Ultrasonographic endometrial thickness measurement is predictive for treatment response in simple endometrial hyperplasia without atypia

Abstract

We sought to determine the predictors of treatment response in simple endometrial hyperplasia without atypia. We prospectively treated 67 women with simple endometrial hyperplasia without atypia who were administered cyclic oral medroxyprogesterone acetate 10 mg/day for 12 days of luteal phase for 3 months and underwent control endometrial sampling after treatment. All subjects were evaluated in terms of age, gravidity, parity, body mass index (BMI), menstrual cycle, endometrial thickness, uterine fibroids, ovarian cysts, serum CA 125 levels, systemic disorders and cigarette smoking. All parameters were used to predict treatment success. Persistent hyperplasia was observed in 11 subjects. Endometrial thickness was significantly correlated with treatment failure (r=0.293, p=0.015). In ROC analysis, endometrial thickness was found to be predictive for persistent hyperplasia (area under curve: 0.724, P=0.019). Optimal cut off value was calculated to be 16.5 mm with 64% sensitivity, 72% specificity and 91% negative predictive value. The number of persistent hyperplasia in women with and without endometrial thickness greater than 16.5 mm was significantly different (7/23 vs. 4/45, p=0.029). Odds ratio of endometrial thickness higher than 16.5 mm for treatment failure was 4.4 (95% CI, 1.2-17.4, p=0.03). Results of this study suggest treatment modification according to the baseline endometrial thickness in patients with simple endometrial hyperplasia without atypia.