ULTRASONOGRAPHIC ENDOMETRIAL THICKNESS AS A PREDICTOR OF THE RISK OF ENDOMETRIAL CANCER IN PATIENTS WITH POSTMENOPAUSAL BLEEDING

Abstract

Objective: to determine the diagnostic performance of endometrial thickness measured by transvaginal sonography in diagnosing endometrial cancer in patients presenting with post-menopausal bleeding. Patients and Methods: The databases of the Department of Gynecological Oncology at the University Clinic of Gynecology and Obstetrics in Skopje, in the period January – December 2015 were searched in order to identify patients that underwent endometrial sampling due to newly-diagnosed postmenopausal bleeding. Results: A total of 158 patient records that met the criteria were identified. The prevalence of endometrial cancer was 15.2%. Endometrial thickness was a statistically significant independent predictor of the presence of endometrial cancer and atypical endometrial hyperplasia (OR 1.19 95% CI 1.09-1.29 for each 1mm increase in thickness, p<0.001). The ROC curve analysis in our study had an AUC of 0.83 (p<0.001) and identified a cut-off level for endometrial thickness of 8mm which was associated to a sensitivity of 88.9%, specificity of 65.6%, PPV of 34.8% and NPV of 96.6% for the detection of endometrial cancer. Using a cut-off for endometrial thickness of ≤3mm achieved 100% sensitivity. Conclusion: None of the analyzed cut-off points for endometrial thickness achieved optimal diagnostic accuracy, as all cut-off values associated to sensitivity rates above 95% had false positive rates of over 60%. Nevertheless, an endometrial thickness cut-off of 3mm, due to the associated high sensitivity, can safely be used to identify women with postmenopausal bleeding who are highly unlikely to harbor endometrial cancer and that can forego initial endometrial sampling.