Endometrial thickness assessed by transvaginal ultrasound insufficiently predicts occurrence of hyperplasia during unopposed oestrogen use.

In the present study, unopposed oestrogens were given to 184 non-hysterectomized women, for a duration varying from four weeks to 24 months, while the endometrial responses were monitored by transvaginal sonography (TVS) and backed up by endometrial sampling (conducted when the endometrial thickness reached 8 mm (double layer) or more, vaginal bleeding occurred during oestrogen administration, or after one year of unopposed oestrogen use). In cases where the endometrial thickness reached 8 mm or more, progestogens were administered for 12 days. In 64% of the women, administration of progestogen could be postponed until at least the fourth month of treatment under the pre-defined decision criteria. Eleven percent of the patients used oestrogens continuously during the two year study period, without any need of additional progestogen. In total, 338 endometrial biopsies were performed; 16 cases of hyperplasia were detected. In three cases, the corresponding endometrial thickness was below 8 mm (in one case, 5 mm). Endometrial thickness could not consistently predict occurrence of hyperplasia. In eight cases, hyperplasia occurred within 4 months of treatment, and in four cases, within only 2 months (of which only one case could possibly be attributed to previous hormone use and none to endogenous oestrogen production). The rapid occurrence of hyperplasia should be taken into account in studies of quarterly progestogen administration with hormone replacement therapy. It is concluded that postponement of progestogen administration with hormone replacement therapy under guidance of TVS only (without biopsies) would not be adequately safe to be recommended for clinical practice.