{"paper_id":"28e68d68-de97-4acf-ae65-996d5b3b0ba3","body_text":"Despite considerable advancements over the past\ndecade in assisted reproduction, poor responders remain\nan important challenge. These patients have more\nproblems in fertilization, embryo quality, and pregnancy.\nPoor response to ovarian stimulation occurs in\n9-18% of assisted reproductive technique (ART) cycles.\nHowever there is no specific definition for poor\nresponders, thus a comparison of outcomes from various\nprotocols is challenging ( 1 - 3 ). The most common\ndefinition of a poor responder is based on increased basal\nFSH, an inadequate ovarian response, low oestradiol\n(E2) levels to ovarian stimulation by FSH/HMG,\nand lower number of retrieved oocytes ( 3 - 6 ).\nSeveral strategies are available to improve ovarian\nstimulation outcome in poor responders, including\nincrease the dose of the gonatropin that\nis being used and administration of gonadotropinreleasing\nhormone (GnRH) analogs (agonists or\nantagonists). The use of clomiphene citrate, aromatase\ninhibitors, growth hormones, transdermal\ntestosterone, corticosteroids, estradiol or aspirin are recommended as adjuvant therapies ( 4 ,  7 - 10 ).\nOne of the most successful protocols for ovarian\nstimulation of poor responders is the microdose flareup\nprotocol ( 11 - 13 ). The basic hypothesis of this approach\ninvolves administration of a minimal dose of\nGnRH-a to stimulate gonadotropin release and minimize\npremature ovulation ( 14 ). GnRH antagonists\nrepresent an alternative in the management of poor responders\n( 15 ). Antagonists act to rapidly block gonadotropin\nreceptors so ovarian stimulation can be initiated\nbefore administration of the GnRH antagonist. As\na result these agents prevent a premature LH surge but\ndo not suppress early follicular development ( 16 - 18 ).\nGnRH antagonists have no flair effect on follicular development\ncompare with GnRH agonists.\nOur hypothesis is to compare the microdose Gn-\nRH-a flare-up protocol with the combined stimulatory\neffect of GnRH agonists and immediate suppression\nof the GnRH antagonist in a unique protocol that\nmay be a valuable new strategy for ovarian stimulation\nof poor responders, causing an improved ART\noutcome. In this study we compare the microdose\nflare-up protocol to the ultrashort GnRH agonist flare\ncombined with the fixed multidose GnRH antagonist\nprotocol in poor responders undergoing ART cycles.\n\nA total of 120 poor responder women who referred\nto the Yazd Fertility and Infertility Center of\nShahid Sadoughi University of Medical Sciences\nfrom June 2007 to July 2009 were enrolled in this\nrandomized clinical trial. This randomized, controlled\nstudy was approved by the Ethics Committee\nof Yazd Fertility and Infertility Center and was\nundertaken in accordance with CONSORT guidelines\n( Fig 1 ). All patients signed a written consent\nform before initiation of the treatment cycles.\nAll included patients had a history of one or more\nfailed IVF cycles with three or less retrieved oocytes.\nThere was no age limitation for participants. We excluded\npatients with: 1. body mass index (BMI) >30,\n2. endocrine or metabolic disorders, 3. history of endometriosis\nor 4.severe male factor (azspermia).\nPatients were randomly allocated into two groups by\nthe use of sealed envelopes. In group I (60 patients)\nthe microdose flare-up regimen was used. Group II (60\npatients) were treated with the ultrashort GnRH agonist\ncombined with fixed GnRH antagonist regimens.\nAll patients received oral contraceptive pills during\ntheir previous menstrual cycle. In group I patients\nreceived 0.05 mg subcutaneous buserelin\n(Suprefact, Serono) injections twice daily from the\nfirst day of the cycle that continued until the day of\nthe HCG injection. Ovarian stimulation was started\nfrom the third day of the patient,s menstrual cycle\nby intramuscular (IM) injections of HMG (Menogon,\nFerring, Germany) at a dose of 300 IU per day.\nFollicular monitoring began from the ninth day of\nthe cycle by serial vaginal ultrasonography and\nmeasurement of serum E2 levels. I.M. injections of\n10000 IU HCG (Pregnyl; NV Organon, Oss, The\nNetherlands) were injected when at least 2 follicles\n≥18 mm were observed on ultrasonography.\nGroup II patients received buserelin (Suprefact, Serono),\n0.5 mg/ subcutaneous (SC) per day from the first\nday of the menstrual cycle, which was continued for\nthree consecutive days. HMG (Menogon, Ferring) at\n300 IU per day was started on day three of the cycle.\nThe GnRH antagonist (Cetrorelix, Serono Laboratories,\nAubonne, Switzerland) at a dose of 0.25 mg SC per day\nwas started when the dominant follicle size reached a\ndiameter of 14 mm. Follicular monitoring by vaginal\nultrasonography and estradiol level measurement began\non the ninth day of the cycle. Patients received 10000 IU\nHCG (Pregnyle, NV Organon, Oss, The Netherlands)\nwhen at least 2 follicles that were ≥17-18 mm in diameter\nwere observed by ultrasonography. In both groups\noocyte retrieval was performed 34-36 hours after the\nHCG injection, using a 17 gauge needle under vaginal\nultrasonography guidance. Conventional IVF or intracytoplasmic\nsperm injection (ICSI) was appropriately\nperformed. Fertilization rate was defined as the ratio\nof number of oocytes with pronuclei observed at least\n18 hours after IVF or ICSI to the number of retrieved\noocytes. A Labotect catheter (Labotect, Gottingen Germany)\nwas used to transfer the embryos at 48-72 hours\nfollowing oocyte retrieval. Luteal phase support began\nwith I.M. injections of progesterone in oil (progesterone,\nAburaihan Co., Tehran, Iran) at a dose of 100 mg daily\non the day of oocyte retrieval and continued until documentation\nof fetal heart activity by ultrasound. Primary\noutcome was the clinical pregnancy rate per cycle.\nClinical pregnancy was identified as observation\nof fetal heart activity by transvaginal ultrasonography\nperformed three weeks after a\npositive β-hCGv (β-hCG >50 IU/L) two weeks\nafter embryo transfer. This means that the ultrasonography\nwas actually 5 weeks after embryo\ntransfer.\nStudy flowchart\nThe Statistical Package for Social Sciences\n(SPSS, version 15.0 for Windows; SPSS Inc., Chicago,\nIL) was used for data analysis. Student’s ttest\nand chi-square test were used to detect significant\ndifference (p<0.05) between the variables. All\ndata were expressed as mean ± SD.\n\nWe randomly recruited 120 patients, with 60 patients\nin each treatment group. There were no significant\ndifferences in mean female age, basal FSH\nand duration of infertility between both groups ( Table 1 ). After randomization, 6 patients in group I did\nnot have embryo transfers. Of these, 2 patients had\nno oocytes in oocyte retrieval and 4 had fertilization\nfailure. In group II, 5 patients did not have embryo\ntransfers, of which 1 patient had no response\nto ovarian stimulation, 1 patient had no oocytes\nobtained during oocyte retrieval, and 3 patients had\nfertilization failure. All 11 patients were part of the\nfinal analysis and not excluded from the study. There\nwere no patients lost to follow up. Table 2 shows the\ncycle characteristics and ART outcomes.\nPatient’s characteristic in two groups (Mean ± SD)\nART outcome in two groups (Mean ± SD)\nThere were no significant differences between\ngroups in the number of used gonadotropin ampoules,\nthe duration of stimulation, the number of\nretrieved oocytes, fertilization rate and the number\nof transferred embryos. The clinical pregnancy rate\n(per cycle) was 10% ( 6 ) in group I and 13,3% ( 8 )\nin group II, which was statistically similar in both\ngroups (p=0.389), although there was a trend toward\na higher clinical pregnancy rate in the ultrashort agonist/\nantagonist protocol.\n\nThe best stimulation protocol for poor responders\nremains controversial. An adequate stimulation\nprotocol should lead to an acceptable rate of cancellation,\nretrieve an adequate number of oocytes, obtain\ngood quality embryos, and eventually achieve\nmaximum pregnancy and live birth rates ( 20 ). Several\nstimulation regimens have been proposed for\npoor responders. Some have improved the ovarian\nresponse to stimulation but none were able to\nsignificantly improve the pregnancy rate ( 21 ). The\nmost common protocols for management of poor\nresponders are the microdose flare-up protocol\nand antagonist protocol. The microdose flare-up\nprotocol benefits from the release of endogenous\ngonadotropin in the early follicular phase of the\ncycle through administration of a low dose GnRH\nagonist to enhance response to ovarian stimulation.\nHowever this approach may lead to a premature\nLH surge and compromise the cycle, which in\nturn can affect oocyte and embryo quality, in addition\nto synchronization between the embryo and\nendometrium ( 19 ). Addition of gonadotropin to an\novarian stimulation protocol prevents premature\nLH surge without suppression of early follicular\ndevelopment ( 22 ).\nIn the present study, we compared the microdose\nGnRH agonist flare-up and ultrashort GnRH agonist\nthat was combined with the fixed multidose\nGnRH antagonist. According to our findings, poor\nresponders demonstrated similar outcomes. The\nnumber of used gonadotropin ampoules, duration\nof simulation, and the number of retrieved oocytes\nwere statistically similar in both groups. Fertilization\nand pregnancy rates per cycles were similar\nin both groups. Antagonist consumption in a poor\nresponder stimulation protocol is associated with\nthe possibility of decreasing the number of gonadotropin\nampoules used and reducing the duration\nof stimulation. However Scott and Navot have\nstudied the microdose GnRH flare-up protocol for\nlow responder women in an ART protocol and reported\na lower cancellation rate, increased number\nof retrieved oocytes, and higher pregnancy rates\nin these patients ( 23 ). A number of previous studies\nhave evaluated the effect of a GnRH antagonist\nin the management of poor responders and determined\nthat these protocols improved implantation\nand pregnancy rates ( 24 ,  25 ). The first study that\nhas compared an agonist-antagonist protocol with\nthe microdose flare-up protocol was reported by\nBerger et al. ( 26 ). They showed that addition of antagonist to an agonist in the ovarian stimulation\nprotocol was associated with reduced gonadotropin\nconsumption and duration of stimulation. However,\nas with our study, they have demonstrated that the\nagonist-antagonist is not inferior to the microdose\nflare-up protocol in poor responders ( 26 ). Erden et\nal. in a pilot study demonstrated that the agonistantagonist\nprotocol compared to microdose flareup\nwas associated with higher peak estradiol levels,\nmore mature and fertilized oocytes, and higher\nclinical pregnancies ( 27 ).\nOrvieto et al. compared an ultrashort GnRH\nagonist combined with a flexible multidose GnRH\nantagonist and microdose flare-up administration\nof GnRH. In contrast to our results, they found\nhigher numbers of mature oocytes and embryos\nin the ultrashort GnRH agonist/ antagonist group.\nPregnancy rate was also significantly higher in this\ngroup ( 28 ). In contrast to our study, they used a\nflexible multidose GnRH antagonist and defined\npoor responder as the retrieval of fewer than five\noocytes in the previous ART cycle.\nThe successful end-point of ART is to obtain a\nlive, healthy infant ( 17 ). Studies have shown similar\nART outcomes or only increased numbers of\nretrieved oocytes and/or obtained embryos. They\ncould not recommend a unique protocol for increasing\nlive births in poor responders ( 21 ,  29 ,  30 ).\n\nAlthough our findings showed no statistically\ndifference in clinical pregnancy rate and ART outcome\nbetween these two protocols, however this\nnew protocol could possibly be considered as a future\novarian stimulation protocol for poor responders.\nAdditional, large randomized prospective studies\nare recommended to further evaluate the role of\nagonist-antagonist in poor responder protocols.","source_license":"CC-BY-4.0","license_restricted":false}