Endometrial thickness and histological abnormalities in women on hormonal replacement therapy: a transvaginal ultrasound/hysteroscopic study

Objectives The aim of this study was to assess if endometrial thickness cou ld be used to select postmenopausal women on hormonal replacement therapy ( HRT) at increased risk for endometrial abnormalities. The secondary aim was to assess if endometrial abnormalities were more likely to occur in patien ts with increased endometrial thickness or in patients with unexpected blee ding. Methods Bi-endometrial thickness was measured by transvaginal ultrasound (T VS) in postmenopausal patients on sequential or combined HRT regimens. Wome n following a sequential regimen underwent TVS examination immediately afte r their withdrawal,al bleed, always between 5 and 10 days after the last pr ogesterone tablet. A hysteroscopy with Endometrial biopsy was performed wit hin 5 days after the TVS Examination, when endometrial thickness was greate r than or equal to 4 nlm, ol when unscheduled bleeding was observed. Results A total of 190 women were recruited. In 138 women on sequential reg imens, the mean value of endometrial thickness was 3.6 mot +/- 1.5, and in 52 women on combined regimens it was 3.2 mm +/- 1.8 (P = n.s.) Twenty-eight patients (15%) had an endometrial thickness >4 mm, 35 patients (18.4%) rep orted unexpected bleeding. The percentage of abnormal endometrial findings (9%; three of 35) in patients selected for unscheduled bleeding was signifi cantly lower? than the percentage of abnormal findings in patients selected for hysteroscopy for endometrial thickness >4 mm (36%; 10 of 28) (P < 0.01 ). All patients with unexpected bleeding and endometrial thickness less tha n or equal to 4 mm (24 cases) were found to have an atrophic endometrium. Conclusions Endometrial thickness in patients on sequential HRT, measured s oon after withdrawal bleeding, is not significantly different from thicknes s measured in patients on combined HRT. Patients on HRT with art endometria l thickness of > 4 mm could be considered for histological sampling. The pr evalence of abnormal endometrial findings in patients with a thick endometr ium is significantly higher than the prevalence observed in patients with u nexpected bleeding.