Gonadotropin-Releasing Hormone-Antagonist in Human In Vitro Fertilization doc

The group of patients receiving0.5 and 0.25 mg/day showed the best ART results in terms of pregnancy andimplantation rates without the risk of a pituitary oversuppression whichoccurred w

Two different compounds are available: the Cetrorelix (Cetrotide1,formerly ASTA Medica, now Serono) and the Ganirelix (Antagon1 orOrgalutran1, Organon) Two different protocols of administration (Fig 1)have been proposed in the literature for using GnRH-nt in controlled

67

ovarian stimulation (COH) In the multiple-dose protocol, small doses(0.25 mg) of the GnRH-nt are injected in the middle of the follicular phase(7–9) In the single-dose protocol, a higher dose (3 mg) is injected during thelate follicular phase, when the LH surge is most feared (10,11).

PHASE II DOSE-FINDING STUDIES

Single-Dose Protocol

In the first investigation with Cetrorelix, we simply reproduced the viously published Nal–Glu protocol consisting of two 5 mg injection 48 hrapart in the late follicular phase (12,13) We therefore proposed twoadministrations of 5 mg Cetrorelix 48 hour apart, the first injection beingadministered on stimulation day 7 We observed that the second injectionwas often unnecessary as hCG was given on the same day We concludedalso that the 5 mg dose induced a deep suppression of LH and that a lowerdose should be tried (10) A single-dose protocol was designed where a singleinjection of 3 mg of the GnRH-nt is performed on stimulation day 7 (11)

pre-To determine the minimal effective dose, we conducted a dose-findingstudy We compared the use of 2 and 3 mg to investigate the ‘‘protectionperiod,’’ the time during which an LH surge is prevented after the antagonistadministration The IVF-ET results were strictly comparable between the

Figure 1 Gonadotropin-releasing hormone antagonist multiple- and single-doseprotocols Fixed day regimens Abbreviations: FSH, follicle-stimulating hormone;hMG, human menopausal gonadotropin; hCG, human chorionic gonadotropin

two doses and the 2 mg dose prevented LH surges for three days in all thepatients However, we observed that the suppression of LH tended to

be reduced three days after the injection in the 2 mg dose and an LH surgewas observed four days after the 2 mg Cetrorelix administration The 3 mgdose was therefore selected as a safer choice, as a ‘‘protection period’’ of

at least four days can be obtained (14) No LH surge was observed in allthe patients treated with the 3 mg dose In some patients, an LH rise(LH > 10 IU/L) was observed on the day of the antagonist administration.The Cetrorelix was able to prevent any further rise in LH, lowering immedi-ately the LH levels and no surge was observed in these patients Moreover,the interruption of LH rises does not seem to have a deleterious effect onIVF-ET results (15) The consumption of hMG was clearly reduced(24–30 hMG ampoules) as compared to the use of GnRH-a in the longprotocol using a depot preparation (11)

The tolerance of the GnRH-nt Cetrorelix was excellent with onlytransient erythema at the injection site in 15% of the patients

Multiple-dose

The two GnRH-nts (Cetrorelix or Ganirelix) were studied in order toachieve the best dose that blocks the premature LH rise and does notoversuppress the pituitary Sommer et al (16) were the first in describingthe suppression of gonadotropin and estradiol secretion after 3 mg ofCetrorelix daily in normal cycling women After that, the dose-finding stud-ies generally initiated the gonadotropins (recombinant or urinary) on day 2

of menstrual cycle and the antagonist daily administration was initiated onstimulation day 6 (7–9,17) The risk for premature LH surge is higher afterthe sixth day of ovarian stimulation, and this day was chosen to initiate theantagonist injections (7) The authors of the different studies compareddifferent doses of Cetrorelix or Ganirelix to achieve the best dose with mostappropriate assisted reproductive technology (ART) results

Comparing the Cetrorelix administration of three and 1 mg or 0.5 and0.25 mg after day 6 of stimulation protocol, the authors showed that allpatients had a decline on LH serum levels The group of patients receiving0.5 and 0.25 mg/day showed the best ART results in terms of pregnancy andimplantation rates without the risk of a pituitary oversuppression whichoccurred with one and 3 mg (7,17) Another study compared the startingdoses of 0.5, 0.25, and 0.1 mg/day (8) The authors demonstrated thatpatients receiving 0.5 or 0.25 mg/day during the follicular phase did notshowed premature LH surge, evidenced by the lower LH serum levels.However, one out of seven patients with 0.1 mg/day protocol showed apremature LH rise with progesterone elevation and the 0.1 mg dose wastherefore abandoned The results were similar between the patients receiving0.25 and 0.5 mg/day in terms of clinical pregnancy and implantation rates

The group of investigators concluded that the minimal necessary and tive dose to prevent premature LH surge was 0.25 mg/day with Cetrorelix.The Ganirelix Study Group (9) also investigated the minimal safe/effective dose to achieve good IVF results This study also showed that,during the multiple dose protocol with Ganirelix, the minimal effective dosewas 0.25 mg/day, inhibiting the premature LH secretion without compro-mising IVF results in stimulated cycles with recombinant FSH This group

effec-of patients, receiving Ganirelix 0.25 mg/day had the highest vital pregnancyrate per transfer (40.3%) as the main clinical outcome if compared with theothers doses (0.0625–2 mg)

Observing very low implantation rates in the groups of daily 1 or 2 mg,Kol et al (18) analyzed the database from the Ganirelix dose-finding study(9) concerning the effect of GnRH-nt in freeze–thaw cycles The authorsconcluded that high doses (1.0 and 2.0 mg/day) of Ganirelix did not affectthe biologic potential of embryos to develop clinical pregnancy

PHASE III RANDOMIZED CONTROLLED TRIALS

AND OPEN STUDIES

One hundred and four patients (90.4%) out of 115 patients receivedonly one 3 mg dose of Cetrorelix If the criteria for triggering of ovulationwere not reached within four days (the protection period), we administered

an additional dose of Cetrorelix (0.25 mg) Only nine (7.9%) of the patientsreceived one additional dose on the morning of the hCG and two pati-ents (1.7%) received two additional doses of 0.25 mg

Moreover, a total of 18 patients of the Cetrorelix group (15.7%)presented an LH rise (LH > 10 IU/L) on the day of Cetrorelix injection.The administration of the Cetrorelix inhibited LH secretion Four of thembecame pregnant (22.2%) These interrupted LH rises seem to have nomeasurable deleterious effect in this study Only one patient in the Triptor-elin group (2.8%) experienced an LH surge None of the 115 patients of theCetrorelix group experienced an LH surge after the Cetrorelix administra-tion No LH surge has been reported so far within the four days followingthe single administration of 3 mg Cetrorelix

The mean length of stimulation was significantly lower in the elix group The mean number of ampoules was significantly higher in the

Cetror-Triptorelin group The E2 levels on the day of hCG were significantly lower

in the Cetrorelix than in the Triptorelin group The total number of follicles

15 and 17 mm was higher in the Triptorelin group (5.0  3.9 vs.3.4 2.6; CI 0.5–2.8) The long GnRH agonist protocol resulted in moreoocytes and more embryos as already demonstrated in the literature whencompared to other stimulation regimens However, the percentage of matureoocytes, fertilization rate, clinical and ongoing pregnancy rates, and mis-carriage rates were not statistically different between the two groups Theincidence of ovarian hyperstimulation syndrome (OHSS) was lower inthe GnRH-nt group This difference did not reach statistical significancebut some patients of the GnRH-a group were cancelled for being at risk ofOHSS Adding these patients brought the difference to significance

In conclusion, this study has confirmed the efficacy of a single dose of

3 mg of Cetrorelix, administered in the late follicular phase, in preventingpremature ovulation as indicated by LH surges The single-dose protocol

is easy to use and assures patient compliance The 3 mg dose of Cetrorelixwas tolerated well, with only mild and transitory reactions at the injectionsite This protocol provides a shorter duration of treatment, uses less gona-dotrophins, and has a lower incidence of OHSS In some of the patientstreated with rec-FSH, a decrease in the E2 level is observed after the injec-tion of the Cetrorelix This was also observed in our first study using ahigher dose of Cetrorelix (5 mg) with human menopausal gonadotrophin(hMG) (11) An increase of the hMG dose, on the day of the antagonistadministration, suppresses most of these E2 decreases, probably related tothe LH suppression but not exclusively (19) However, no difference isobserved in our experience in the IVF-ET results in the patients with orwithout an E2 decrease following the Cetrorelix administration (unpub-lished results) One study done in an oocyte donor model (20) found a lowerimplantation rate of embryos coming from oocytes collected in patientswith an E2 drop as compared to patients with continuous rise of E2

Multiple-Dose Protocol

In all the studies presented, the multiple-dose protocol uses 0.25 mg/day ofCetrorelix or Ganirelix To compare the antagonist multiple dose protocol(0.25 mg/day) to the GnRH-a in IVF cycles, the European Cetrorelix StudyGroup (21) published the results of an open randomized trial They studied

188 patients treated with Cetrorelix and 85 patients treated with the long(Buserilin) agonist protocol; both groups received hMG The authors trans-ferred embryos in 83.5% of Cetrorelix group versus 79% of Buserelin group.The clinical pregnancy rate was 22.3% and 25.9% per started cycle in theCetrorelix and Buserilin groups, respectively; these differences were notstatistically significant The duration of treatment with gonadotropins andthe estradiol serum levels on the day of hCG were lower in the antagonist

group The incidence of ovarian hyperstimulation syndrome (OHSS II andIII) was higher in patients using agonist treatment.

The European Ganirelix Study Group (22) also performed acontrolled, multicentric, randomized trial to compare two treatment regi-mens for ovarian stimulation (multiple-dose antagonist vs long-agonist)

in women receiving recombinant FSH A total of 672 patients were gated and randomized The total dose of FSH administered was higher inthe Buserilin group (1500 and 1800 IU) In addition, patients receiving antag-onist had a shorter stimulation duration than the agonist group The estradiolserum levels on the day of hCG administration were higher in patients usingBuserelin than Ganirelix and the incidence of OHSS was higher (5.9 vs 2.4%)

investi-in the Buserelinvesti-in group Otherwise, the number of good quality embryos, tilization rate (62.1% in both groups), and replaced embryos were similarbetween the two treatments schemes The implantation rate was lower inthe Ganirelix group (15.7%) than in Buserelin group (21.8%), however theclinical pregnancy rates per attempt were not statistically significant.The North American Ganirelix Study Group (23) was organized toevaluate the efficacy and safety of Ganirelix (multiple-dose protocol) versusleuprolide (long-protocol) in IVF patients This multicenter (United Statesand Canada) trial demonstrated that the mean number of retrieved oocyteswas similar between the groups (11.6 in antagonist group and 14.1 in agonistgroup) Moreover, the fertilization rates (62.4% and 61.9%) and implan-tation rates (21.6% and 26.1%) were also similar in both groups Theongoing pregnancy rates per attempt were 30.8% in ganirelix group and36.4% in leuprolide group; however, the antagonist group showed fewerlocal site reactions after injection administration (12.5%) than the leuprolidegroup (25.5%) The authors proved the effectiveness and safety of multipleantagonist drug protocol with a shorter stimulation period and fewer sideeffects when compared with the long agonist (leuprolide) protocol

fer-Another multicentric European (The European and Middle East)Orgalutran Study Group (24) trial, comparing two treatment schemes(Ganirelix and Triptorelin) in 337 women, demonstrated that the mediandose of FSH recombinant was lower in the antagonist protocol The authorsshowed also that the estradiol serum levels were lower in Ganirelix group onthe day of hCG The fertilization rates (64% Ganirelix and 64.9% Triptor-elin), the mean number of good quality embryos (2.7 and 2.9, respectively),the implantation rates (22.9% both treatments), and finally the ongoingpregnancy rate per attempt were similar between the two treatments (31%and 33.9%, Ganirelix and Triptorelin, respectively)

The multiple-dose protocol, compared with the long-agonist regimen,offers a simple, safe, and efficient option, with comparable IVF results TheOHSS risk is decreased (25), the total dose of gonadotrophin needed tostimulate the ovulation is lower, and the stimulation period is also shorterthan in the long protocol Patients receiving antagonist treatment had lower

estradiol serum levels at the time of hCG administration, probably because

of the lower number of follicles The impact of this finding in implantationrates is disputed and unknown

In the multiple-dose protocol, there is a small incidence of LH surge(between 1% and 2.5%) These surges were often associated with a lack ofcompliance by the patients, forgetting one antagonist administration Thispoint is important to stress to the patients More recently, some centers haveobserved a higher incidence of LH surge in poor responders using themultiple-dose protocol (unpublished data) These reports should be con-firmed and documented The dose of 0.25 mg might not be always sufficient.This dose might also have to be adapted to the weight of the patients.The follicular development was also studied in a controlled random-ized multicentric study, in patients using Ganirelix with different doses(0.0625–2.0 mg/day) Patients received recombinant FSH after day 2 ofmenstrual cycle and Ganirelix were administrated daily after day 6 of ovar-ian stimulation protocol (26) Overall, 311 patients were studied andcompared in terms of number of follicles, total follicular surface area, serumgonadotropin, and steroid hormones levels Increasing GnRH-nt dosesdemonstrated an additional suppressive action on estradiol and androstene-dione serum levels, probably by an important inhibition of LH secretion,which may have exerted a harmful effect The follicular growing patternwas not affected by the dose of GnRH-nt The decreased secretion ofandrostenedione and estradiol was not totally explained by the LHinhibition Other mechanisms could be involved in GnRH-nt action andinfluence the cycles stimulated with this regimen protocol

Wikland et al (27), in a prospective randomized trial, evaluate twostarting doses (150 vs 225 IU) of recombinant FSH with the multiple-doseCetrorelix protocol The purpose was to increase the number of follicles,oocytes, and embryos to increase the pregnancy rates Despite a highernumber of recovered oocytes in the group of patients receiving 225 IU ofrecombinant FSH, pregnancy and implantation rates were similar

The effect of GnRH-nt on oocyte and embryo quality could also be red by studying the implantation and pregnancy rates after cryopreservation ofpronuclear oocytes or embryos The first study to evaluate this aspect was con-ducted with 62 patients divided into two groups (28) One group received themultiple GnRH-nt dose protocol (group I) and the other group the conventionalGnRH long protocol (group II) The implantation and pregnancy rates, afterfrozen-thawed procedure in pronucleous oocyte stage, were similar betweenthe groups (3.26% and 8.33% for group I; 3.73% and 10.25% for the group II)

measu-GnRH-nt IN MILD STIMULATION

Mild stimulation protocol is aimed at reducing the intensity of stimulation

to obtain a lower number of oocytes The advantages of those regimens

are a reduction in the intensity of side effects, a reduction in the incidence ofOHSS, and a reduction in costs and complication Other advantages lie inthe potential adverse effects on endometrium of heavy stimulation regimensand the possible benefit of a natural selection of the best oocytes.

Those protocols include mainly the association between citrate of miphene and gonadotrophins, the modified natural cycle, and the reducedamount of gonadotrophin dose In both these protocols, the GnRH-ntswere used to prevent LH surges

clo-Very few data are available on the use of CC-hMG/rec-FSH andantagonists in IVF (29,30) Pregnancy rates appear satisfactory but a highincidence of LH surges was observed (31) It was demonstrated in an animalmodel that CC increases the sensitivity of the pituitary to GnRH and thatthe dose of antagonists might need to be increased when CC is used Aprospective study found a lower pregnancy rate in CC-hMG cycles anddid not support the interest in this protocol (32)

The revival of the natural cycle was proposed as the use of GnRH-ntscould prevent premature LH surges The objective of this regimen is to com-bine the possible prevention of an LH surge by the administration of theGnRH-nt and the simplicity of the natural cycle with minimal stimulation(Fig 2) We investigated the administration of the Cetrorelix in the latefollicular phase of minimally stimulated cycles in women of good prognosis.These patients had an age between 26 and 36 years old (mean 34.1 1.4),normal menstrual cycles, day 3 FSH < 8 UI/l, day 3 E 2 < 50 pg/ml, lessthan three previous IVF procedures, male factor infertility requiring IVF,and ICSI (33) A single subcutaneous injection of 1 or 0.5 mg Cetrorelixwas administered when plasma estradiol levels reached 100–150 pg/ml,and a lead follicle was between 12 and 14 mm to assess the minimal effectivedose Since studies with Nal–Glu (34) and Cetrorelix (35) demonstrated thatestradiol secretion can be reduced after the GnRH-nt administration, dailyadministrations of 150 IU of hMG were performed at the time of the firstinjection of Cetrorelix and repeated thereafter until hCG administration

Figure 2 Modified natural cycle using Gonadotropin-releasing hormone nists Abbreviations: hCG, human chorionic gonadotropin; FSH, follicle-stimulatinghormone; hMG, human menopausal gonadotropin; OPU, oocyte pick-up

antago-This treatment scheme is not a complete natural cycle since a lowgonadotrophin support is associated (minimal stimulation) Triggering ofovulation (5000 IU of hCG) was decided when the lead follicle reached16–20 mm and estradiol values were above 200 pg/ml Oocyte pick-up wasperformed 36 to 40 hours later without anesthesia (36).

A total of 33 patients (44 cycles) were included The mean number ofhMG ampoules was 4.7 1.4 and the mean time between the Cetrorelix andhCG administration was 2.0 0.7 days We canceled four cycles (9.0%).Follicular growth and E2 secretion were not altered by the Cetrorelixadministration A total of 40 oocytes retrievals leading to 22 transfers(55%) were performed In 10 cycles, no oocyte was obtained Fertilizationfailure occurred in six cycles, and in two patients the transfer was notperformed because of a developmental arrest of the embryo at the two pronu-clear stages The fertilization rate was 80% (24 embryos/30 oocytes) A total

of five clinical pregnancies were obtained (32.0%/transfer, 17.5%/retrieval) ofwhich four are ongoing

The number of patients in whom the cycle was canceled for premature

LH surge was very low (9.0%) as compared to previous reports on naturalcycles, confirming the efficacy of the antagonist administration In addition,the pregnancy rate (17.5% clinical pregnancy rate per retrieval; 32.0% pertransfer) seems interesting, even though it has to be confirmed in larger ser-ies The same protocol was used in a recent study with rec-FSH as folliculargrowth support (37)

Another form of mild stimulation was proposed recently by delaying thestart of gonadotrophin stimulation The mild ovarian stimulation protocolresulted in pregnancy rates per started IVF cycle similar to those observedafter profound stimulation with GnRH agonist cotreatment, despite shorterstimulation and a 27% reduction in exogenous FSH A higher cancellation ratebefore oocyte retrieval was compensated by improved embryo quality con-comitant with a higher chance of undergoing ET (38)

The high burden and drawbacks of ‘‘heavy’’ (stimulated) COHprotocol (side effects, multiple pregnancies, and potential serious healthcomplications) make a clear demand for softer protocols (39,40), a ‘‘friendlyIVF’’ (41)

If these preliminary results with spontaneous cycle and hMG port are confirmed on large numbers, the repetition of two or three ofthese cycles could lead to acceptable cumulative pregnancy rates withoutthe potential adverse effects of COH (42–44) and be more cost-effective (45)

sup-Another use of natural cycle with GnRH-nt was proposed recently.The protocol was proposed in poor responders (46) The number of studiesevaluating this approach is very low so far and results are controversial (47).Further studies are needed to analyze the possible interest of the modifiednatural cycle in poor responders

REMAINING QUESTIONS ON GnRH-nt

Pregnancy Rates

There is a trend in most of the controlled studies using GnRH-nt (with bothcompounds and protocols) to find slightly lower pregnancy rates as compared

to the GnRH-a long protocol This led to the questioning of IVF-ET results

of GnRH-nts A meta-analysis concluded there were significantly lower nancy rates in GnRH-nt cycles as compared to GnRH agonists (48) However,the difference was very close to being nonsignificant (OR 0.79; 95% CI 0.63–0.99) Adding one study, Ludwig et al (49) did not find a significant difference

preg-in pregnancy rates for Cetrorelix A recent meta-analysis presented by Daya didnot find a difference in ongoing pregnancy rate (unpublished data) Care should

be taken in drawing conclusions on these observations Some population factorswere not equivalent in the groups despite randomization In addition, the learn-ing curve, inherent to the use of new treatment schemes, could have influencedsome of the studied outcome The trend towards higher pregnancy rate (PR) inthe GnRH-a group may be associated with the relatively higher number ofobtained embryos due to the higher number of oocytes This hypothesis wasnot confirmed by the study presented by Wikland et al (27) The differencecould be related to the absence of desensitization of the previous luteal phase

In fact, a difference of the same magnitude in the pregnancy rates is foundbetween the short and long protocols (50) The clear reasons for this differenceare still not clear The potential deleterious effect of the GnRH-nt on the endo-metrium or even on the fertilization process has been presented (51) However,there is no clinical data to confirm these hypotheses in the human species (52,53).Therefore, a careful analysis is needed before drawing conclusions on the PR Infact, a comparative study designed to assess a 5% difference for PR situated inthe region of 20% will require over 1200 patients in each treatment group.The important issue about GnRH-nt pregnancy rates is that there arestill major questions to be answered on the best way to use them

Single Versus Multiple-Dose Protocols

The two protocols have not been prospectively compared in large studies.The only large available study presents the results of two large multicentricstudies using the single- and multiple-dose Cetrorelix protocols (54) Nodifferences are observed between the two regimens but as this is not a pro-spective randomized study; firm conclusions cannot be drawn from thosedata A propspective randomized study compared the Ganirelix multipledose protocol to the Cetroreliw single dose regimen and did not find a stat-istical difference in the PR between the two protocols (55,56)

Fixed Versus Flexible Administration of the Antagonist

In the first study using GnRH-nt, the protocol included a fixed day ofGnRH-nt administration In the multiple-dose protocol, the antagonist

was started on day 6 In the single dose, the 3 mg administration of elix was proposed on day 7 We proposed a more flexible approach byadapting the moment of antagonist administration to ovarian response(Fig 3) (11) The antagonist could be administered when the leading folliclereached 14 mm and/or E2 reached 600 pg/ml The same approach was laterproposed by Ludwig et al (57) in the multiple-dose protocol (Fig 3) Someauthors have predicted a lower pregnancy rate when the flexible approachwas proposed (58) Data remain controversial on this question A recentmeta-analysis did not find differences in the pregnancy rates between theflexible and fixed approaches (59).

Cetror-A very early start of the antagonist, injected at the beginning of thestimulation, was proposed to reduce the LH levels at the beginning ofthe follicular phase A prospective randomized study failed to demonstrate

an advantage to this proposal (60) Moreover, a long period of injection ofantagonist will really alter the main benefit of GnRH-nts

aspects of PCOS patients is the increased LH tone secretion This group ofpatients is characterized by anovulation, and ovarian ovulation induction isusually performed using clomiphene citrate, FSH associated or not withGnRH agonists The rationale for the use of GnRH-nt in PCOS patients

is the fact that the LH/FSH ratio will be decreased since LH secretion ismore affected by the antagonist administration than FSH secretion (2) InIVF, another clear advantage is the reduced incidence of OHSS with the uti-lization of GnRH-nt The use of GnRH-nt protocol allows also to inducethe oocyte final maturation with GnRH agonist, to elicit an endogenous LHsurge and, subsequently, decreasing the risk of OHSS (61) However, a largeprospective trial is necessary to confirm these physiological hypotheses (seeinfra) Lubin et al (62) described two case reports of PCOS patients treatedwith GnRH-nt before the treatment with GnRH agonist to induce ovula-tion The patients showed a normalization of LH and testosterone serumlevels, however, the authors failed to induce an appropriate ovarian response.Two recent prospective randomized studies compared GnRH-nt toagonists in PCO patients (63,64) Those studies found similar pregnancy ratebut failed to find a significant advantage to GnRH-nts

Larger studies are needed to further evaluate the potential benefits ofthe association of GnRH-nt in PCOS patients

POOR RESPONDERS

The definition of poor responders and the heterogeneity of this group ofpatients cause an important bias in published series The rational for usingovarian stimulation protocols with GnRH-nt in poor responders is thatGnRH-nts do not require desensitization and are not causing an importantdepression on gonadotropin secretion during the stimulation

Forty-two patients who are poor responders were divided into twogroups for ICSI treatment (long GnRH agonist or Cetrorelix multiple-doseprotocols) (65) The stimulation protocol also included, in some patients,clomiphene citrate associated with gonadotropins Age, number of oocytesretrieved, number of fertilized oocytes, transferred embryos, score ofembryo quality, and clinical pregnancy were not significantly differentbetween the groups A trend was observed in the pregnancy rates (14.28%for Cetrorelix vs 9.52% for GnRH agonist treatment) but the differencewas not significant The authors discussed the sample size utilized Probably,with an adequate power calculation this difference in terms of pregnancyrate can become statistically important

With the same objective of the above-mentioned paper, Akman et al.(66) presented a randomized trial comparing the microdose flare-up GnRH-aprotocol versus the antagonist multiple dose protocol Forty-eight patientswere divided into the two regimen protocols The implantation rates (15.07%for flare-up protocol and 11.36% for Cetrorelix) and the ongoing pregnancy