lesions and some pre-malignant conditions due to its low sensitivity. Therefore, EB results should be supported by advanced diagnostic imaging methods, especially when clinical suspicion exists during the decision-making process for hysterectomy.