{"paper_id":"4b5c08a2-67d7-4c3f-bf33-9efad7ef2a54","body_text":"Endometriosis, an estrogen-dependent inflammatory disease \naffecting 10-25% of women, is associated with significant \nreductions in fertility and is one of the most common \nbenign gynecological diseases. Retrograde menstruation \nwith subsequent adhesion formation, invasion, and neovascularization \nare believed to give rise to the occurrence \nof endometriosis lesions. The most common locations for \nendometrial implants are the ovaries, fossa ovarica, uterosacral \nligaments, and posterior cul-de-sac ( 1 ).\nAlthough different medications are used to control endometriosis, \ntheir adverse effects following long-term use \nand recurrence of disease after discontinuation of therapy \nlimit their applications. Additionally, these medications \nare not useful for endometriosis-associated infertility \n( 2 ). Regarding the fact that no ideal medical treatment is\navailable to control endometriosis, introducing new medical \nagents with minimal side effects and improved effectiveness \nfor infertility treatment, are required.\nGonadotropin releasing hormone agonists (GnRHa) \nsuch as diphereline, as standard medications for the treatment \nof endometriosis, are able to induce inactivation and \ndegeneration of endometrial implants via suppression of \nhypothalamic-pituitary-gonadal axis and ovarian estrogen \nproduction ( 3 ). GnRHa not only induces amenorrhea, but \nalso may cause hot flush, depression, headache, hair loss, \nmusculoskeletal stiffness, vaginal dryness and bone loss ( 4 ).\nIt is known that in women with endometriosis, the growth \nof endometrial cells within the peritoneal cavity is induced by \ninflammatory mechanisms ( 5 ); so, anti-inflammatory drugs \nare suggested to control endometriosis growth. Cyclooxygenase \nenzymes (COXs), known as prostaglandin-endoperoxide \nsynthase, are responsible for formation of inflammatory \nmediators such as prostaglandins. COX-1 is expressed in almost \nall cells for maintenance of cell. COX-2 is produced at \nsites of inflammation, angiogenesis, and estrogenic cellular \nprocesses. Pharmacological inhibition of COX-2 was able \nto reduce the survival and growth of endometrial tissues at \nectopic sites ( 6 ). NSAIDs (non-steroidal anti-inflammatory \ndrugs) such as celecoxib, inhibit cyclooxygenase isoforms \nand induce gastrointestinal side effects ( 7 ).\nUsing herbal medicine has always played a significant \nrole in Iranian culture and civilization and some of these \nherbs have been recommended for treatment of infertility-\nrelated diseases. Licorice ( Glycyrrhiza glabra ), is one of \nthe most widely used herbal drugs in Iranian traditional \nmedicine. Licorice root contains triterpene, saponins, flavonoids, \nisoflavonoids, hydroxycoumarins, steroids and \nvolatile oil. Licoricidin, is a potent compound isolated \nfrom licorice root ( 8 ). Studies showed that licoricidin is a \nselective COX-2 inhibitor and inhibits phospholipase A2 \nactivity that is a critical enzyme involved in numerous inflammatory \nprocesses ( 9 ,  10 ). Licorice root with its anti-\ninflammatory/anti-platelet, antiviral, antifungal and mineralocorticoid \nfunctions has been used for the treatment \nof gastric ulcers, cough and bronchitis since the ancient \ntimes. Licorice is not recommended to be used for more \nthan 6 weeks. Complications such as hypokalemia, hypernatremia, \nedema, hypertension and cardiac complaints are \nassociated with long-term time use of licorice ( 8 ).\nWe hypothesized that licorice or celecoxib might be \ngood candidates for treatment of endometriosis as an inflammatory \ncondition. In the present study, we compared \nthe effects of licorice, celecoxib and diphereline on the \ngrowth of endometrial implants in rats.\n\nIn this experimental study, 48 mature female Sprague-\nDawley rats (almost 8 weeks old, weighting 220 ± 20 g) \nwere purchased from the Center of Comparative and Experimental \nMedicine at Shiraz University of Medical Sciences \n(SUMS), Shiraz, Iran. The animals were kept on 12 \nhours light: 12 hours dark cycles at a controlled temperature \nwith free access to water and food. The animal experiments \nwere performed according to the principles of \nthe care and use of laboratory animals established by the \nNational Institutes of Health, Bethesda, MD, USA, and \napproved by the Institutional Animal Ethics Committee at \nSUMS (No. 92-01-01-6869). These animal experiments \nwere performed in the animal house of Shiraz University \nof Medical Sciences. To select the rats with normal estrous \ncycle, daily vaginal smears were taken and evaluated \nby a light microscope. Rats with three normal estrous \ncycles were used in the next steps.\nLicorice roots were purchased from herbal stores in \nShiraz, Iran).  Glycyrrhiza glabra  was preserved in herbarium \nafter authentication by a botanist (Voucher No. \nPM 684). L. Licorice roots were air-dried, powdered and \nan alcoholic extract was produced by using ethanol (80%) \nand percolation method. Then, solvent was completely \nremoved by drying under reduced pressure in a rotary \nevaporator. The extract was stored at 4°C until use.\nEndometriosis was induced surgically using the method \ndescribed by Vernon and Wilson with little modifications \n( 11 ) ( Fig .1 ). It should be mentioned that as the growth \nof endometriosis is estrogen-dependent, if induction of \nendometriosis is performed in an ovariectomized animal, \nestrogen supplementation is mandatory ( 11 ,  12 ). However \nsimilar to the previous researches, since in our study adult \nintact rats were used, we did not use exogenous estrogen \nfor induction of endometriosis ( 11 - 13 ).\nBriefly, all the female rats were anesthetized using ketamine \nhydrochloride 10% (100 mg/kg, Alfasan, Netherlands) \nand xylazine 2% (10 mg/kg, Alfasan, Netherlands). \nThen, rats’ abdomen was opened through a 4 cm \nmidline incision starting from 1 cm below the xiphoid. \nThe left horn of uterus was ligated at both the uterotubal \nand cervical junction ends and removed. A longitudinal \ncut was made through the uterine horn. By a punch biopsy, \n4 round pieces of the distal part of uterine tissue \nwere excised (4×4×1 mm) and placed in warm sterile \nsaline 0.9%. Two implants were sutured with proline \n5-0, one on the left and the other on the right side of the \nperitoneal cavity on the areas of well-visible vasculature \nwith endometrial surface facing the peritoneum. Finally, \nthe abdominal muscles, fascia and skin were sutured. \nThen, chlortetracycline (Cyclo Spray, Eurovet, UK) was \nsprayed on the incisions site and animals were allowed \nto recover from anesthesia. Two rats died due to hemorrhage \nat this stage. Six weeks after the first surgery, a \nsecond look laparotomy was performed and the viability \nof endometrial implants was confirmed by observation \nof good vascular supply and pinkish colored tissue \nin contrast to necrosis and fibrosis seen in two rejected \ncases as showed in Figure 1.\nThe flow diagram of the study.\nAt this stage, 44 female rats were divided into 4 groups \n(11 rats in each group). The control group was treated by \n0.5 ml of saline 0.9%/day, the second group by licorice \nroot extract (3000 mg/kg/day) and the third group took \ncelecoxib (Damloran Razak Pharmaceutical Co., Iran, 50 \nmg/kg, twice a day, dissolved in 0.5 ml of saline 0.9%) for \nthe next 6 weeks. All the treatments of these three groups \nwere administered by oral gavage. The fourth group received \na single IM injection of diphereline S.R. 11.25 mg \n(3 mg/kg, Ipsen, France). Six weeks after the treatments \nthe rats were sacrificed and endometrial implants were \nevaluated as shown in Figure 2.\nEndometrial implants in different times and groups. A. The first \nlaparotomy: auto-transplant of endometrial implants on the peritoneum, \nB. The second look surgery: shows the adhesion bands and endometrial \nimplants six weeks after induction of endometriosis, C. The third surgery: \nnecropsy of a rat in the control group showing growth of implanted lesions \nof endometriosis, and D. The third surgery: necropsy of a rat in diphereline \ngroup showing regression of the implants.\nThe length, width and height of each implant were \ncarefully measured using collis rulers by one researcher \nwho was blinded to the treatment arms. The area \nof the endometrial implants in four groups was measured \nby multiplying length by width), and the volume \nwas calculated by ellipsoid volume formulation \n(π/6×length×width×height).\nAll of the endometrial implants were fixed in formalin, \nplaced in paraffin, cut into 5 µm sections, stained with \nhematoxylin-eosin and evaluated by the same pathologist. \nPhotographs were taken by a digital camera (Sony, \nJapan). To classify the persistence of epithelial cells in \ngrafts, the scoring system applied by Keenan et al. ( 14 )\nwas used with score 0 showing no epithelial layer, and \nscores 1, 2 and 3 show poorly, moderately and well-preserved \nepithelial layers, respectively. The percentage of \nhemosiderin-laden macrophages (HLMs) was also measured \nin all of the sections. The surgeon, pathologist, and \nthe lab technicians were blinded to the groupings, medications, \nand specimens.\nFor statistical analysis, the software SPSS 15 (SPSS \nInc., Chicago, USA) was employed. To compare the mean \narea and the mean volume, ANOVA followed by Tukey \nHSD test was performed. To compare the histopathologic \nscoring, Kruskal-Wallis test and Mann-Whitney U test \nwith Bonferroni correction were applied. A P<0.05 was \nconsidered significant.\nThe mean area, volume and pathologic scores of implants in control, licorice, celecoxib and diphereline groups\nData was shown as mean ± SD. P<0.05 were considered statistically significant.  a ; Statistically significant differences between licorice and the control group,  b ; Statistically significant differences between diphereline and the control group,  c ; Statistically significant differences between control and other groups, and  d ; Statistically significant differences between Licorice, diphereline and celecoxib group.\n\nTwo rats died during laparotomy due to hemorrhage \nand in two other rats the implants did not grow. The \nremaining 44 rats were divided into 4 groups and \ntreated. In licorice group, the mean area and volume \nvalues of endometrial implants were significantly \nlower than those of the control group (P=0.042 and \nP<0.001, respectively) ( Table 1 ). The mean area and \nvolume of endometrial implants in the celecoxib \ngroup were lower compared to the control group, but \nthe differences were not statistically significant (P= \n0.953 and P=0.818, respectively). The mean area and \nvolume of diphereline group were significantly lower \ncompared to the control group (P<0.001 and P<0.001, \nrespectively). The pathologic scores of the licorice \nand celecoxib groups were lower than those of the \ncontrol group, but the differences were not statistically \nsignificant (P=0.221 and P=0.960, respectively). \nPoorly preserved epithelial layers were observed in \ndiphereline group and the mean pathological score in \nthis group, was significantly lower compared to the \ncontrol group (P<0.001,  Fig .3 ).\nSpecimens of the treated groups stained by hematoxylin-eosin.  A . \nWell preserved epithelial layer of endometrial implants in the control group \n(grade 3) (scale bar: 50 µm),  B . Poorly preserved epithelial layer of endometrial \nimplants in the licorice group (grade 1) (scale bar: 50 µm),  C . Moderately \npreserved epithelial layer of endometrial implants in licorice group \n(grade 2) (scale bar: 100 µm), and D. Poorly preserved epithelial layer of \nendometrial implants in diphereline group (grade 1) (scale bar: 100 µm). \nArrows demonstrate epithelial layer of endometrial implants.\nThe percentage of HLMs in endometrial implants of \nrats in celecoxib, licorice and diphereline group was significantly \nlower than that of the control group (P=0.004, \nP=0.000 and P=0.000, respectively). Also, the percentage \nof HLMs was significantly lower in licorice and diphereline \ngroup compared to celecoxib group (P<0.001 and P<0.001, \nrespectively). The percentage of HLMs was not different \nbetween licorice and diphereline group (P=1.000,  Fig .4 ).\nHemosiderin-laden macrophages (HLMs) in the specimens of different \ngroups.  A.  Control group and  B.  Licorice group (scale bar: 10 µm). \nArrows indicate hemosiderin - laden macrophages.\n\nWe compared the effects of licorice, celecoxib, and \ndiphereline in a rat model of endometriosis induced by \nauto-transplantation of endometrium on the peritoneal \nsurface as a well-established method ( 11 ). Licorice decreased \nthe growth of endometrial implants; celecoxib \nhad no significant effect and diphereline had the highest \npotency in suppression of the endometrial growth. According \nto our knowledge, this is the first study on the \neffect of licorice on the endometrial implants.\nPrevious studies showed that glycyrrhetinic acid as a \nconstituent of licorice extract, inhibits thrombin-induced \nplatelet aggregation ( 9 ,  15 ) and has steroid-like anti-inflammatory \neffects similar to glucocorticoids ( 16 ,  17 ). \nPark and colleagues showed that administration of hexane/\nethanol extract of Glycyrrhiza uralensis to mice decreases \ncell proliferation, inhibits the expression of angiogenic \nand inflammatory proteins and induces cell cycle \narrest or apoptosis ( 18 ). Also, they observed that licoricidin \nreduces macrophages number and tumor growth in the \ntumor microenvironment. In another study, it was shown \nthat licoricidin inhibits the metastatic and invasive capacity \nof malignant prostate cancer cells in vitro ( 19 ). La et \nal. ( 20 ) reported that licoricidin suppresses the production \nof inflammatory cytokines. The anti-inflammatory property \nof licoricidin is due, in part, to the inhibition of phospholipase \nA2 activity, resulting in inhibition of cyclooxygenase \nactivity and prostaglandin formation ( 9 ,  16 ,  17 ). \nLicoricidin also inhibits an isomer of platelet-activating \nfactor and acetyltransferase resulting in an anti-inflammatory \nactivity ( 21 ).\nCOX-2 overexpression has been detected in both eutopic \nand ectopic endometrium, and also in peritoneal \nmacrophages derived from women with endometriosis \n( 22 ,  23 ). In the family of selective COX-2 inhibitors, rofecoxib \nand valdecoxib, are no longer used because of \ntheir side effects but celecoxib with lower gastrointestinal \nproblems is still used.\nHisthopathologic slides prepared from endometriosis \nshowed higher counts of HLMs that serve as an indirect \nevidence for diagnosis of endometriosis. As shown in our \nstudy, endometrial implants had normal growth with intact \nendometrial lining and more scattered foci of HLMs \nafter treatment with celecoxib or normal saline. However, \nafter taking licorice or diphereline, growth of endometrial \nimplants were highly limited with lower HLMs. These \nfindings are in favor of the potential therapeutic effect of \nlicorice in suppression of endometrial implants growth.\nIn this study, celecoxib did not significantly reduce \nthe growth of endometrial implants that was against our \nprimary hypothesis. There are studies that showed that \ncelecoxib, dexketoprofen trometamol or rofecoxib were \nable to cause regression and atrophy of endometriosis lesions \n( 7 ,  21 ,  24 ). However, Hull et al. ( 25 ) showed that \nsubcutaneous injection of nimesulide, a COX-2 inhibitor \ndid not reduce the size and number of the endometriosis \nlesions that is in agreement with our results. We believe \nthat ineffectiveness of celecoxib in our study might \nhave been because of two reasons. First, it was previously \nshown that COX-2 immunostaining density was greater \nin ovarian endometrioma than in peritoneal implants and \nit was concluded that celecoxib might influence ovarian \nendometrioma more than peritoneum ( 26 ). Second, the \ncelecoxib brand that we used was different from that of \nother studies.\nSince endometriosis is a chronic disease and needs \nlong-term treatment, complications of prolonged use of \nlicorice such as hypokalemia, hypernatremia, edema, hypertension \nand cardiac complaints should be kept in mind \nbefore human application. It should be considered that \nthe maximum permitted dosage of licorice root is 5 to \n15gr/day and the duration of treatment should not exceed \n6 weeks in humans ( 8 ). Studies on all components and \nfractions of licorice are also needed to discover its active \ncomponent(s) and exact mechanism(s) of action to introduce \na safe pharmacological agent with targeted effects \nand without adverse effects.\nAs a limitation of our study, we did not evaluate inflammatory \nmarkers such as white blood cells counts, nor \ninterleukin-6 (IL-6), vascular endothelial growth factor \n(VEGF) and tumor necrosis factor-alpha (TNF-a) levels \nin the peripheral blood and peritoneal fluid before and after \ninterventions to assess the anti-inflammatory properties \nof licorice.\nWe believe that licorice might have the potency to be \nused as a novel and excellent alternative in the management \nof endometriosis after in-depth investigations in animals \nand humans.\n\nLicorice decreased the growth and histopathologic \ngrades of auto-transplanted endometrial implants. However, \ncelecoxib had no significant effect and diphereline \nhad the highest potency in reduction of the endometrial \ngrowth.","source_license":"CC0","license_restricted":false}