Objective
We sought to test if metformin could regress endometriotic explants in rats.
Study Design
After inducing endometriotic implants and randomization of female Wistar albino rats, they were given 25 and 50 mg/kg/day of oral metformin in group A (n = 9) and B (n = 8), respectively, for 28 days. Group C (n = 9) was given saline as placebo.
Results
Mean volume, weight, and histologic score of implants in groups A ( P < .01, P < .05, and P < .05, respectively) and B ( P < .01, P < .05, and P < .05, respectively) were significantly lower than in group C. The activity of superoxide dismutase and tissue inhibitor of metalloproteinase-2 staining in groups A ( P < .05 and P < .01, respectively) and B ( P < .01 and P < .01, respectively) was significantly higher than in the control group. Moreover, there were more significant reductions in implant levels of vascular endothelial growth factor and matrix metalloproteinase-9 in groups A (both P < .001) and B (both P < .001) than in group C.
Conclusion
Metformin causes regression of endometriotic implants in rats.
Endometriosis is an estrogen-dependent disorder defined as the presence of endometrial tissue outside of the uterine cavity. Although the pathogenesis of endometriosis still remains unclear, the most accepted theory assumes that the endometriotic implants originate by the migration of eutopic endometrial cells through retrograde transtubal flow into the peritoneal cavity of menstrual debris, with future implantation and growth on an ectopic site.
Oxidative stress has been proposed as a potential factor involved in the pathophysiology of the endometriosis. Reactive oxygen species exert their cytotoxic effect by causing peroxidation of membrane phospholipids, which results in increased cell membrane permeability, loss of membrane integrity, enzyme activation, structural damage to DNA, and, in effect, cell death. Malondialdehyde (MDA) is the breakdown product of the major chain reactions leading to oxidation of polyunsaturated fatty acids and thus serves as a reliable marker of lipid peroxidation mediated by oxidative stress. Moreover, superoxide dismutase (SOD) is one of the enzymes that prevents the adverse effects of oxidative stress by breaking down reactive oxygen species and inhibiting generation of the highly toxic hydroxyl radical.
Many factors are important for the degradation of extracellular matrix and the implantation of endometrial tissue in ectopic sites—notably, matrix metalloproteinases (MMPs). It has been shown that blocking MMP activity inhibits the formation of ectopic lesions in experimental models. Vascular endothelial growth factor (VEGF) has been implicated as inducer of attachment, proliferation, and neovascularization. Moreover, antiangiogenic agents inhibited the growth of explants in an in vivo model of endometriosis by disrupting the vascular supply.
Metformin is one of the oldest and most widely used oral agents in the treatment of type 2 diabetic subjects, with no effect on insulin secretion. Metformin has not only been shown to reduce the risk for vascular complications but also has protective effects largely independent of its well-known antihyperglycemic action. It has antioxidant properties and has beneficial effects on VEGF and MMPs. However, there is only 1 publication in the literature evaluating the effectiveness of metformin as an antiendometriotic drug, but no study was designed to determine the effects of metformin on experimentally induced endometriosis in rats.
In this study, the rat model of endometriosis was used to test whether metformin could cause regression of endometriotic implant and improve implant levels of SOD, MDA, VEGF, tissue inhibitor of metalloproteinase (TIMP)-2, and MMP-9.
Materials and Methods
Animals
The study was approved by the institutional review board of Ankara Education and Research Hospital and carried out at the Animal Research Center of Ankara Education and Research Hospital, Ankara, Turkey. Thirty mature, nonpregnant female Wistar albino rats (aged between 10–12 weeks) weighing between 190–230 g were used as a model for experimental induction of endometriosis. They were caged in a controlled environment of 22°C with 12-hour light/dark cycles. All rats were observed for 1 week to ascertain health before surgery. The guidelines for care and use of animals approved by the institutional review board were followed.
Surgical procedures
Operation 1
Endometriosis was surgically induced in 32 rats during laparotomy under ketamine hydrochloride (4 mg/kg intramuscularly) and xylazine hydrochloride (0.04 mg/kg intramuscularly) anesthesia by transplanting an autologous fragment of uterine tissue onto the inner surface of the abdominal wall, as described by Vernon and Wilson with minor modifications by Lebovic et al. All procedures were carried out under sterile conditions. A 3-cm vertical midline incision was made and both uterine horns were exposed. The right uterine horn was ligated with polypropylene 4-0 suture and a 1-cm segment was removed. The excised horn was immersed in sterile phosphate-buffered saline (PBS) solution and opened longitudinally. Then, the uterine segment was divided into 5- × 5-mm pieces. The endometrial fragment was transplanted without removing the myometrium onto the inner surface of the right abdominal wall with the serosal surface apposed and secured with nonabsorbable 4-0 polypropylene sutures at 2 edges ( Figure 1 , A). Before closure of the abdominal wall, 2 mL of saline was administered into the abdominal cavity to prevent drying and minimize adhesion formation. The midline incision was closed in 2 layers with the use of a simple interrupted 2-0 polyglactin 910 suture for the peritoneum-fascia and for the skin. All of the operations were performed by the first 3 authors. After the first surgical operation, all rats were caged individually and observed for 4 weeks without any medication. Body weight of the rats was monitored regularly, and they were not treated with estrogen.
Operation 2
In the 4 weeks after the first operation, 4 rats died. Two of the remaining 28 rats underwent a second exploratory laparotomy in which the viability of endometrial implants was checked, as in the previous 2 studies. Ectopic uterine tissues were identified and measured in 3 dimensions (length × width × height in millimeters) using a caliper by the fourth author. The spherical volume of each ectopic uterine tissue was calculated using the prolate ellipsoid formula: V (mm 3 ) = 0.52 × A × B × C, where A, B, and C denote width, length, and height, respectively. Tissues were photographed using a digital camera and measurements were recorded. The remaining 26 rats then were randomly allocated to 3 groups blinded to the surgeons using a computer-generated randomization (Random Numbers Generator Pro trial version; Segobit Software, Issaquah, WA). The rats in groups A (metformin 25-mg group, 9 rats) and B (metformin 50-mg group, 8 rats) were given 25 and 50 mg/kg/day of oral metformin (Glucophage; Merck Sante S.A.S., Istanbul, Turkey), respectively, as in the previous studies. Group C (control group, 9 rats) was given 2 mL of saline as placebo. The medications were given via orogastric tubes after dissolving in 2 mL of saline for 28 days. Every day each rat was held at the back of the neck region by an experienced technician. The head was gently squeezed and fixed with the help of the thumb and the index finger and an orogastric tube was inserted into the stomach. All the rats were observed for 4 weeks.
Operation 3
At 24 hours after the end of the medical treatments, the third laparotomy was performed, and the rats were killed by ketamine anesthesia. In the third laparotomy, the length, width, and height of the implants were measured by the fourth author, who was blinded to the groups. The endometrial explants were quickly excised by the first 3 authors and weighed (in milligrams) by the laboratory technician (Serkan Caliskan). The endometrial explants were divided equally into 2 longitudinal sections. One half was placed in formaldehyde solution for routine histopathologic and immunohistochemical examination. The other half of the endometrial explant was washed with physiologic saline for biochemical analyses. The histopathologist and biochemist assessing the samples were also blinded to the treatment groups.
Histopathologic examination
The histologic diagnosis of endometriosis was based on the morphologic identification of endometrial glandular tissue and stroma: glands and stroma of the endometrial type, with epithelial lining and luminal formation as in the rat endometriosis study. The persistence of epithelial cells in endometrial implants was evaluated semiquantitatively: 3 = well-preserved epithelial layer; 2 = moderately preserved epithelium with leukocyte infiltrate; 1 = poorly preserved epithelium (occasional epithelial cells only); and 0 = no epithelium.
Immunohistochemical examination
After a 1-hour application of primary antibody, including VEGF (rabbit polyclonal antibody, catalog no. RB-222-PO; Labvision/NeoMarkers Corp, Fremont, CA), TIMP-2 (catalog no. MS-1485-P; Thermo Scientific, Fremont, CA), and MMP-9 (catalog number RB-9234-P; Thermo Scientific), the samples were washed with PBS and a post-PBS level was applied. After rewashing with PBS, the specimens were placed in 10-minute 3-Amino-9-Ethylcarbazole chromogen. Finally, the counterstain with Mayer hematoxylin was performed for 5 minutes. Dilutions of the primary antibody were 1:100, 1:200, and 1:500 for VEGF, TIMP-2, and MMP-9, respectively. Amount of antibody added to each sample was 100 μL, and large-volume UltrAb Diluent was used for dilution (TA-125-UD; Thermo Scientific). Biotinylated goat antipolyvalent was used as second antibody (TP-125-BN; Thermo Scientific). Ten random fields and ×400 magnification were used to evaluate the samples. More than 80% of whole slide areas were evaluated. All slides were evaluated with a light microscope (DMI 4000 B; Leica, Wetzlar, Germany). To prevent interindividual bias, all tissues were evaluated by the same histologist (N.L.), who was blinded to the origin of the samples. The relative intensity of immunoreactivity staining was assessed quantitatively, as previously described by McCarty et al, taking into account both the intensity and the distribution of a specific staining. A value of histologic score was derived from the sum of the percentages of positively stained epithelial cells multiplied by the weighted intensity of staining: histological score = ΣPi(I + 1), where I represents staining intensity (0 = no expression, 1 = mild, 2 = moderate, and 3 = intense) and Pi is the percentage of stained cells for each intensity.
Biochemical analysis
MDA determination
Samples were subsequently homogenized in buffer and assayed for MDA content using the thiobarbituric acid reaction, as described by Mihara and Uchiyama. MDA content was then expressed as nmol/g of tissue. The MDA assay was performed for each animal at least in duplicate. All samples for MDA from all of the animals were done in 1 assay run; thus, interassay variation was avoided and coefficient of intraassay variation was 3.8%.
SOD activity determination
Uterus-derived SOD activity (expressed as U/mg protein) was determined using the Bioxytech SOD-525 assay (OxisResearch, Portland, OR). The SOD assay is based on change in absorbance (525 nm) that results from an increased rate of SOD-mediated autoxidation of 5,6,6a,11b-trihydroxybenzo[c]fluorene under alkaline conditions. The total protein content was measured with the method of Bradford. The SOD assay was performed for each animal at least in duplicate. All samples for SOD from all of the animals were done in 1 assay run; thus, interassay variation was avoided. Intraassay coefficient of variation was 7.32%.
Statistical analysis
All values are given as median (minimum-maximum) and mean ± SD. Normality was tested by Kolmogorov-Smirnov test. The Kruskal-Wallis test was used for comparison of the variables among the groups. The Mann-Whitney U test with Bonferroni correction was used for multiple comparisons if the difference was significant. Statistical significance was considered as P < .05.
Results
All laparotomy sites were intact, and none of the animals had an incisional hernia. The general appearance of the rats and weights were observed. The treatment had no adverse effects on the general health of the rats, and no deaths resulted from the treatment in any of the groups. Four rats died after the first surgery, possibly as a result of complications related to surgery.
Four weeks after autotransplantation of endometrial tissues, 2 rats were killed. Uterine autografts appeared as cystic structures containing serous fluid. The explants had an average spherical volume of 183 mm 3 and a weight of 217 mg. Histologically, the epithelia were well preserved, with a score of 3 for both samples.
After the third surgery, it was determined that endometrial explants produced viable implants in all of the animals. Representative in vivo images of the explants after 1 month of respective treatment revealed a cystic implant in the control group ( Figure 1 , B) in contrast to the degenerative, pellet appearance in a majority of the metformin-treated cohort of rats ( Figure 1 , C).
The Table exhibits the treatment results regarding mean volume and weight of the implants, as well as biochemical, histopathologic, and immunohistochemical findings. At the end of the treatment, the mean volume of the explants in group A (23 ± 21) and group B (92 ± 122) were significantly lower ( P < .01 and P < .01, respectively) when compared with the control group (396 ± 491). Likewise, the mean weight (±SD) of the implants was lower in groups A and B, when compared with group C (98 ± 72 vs 467 ± 473, P < .05, and 139 ± 188 vs 467 ± 473, P < .05, respectively).
