Metformin decreases the incidence of ovarian hyperstimulation syndrome: an experimental study

Abstract

Background: In assisted reproduction cycles, gonadotropins are administered to obtain a greater number of oocytes. A majority of patients do not have an adverse response; however, approximately 3-6% develop ovarian hyperstimulation syndrome (OHSS). Metformin reduces the risk of OHSS but little is known about the possible effects and mechanisms of action involved.

Objective: To evaluate whether metformin attenuates some of the ovarian adverse effects caused by OHSS and to study the mechanisms involved.

Material and methods: A rat OHSS model was used to investigate the effects of metformin administration. Ovarian histology and follicle counting were performed in ovarian sections stained with Masson trichrome. Vascular permeability was measured by the release of intravenously injected Evans Blue dye (EB). VEGF levels were measured by commercially immunosorbent assay kit. COX-2 protein expression was evaluated by western blot and NOS levels were analyses by immunohistochemistry.

Results: Animals of the OHSS group showed similar physiopathology characteristics to the human syndrome: increased body weight, elevated progesterone and estradiol levels (P<0.001), increased number of corpora lutea (P<0.001), higher ovarian VEGF levels and vascular permeability (P<0.001 and P<0.01); and treatment with metformin prevented this effect (OHSS+M group; P<0.05). The vasoactive factors: COX-2 and NOS were increased in the ovaries of the OHSS group (P<0.05 and P<0.01) and metformin normalized their expression (P<0.05); suggesting that metformin has a role preventing the increased in vascular permeability caused by the syndrome.

Conclusion: Metformin has a beneficial effect preventing OHSS by reducing the increase in: body weight, circulating progesterone and estradiol and vascular permeability. These effects of metformin are mediated by inhibiting the increased of the vasoactive molecules: VEGF, COX-2 and partially NOS. Molecules that are increased in OHSS and are responsible for a variety of the symptoms related to OHSS.

Figure 1

Effect of the treatment with Metformin on the ovarian morphology in an OHSS rat model. This is a quantitative morphometric analysis of the ovarian sections after Masson trichrome staining. Follicles were classified according to the stage of development indicated in materials and methods. Values are expressed as the number of follicles per 10 mm². Data represent the mean ± S.E.M for 10 ovaries from different animals with the same treatment, and each value represents the mean of 10 sections from the same ovary. ** P< 0.01 and *** P< 0.001 respect to the control group; ^# P< 0.05 between the pointed groups.

Figure 2

Vascular permeability measured as concentration of Evans Blue (EB) in ovaries homogenates and in peritoneal fluid. Each bar represents the mean ± S.E.M and the P values were determined by one way ANOVA, followed by Tukey’s test. * P< 0.05, ** P< 0.01 and *** P< 0.001 respect to the control group; ^# P< 0.05 between the pointed groups.

Figure 3

Concentration of the vascular endotelial growth factor (VEGF) in ovaries homogenates and in serum. Each bar represents the mean ± S.E.M and the P values were determined by one way ANOVA, followed by Tukey’s test. *** P< 0.001 respect to the control group and # P< 0.05 between the pointed groups.

Figure 4

Protein expression of Cycooxygenase type 2 (COX-2) in the ovaries. Each bar represents the mean ± S.E.M and the P values were determined by one way ANOVA, followed by Tukey’s test. * P< 0.05 respect to the control group and ^# P< 0.05 between the pointed groups.

Figure 5

NOS immunoreactivity in control and treated rat ovaries. (A) Representative ovarian sections showing the immunolocalization of NOS from: upper panel: a control rat; middle panel: a OSSH rat; and lower panel: a OHSS+MET rat. A, C and E shows developed follicles (×100) while B, D and F (×100) shows corpora lutea. (B) Densitometric analyses of NOS immunoreactivity in the rat ovaries. Each bar represents the mean ± S.E.M and the Pvalues were determined by one way ANOVA, followed by Tukey’s test. ** P< 0.01 respect to the control group and #P< 0.0001 between the pointed groups.