High dose estrogen therapy has been used effectively in the treatment of human breast cancer. To understand the mechanisms involved, the effects of high concentrations (5–100 µm) of estrogens were studied in estrogen receptor (ER) positive (T-47D and MCF-7) and ER negative (MDA-MB-330) human breast cancer cell lines in vitro. Inhibition of cellular proliferation was seen with the synthetic estrogen diethylstilbestrol (DES) at concentrations >10 µm in each of the three cell lines. In T-47D cells DES was shown by clonogenic survival assays to be cytotoxic. This effect was evident in both plateau phase and exponentially growing cultures, in contrast to the effects of the antiestrogen tamoxifen, which has minimal effects on plateau phase cells. The effects of DES on the proliferation of exponentially growing cultures were accompanied by changes in cell cycle parameters which included an increase in the percentages of S-phase, G₂ + M, and polyploid cells and a corresponding decrease in the percentage of G₀-G₁ cells. These changes, which contrasted with the known effects of tamoxifen, were not seen in the non- or slowly cycling plateau phase T-47D cells. Such results are consistent with two mechanisms of action of high dose estrogen in vitro: a cell cycle phase-specific effect and cell cycle-independent cytotoxicity.

The stereoisomers 17α-estradiol and 17β-estradiol had similar potency to DES in inhibiting cell proliferation and inducing these changes in cell cycle parameters in both MCF-7 and MDA-MB-330 cells. The high-dose estrogen effect was ligand specific in that estrone and estriol were less potent than DES, 17α-estradiol and 17β-estradiol in inhibiting cell proliferation, and the characteristic cell cycle changes were produced only by concentrations of estriol >75 µm and not at all by estrone at concentrations up to 100 µm. The androgens testosterone and dihydrotestosterone were similar in effect to estrone.

The cell cycle changes associated with estrogen-induced growth inhibition in vitro are identical to those observed during regression of ER positive but not ER negative human tumor xenografts in nude mice. However, the role of ER in mediating estrogen-induced regression of ER positive tumors in vivo remains undefined.