Tamoxifen-induced hyperplasia of the endometrium

Greatly thickened endometria are often observed in vaginal sonography in patients taking tamoxifen. The incidence of type I carcinoma of the endometrium is also raised under tamoxifen. Both effects are ascribed to the known partial estrogenic effect of tamoxifen. However, if sonographically hyperplastic endometria are abraded, mostly only atrophic material is obtained. To clarify this contradiction, 89 patients treated with tamoxifen were investigated by means of vaginal sonography at intervals of 6 months. In 33 of them (37%), an endometrial thickness > 8 mm was found. In order to investigate the endometrial thickness throughout, a hysteroscopic loop resection (polyps n = 18, diffuse hyperplasia n = 15) and afterwards histological work-up of the endometrium in serial sections were carried out in all these patients. Hyperplasia of the endometrial stroma without cellular atypia was found in all cases. However, the glandular epithelium was always atrophic. This explains the contradiction between sonographically thick endometrium (stromal hyperplasia) and the scanty abrasion material (atrophy of the glandular epithelium). The 'hyperplasia' found sonographically is not necessarily estrogen-dependent hyperplasia of the glandular epithelium leading to a carcinoma. This raised the question as to the etiological reasons for the increased incidence of carcinomas. We discuss the hypothesis that tamoxifen may also promote the conversion of pre-existent subclinical lesions into manifest carcinomas.