A comparison of three gonadotrophin-releasing hormone analogues in an in-vitro fertilization programme: a prospective randomized study

Transcription

1 Human Reproduction vol.14 no.2 pp , 1999 A comparison of three gonadotrophin-releasing hormone analogues in an in-vitro fertilization programme: a prospective randomized study T.Dada 1, O.Salha, H.S.Baillie and V.Sharma Assisted Conception Unit, St James s University Hospital, Leeds LS9 7TF, UK 1 To whom correspondence should be addressed The use of gonadotrophin-releasing hormone analogues (GnRHa) has resulted in improved pregnancy rates in invitro fertilization (IVF) treatment cycles. Traditionally, short-acting analogues have been employed because of concerns over long-acting depot preparations causing profound suppression and luteal phase defects adversely affecting pregnancy and miscarriage rates. We randomized 60 IVF patients to receive a short-acting GnRHa, nafarelin or buserelin, or to receive a depot formulation, leuprorelin, all commenced in the early follicular phase and compared their effects on hormonal suppression and clinical outcome. We found that on day 15 of administration there was a significant difference in the suppression of oestradiol from initial concentrations, when patients on buserelin were compared with patients on nafarelin or leuprorelin (54 versus 72 and 65%; P < 0.05) and also in the number of patients satisfactorily suppressed, (80 versus 90 and 90%; P < 0.05), though there were no differences between the analogues by day 21. Similarly there was no difference in hormonal suppression during the stimulation phase or in implantation, pregnancy or miscarriage rates in comparing the three agonists. We conclude that with nafarelin and leuprorelin, stimulation with gonadotrophins may begin after 2 weeks of suppression and that long-acting GnRHa are as effective as short-acting analogues with no detrimental effects on the luteal phase. Key words: GnRHa/hormonal suppression/ivf/pregnancy Introduction Gonadotrophin-releasing hormone agonists (GnRHa) are modifications of the parent decapeptide, GnRH, and their action is based upon reversible blockade of pituitary gonadotrophin release, resulting in suppression of ovarian function. Over the last decade, almost all in-vitro fertilization (IVF) centres have employed GnRHa for pituitary suppression before human menopausal gonadotrophin (HMG) administration. This has been shown to result in superior pregnancy rates when compared with cycles not utilizing analogues (Meldrum et al., 1989; Abdalla et al., 1990; Hughes et al., 1992). Furthermore, pre-treatment with GnRHa increases the number of developing follicles (Awadalla et al., 1987; Serafini et al., 1988), prevents premature luteinizing hormone (LH) surges (Leung et al., 1983; Meldrum et al., 1984; Smitz et al., 1992; Testart et al., 1993), benefits poor responders (Loumaye et al., 1988; Belaisch-Allart et al., 1988; Ben-Rafael et al., 1990) and reduces the number of cancelled cycles (Kerin, 1989; Loumaye, 1990; Hughes et al., 1992). Finally, with analogue use, better timing of the cycle is realized (Juha et al., 1993), allowing for improved patient and clinician convenience (Porcu et al., 1995). Most reports comparing the use of analogues in the long protocol have analysed cycles with the analogues being started in the mid-luteal phase. The aim of this study was to compare three GnRH analogues: buserelin as a s.c. injection, nafarelin as an intranasal spray, both used routinely in this unit, and leuprorelin, as a longacting depot preparation, all commenced in the early follicular phase in a long protocol regime. We particularly wanted to compare any evidence of escape from suppression between the long-acting depot with the shorter-acting analogues and also to investigate whether leuprorelin would have any detrimental effects in the luteal phase, because of its longer duration of action. Materials and methods Sixty patients undergoing IVF embryo transfer treatment at the Assisted Conception Unit, St James s University Hospital, Leeds, were evaluated for this study. Patients were required to be 38 years old, have had no more than two previous IVF attempts, a normal semen analysis as defined by World Health Organization (WHO, 1987) criteria, no ovarian surgery, no baseline ovarian cysts 20 mm diameter, a normal menstrual history and no major systemic disease or uterine abnormality. Recruited patients were randomized into three groups, receiving buserelin, nafarelin or leuprorelin for pituitary suppression. GnRHa regime Group A patients (n 20), received s.c. buserelin acetate (Hoechst, Hounslow, Middlesex, UK), at a dose of 0.5 mg daily. Group B patients (n 20), received nafarelin intranasal spray (Searle, High Wycombe, UK), 100 µg 8-hourly. Both drugs were started on the first day of the menstrual period (day 1), and continued throughout the cycle until the day human chorionic gonadotrophin (HCG) was administered. Patients in group C (n 20) received a single s.c. depot injection of leuprorelin acetate (Prostap SR; Searle, Lederle Laboratories, Gosport, Hants, UK) 3.75 mg on day 1. Ovarian stimulation in all groups commenced after 21 days of pituitary desensitization. Patients were considered desensitized when an ultrasound scan confirmed an endometrial thickness 3 mm. Clinical protocol Multiple follicular stimulation was achieved with highly purified urofollitropin Metrodin HP (Serono, Welwyn Garden City, Herts, UK). 288 European Society of Human Reproduction and Embryology

2 GnRH analogues for IVF In our unit the regimen for gonadotrophin (Metrodin) administration is a sliding scale based on the age of the patient. Women 34 years of age receive 3 ampoules (225 IU) daily); years, 4 ampoules (300 IU) daily; and 38 years, 6 ampoules (450 IU) daily. Tracking of follicular growth was performed using transvaginal ultrasonography (Kretz, Technik, Combison 410, Tiefenbach, Austria) on day 6 of gonadotrophin stimulation and again on day 8. The mean maximal follicular diameter was calculated from four measurements of the leading follicle, in two planes at 90 to each other. When two or more follicles had a mean diameter of mm, HCG (Pregnyl; Organon, Cambridge, UK) was administered at a dose of 5000 IU i.m., 36 h prior to oocyte retrieval. If follicular measurements did not meet the criteria described above then a further scan was performed after 48 h, or a projected follicular growth of ~2 mm/day was employed to predict the time of HCG injection. Transvaginal oocyte retrieval under ultrasound guidance was performed 36 h post HCG. Once identified the oocyte was placed in 1 ml culture medium (bicarbonate-buffered Earle s balanced salt solution 0.03 M sodium pyruvate 15% human serum albumin penicillin gentamycin), in a Nunclon 4-well dish (Nunc, Copenhagen, Denmark) and incubated in a humid atmosphere at 37 C, with 5% CO 2. The oocytes were inseminated h post HCG with ~ spermatozoa per well. The appearance of pronuclei was checked h post insemination. Embryo transfer was performed on day 2 when the embryos were at the 2 4-cell stage. Luteal support was achieved with either HCG (Pregnyl) 2500 IU on the day of transfer and 72 h later, or daily i.m. administration of 100 mg progesterone (Gestone; Paines & Byrne, Surrey, UK). In this unit we administer luteal support with GnRHa cycles based on risk assessment of ovarian hyperstimulation syndrome. Patients who recruit 20 follicles or who have 15 oocytes collected are considered at high risk and are given i.m. progesterone for luteal support, whereas those with fewer follicles and eggs receive i.m. HCG. Pregnancy tests were performed on day 14 (HCG support) or day 16 (progesterone support) post embryo transfer, depending on the method of luteal support, using a commercial urinary kit. Clinical pregnancies were confirmed by transvaginal ultrasound scan at 6 7 weeks gestation. A venous blood sample was taken from each patient on the first day of GnRHa administration and every third day of the downregulation and stimulation phases of the cycle, up until the day of HCG administration. These samples were analysed for serum concentrations of LH, follicle stimulating hormone (FSH), oestradiol and progesterone. Hormonal assays Serum FSH and LH were measured in all samples using a Technicon Immuno 1 automated clinical analyser employing a sandwich immunoassay format, with a detection limit of 0.1 miu/l. Progesterone and oestradiol assays used a similar kit employing a competitive immunoassay format with a detection limit of 0.4 nmol/l and 37 pmol/l respectively. Serum HCG concentrations were determined with a commercial kit (AXSYM; Abbot Laboratories, Abbot Park, IL, USA) based on microparticle enzyme immunoassay (MEIA) technology. Sensitivity of the assay was 2.0 miu/ml. Statistical analysis All data were expressed as mean SEM, and analysed using SPSS version 7.0. One-way analysis of variance (ANOVA), or multiple ANOVA (MANOVA) were used to assess results. Statistical significance was defined as P 0.05 for all tests. Results Statistical analysis showed no significant difference in the mean age of patients in the three groups, A, B or C (30.9, 32.4, 30.8 years repectively), nor in the mean gonadotrophin stimulation dose (42.6, 48.6, 42.1 ampoules respectively). Figure 1a d, show serum hormone concentrations of oestradiol, FSH, LH and progesterone, expressed as mean SEM against time (days), in each group, in the down-regulation phase of the cycle. The curve for each group was analysed independently and then compared at each time-point with the curves for the other two groups. Cut-off limits for satisfactory pituitary/ovarian suppression were set at 2.5 U/l, 0.5 U/l, 130 pmol/l and 2 nmol/l for FSH, LH, oestradiol and progesterone respectively. There was no significant difference between the curves for FSH, LH or progesterone. However, for oestradiol (Figure 1a), the suppression curve of oestradiol with buserelin was significantly different (P 0.05) when compared with the curves for nafarelin or leuprorelin. Patients on buserelin took longer for their oestradiol to be suppressed to the cut-off limit of 130 pmol/l, though by day 21 of the down-regulation phase, all three analogues had achieved comparable suppression. Table I shows the percentage reduction in serum concentrations of oestradiol from initial values to the defined cut-off limit. All three analogues achieved comparable reduction of oestradiol by day 21 of the cycle. However, on day 6 and day 15 there was a significant difference in the percentage reduction of oestradiol when patients using buserelin were compared with patients on nafarelin or leuprorelin (25 versus 44 and 52% respectively; P 0.05) (54 versus 72 and 65% respectively; P 0.05). Furthermore, by day 15, 95% of patients using nafarelin or leuprorelin had suppressed oestradiol to 130 pmol/l, compared with only 80% of patients using buserelin (P 0.05) (Table II). Figure 2a d shows the effect of the agonists on serum concentrations of FSH, LH, oestradiol and progesterone, during the stimulation phase of the cycle. There was no significant difference in escape from suppression between the three analogues, as indicated by continually suppressed serum LH concentrations, nor was there any evidence of profound suppression in any of the groups as shown by poorly rising or depressed serum concentrations of oestradiol. The clinical result of IVF treatment cycles is compared in Table III. Gonadotrophin dose (ampoules) was similar in all groups, as was the duration of stimulation. There was no significant difference between the analogues in follicular recruitment, numbers of oocytes retrieved and fertilized, nor in the number of embryos formed, cleaved or transferred. Figure 3 shows the implantation, pregnancy and miscarriage rates in patients in the three groups, with no significant difference demonstrated in any of the parameters investigated. Discussion A number of studies investigating GnRH analogues have compared their use in various regimes. The ultrashort and short protocols employ analogues commenced at the start of the menstrual cycle and continued for 3 days with the ultrashort regime and until the day of HCG administration with the short protocol. Thereafter multiple follicular development is achieved with gonadotrophins. These two regimes utilize the 289

3 T.Dada et al. Figure 1. (a) The effect of nafarelin, buserelin and leuprorelin on serum oestradiol in the down-regulation phase. (b) The effect of nafarelin, buserelin and leuprorelin on serum follicle stimulating hormone (FSH) in the down-regulation phase. (c) The effect of nafarelin, buserelin and leuprorelin on serum luteinizing hormone (LH) in the down-regulation phase. (d) The effect of nafarelin, buserelin and leuprorelin on progesterone in the down-regulation phase. Table I. Percentage reduction of oestradiol, on days 6, 15 and 21 achieved by nafarelin, leuprorelin and buserelin Days from start of menstrual period Table II. Percentage of patients down-regulated on days 12, 15 and 21 after treatment with nafarelin, leuprorelin or buserelin Days from start of menstrual period Nafarelin Leuprorelin Buserelin P value NS initial stimulatory effect of endogenous, stored, and subsequently released pituitary gonadotrophins, as an adjunct to HMG for follicular growth (Tan, 1996). The long protocol of down-regulation, however, allows for complete suppression of oestradiol before ovarian stimulation is begun. This suppression process may take at least 14 days (Neveu et al., 1987; Kerin, 1989). Although some authors have found no difference in clinical outcome between the long and short protocols, others Nafarelin Leuprorelin Buserelin P value 0.05 a 0.05 a NS NS not significant. a Comparing buserelin with nafarelin and leuprorelin. including Kerin (1989), Smitz et al. (1992), Tarlatzis et al. (1994a) and Tan et al. (1996) have shown higher implantation and pregnancy rates with the long protocol. This has been attributed to more effective LH suppression, higher oocyte retrieval per cycle, more embryos developed and a higher chance of pregnancy from the use of cryopreserved embyos

4 GnRH analogues for IVF Figure 2. (a) The effect of nafarelin, buserelin and leuprorelin on serum oestradiol in the stimulation phase. (b) The effect of nafarelin, buserelin and leuprorelin on serum follicle stimulating hormone (FSH) in the stimulation phase. (c) The effect of nafarelin, buserelin and leuprorelin on serum luteinizing hormone (LH) in the stimulation phase. (d) The effect of nafarelin, buserelin and leuprorelin on serum progesterone in the stimulation phase. Table III. Clinical results after treatment with nafarelin, buserelin and leuprorelin (all mean values SD) Nafarelin Buserelin Leuprorelin No. of cycles Incidence (%) of cysts 20 mm No. of HMG ampoules Day of HCG No. of follicles No. of oocytes No. of viable oocytes No. of oocytes fertilized No. of embryos No. of embryos cleaved No. of embryos transferred There were no significant differences between groups. HMG human menopausal gonadotrophin; HCG human chorionic gonadotrophin. in the same cycle. The long protocol of down-regulation initiated in the mid-luteal phase produces more prompt and profound suppression with a lower resultant incidence of cyst formation (Martin et al., 1992), by reducing the flare response (Lockwood et al., 1995). An incidence of ~5% cyst formation Figure 3. Implantation, pregnancy and miscarriage rates after treatment with the three analogues, nafarelin, buserelin and leuprorelin. has been found in IVF cycles and possibly higher if downregulation is started in the early follicular phase (Jenkins, 1996). However, studies comparing ovarian cysts in IVF cycles pretreated with GnRHa are plagued by problems of definition and true influence, with only steroidically active cysts being found detrimental to IVF outcome (Jenkins et al., 1992). 291

5 T.Dada et al. Nevertheless, Serafini et al. (1988a), Meldrum et al. (1988), Ron El et al. (1990b) and Daya (1993), comparing the long protocol commenced in the mid-luteal phase with early follicular phase starts, have shown higher follicular recruitment, oocyte retrieval and pregnancy rates with follicular protocols of GnRHa administration. Since the establishment of this unit in 1991, we have utilized early follicular phase administration of GnRH agonists with good results (Assisted Conception Unit, 1997). Our results agree with those of Penzias et al. (1992), Balasch et al. (1992), Tarlatzis et al. (1994a) and Porcu et al. (1995), i.e. long-acting depot preparations are as effective in pituitary suppression as the short-acting analogues. We found no significant difference in suppression of FSH, LH, oestradiol or progesterone with any of the three analogues after 21 days of administration. We also found that the number of patients suppressed as well as the percentage suppression of serum oestradiol to defined cut-off values, by day 15, was superior with nafarelin or leuprorelin when compared to patients on buserelin. The incidence of cyst formation in all three groups was similar (nafarelin 10%; buserelin 10%; leuprorelin 15%). Tarlatzis et al. (1994b) have shown that cyst formation with GnRHa use is associated with elevated serum oestradiol concentrations. The delayed suppression with persistence of cysts observed with the patients on buserelin could indicate failure of buserelin to deal effectively with this adjunctive oestrogen production. We found no evidence of escape from suppression with any of the three analogues, comparing the number of viable oocytes retrieved or fertilized. In addition, leuprorelin did not cause profound suppression by comparable gonadotrophin dose and duration of stimulation. One of the recognized complications of analogue use in a long protocol is the risk of luteal phase dysfunction from premature luteolysis. Inhibition of luteal steroidogenesis by GnRHa has been described by Smitz et al. (1987). Shortacting analogues such as buserelin or nafarelin have been traditionally favoured because of their brevity of action. At cessation of administration they rapidly disappear from the circulation and a normal pituitary response can be shown within a few days (Porcu et al., 1995). Any effect on the luteal phase is thus minimal. Depot preparations have a longer duration of action, ~7 weeks (Broekmans et al., 1992; Porcu et al., 1995), with potentially adverse consequences on corpus luteal function. A report by Bourgain et al. (1994) investigated the influence of GnRHa on the endometrium and endometrial development. Endometrial histological maturation, ultrastructure, oestrogen and progesterone receptor status were analysed in the mid-luteal phase in GnRHa/HMG cycles with and without luteal supplementation. They found that supplementation with HCG or progesterone produced fewer signs of luteal phase deficiency compared to non-supplemented cycles. All the patients in this study received luteal phase support with either HCG or progesterone, and we found no difference in the miscarriage rates of patients on leuprorelin as compared with either nafarelin or buserelin. Filicori et al. (1996) have noted that depot GnRHa administration that provides measurable drug concentrations in early pregnancy is not associated with an increased abortion rate. Moreover 292 we did not find any difference in implantation rates or pregnancy rates between the three agonists. Although the presence of GnRH and its receptor has been established in endometrium, until recently it has been unknown whether these receptors are present on preimplantation embryos. Inadvertent exposure of concepti to GnRHa does not appear to have a deleterious effect (Tolis et al., 1981; Skarin et al., 1982; Testart et al., 1993). Gartner et al. (1997) have even suggested that analogues may be beneficial to implantation. A recent study by Raga et al. (1998) has demonstrated the presence of GnRHa receptors on preimplantation embryos at the mrna and protein levels. Their randomized study between administration or omission of GnRHa until 6 weeks of gestation, with luteal support, showed that administration of GnRHa throughout the luteal phase and early pregnancy had a positive influence on pregnancy and implantation outcome. Direct effects of longacting depot preparations on the embryo can now be viewed in a fuller context and seem unlikely to be detrimental. Furthermore we noted no difference in the mean numbers of embryos formed or cleaved in those patients on leuprorelin compared to the shorter acting analogues. Our study shows that leuprorelin, nafarelin and buserelin are equally effective when pituitary suppression, ovarian stimulation and IVF outcome are compared. Although satisfactory suppression with nafarelin and leuprorelin occurred after 2 weeks, buserelin required an additional week to achieve equivalent results. However, after 3 weeks there was no significant difference between the three analogues. GnRH antagonists, particularly the cyclic compounds which have a potency comparable with the linear agonists, may become more widespread in general clinical use. Their advantages include no initial flare response due to an immediate competitive pituitary receptor occupancy (Rivier et al., 1996). Further work in rat models may ultimately lead to the availability of GnRH antagonists with oral activity and minimal histamine-releasing side-effects (Rivier et al., 1996), making them eminently more attractive. At present, however, the decision on which GnRH agonist to use will ultimately depend on patient and clinician preference, side-effects, availability and cost. References Abdalla, H.I., Ahuiakk, L.T., Morris, N.N. et al. (1990) Comparative trial of luteinising-hormone releasing hormone analogue/human menopausal gonadotrophin in an assisted conception program. Fertil. Steril., 53, Assisted Conception Unit (1997) Annual Report. St James s University Hospital, Leeds, UK, pp Awadalla, S.G., Friedman, C.I., Chin, N.W. et al. (1987) Follicular stimulation for in vitro fertilisation using pituitary suppression and human menopausal gonadotrophins. Fertil. Steril., 48, Balasch, J., Jové, C.I., Moreno, V. et al. (1992) The comparison of two gonadotrophin-releasing hormone agonists in an in vitro fertilization program. Fertil. Steril., 58, Belaisch-Allart, J., Testart, J. and Frydman, R. (1988) Utilisation of GnRHa for poor responders in an IVF programme. Hum. Reprod., 4, Ben-Rafael, Z., Menashe, Y., Mimon, R. et al. (1990) Limitations in the use of combined gonadotrophin releasing hormone analog and human menopausal gonadotrophin for in vitro fertilisation. In Maschiach, S., Ben- Rafael, Z., Laufer, N. and Schenker, J.G. (eds), Advances in Assisted Reproductive Technologies. Plenum Press, New York, pp

6 GnRH analogues for IVF Bourgain, C., Smitz, J., Camus, M. et al. (1994) Human endometrial maturation is markedly improved after luteal supplementation of gonadotrophinreleasing hormone analogue/human menopausal gonadotrophin stimulated cycles. Hum. Reprod., 9, Broekmans, F.J., Bernardus, R.E., Berkhout, G. and Shoemaker, J. (1992) Pituitary and ovarian suppression after early follicular and mid-luteal administration of a LHRH agonist in a depot formation: decapeptyl CR. Gynaecol. Endocrinol., 6, Daya, S. (1993) Short versus long protocol using gonadotrophin releasing hormone agonists (GnRHa) for in vitro fertilization cycles a metaanalysis. American Fertility Society Program Supplement, Abstracts of the Scientific Oral and Poster Session, 058, Filicori, M., Cognigni, G.E., Arnone, R. et al. (1996) Role of different GnRH agonist regimens in pituitary suppression and the outcome of controlled ovarian hyperstimulation. Hum. Reprod., 11 (Suppl. 3), Gartner, B., Moreno, C., Marinaro, A. et al. (1997) Accidental exposure to daily long-acting gonadotrophin-releasing hormone analogue administration and pregnancy in an in-vitro fertilisation cycle. Hum. Reprod., 12, Hughes, E.G., Ferdokow, D.M., Daya, S. et al. (1992) The routine use of gonadotrophin-releasing agonists prior to in vitro fertilisation and gamete intra-fallopian transfer: a meta-analysis of randomised controlled trials. Fertil. Steril., 58, Jenkins, J.M. (1996) The influence, development and management of functional ovarian cysts during IVF cycles. Hum. Reprod., 11 (Suppl. 4), Jenkins, J.M., Davies, D.W., Anthony, F.W. et al. (1992) The detrimental influence of functional ovarian cysts during in-vitro fertilization cycles. Hum. Reprod., 7, Juha, T., Outi, H., Kaisa, J. et al. (1993) Subcutaneous goserelin versus intranasal buserelin for pituitary down-regulation in patients undergoing IVF: a randomized comparative study. Hum. Reprod., 8, Kerin, J.F. (1989) The advantages of a gonadotrophin releasing hormone agonist (leuprolide acetate) in conjunction with gonadotrophins for controlled ovarian hyperstimulation in IVF and GIFT cycles. Arch. Gynecol. Obstet., 246, S45 S52. Leung, P.C.S., Lopata, A., Kellow, G.N. et al. (1983) Histochemical study of cumulus cells for assessing the quality of preovulatory oocytes. Fertil. Steril., 39, Lockwood, G.M., Pinkerton, S.M. and Barlow, D.H. (1995) A prospective randomised single blind comparative trial of nafarelin acetate with buserelin in long-protocol gonadotrophin-releasing hormone analogue controlled in vitro cycles. Hum. Reprod., 10, Loumaye, E. (1990) The control of endogenous secretion of LH by gonadotrophin-releasing hormone agonists during ovarian hyperstimulation for in-vitro fertilisation and embryo transfer. Hum. Reprod., 5, Loumaye, E., de Cooman, S., Anoma, M. et al. (1988) Short term utilisation of a gonadotrophin releasing hormone agonist (Buserelin) for induction of ovulation in an in vitro fertilization program. Ann. NY Acad. Sci., 541, Martin, M.C., Willman, S., Monroe, S.E. et al. (1992) Pretreatment with the GnRH agonist nafarelin acetate for in vitro fertilization. In Nafarelin Recent Advances. Selected proceedings of the Second International Symposium on GnRH analogues in Cancer and Human Reproduction, Parthenon, Carnforth, UK, pp Meldrum, D.R., Tsao, Z., Monroe, S.A. et al. (1984) Stimulation of LH fragments with reduced bio-activity following GnRH agonist administration in women. J. Clin Endocrinol. Metab., 58, Meldrum, D.R., Wisot, A., Hamilton, F. et al. (1988) Timing of initiation and dose schedule of leuprolide influence the time course of ovarian suppression. Fertil. Steril., 50, Meldrum, D.R., Wisot, A., Hamilton, F. et al.(1989) Routine pituitary suppression with leuprolide before ovarian stimulation for oocyte retrieval. Fertil. Steril., 51, Neveu, S., Hedon, B., Bringer, J. et al. (1987) Ovarian stimulation by a combination of gonadotrophin-releasing hormone agonist and gonadotropins for in vitro fertilisation. Fertil. Steril., 47, Penzias, A.S., Shamma, F.N., Gutmann, J.N. et al. (1992) Nafarelin versus leuprolide in ovulation induction for in vitro fertilisation: a randomised clinical trial. Obstet. Gynecol., 79, Porcu, E., Filicori, M., Dal Prato, L. et al. (1995) Comparison between depot leuprorelin and daily buserelin in IVF. J. Assist. Reprod. Genet., 12, Raga, F., Casa, E. M., Ruiz, A. et al. (1998) Gonadotrophin-releasing hormone agonist administered throughout the luteal phase and early pregnancy increases fecundity in IVF embryo transfer patients. Hum. Reprod., 13 (Abstract Book), O-184, p. 93. Rivier, J., Jiang, G-C., Lahrichi, S.L. et al. (1996) Dose relationship between GnRH antagonists and pituitary suppression. Hum. Reprod., 11 (Suppl. 3), Ron-El, R., Herman, A., Golan, A. et al. (1990b) The comparison of early follicular and mid luteal administration of long acting gonadotrophinreleasing hormone agonist. Fertil. Steril., 54, Serafini, P., Stone, B., Kerin, J. et al. (1988) An alternative approach in controlled ovarian hyperstimulation in poor responders : pretreatment with a gonadotrophin-releasing hormone analog. Fertil. Steril., 49, Skarin, G., Nillius, S. J. and Wide, L. (1982) Failure to induce abortion by huge doses of LHRH agonist in women. Contraception, 26, Smitz, J., Devroey, P., Broekmans, P. et al. (1987) Management of failed cycles in an IVF/GIFT programme with the combination of a GnRH analog and hmg. Hum. Reprod., 2, Smitz, J., Ron-El, R. and Tarlatzis, B.C. (1992) The use of gonadotrophin releasing agonists for in vitro fertilisation and other assisted procreation techniques: experience from three centres. Hum. Reprod. (Suppl.), 7, Tan, S.L. (1996) Gonadotrophin-releasing hormone agonists in assisted reproductive therapy. Hum. Reprod., 11 (Suppl. 4), Tarlatzis, B.C., Grimbizis, G., Pournaropoulos, F. et al. (1994a) Evaluation of two gonadotrophin-releasing hormones (GnRHa) analogues (leuprolide and buserelin) in short and long protocols for assisted reproduction techniques. J. Assist. Reprod. Genet., 11, Tarlatzis, B.C., Bili, H., Boutis, J. et al. (1994b) Follicle cyst formation after administration of different gonadotrophin- releasing hormone analogues for assisted reproduction. Hum. Reprod. 9, Testart, J., Lefevre, B. and Gongeon, A. (1993) Effects of gonadotrophin releasing hormone agonists, GnRHa on follicle and oocyte quality. Hum. Reprod., 8, Tolis, G., Comaru-Schally, A.M., Mehta, A.E. and Schally A.V. (1981) Failure to interrupt established pregnancy in humans by D-trytophan-6-luteinising hormone-releasing hormone. Fertil. Steril., 36, World Health Organization (1987) Laboratory Manual for the Examination of Human Semen and Semen Cervical Mucus Interaction. Cambridge University Press, Cambridge, p. 27. Received on March 27, 1998; accepted on October 28,