Metformin attenuates the production and proliferative effects of prolactin induced by medroxyprogesterone acetate during fertility-sparing treatment for endometrial cancer

Abstract

Background: Progestin is used for fertility-sparing treatment in cases of endometrial cancer (EC). Progestin can induce hyperprolactinemia by increasing pituitary secretion and endometrial decidualization. However, progestin induces prolactin (PRL) secretion, which stimulates cell proliferation and deleteriously affects treatment. To date, the detrimental effect of PRL, the secretion of which is induced by medroxyprogesterone acetate (MPA) during fertility-sparing treatment, has not yet been fully elucidated. Therefore, we aimed to assess the effects of PRL on EC cells during combined treatment with progestin and metformin.

Methods: In total, 71 patients with EC/endometrial atypical hyperplasia who underwent fertility-sparing treatment at our institution from 2009-2019 were enrolled. Serum PRL levels were determined using enzyme immunoassays; mRNA levels in endometrial tissues were determined using quantitative reverse-transcription PCR. To evaluate MPA-induced decidualization, cancer-associated stromal cells were enzymatically released from surgically removed specimens of six patients with EC. To examine PRL-induced cell proliferation, the EC cell lines Ishikawa, HEC1B, and HEC265 were used. In vitro cell proliferation was evaluated using the WST assay; protein levels of signaling molecules were determined using western blotting.

Results: MPA administration significantly increased serum PRL levels at 3 and 6 months and upregulated IGFBP-1 and PRL mRNA expression in tissues at 3 months of fertility-sparing treatment. Metformin significantly reduced MPA-induced IGFBP-1 and PRL mRNA expression during fertility-sparing treatment and significantly inhibited the upregulation of IGFBP-1 and PRL mRNA and PRL levels due to decidualization induced by MPA and cAMP treatment in primary cultured EC stromal cells. In vitro, PRL increased cell proliferation and ERK1/2 phosphorylation levels, whereas metformin attenuated these increases.

Conclusions: MPA upregulated PRL levels in serum and endometrial tissues during fertility-sparing treatment. Metformin co-administration reduced PRL production and attenuated PRL-induced cell-proliferation activity. This study may provide valuable insights on the application of metformin to improve the outcomes of fertility-sparing treatment.

Keywords: Endometrial neoplasms; Fertility preservation; Medroxyprogesterone acetate; Metformin; Prolactin.

Fig. 1

Serum prolactin levels during fertility-sparing treatment with MPA in patients with EC. Columns and error bars represent the mean ± standard deviation of the mean before treatment and after 3 and 6 months of MPA treatment. Wilcoxon signed-rank test was used and PRE indicates MPA pretreatment. (*P < 0.001). EC, endometrial cancer; MPA, medroxyprogesterone acetate

Fig. 2

IGFBP-1 a, PRL b, PgR c, and ERα d mRNA levels before and after MPA treatment. PRE: MPA pretreatment; POST: 3 months after MPA treatment; NS, not significant. (*P < 0.01, ** P < 0.05). ERα, estrogen receptor-α; IGFBP-1, insulin-like growth factor-binding protein 1; MPA, medroxyprogesterone acetate; PgR, progesterone receptor

Fig. 3

Decidualization of stromal cell by MPA and suppression by metformin. Stromal cell treatment with/without MPA (10^–6 M), cAMP (0.5 mM), and with/without metformin (1 mM) (8 days). IGFBP-1 a, PRL b, PgR c, and ERα d mRNA levels e. Columns, error bars represent the mean ± standard deviation of the mean. Met, metformin; NS, not significant (*P < 0.01, ** P < 0.05). MPA, medroxyprogesterone acetate

Fig. 4

Viability of Ishikawa, HEC1B, and HEC265 cells exposed to 500 ng/mL PRL for various durations. Cell proliferation at 72 h after prolactin (PRL) treatment a. Effect of metformin on PRL-induced cell proliferation at 72 h b. Columns, error bars represent the mean ± standard deviation of the mean. (*P < 0.01, **P < 0.05)

Fig. 5

PRL induced the activation of ERK1/2 and rpS6 and metformin attenuated this effect. We preliminary observed that phospho-ERK1/2 levels were the most elevated at 6 h after the addition of PRL, and this increase was time-dependent (data not shown).Total protein extracted from HEC1B cells a and HEC265 cells b exposed to 500 ng/mL of PRL with or without 1 mM metformin for 6 h was subjected to western blotting. Met indicates metformin