Assessment of Risk Scoring Model for Prediction of Endometrial Cancer Among Symptomatic Postmenopausal Women (A Prospective Cohort Study)

Abstract

Abstract Background Endometrial carcinoma is the most common gynecological malignancy in the developed countries and the third common gynecological malignancy in Egypt after breast and ovarian cancers. Aim of the Work to evaluate this risk scoring model on Egyptian patients and to study the effect of adding other patient characteristics (DM, BMI and relevant family history) on the sensitivity and specificity of RHEA scoring model. Patients and Methods The current study was conducted in Ain Shams University Maternity Hospital in the period between September 2017 and December 2018. A total of 100 women with postmenopausal bleeding and endometrial thickness > 4mm were included in the study. Results Histological examination revealed that benign pathology (n = 65) (73%) was found to be: most common cause was endometrial hyperplasia without atypia (20.3%) followed by chronic endometritis (13.5%), then endometrial polyp (11.3%), cystic atrophy of endometrium (8.9%), proliferative endometrium (8.9%), endometrial hyperplasia with atypia (6.7%) and lastly mucous polyp (3.4%) while malignant histopathology(n = 24)(27%) which is significantly higher than the international rates showed: Endometriod adenocarcinoma (n = 19)(21.3%), papillary serous carcinoma (n = 4)(4.5%) and undifferentiated carcinoma (n = 1)(1.1%). The current study showed that RHEA score performs in our study population with a comparable validity to that reported by its inventors with sensitivity 79.2% (57.8% - 92.9%) vs. 87.5% and specificity 84.6% (73.5% - 92.4%) vs. 80.1% respectively. In results of the current study it was found that the time since onset of menopause rather than age was associated with endometrial cancer with the optimum cut-off for postmenopausal duration was estimated to be 9 years achieving a sensitivity of 87.5% and a specificity of 60.0%, but it needs multivariate analysis on larger and more representative sample size to confirm this association, A statistically significant regression model was including only postmenopausal duration, recurrent bleeding and endometrial thickness. None of age, BMI, family history or hypertension proved a statistically significant predictive effect after adjustment for other predictive variables. Conclusion Taking in consideration the higher prevalence of endometrial carcinoma in the sample of the current study, the wide 95% confidence intervals for the different validity indices for the RHEA scores derived from this study, it seems that RHEA score performs in this study population with a comparable validity to that reported by its inventors.