Will GnRH antagonists provide new hope for patients considered `difficult responders' to GnRH agonist protocols?

Ian Craft1, Amin Gorgy, Jennifer Hill, David Menon and Barbara Podsiadly

London Gynaecology & Fertility Centre, Cozens House, 112a Harley Street, London W1N 1AF, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We have assessed the use of cetrorelix, a gonadotrophin releasing hormone (GnRH) antagonist, in conjunction with clomiphene citrate and gonadotrophin in 31 in-vitro fertilization (IVF)/gamete intra-Fallopian transfer (GIFT) cycles for 25 difficult responders. Group I included 18 poor responders (24 cycles) with no live birth in 23 previous IVF cycles with GnRH agonists. Group II included seven patients (seven cycles) with polycystic ovaries. Thirteen previous IVF/GIFT cycles with GnRH agonists had resulted in one live birth and three of these patients had developed ovarian hyperstimulation syndrome (OHSS). The treatment protocol involved a daily dose of clomiphene citrate 100 mg for 5 days and gonadotrophin injections from cycle day 2. Cetrorelix 0.25 mg/day was started when the leading follicle reached 14 mm. The outcome in both groups was favourable compared to previous treatment with GnRH agonists. In group I the abandoned cycle rate was 29 versus 57% (P = 0.06). More oocytes were produced (6.4 versus 4.7 oocytes/cycle) at a lower dose of follicle-stimulating hormone (FSH) (709 versus 1163 IU/oocyte; P = 0.08) and two live births resulted (11.8%). In group II fewer oocytes were produced (10.2 versus 14.5 oocytes/cycle), using a lower dose of gonadotrophin (170 versus 189 IU/oocyte) and resulted in one ongoing pregnancy. No patients experienced OHSS. This report is preliminary and a further controlled randomized study is required.

Key words: GnRH antagonist/ovarian hyperstimulation syndrome/polycystic ovary/protocols/difficult responders


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
It was the report by Fleming et al. (Fleming et al., 1982Go) that first prompted one of us to apply the use of luteinizing hormone-releasing hormone (LHRH) analogues to assisted conception (Porter et al., 1984Go). It was in the hope that more mature oocytes would result from gonadotrophin stimulation and that down-regulation of the pituitary would eliminate premature luteinizing hormone (LH) surges and the risk of ovulation occurring prior to planned oocyte recovery, as was associated with the clomiphene citrate and gonadotrophin combination, which was then in vogue.

Time has proven the validity of this approach from its ubiquitous use worldwide with the long protocol becoming the preferred option over the short, or ultrashort, alternatives. However, as with most advances, there has been a `trade off' with some adverse aspects, including the cost of the increased amount of gonadotrophin required, the risk of the ovarian hyperstimulation syndrome (OHSS) and its sequelae, especially in polycystic ovary (PCO) patients, and the speculative unproven reports of an association with ovarian tumours.

Not all patients respond well and some may be categorized as `difficult responders' either because of adverse consequences of an exaggerated response leading to OHSS, or because of a failure to respond to increasing doses of gonadotrophins, sometimes administered to very high levels, when used in conjunction with the analogue given on the long, or short, protocol basis. Whilst the latter response may be more frequent in those of older age, some younger women who have quite normal day 3 follicle-stimulating hormone (FSH) values may also be difficult to stimulate.

We have considered that the failure to respond may in certain instances be due to the LHRH analogue having a direct effect on the ovary, with the assumption that the ovary, like the pituitary, also has receptors to LHRH, whose expression are not normally manifested to a degree to prevent an adequate ovulatory response. The recent finding of LHRH receptors in the human ovary lends support to our philosophy (Leung, 1999Go).

We came to this conclusion over a decade ago when an in-vitro fertilization (IVF) birth resulted following withdrawal of nasal Buserelin® (a GnRH agonist) in a 37 year old woman whose response to serial increases in gonadotrophin dosage was profoundly poor. Continuation of ultrasound monitoring, after this cycle had been interrupted and the nasal spray stopped, indicated a subsequent marked increase in follicle growth with later oocyte recovery, fertilization, embryo transfer and establishment of a clinical pregnancy. This effect would not normally be apparent with depot analogue use, and, in any event, follicle monitoring is usually discontinued once a decision has been made to cancel a cycle. However, we have observed this response after discontinuing nasal Buserelin, and more recently, Nafarelin.

The concept that the human ovary may have its response to gonadotrophins suppressed to a variable degree by the direct effect of LHRH analogues on the ovary in certain patients prompted an interest in Gary Hodgen's early work on LHRH antagonists aimed at developing clinical protocols using the normal follicle recruiting system and augmentation by gonadotrophins with gonadotrophin releasing hormone (GnRH) antagonists suppressing LH surges, once recruitment is promoted (Kenigsberg et al., 1984aGo, bGo).

We have recently assessed the use of cetrorelix (GnRH antagonist; Asta Medica, Frankfurt, Germany) used in conjunction with clomiphene citrate and gonadotrophin for ovarian stimulation in a group of patients considered to be `difficult responders' to conventional GnRH analogue and gonadotrophin protocols. Two categories of patients were assessed. The majority was patients who did not respond adequately to increasing doses of gonadotrophins and a minority had been overstimulated.

The concept behind using this clinical protocol has been to establish whether a better-controlled response could be achieved with the normal pituitary–ovarian axis in operation in those in whom a sub-optimal response occurred with a LHRH analogue. Before starting this clinical study we were aware that some patients would not benefit because of a decreasing ovarian reserve due to increasing age, but that a positive response would indicate an alternative stimulation regime and the potential value of this protocol for use in normal responders. This would raise the prospect of reducing the dose and cost of gonadotrophins per cycle, reducing the risk of OHSS (especially in PCO patients), and increasing patient acceptability, with particular relevance to healthy young women, including those acting as egg donors.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We have evaluated the use of cetrorelix in 31 assisted conception cycles in 25 patients who received clomiphene citrate and gonadotrophin, since they had been `difficult responders' to the combination of a GnRH analogue and gonadotrophins. The cetrorelix was used on a `named-patient' basis as it was, as yet, not licensed in the UK.

Patients
The difficult responders were divided into two groups. Group I included 18 patients (24 cycles); 16 patients (22 cycles) were known poor responders from their previous treatment on a GnRH analogue protocol. The remaining two patients had not had previous IVF or gamete intra-Fallopian transfer (GIFT), but were included as their day 2–3 serum FSH levels were 18 and 34 IU/l respectively. The poor responders were patients who had cancelled cycles with less than five progressing follicles following 9 days of FSH therapy given at an appropriate dosage for age and weight or those who have previously required a total dose of more than 600 IU of gonadotrophins per oocyte retrieved. Twelve patients (16 cycles) of the poor responders group were more than 37 years old. In eight patients (10 cycles) serum FSH level was more than 9 IU/l. In 23 previous IVF attempts (16 patients), 13 cycles had been abandoned because of a poor response following a mean dose of 3170 IU of FSH per cycle. Five of these 13 abandoned cycles had been converted to intrauterine insemination (IUI) or timed sexual intercourse (TSI) while the remaining eight had been totally cancelled. The remaining 10 of the previous 23 assisted conception cycles with a GnRH analogue had been completed using a mean dose of 5468 IU of FSH/cycle. A total of 47 oocytes had been retrieved with an average of 4.7 oocytes/cycle and an average total dose of 1163 IU of FSH was required per oocyte retrieved. One clinical pregnancy had been achieved and ended in a miscarriage. Group II included seven patients (seven cycles) who had PCO. One patient had not previously had IVF and the other six underwent 13 previous attempts at IVF/GIFT with a GnRH agonist. One of these 13 cycles had been converted to IUI due to a suboptimal response with cautious low FSH dosage (1200 IU). The other 12 cycles had been completed using an average dose of 2733 IU of FSH/cycle. A total number of 174 oocytes were retrieved with a mean of 14.5 oocytes/cycle. A total of 189 IU of gonadotrophins had been required per each retrieved oocyte. Following these 13 previous treatment cycles, three patients developed ovarian hyperstimulation syndrome requiring hospital admission and two patients became pregnant. One patient later miscarried and the other one had a live birth.

Drug protocol
All patients received clomiphene citrate and gonadotrophin injections following a baseline vaginal scan performed on day 2 of the cycle. Clomiphene citrate was given as a daily dosage of 100 mg from day 2 for 5 days. Gonadotrophin injections were either given daily, or on alternate days. The dose and the frequency of gonadotrophin injections was determined by the patient's age, body weight, FSH level and her response in previous treatment cycles, e.g., alternate days in those who have previously hyperstimulated, as opposed to daily in those who have been profoundly poor responders.

Cetrorelix was administered as a daily 0.25 mg s.c. injection. Initially the injection was started on the 5th day of gonadotrophin stimulation, i.e., on day 6 of the cycle. However, very few patients developed follicles of significant size by this stage, so we delayed the starting day of cetrorelix until day 6 of gonadotrophin injections, i.e. day 7 of the cycle, and assessed the ovarian response by ultrasound on that day. The latter has been our standard method of follicle monitoring for the past 15 years. We later modified our protocol by scanning the patients on day 8 of the cycle and starting the cetrorelix injections on the day that the leading follicle reached 14 mm in diameter irrespective of the day of the cycle, or of how many days gonadotrophins had been administered for. This observation may have particular relevance to the group considered to be `difficult responders' to previous GnRH and gonadotrophin stimulation, since we consider cetrorelix could be started unnecessarily early in such situations. Cyclogest pessaries 400 mg (Cox, Barnstaple, UK) were used twice daily for luteal phase support.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In group 1, seven out of 24 cycles (29.2%) were abandoned from the intended IVF/GIFT after an average total dose of 3121 IU of FSH/cycle: four were converted to IUI/TSI because of a further sub-optimal response, which however, was not so profound as to indicate that alternative IUI treatment should not proceed, and the other three were cancelled altogether. Seventeen IVF/GIFT cycles were completed (70.2%) with an average dose of 4506 IU of FSH/cycle. A total of 108 oocytes were retrieved with a mean of 6.4 oocytes per cycle and a total dose of 709 IU of FSH/oocyte retrieved. Four out of the completed 17 IVF/GIFT cycles resulted in a clinical pregnancy (23.5% per completed IVF/GIFT cycle). Two pregnancies ended in miscarriage while the other two resulted in two twin live births with a live birth rate of 11.8% per completed cycle. For results see Table IGo.


View this table:
[in this window]
[in a new window]
 
Table I. Group I: outcome of cycles on the cetrorelix–clomiphene citrate–gonadotrophin protocol as compared to that of previous cycles on luteinizing hormone-releasing hormone (LHRH) analogue protocols in poor responders
 
One of the seven cycles in group II was converted to IUI (14.3%) due to a suboptimal response following a total dose of 1400 IU of FSH. The remaining six cycles were completed (85.7%) with an average total dose of 1733 IU of FSH/cycle. A total of 61 oocytes were retrieved with a mean of 10.2 oocytes/cycle. A total dose of 170 IU of FSH was required/oocyte retrieved. Three pregnancies were achieved: one pregnancy was ectopic, one pregnancy ended in a miscarriage and the third one resulted in a twin live birth (16.7%) per completed cycle. For results see Table IIGo.


View this table:
[in this window]
[in a new window]
 
Table II. Group II: outcome of cycles on the cetrorelix–clomiphene citrate–gonadotrophin protocol as compared to that of previous cycles on LHRH analogue protocols in polycystic ovary (PCO) patients
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We are encouraged by the response in these groups of patients and acknowledge that those with a limited ovarian reserve may not benefit to the same degree as others who have responded poorly to LHRH analogue and gonadotrophin protocols because of an adverse effect of the analogue on the ovary and in whom one can expect a totally different response with antagonist use. In group I, the success rate for poor responders with previous failed treatment is relatively good. In addition, the response to gonadotrophin stimulation with the antagonist improved compared to that with the agonist as evidenced by a reduced rate of cancelled cycles (29 versus 57%) and improved oocyte production (6.4 versus 4.7 oocytes/cycle) with a lower FSH dosage (4506 versus 5468 IU /completed cycle and 709 versus 1163 IU /oocyte retrieved). This may be due to the release of natural gonadotrophins, together with the stimulating effect of clomiphene citrate and lack of any direct inhibitory effect of GnRH analogues on the ovary. The difference in the rate of cancelled cycles and in the total dose of FSH required per each retrieved oocyte were not quite significant at P values of 0.06 and 0.08 respectively in favour of the cetrorelix protocol. Other data did not show statistically significant differences. This was probably due to small sample size. As many of our patients were slow responders, we delayed the start of the daily cetrorelix injections until the leading follicle was 14 mm in diameter. Olivennes et al. (1998) had a similar observation and delayed giving the single dose injection of cetrorelix in cases of slow follicular development until the serum oestradiol level reached 400 pg/ml. We are encouraged that this protocol will be appropriate for normal responders. This benefit needs to be further clarified by studying the use of the clomiphene citrate–gonadotrophin–cetrorelix protocol in normal responders, compared with that obtained with down-regulation using a GnRH analogue.

In group II, the patients with PCO produced fewer oocytes (10.2 versus 14.5 oocytes/cycle) using a lower dose of FSH (1733 versus 2733 IU/cycle and 170 versus 189 IU/oocyte) and none developed OHSS. Although the trends were in favour of the cetrorelix protocol, the differences did not reach statistical significance. Again, this was probably due to small sample size. The use of down regulation protocols with a GnRH agonist is known to be associated with an increased risk of OHSS, particularly in patients with PCO. However, we believe that the clomiphene citrate–gonadotrophin–cetrorelix protocol reduces this risk as there is no down-regulation prior to ovarian stimulation and one is able to administer cautiously lower doses of gonadotrophins, either on an alternate day, or on a daily basis. The requirement of lower doses of gonadotrophin, with the use of GnRH antagonist, was previously suggested by Olivennes et al. (1994). Furthermore, human chorionic gonadotrophin (HCG) administration for induction of ovulation can be replaced by native GnRH or a GnRH agonist, and the luteal support might be unnecessary, both reducing the risk of ovarian hyperstimulation (Emperaire and Ruffie, 1991Go).

Our experience of using clomiphene citrate–gonadotrophin–cetrorelix for `difficult responders' suggests that this protocol will be of value for some patients who have not responded well to GnRH analogues and gonadotrophins. We predict that an improved response will occur in some, and the required amount of gonadotrophin injections will be reduced and hence the cost. We also believe that cetrorelix reduces the risk of OHSS, since a lower dose of gonadotrophins may be found effective in stimulating a moderate number of follicles, rather than an excessive cohort, particularly in patients with PCO because of lower stimulation. A properly controlled randomized study is required to confirm our preliminary observation which is based on a small number of cycles.

Cetrorelix therapy, we believe, will have relevance to IUI treatment cycles, and those which are converted to IVF or GIFT, for frozen embryo replacement in a natural cycle, and possibly even to IVF in natural cycles in young women with favourable features. We also consider it will have a significant place in the management of oocyte donation, both from the perspective of donors who will require less gonadotrophins, and therefore have less risk of OHSS, and also from the recipient's perspective, since donation is likely to proceed with less risk of cycle cancellation.


    Notes
 
1 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Emperaire, J.C. and Ruffie, A. (1991) Triggering ovulation with endogenous luteinizing hormone may prevent the ovarian hyperstimulation syndrome. Hum. Reprod., 6, 506–510.[Abstract]

Fleming, R., Adam, A.H., Barlow, D.H. et al. (1982) A new systematic treatment for infertile women with abnormal hormone profiles. Br. J. Obstet. Gynaecol., 89, 80–83.[ISI][Medline]

Kenigsberg, D., Littman, B.A. and Hodgen, G.D. (1984a) Medical hypophysectomy: I. Dose-response using a gonadotrophin releasing hormone antagonist. Fertil. Steril., 42, 112–115.[ISI][Medline]

Kenigsberg, D., Littman, B.A., Williams, R.F. and Hodgen, G.D. (1984b) Medical hypophysectomy: II. Variability of ovarian response to gonadotrophin therapy. Fertil. Steril., 42, 116–126.[ISI][Medline]

Leung, P.C.K. (1999) GnRH receptor and potential action in human ovary. Gynaecol. Endocrinol., 13, 10.

Olivennes, F., Fanchin, R., Bouchard, P. et al. (1994) The single or dual administration of the gonadotrophin-releasing hormone antagonist Cetrorelix is an in vitro fertilization–embryo transfer programme. Fertil. Steril., 62, 468–476.[ISI][Medline]

Olivennes, F., Alvarez, S., Bouchard, R. et al. (1998) The use of a GnRH antagonist (cetrorelix) in a single dose protocol in IVF–embryo transfer; a dose finding study of 3 versus 2 mg. Hum. Reprod., 13, 2411–2414.[Abstract]

Porter, R.N., Smith, W., Craft, I.L. et al. (1984) Induction of ovulation for in vitro fertilization using Buserelin and gonadotrophins. Lancet, ii, 1284–1285.

Submitted on April 12, 1999; accepted on September 14, 1999.