GnRH antagonist versus long GnRH agonist protocol in poor responders undergoing IVF: a randomized controlled trial

Transcription

1 Human Reproduction Vol.20, No.3 pp , 2005 Advance Access publication December 17, 2004 doi: /humrep/deh668 GnRH antagonist versus long GnRH agonist protocol in poor responders undergoing IVF: a randomized controlled trial Lai-Ping Cheung 1,2, Po-Mui Lam 1, Ingrid Hung Lok 1, Tony Tak-Yu Chiu 1, Sum-Yee Yeung 1, Ching-Ching Tjer 1 and Christopher John Haines 1 1 Department of Obstetrics and Gynecology, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong SAR, China 2 To whom correspondence should be addressed: lpcheung@cuhk.edu.hk BACKGROUND. This is the first published report of a prospective, randomized, controlled trial comparing a fixed, multi-dose GnRH antagonist protocol with a long GnRH agonist protocol in poor responders undergoing IVF. METHODS. Sixty-six poor responders were randomized into two groups: the study group received 0.25 mg of cetrorelix daily starting on day 6 of stimulation; the control group received 600 mg of buserelin acetate daily starting in the mid-luteal phase of the preceding cycle. Both groups were given a fixed dose of recombinant FSH (300 IU daily) for stimulation. RESULTS. There were no significant differences in the cycle cancellation rates, duration of stimulation, consumption of gonadotrophins, and mean numbers of mature follicles, oocytes and embryos obtained. The implantation rates were similar, but the number of embryos transferred was significantly higher for the antagonist group ( versus ; P ). The pregnancy rates were also higher in the antagonist group, but the difference was not statistically significant. CONCLUSION. A fixed multi-dose GnRH antagonist protocol is feasible for patients who are poor responders on a long agonist protocol; however, our study failed to demonstrate an overall improvement in ovarian responsiveness. Clinical outcomes may be improved by developing more flexible antagonist regimens, an approach that requires further evaluation. Key words: GnRH agonist/gnrh antagonist/ivf/poor responder/randomized controlled trial Introduction Nine to twenty-four percent of infertile women undergoing assisted reproduction have a poor response to ovarian stimulation (Keay et al., 1997). Various strategies have been investigated in an attempt to improve ovarian response, including the use of high doses of gonadotrophins (Karande et al., 1990; Hofmann et al., 1993; Land et al., 1996), the change to a flare-up protocol (Karande et al., 1997), and the use of growth hormone or growth hormone-releasing factor (Dor et al., 1995; Howles et al., 1999), L-arginine (Battaglia et al., 1999) or aspirin (Lok et al., 2004) as adjunct therapies. However, most of these interventions have met with only limited success and the optimum stimulation protocol for poor responders is still unknown. Most IVF programmes use long GnRH agonist protocols for ovarian stimulation. By inducing hypophyseal desensitization, GnRH agonist protocols prevent premature ovulation and luteinization, and significantly reduce the cycle cancellation rate compared with cycles where gonadotrophins are administered alone (Hughes et al., 1992). However, the disadvantages are that a higher overall dose of gonadotrophins and a longer duration of stimulation are often required to achieve an adequate ovarian response (Hughes and Durnerin, 1998). This is thought to be related to the extreme pituitary suppression (Ben-Rafael et al., 1991) caused by these agents and a direct inhibitory effect on the ovaries (Sheehan et al., 1982). The reduced sensitivity to gonadotrophins associated with a long GnRH agonist protocol has led some authors to reduce the dose and duration of agonist in patients classified as poor responders. Such protocols include cessation or reducing the dose of GnRH agonist after initial downregulation (Feldberg et al., 1994; Faber et al., 1998) and the use of a microdose flare GnRH agonist protocol (Scott and Navot, 1994). In these studies, lower cycle cancellation rates, less gonadotrophin consumption and an increased number of oocytes and embryos were frequently reported. It therefore appears that reducing the dose and duration of GnRH agonist in patients who are poor responders may be an effective approach to enhance ovarian responsiveness. More recently, GnRH antagonists have become available. GnRH antagonists competitively block pituitary GnRH receptors, inducing a rapid, reversible suppression of gonadotrophin secretion (Diedrich et al., 1994). Due to their distinct 616 q The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.

2 GnRH antagonist versus GnRH agonist in poor responders pharmacological mode of action, GnRH antagonists can be administered at mid-cycle to prevent a premature LH surge while not causing any suppression in the early follicular phase, which is a crucial time for follicular recruitment. This is particularly important in those patients who have decreased ovarian reserve, as lack of downregulation early in the cycle may allow the ovaries to respond maximally to the administrated gonadotrophins. We therefore postulate that the use of the antagonist protocol in poor responders may improve the ovarian responsiveness to gonadotrophin stimulation compared with the conventional long GnRH agonist regimen. This hypothesis is supported by preliminary comparisons between the two regimens which showed that the use of GnRH antagonists in poor responders was associated with a lower consumption of gonadotrophins (Nikolettos et al., 2001), a shorter duration of stimulation (Nikolettos et al., 2001) and a non-significant trend for a greater number of oocytes retrieved and a lower cycle cancellation rate (Craft et al., 1999). However, these studies were either retrospective, comparing patients with their prior failed cycles (Craft et al., 1999), or else compared only with nonrandomized controls (Nikolettos et al., 2001). The present study was therefore set up as a prospective, randomized controlled trial to compare a fixed, multi-dose GnRH antagonist protocol with a standard long GnRH agonist protocol in poor responders undergoing IVF. Materials and methods Subjects were selected from patients attending the assisted reproductive technology (ART) unit of the Prince of Wales Hospital, Hong Kong, during the period April 2001 to December Poor responders were classified as patients who had exhibited a poor ovarian response with,3 mature follicles on a long GnRH agonist protocol in their previous IVF cycles, or those with repeated high basal levels of FSH. 10 IU/l. Patients with polycystic ovaries were excluded from the study. The Clinical Research Ethics Committee of the Chinese University of Hong Kong approved the study. After informed consent, patients were randomly allocated to receive either GnRH antagonist or agonist using computer-generated random numbers concealed in opaque envelopes. A research nurse coordinated the randomization process and distribution of medication throughout the treatment cycles. Doctors and embryologists involved in the study were blinded to the treatment allocation. In all patients, serum FSH, LH and oestradiol (E 2 ) were measured on day 3 of the cycle preceding ovarian stimulation. The oral contraceptive pill (Nordette, Wyeth; 30 mg of ethinylestradiol Figure 1. Study flow chart and patient outcomes. 617

3 L.-P.Cheung et al. and 150 mg of levonorgestrel) one tablet daily was given for the rest of the cycle for a total of 21 days. The study group received a fixed, multi-dose GnRH antagonist protocol. Transvaginal ultrasound and serum E 2 were arranged on day 2/3 of the period after oral contraceptive pre-treatment. After confirmation of quiescent ovaries and low E 2 level, s.c. recombinant FSH (Gonal-F, Serono Laboratories, Aubonne, Switzerland) was commenced at 300 IU daily for ovarian stimulation. S.c. cetrorelix (Cetrotide, Serono Laboratories, Aubonne, Switzerland) 0.25 mg daily was then given as a fixed protocol starting on day 6 of the stimulation until the day of HCG The control group received a long GnRH agonist protocol. Buserelin acetate nasal spray (Suprecur, Hoechst AG, Frankfurt am Main, Germany) was given at a daily dose of 600 mg starting at the mid-luteal phase of the preceding cycle, and co-administered with the final week of oral contraceptive pre-treatment. Two weeks later, after confirmation of adequate pituitary desensitization, ovarian stimulation with s.c. recombinant FSH (Gonal-F) 300 IU daily was commenced. Buserelin was continued until the day of HCG In both groups, serum FSH, LH and E 2 were monitored on the first day of stimulation, and then every 3 days including the day of HCG Serial transvaginal ultrasound examinations were performed to monitor the follicular growth. Endometrial thickness was also assessed on the day of HCG Ovulation was triggered with IU of i.m. HCG (Profasi, Serono Laboratories, Aubonne, Switzerland) when the leading follicles reached mm together with at least three mature follicles $16 mm detected on ultrasound scan. This was followed by transvaginal ultrasound-guided oocyte retrieval, 36 h later. Cycles in which,3 mature follicles developed, or if the ovaries failed to respond after 10 days of stimulation, were either cancelled or converted to intra-uterine insemination in patients with patent tube(s). ICSI was performed only in cases with severe male factor or previous fertilization failure. Depending on the number of embryos available, up to three embryos were transferred on day 3 after oocyte retrieval. The surplus transferable embryos, if any, were then frozen. All patients received luteal phase support with i.m. HCG (Profasi) 2000 IU given every 3 days for four doses starting on the day of oocyte retrieval. A clinical pregnancy was established when there was a gestational sac seen on ultrasonography. The main outcome measures were duration of stimulation, consumption of gonadotrophins, cycle cancellation rate, and the number of mature follicles recruited and total oocytes retrieved. The hormone levels throughout the cycle, laboratory outcomes and clinical pregnancy rates were also reviewed. Sample size and statistics The sample size was estimated on the basis of the expected number of oocytes retrieved. The reported mean number of oocytes retrieved in poor responders using a long GnRH agonist protocol is 3.27 (SD ¼ 2.3) (Howles et al., 1999). Assuming an increment of two oocytes per retrieval as being clinically significant, 22 patients in each arm would be required to provide 80% power at the 5% significance level. Approximately 30 patients were therefore required to be recruited in each arm assuming a recruitment rate of 80% and a drop-out rate of 10%. Statistical analysis was performed using the Statistical Packages for Social Sciences for Windows Version 11.0 (SPSS Inc., IL). 618 Data were presented as mean ^ SD number (percentage) or median (interquartile range), and analysed by Student s t-test, x 2 or Fisher s exact tests, and Mann Whitney U-test as appropriate. A P-value of, 0.05 was considered statistically significant. Results Figure 1 shows the study flow chart and patient outcomes. A total of 66 patients were recruited to the study, with 33 randomized to each treatment arm. Two patients (one in each group) later declined to participate and withdrew from the study, while another patient in the antagonist group was found to have a dermoid cyst during the baseline scan and was excluded, leaving 32 patients in the agonist and 31 patients in the antagonist group. The patient demographic variables are compared in Table I. There were no significant differences between the two groups in any of the baseline characteristics including female age, basal FSH levels on cycle day 3 and proportion of patients with previous cycle cancellation due to poor ovarian response. The stimulation and cycle outcomes are compared in Table II. High cancellation rates were found in both groups (antagonist versus agonist: 38.7 versus 34.4%, P ¼ 0.9). Most of these cycles were cancelled for insufficient ovarian response, but two of the antagonist cycles were cancelled with evidence of premature LH surges. The two cancelled cycles had a LH rise to 16.5 IU/l at day 14 of stimulation and 33.4 IU/l at day 13 of stimulation, respectively. There were no significant differences between the two groups in the cycle cancellation rates, and the number of mature follicles and total oocytes obtained. Although the treatment duration (10.5 ^ 2.7 versus 11.5 ^ 2.4 days, P ¼ 0.13) appeared shorter and the gonadotrophin consumption (3150 ^ 813 versus 3445 ^ 730 IU, P ¼ 0.13) appeared lower in the antagonist group, the differences were not significant. Table I. Patient demographic characteristics Antagonist (n ¼ 31) Agonist (n ¼ 32) Age (years) 36.0 ^ ^ 3.0 Body mass index (kg/m 2 ) 21.6 ^ ^ 2.8 Basal FSH on cycle day 3 (IU/l) Mean ^ SD 11.1 ^ ^ 3.4 Median (interquartile range) 11.0 (4.5) 11.2 (4.2) No. of cycles with day 3 FSH.10 IU/l 19 (61.3%) 23 (71.9%) No. of cycles where previous cycle 15 (48.4%) 13 (40.6%) cancelled due to poor response Duration of infertility (years) 6.7 ^ ^ 3.3 Main causes of infertility (%) Tubo-peritoneal disease 18 (58.0%) 13 (40.6%) Endometriosis 6 (19.4%) 6 (18.8%) Male factor 4 (12.9%) 5 (15.6%) Unexplained 3 (9.7%) 7 (21.9%) Others 0 1 (3.1%) Values are expressed as mean ^ SD or number (percentage) unless otherwise specified, and analysed by t-test, x 2 or Fisher s exact test, and Mann Whitney U-test where appropriate. No significant differences were found in any of the parameters.

4 GnRH antagonist versus GnRH agonist in poor responders Table II. Stimulation and cycle outcomes Antagonist (n ¼ 31) Agonist (n ¼ 32) Total stimulation 10.5 ^ ^ 2.4 NS duration (days) Total gonadotrophin 3,150 ^ 813 3,445 ^ 730 NS consumption (IU) No. of mature 3.39 ^ ^ 3.33 NS follicles $16 mm Endometrial thickness 10.6 ^ ^ 2.8 NS on day of HCG (mm) Cancellation rate (%) 12/31 (38.7%) 11/32 (34.4%) NS Cycles cancelled for NS poor response Cycles cancelled for 2 0 NS premature LH surge No. of cycles with NS oocyte retrieval No. of cycles with ICSI 8 4 NS No. of cycles with NS embryo transfer No. of cycles with surplus embryos frozen 2 4 NS P-value Values are expressed as mean ^ SD or number (percentage), and analysed by t-test, x 2 or Fisher s exact test where appropriate. Table III. Serum hormone levels during the stimulated cycles Hormone level Antagonist Agonist P-value FSH (IU/l) Day 1 of stimulation 9.6 (6.0) 5.7 (2.4),0.001 Day 6 of stimulation 16.9 (8.1) 17.6 (4.6) NS Day of HCG 15.9 (7.8) 17.3 (6.6) NS LH (IU/l) Day 1 of stimulation 3.6 (1.7) 1.9 (1.6),0.001 Day 6 of stimulation 1.9 (1.3) 1.3 (1.7) Day of HCG 2.0 (2.2) 1.7 (1.8) NS E 2 (pmol/l) Day 1 of stimulation 88 (63) 63 (28) 0.03 Day 6 of stimulation 1488 (2020) 416 (719) 0.03 Day of HCG 5275 (4822) 4358 (5111) NS Values are expressed as median (interquartile range) and analysed by Mann Whitney U-test. Table III compares the median serum hormone levels (FSH, LH and E 2 ) between the two protocols during the stimulated cycles. The endogenous LH and E 2 levels were significantly higher in the antagonist protocol prior to the administration of cetrorelix on day 6 of the cycle. No significant differences in the hormone measurements were then observed in the subsequent cycle including the day of HCG The laboratory and pregnancy outcomes are presented in Table IV. There was no statistically significant difference between the groups in the number of mature, immature and degenerate oocytes, fertilization rates and total embryos obtained. The implantation rates were similar between the two groups, but the number of embryos transferred was significantly higher for the antagonist group (2.32 ^ 0.58 versus 1.50 ^ 0.83; P ¼ 0.01). The clinical pregnancy rates per cycle initiated (16.1 versus 9.4%; P ¼ 0.22), per retrieval (26.3 versus 14.3%; P ¼ 0.20) and per transfer (26.3 versus Table IV. Laboratory and pregnancy outcomes Antagonist Agonist P-value No. of oocytes per retrieval 5.89 ^ ^ 4.17 NS No. of mature oocytes 5.32 ^ ^ 3.26 NS (% per retrieval) (90.2%) (89.0%) No. of immature oocytes 0.47 ^ ^ 0.75 NS (% per retrieval) (8.0%) (7.6%) No. of degenerate oocytes 0.11 ^ ^ 0.60 NS (% per retrieval) (1.8%) (3.4%) No. of fertilized oocytes 3.89 ^ ^ 2.92 NS (fertilization rate %) (71.2%) (53.6%) No. of embryos transferred 2.32 ^ ^ No. of embryos frozen 0.21 ^ ^ 1.15 NS Total embryos obtained 2.53 ^ ^ 1.61 NS Implantation rate (%) 6/44 (13.6%) 4/30 (13.3%) NS Clinical pregnancy per 5/31 (16.1%) 3/32 (9.4%) NS cycle initiated (%) Clinical pregnancy 5/19 (26.3%) 3/21 (14.3%) NS per retrieval (%) Clinical pregnancy per transfer (%) 5/19 (26.3%) 3/17 (17.6%) NS Values are expressed as mean ^ SD or number (percentage), and analysed by t-test, x 2 or Fisher s exact test where appropriate. 17.6%; P ¼ 0.26) were also higher in the antagonist group, but the differences were not statistically significant. Discussion With the discovery of extrapituitary GnRH receptors in the human ovaries (Janssens et al., 2000), there has been some concern that the non-physiological level of GnRH agonist given to achieve downregulation may have a direct, deleterious effect on the ovary, and contribute to the poor response to ovarian stimulation in some patients undergoing IVF. Some authors therefore consider a GnRH antagonist protocol as a reasonable option for patients who have responded poorly in previous stimulation cycles. Of the few randomized controlled trials that have addressed this issue, a study by Akman et al. (2000) compared two ovarian stimulation protocols in 40 poor responder patients in which no GnRH agonist was given; one of these groups received the GnRH antagonist cetrorelix during the late follicular stage. There was a non-significant trend for an improvement in clinical pregnancy and implantation rates in the antagonist group. A later study by the same author compared a GnRH antagonist multiple-dose protocol with a microdose GnRH agonist flare-up protocol in 48 poor responders (Akman et al., 2001). Similar pregnancy and implantation rates were found between the two groups. However, so far, no published randomized controlled trial has directly compared a GnRH antagonist protocol with a standard long GnRH agonist protocol in poor responders undergoing IVF. Our study, to the best of our knowledge, is the first prospective randomized trial of this kind. In a previous meta-analysis comparing GnRH antagonist with GnRH agonist protocols in non-selected IVF patients, the use of GnRH antagonists was associated with significantly less gonadotrophin consumption and shorter treatment duration, while having an equal effectiveness in the prevention of a premature LH surge (Al-Inany and Aboulghar, 619

5 L.-P.Cheung et al. 2002). Unlike these normally responding IVF patients, the use of GnRH antagonists in our study population did not result in a significant reduction in treatment time or gonadotrophin usage. A high gonadotrophin requirement was not a surprising feature in poor responders which were characterized further by high cancellation rates and low numbers of oocytes retrieved. Despite encouraging results from early poor responder studies (Craft et al., 1999; Nikolettos et al., 2001), our trial also failed to demonstrate any improvement in ovarian responsiveness with the use of GnRH antagonists in our study population. One explanation for our negative findings was that some of our patients could have a genuine depletion in ovarian reserve which might be instinctly uncorrectable. The heterogeneity of our study population was reflected by the differences in the basal FSH concentrations. One of the common problems faced in the poor responder studies is the insufficient patient numbers and the lack of a universally accepted definition for poor responders. Previous cycle cancellation due to poor response and high basal FSH levels are two of the most commonly used selection criteria for poor responder patients (Surrey and Schoolcraft, 2000; Tarlatzis et al., 2003). Both criteria were adopted in the present study to increase the recruitment rate but, conversely, this had led to a heterogeneous study population. Although we have attempted to subanalyse the results according to the basal FSH levels, the number of patients in each subgroup was too small for us to draw any conclusions. Nonetheless, the results obtained for poor responders with a fixed antagonist protocol were at least comparable with those obtained using a long agonist protocol with similar numbers of mature follicles and oocytes obtained. In addition, the findings of a higher number of embryos transferred and suggestion of a slightly higher pregnancy rate (although not significant) were encouraging and should warrant further investigation. There were a few factors in our study design which may have affected the outcome. First, oral contraceptive pretreatment was included in our antagonist protocol. Oral contraceptive pre-treatment in non-downregulation protocols helps to abolish corpus luteum rescue and synchronize follicular development during IVF. Improvement in stimulation and pregnancy outcomes has been reported in poor responders with oral contraceptive pre-treatment in various non-downregulation protocols including those with no pituitary suppression (Gonen et al., 1990), microdose flare GnRH agonist protocols (Surrey et al., 1998) and GnRH antagonist protocols (Copperman, 2003). However, contradictory results have been found in antagonist protocols (Shapiro et al., 2002) and there is still some concern about the use of oral contraceptive pre-treatment in poor responders as their ovarian reserve may be especially sensitive to the oversuppression of endogenous gonadotrophins. Nevertheless, hormone profiles in our study do not suggest that there was profound suppression, and hence inclusion of oral contraceptive pre-treatment in our antagonist protocol was unlikely to cause any adverse effect. Secondly, our stimulation regimen did not include any exogenous LH in the form of HMG or recombinant 620 LH. Individualization of LH supplementation after GnRH antagonist administration may influence the outcome. When administered in the late follicular phase to prevent an LH surge, GnRH antagonist can induce a sharp decrease in serum LH level which may be detrimental if the level falls below a threshold. The threshold level of LH necessary for optimal follicular stimulation is still a topic of debate, although some LH activity is certainly required for optimal estrogen production and development of the endometrium. Another controversial issue is the LH requirement during IVF treatment in normogonadotrophic women suppressed with antagonist or agonist protocols and stimulated by FSH alone. In a recent prospective randomized study in unselected patients undergoing IVF and receiving a GnRH antagonist protocol, there was no evident benefit of LH supplementation (Cédrin-Durnerin et al., 2004). However, other studies in which LH levels were profoundly suppressed in women undergoing assisted reproduction argue that LH supplementation may be of benefit in LH-deficient patients (Westergaard et al., 2000; Loumaye, 2002). In the present study, as a low daily dose of GnRH antagonist was used, and the LH levels measured throughout the study were not profoundly suppressed as evidenced by no difference in the estradiol concentration and endometrial thickness on the day of HCG, administration of exogenous LH is unlikely to improve the outcome further. Thirdly, despite the current move towards a flexible antagonist regimen, a fixed antagonist protocol was chosen in our study. A fixed protocol facilitates blinding of doctors to the treatment allocation; however, introduction of a GnRH antagonist too early in a cycle may be counterproductive as it may lead to a shut off of potentially helpful endogenous gonadotrophins and interfere with early follicular recruitment. In the flexible protocol, the GnRH antagonist is usually administered later, based on the size of the leading follicle. There is some evidence that flexible antagonist protocol tailoring to an individual s needs can optimize ovarian stimulation and improve the yield of oocytes retrieved, despite a lower dose of gonadotrophins required for stimulation (Ludwig et al., 2002). It therefore appears that the clinical outcome may be improved further by adopting more flexible antagonist regimens, an approach that requires further evaluation. In conclusion, a protocol including a GnRH antagonist appears at least as effective as one using a GnRH agonist in patients who are poor responders on a long agonist protocol, and may be easier or more convenient to administer. However, much work remains to be done in optimizing the GnRH antagonist protocols and individualizing these to different cycle characteristics. In the meantime, GnRH antagonist treatment may well be considered for patients not responding to a long GnRH agonist protocol. Acknowledgements The authors wish to thank Ms Mei-Chun Cheung, Ms Stella Tang and Ms Pou-Lin Kong in the ART unit for their help during the study; and Serono HK Ltd for providing samples of cetrorelix (Cetrotide w ).

6 GnRH antagonist versus GnRH agonist in poor responders References Akman MA, Erden HF, Tosun SB, Bayazit N, Aksoy E and Bahceci M (2000) Addition of GnRH antagonist in cycles of poor responders undergoing IVF. Hum Reprod 15, Akman MA, Erden HF, Tosun SB, Bayazit N, Aksoy E and Bahceci M (2001) Comparison of agonistic flare-up-protocol and antagonistic multiple dose protocol in ovarian stimulation of poor responders: results of a prospective randomized trial. Hum Reprod 16, Al-Inany H and Aboulghar M (2002) GnRH antagonist in assisted reproduction: a Cochrane Review. Hum Reprod 17, Battaglia C, Salvatori M, Maxia N, Petraglia F, Facchinetti F and Volpe A (1999) Adjuvant L-arginine treatment for in-vitro fertilization in poor responder patients. Hum Reprod 14, Ben-Rafael Z, Lipitz S, Bider D and Mashiach S (1991) Ovarian hyporesponsiveness in combined gonadotropin-releasing hormone agonist and menotropin therapy is associated with low serum follicle-stimulating hormone levels. Fertil Steril 55, Cédrin-Durnerin I, Grange-Dujardin D, Laffy A, Parneix I, Massin N, Galey J, Théron L, Wolf JP, Conord C, Clément P et al. (2004) Recombinant human LH supplementation during GnRH antagonist administration in IVF/ICSI cycles: a prospective randomized study. Hum Reprod 19, Copperman AB (2003) Antagonists in poor-responder patients. Fertil Steril 80 (Suppl 1),S16 S24. Craft I, Gorgy A, Hill J, Menon D and Podsiadly B (1999) Will GnRH antagonists provide new hope for patients considered difficult responders to GnRH agonist protocols? Hum Reprod 14, Diedrich K, Diedrich C, Santos E, Zoll C, Al-Hasani S, Reissmann T, Krebs D and Kingmuller D (1994) Suppression of the endogenous lutenizing hormone surge by the gonadotrophin-releasing hormone antagonist Cetrorelix during ovarian stimulation. Hum Reprod 9, Dor J, Seidman DS, Amudai E, Bider D, Levran D and Mashiach S (1995) Adjuvant growth hormone therapy in poor responders to in-vitro fertilization: a prospective randomised placebo-controlled double-blind study. Hum Reprod 10, Faber BM, Mayer J, Cox B, Jones D, Toner JP, Oehninger S and Muasher SJ (1998) Cessation of gonadotropin-releasing hormone agonist therapy combined with high-dose gonadotropin stimulation yields favorable pregnancy results in low responders. Fertil Steril 69, Feldberg D, Farhi J, Ashkenazi J, Dicker D, Shalev J and Ben-Rafael Z (1994) Minidose gonadotropin-releasing hormone agonist is the treatment of choice in poor responders with high follicle-stimulating hormone levels. Fertil Steril 62, Gonen Y, Jacobson W and Casper RF (1990) Gonadotropin suppression with oral contraceptives before in vitro fertilization. Fertil Steril 53, Hofmann GE, Toner JP, Muasher SJ and Jones GS (1993) High-dose follicle stimulating hormone (FSH) ovarian stimulation in low responder patients for in-vitro fertilization. J In Vitro Fertil Embryo Transf 6, Howles CM, Loumaye E, Germond M, Yates R, Brinsden P, Healy D, Bonaventura LM and Strowitzki T (1999) Does growth hormone-releasing factor assist follicular development in poor responder patients undergoing ovarian stimulation for in-vitro fertilization? Hum Reprod 14, Hughes EG, Fedorkow DM and Daya S (1992) The routine use of gonadotropin releasing hormone agonists prior to in-vitro fertilization and gamete intrafallopian transfer: a meta-analysis of randomised controlled trials. Fertil Steril 58, Hughes JN and Durnerin IC (1998) Revisiting gonadotrophin-releasing hormone agonist protocols and management of poor ovarian responses to gonadotrophins. Hum Reprod Update 4, Janssens RMJ, Brus L, Cahill DJ, Huirne JAF, Schoemaker J and Lambalk CB (2000) Direct ovarian effects and safety aspects of GnRH agonists and antagonists. Hum Reprod Update 6, Karande VC, Jones GS, Veeck LL and Muasher SJ (1990) High-dose follicle-stimulating hormone stimulation at the onset of the menstrual cycle does not improve the in vitro fertilization outcome in low-responder patients. Fertil Steril 53, Karande V, Morris R, Rinehart J, Miller C, Rao R and Gleicher N (1997) Limited success using the flare protocol in poor responders in cycles with low basal follicle-stimulating hormone levels during in vitro fertilization. Fertil Steril 67, Keay SD, Liversedge NH, Mathur RS and Jenkins JM (1997) Assisted conception following poor ovarian response to gonadotrophin stimulation. Br J Obstet Gynaecol 104, Land JA, Yarmolinskaya MI, Dumoulin JCM and Evers JLH (1996) Highdose human menopausal gonadotrophin stimulation in poor responders does not improve in vitro fertilisation outcome. Fertil Steril 65, Lok IH, Yip SK, Cheung LP, Leung PHY and Haines CJ (2004) Adjuvant low-dose aspirin therapy in poor responders undergoing in vitro fertilization: a prospective, randomised, double-blind, placebo-controlled trial. 81, Loumaye E (2002) Ovarian stimulation: is exogenous LH necessary in all patients? Gynecol Obstet Fertil 30, Ludwig M, Katalinic A, Banz C, Schröder AK, Löning M, Weiss JM and Diedrich K (2002) Tailoring the GnRH antagonist cetrorelix acetate to individual patients needs in ovarian stimulation for IVF: results of a prospective, randomized study. Hum Reprod 17, Nikolettos N, Al-Hasani S, Felberbaum R, Demirel LC, Kupfer W, Montzka P, Xia YX, Schopper B, Sturm R and Diedrich K (2001) Gonadotrophinreleasing hormone antagonist protocol: a novel method of ovarian stimulation in poor responders. Eur J Obstet Gynecol Reprod Biol 97, Scott RT and Navot D (1994) Enhancement of ovarian responsiveness with microdoses of gonadotropin-releasing hormone agonist during ovulation induction for in vitro fertilization. Fertil Steril 61, Shapiro D, Carter M, Mitchell-Leef D and Wininger D (2002) Plateau or drop in estradiol (E2) on the day after initiation of the GnRH antagonist Antagone in in vitro fertilization (IVF) treatment cycles does not affect pregnancy outcome. Fertil Steril 78 (Suppl 1),S22. Sheehan KL, Casper RF and Yen SS (1982) Luteal phase defects induced by an agonist of luteinizing hormone-releasing factor: a model for fertility control. Science 215, Surrey ES and Schoolcraft WB (2000) Evaluating strategies for improving ovarian response of the poor responder undergoing assisted reproductive techniques. Fertil Steril 73, Surrey ES, Bower J, Hill DM, Ramsey J and Surrey MW (1998) Clinical and endocrine effects of a microdose GnRH agonist flare regimen administered to poor responders who are undergoing in vitro fertilization. Fertil Steril 69, Tarlatzis BC, Zepiridis L, Grimbizis G and Bonties J (2003) Clinical management of low ovarian response to stimulation for IVF: a systematic review. Hum Reprod Update 9, Westergaard LG, Laursen SB and Andersen CY (2000) Increased risk of early pregnancy loss by profound suppression of luteinizing hormone during ovarian stimulation in normogonadotrophic women undergoing assisted reproduction. Hum Reprod 15, Submitted on October 8, 2004; resubmitted on November 9, 2004; accepted on November 18,