Ovulation Induction and Controlled Ovarian Stimulation, 2st ed.

16. Controlled Ovarian Stimulation for IVF/ICSI

Roy Homburg1


Homerton Fertility Centre, Homerton University Hospital, London, UK


The aim of controlled ovarian hyperstimulation (COH) for IVF is to produce multiple follicular development in order to harvest a suitable number of oocytes which can be fertilised and allow a selection of embryos which can be replaced into the uterus. Premature luteinisation is avoided by suppressing pituitary gonadotrophin production by co-treating with a GnRH agonist or antagonist, each with its own pros and cons. Individually tailoring the programme attempts to achieve the best live birth rates while taking into account the avoidance of ovarian hyperstimulation syndrome and the promotion of patient comfort, performed by utilizing several known facts such as ovarian response in any previous cycles, age, AMH or other predictors of ovarian reserve. The choice of gonadotrophin preparation for COH makes little if any difference regarding live birth rates while the antagonist is preferred over the agonist as regards patient comfort. Pre-treatment predictions of high, low and normal responders (using age, AMH and/or AFC) determines the protocol to be used and the starting dose of stimulation.


Controlled Ovarian StimulationIVFICSIOvarian hyperstimulationFollicular developmentOocytesEmbryosUterusOvarian stimulationOvulation inductionAnovulatory patientsMonofollicular ovulationGonadotrophinsOestradiolLHLuteinisationFolliclesGnRHOvarian hyperstimulation syndromeOHSSFrozen-thawed cyclesrecFSHrecLHARTAMHFSHhMGMinimal stimulationNatural cycleFollitropin-αFollitrophin-βHypogonadotrophic-hypogonadismFolliclesAntral follicleLuteal phaseOestrogenOral contraceptivesOvarian cystsCetrorelixGanirelixPremature luteinisation

16.1 Principles

The aim of controlled ovarian hyperstimulation (COH) for IVF is to produce multiple follicular development in order to harvest a suitable number of oocytes which can be fertilised and allow a selection of embryos which can be replaced into the uterus. This is in sharp contrast to ovarian stimulation for ovulation induction in anovulatory patients in which the aim is to produce a monofollicular ovulation. The difference is therefore in the strength of stimulation needed and, obviously, more gonadotrophins are needed to produce more follicles for COH in IVF.

Because of the high oestradiol concentrations produced by COH, in about 15–25 % of patients a positive feedback mechanism will produce a premature LH rise causing premature luteinisation of the developing follicles and abandonment of the cycle. Today this is avoided by suppressing pituitary gonadotrophin production by co-treating with a GnRH agonist or antagonist. The use of the agonist brought with it a number of advantages and some disadvantages and various protocols involving the use of a GnRH agonist have been developed to tailor individual requirements.

Later, GnRH antagonists appeared on the market for the same purpose of providing a temporary reduction of, primarily, LH production during COH with gonadotrophins. Their properties differ in several respects from those of GnRH agonists and each has its own pros and cons.

These improvements in protocols for COH in preparation for IVF have given us a variety of programmes from which to choose. While some centres use a permanent protocol for all patients, most individually tailor their programme in an attempt to achieve the best live birth rates while taking into account the avoidance of ovarian hyperstimulation syndrome and the promotion of patient comfort. Tailoring is performed by utilizing several known facts such as ovarian response in any previous cycles, age, AMH or other predictors of ovarian reserve, etc. All these can help determine the type of GnRH analogue to be used and the starting dose of FSH or hMG for stimulation. The choice of protocol may range from natural cycle IVF, modified natural cycle, minimal stimulation, so-called ‘soft’ stimulation protocols to full-blown COH. Each has its own advantages and disadvantages for the individual patient.

16.2 Gonadotrophins in COH

16.2.1 The Choice of Gonadotrophin Preparation

A number of gonadotrophin preparations are available for COH in assisted reproductive technologies. They all contain FSH whether derived from menopausal urine and purified or from recombinant DNA technology, or urinary preparations of highly purified human menopausal gonadotrophins (hMG) containing both FSH and LH or a combination of recombinant FSH and LH. The attributes of the various preparations have been discussed in respect to ovulation induction in Chap. 13 and the basic properties are, of course, applicable to COH for IVF.

Much discussion, debate, clinical research and scientific papers have been presented in the last few years regarding the suitability of these various preparations for COH in assisted reproductive technologies. As IVF is now so widespread and relatively expensive then commercial interests have also been involved. The main debates have centred around urinary versus recombinant and LH (or hCG) containing versus pure FSH preparations. The most poignant end-point to use in this discussion is that of the live birth rate. Here the differences, if any, are extremely subtle. As so much individual research has been published, it is probably wisest to heed large independent meta-analyses such as that performed by the National Institute of Clinical Excellence of the UK in 2004 [1]. A total number of 21 RCT’s involving 4,727 women were examined. When comparing recombinant FSH to any urinary derived FSH-containing preparation using a long GnRH agonist protocol, no significant differences were found in live birth rates, ongoing or clinical pregnancies. Recombinant FSH seems to be more efficient than urinary FSH in that, in the majority of studies, less ampoules are needed, i.e. 1 IU of recFSH is more potent than 1 IU of urinary FSH. This difference may be offset by the increased price of the recombinant product but its superiority as regards purity and batch-to-batch consistency are not disputed.

Meta-analyses of RCT’s comparing FSH with hMG preparations have shown contrasting results. Using clinical pregnancy rates as the end-point, one study from 2000 [2] showed in favour of hMG but the same authors updated this analysis 10 years later and, following the myriad studies on this subject which are still appearing (almost ad nauseam!), we have put our faith in this, one of the latest and most comprehensive of the Cochrane reviews by van Wely et al. in 2012 [3]. Comparing the effectiveness of recombinant FSH (recFSH) with three urinary gonadotrophins; hMG, purified and highly purified FSH, there was no difference in pregnancy outcomes, live birth rates or incidence of OHSS both in fresh and frozen-thawed cycles. Although recFSH was slightly inferior to hMG regarding live birth rate, the authors concluded that all available gonadotrophins seem equally effective and safe.

The overall conclusion regarding the use of the various gonadotrophin preparations in COH for ART must be that there is little if any difference between them regarding clinical pregnancy and live birth rates. As costs vary in each country, it is difficult to comment on cost efficiency. Finally, unsurprisingly, no difference has been found between the two clinically available recombinant FSH preparations, follitropin-α and follitrophin-β [4].

16.2.2 LH Content

The need for some LH content in COH protocols, whether of recombinant or urinary origin, is still under discussion. The fact that some LH, albeit in small amounts, is necessary both physiologically and for efficient induction of ovulation for women who have hypogonadotrophic-hypogonadism and virtually no endogenous LH, raised the question of whether over-suppression of pituitary LH secretion can affect the results of COH with FSH alone. This is a contentious issue as excessive suppression of LH concentrations with GnRH agonists has been associated with a detrimental effect on the outcome of IVF [5], while others comparing urinary FSH with hMG suggested that resting levels of LH following down-regulation are sufficient to support development and maturation of follicles and oocytes in normo-gonadotrophic women [6]. The truth probably lies somewhere in between as some studies [7], suggest that mid-follicular LH serum concentrations of <0.5 IU/l are likely to be detrimental to IVF outcome. This probably only occurs in a very small minority of cases as only 1 % of LH receptors need to be occupied in order to produce a full LH effect but nevertheless, opens the door for less aggressive doses of GnRH analogues or, possibly, the addition of LH to COH protocols, whether recombinant or urinary, in certain cases. A Cochrane review [8] included a heterogeneous collection of studies examining the effectiveness of a combination of recFSH and recLH compared with recFSH in both GnRH agonist and antagonist cycles. When this data was pooled, no significant difference in live birth rates emerged and no benefit of LH supplementation for COH was seen in an unselected population. Although there was some suggestion that higher concentrations of LH in the mid- and late follicular phase of cycles stimulated with pure FSH under GnRH analogue pituitary suppression are beneficial to the older age groups or poor responders undergoing COH for IVF [8] subsequent large RCT’s comparing FSH + LH and FSH alone (now in press) have not substantiated this claim.

16.2.3 Starting Doses

The starting dose for gonadotrophin stimulation of the ovaries for the first IVF cycle should be determined by the AMH serum concentrations and/or antral follicle count in combination with age and BMI. This is a far more accurate method, certainly for the first cycle, than guessing on the strength of age alone. These are particularly good for predicting both high and low responders, which after all, are the two groups which are important to predict ahead of treatment. For the predicted high responders, a low dose of FSH should be used in the first cycle (not more than 150 IU) in order to avoid multiple follicular development and possible OHSS. For predicted low responders, a relatively high dose of FSH can safely be used in the hope of cutting down the duration of stimulation. For predicted normal responders (who do not have polycystic ovaries), a starting dose of 150 or 250 IU makes little difference in the number of oocytes retrieved or pregnancies achieved. See Chaps. 17 and 18 for more details on the management of poor and high responders respectively.

16.2.4 Patient Comfort

Regarding patient comfort, an important advance has been made in delivery systems. The development of injection ‘pens’ by three major companies has provided an almost foolproof system by which the patient can self-inject the prescribed amount of daily FSH with much greater accuracy and without recourse to daily advice from the medical staff. A further advance to enhance patient convenience is the use of a long-acting FSH preparation whose aim is to reduce the number of injections required to achieve the required stimulation in COH. These novel delivery systems are more fully described in Chap. 22.

16.2.5 Triggering Ovulation

Traditionally hCG is used for the triggering of ovulation following COH when the appropriate criteria are attained. In a single dose of 5,000–10,000 IU, urinary hCG has proved an excellent substitute for the LH surge. With the emergence of recombinant technology, recombinant LH has been compared with the standard doses of urinary hCG for triggering of ovulation. A single dose of recombinant LH of between 15,000 and 30,000 IU was found to be equally effective as 5,000 IU of hCG in inducing final follicular maturation and luteinisation [9]. Because of its much shorter half-life than hCG, the use of recombinant LH for this purpose reduces the incidence of ovarian hyperstimulation syndrome but it is no longer used for triggering ovulation as large repeated doses are needed to support the luteal phase and this is much more expensive than hCG and standard progestin vaginal pessaries. .

When a GnRH antagonist is used for pituitary down-regulation, then a single injection of an agonist can also be utilized to trigger ovulation. This employs the immediate temporary discharge of LH following the injection and with the aim of preventing OHSS in susceptible cases, it has found a place in the armamentarium. Heavy luteal phase support is needed to maintain pregnancy rates. Alternatively, a ‘freeze all’ policy with replacement in a consequent cycle has produced the best results. See Chap. 18 for a detailed description of management of the predicted high responder.

16.3 GnRH Agonists

16.3.1 Protocols

The use of GnRH agonists concomitant with gonadotrophin stimulation for IVF has the primary purpose of blocking the oestrogen initiated positive feedback mechanism and preventing release of LH. Following a short initial discharge of both FSH and LH, it induces down-regulation and densensitisation of the GnRH receptors in the anterior pituitary. A GnRH agonist can be administered daily by the sub-cutaneous or intra-nasal routes. A depot preparation (microcapsules) given by deep intra-muscular injection is also available which has a duration of action of about 28 days and, although saving daily use, may induce a more severe suppression than is really required and hence, increased gonadotrophin requirements and duration of stimulation.

There are many ways of using these compounds, the two most popular of which, the long (luteal) and the short protocol, are illustrated in Fig. 16.1.


Figure 16.1

The two most widely used protocols involving GnRH agonists in controlled ovarian stimulation for IVF. (a) The long (luteal) protocol, (b) the short protocol. The arrows denote the days of GnRH agonist administration

In the long (luteal) protocol, GnRH agonist is started on day 21 of the cycle preceding treatment and continued in a constant dose until the day of hCG administration. It is continued in parallel with gonadotrophin treatment which is usually started on the first days of an ensuing menstruation, after two weeks of agonist treatment or following demonstration of pituitary down regulation by measuring low (<200 pmol/l) oestradiol levels. Although this is the most widely used long protocol, some variations on this theme are being used. In the long follicular protocol, the GnRH agonist is started on the first day of the preceding cycle. Some centres prefer to reduce the dosage of the agonist once gonadotrophin stimulation is started rather than giving a constant dose throughout in an attempt to cut down gonadotrophin dosage, the ‘early cessation’ protocols.

The short or ‘flare-up’ protocol of GnRH administration is an attempt to utilize the initial, temporary FSH releasing properties of the agonist to promote follicular recruitment during menstruation before the suppressive action takes over. The agonist is given from day 2 of the cycle until the day of hCG administration in a constant dose and gonadotrophins are started on cycle day 3. The short protocol is mainly used for the older patients or poor responders. A further alternative, touted for poor responders, is the ultra-short protocol in which the GnRH agonist is given only for the first 3 days of the cycle and then discontinued and gonadotrophins are started on cycle day 3.

The long protocol starting in the mid-luteal phase has consistently been reported to be more effective than the short and ultra-short protocols as far as pregnancy rates are concerned [1011] and the most recent Cochrane review [12] has confirmed these findings. The higher clinical pregnancy rate achieved with the long protocol comes with a higher cost as more gonadotrophins are needed to reach the criteria for the triggering of ovulation. There is no obvious difference in the results of the long protocol between intra-nasal or sub-cutaneous administration, luteal or follicular start or the use of a stop or reduced dose protocol. For poor responders, the short and early cessation protocols may have a place. The advantages and disadvantages of the long protocol for administration of a GnRH agonist are listed in Table 16.1.

Table 16.1

Use of GnRH agonist in ART: advantages (+) and disadvantages (−)


Prevent LH surges and premature luteinisation

Long duration of administration

Increase number of available oocytes

Increase FSH requirements

Allow programming

Increase incidence of OHSS

Improve pregnancy rates

Oestrogen withdrawal symptoms

16.3.2 Oral Contraceptives and the Long GnRH Protocol

Pre-treatment with the oral contraceptive pill, before starting the GnRH agonist, may offer several advantages. It is effective in preventing ovarian cysts, which may be a troublesome side effect of GnRH agonist therapy in some patients. This dual suppression of pituitary gonadotrophin release has shown good results in high responder patients and also allows accurate programming, especially important in a busy IVF centre.

16.3.3 Doses

The doses of GnRH agonists used in IVF today are, in the main, probably much higher than needed for their principal purpose, i.e. to merely suppress LH concentrations enough to prevent premature luteinisation. There is a curious lack of dose finding studies and overdosing increases gonadotrophin requirements and the duration of administration. The optimal minimal effective dose of each agonist preparation needs to be sought in order to make their use more efficient and effective.

16.4 GnRH Antagonists

16.4.1 Principles

After many years involved in their development, two GnRH antagonists, cetrorelix and ganirelix, are now available for clinical use. The mode of action of the antagonists is completely different from that of the agonists in that they cause pituitary suppression of gonadotrophin release by receptor competition. The GnRH antagonists bind competitively to GnRH receptors so that endogenous GnRH cannot act. The result is an immediate decrease in gonadotrophin concentrations, within hours after administration, without any flare-up effect as seen with the agonists. Because the principle of the GnRH antagonists’ action is competition for a place on the GnRH receptors, their duration of action is dose dependent. The differences in the mode of action and properties of the GnRH agonists and antagonists make for differences in their use for gonadotrophin suppression during ovarian stimulation for IVF/ICSI.

Table 16.2 contrasts agonists and antagonists with respect to their mode of action and use in ART. The theoretical and proven advantages of the antagonist compared with the agonist, shown in Table 16.3, include a shorter stimulation period and a decreased requirement for gonadotrophins, a reduced incidence of OHSS, no flare-up effect and an absence of oestrogen withdrawal symptoms.

Table 16.2

Comparison of GnRH agonists and antagonists



Act by down-regulation and desensitization of GnRH receptors

Act by GnRH receptor competition

Suppress LH release after flare-up

Suppress LH release, immediate

No flare-up

Administered by subcutaneous daily injection or depot preparation i.m.

Administered subcutaneously as depot or daily injection

Best results starting mid-luteal phase, long protocol

Given on day 5–6 of stimulation (fixed) or when lead follicle is 14 mm (flexible)

Table 16.3

Comparison of GnRH agonists and antagonists: advantages perceived are signified as +




Duration of action dose dependent



FSH requirement



Number of retrievable oocytes



Incidence of OHSS






Side effects



Clinical pregnancy rates


Triggering of ovulation with GnRH or GnRH agonist









16.4.2 Protocols

Two issues are relevant to the design of protocols for the use of the GnRH antagonists during COH for IVF/ICSI:



The possibilities and combinations which have been suggested and are in use are presented in Figs. 16.2 and 16.3.


Figure 16.2

The use of a GnRH antagonist in a fixed day protocol. (a) With a single, slow release injection of cetrorelix; (b) a multiple dose protocol using either cetrorelix or ganirelix


Figure 16.3

The use of a GnRH antagonist in a flexible protocol. (a) With a single, slow release injection of cetrorelix; (b) a multiple dose protocol using either cetrorelix or ganirelix

Either cetrorelix or ganirelix may be used in the multiple dose protocol whereas only cetrorelix is presently available as a depot preparation which can be used in a single dose scheme. In the single dose regimen, the optimal dose to prevent a premature rise of LH seems to be 3 mg of cetrorelix. This is given either on a fixed day, usually day 6–8 of gonadotrophin stimulation or, in a flexible protocol, usually when the leading follicle reaches a diameter of 14 mm. A repeat injection is sometimes required if the criteria for administration of hCG have not been reached within 72–96 h following the initial injection.

For the more popular multiple dose protocol, the optimal daily dose for either cetrolerix or ganirelix is 0.25 mg given sub-cutaneously. This again may be applied from a fixed day of gonadotrophin stimulation, usually day 5 or 6, or more flexibly, according to the diameter of the leading follicle, usually 14 mm. In either case, the antagonist is continued up to and including, the day of hCG administration.

16.4.3 Single or Multiple Doses, Fixed Day or Flexible?

Two studies assessing the multiple dose and the single dose use of cetrorelix have been analysed [13], the first in a dose of 0.25 mg/day (n = 1,066) and the second in a single dose of 3 mg (n = 541). The results of each were comparable as regards pregnancy rates/embryo transfer (27 % vs 28 % respectively), numbers undergoing oocyte retrieval and embryo transfer and the number of embryos obtained and transferred. The prevalence of OHSS in each case was negligible. Of the patients in the single dose protocol, 27 % needed a repeat injection. Although it would seem that, all things being equal, a single injection protocol is preferable to daily injections, daily injections are much more widely used.

A fixed day of administration (usually day 5 or 6 of gonadotrophin stimulation) or a flexible protocol in which the antagonist is usually given when the dominant follicle reaches a diameter of 14 mm, have little to choose between them as far as results are concerned. A comparison of these two alternatives in a multiple dose protocol [14] found no significant differences in efficacy. Pregnancy rates were equally high (44.4 and 51 % in day 6 and flexible protocols respectively) but, as the authors pointed out, this was a selected group of ovulating healthy patients under the age of 36 years. Lately, my personal preference is the fixed protocol (starting on day 5 of stimulation) which had a slight edge in results compared with flexible and has the added advantage that it is simpler for patient and practitioner alike to predetermine the starting day. There is a modern tendency to start the antagonist as early as day 2 of stimulation which, although more expensive, is said to involve less premature luteinisation and better results.

Agonist Versus Antagonist

Since its inception, pregnancy rates with the use of the antagonist were consistently reported to be inferior to those when using a long agonist protocol and this deterred many practitioners from using them. In the first edition of this book (2005), I suggested that the most likely reason for this was the fact that we were still on a learning curve regarding the use of the antagonist and that as our experience with their use mounts, the antagonist will be equally, if not more effective than the agonists and their added advantages will be more widely appreciated. I am pleased to see that this observation has come to fruition. The latest Cochrane analysis has shown no significant difference in live birth rates between a long agonist and an antagonist protocol and a significantly lower risk of OHSS using an antagonist protocol [15]. This, together with the advantages of the antagonist mentioned above (particularly a shorter cycle, less gonadotrophin consumption and more patient comfort) has persuaded many centres to prefer an antagonist protocol for the majority of their treatment cycles.



National Institute for Clinical Excellence Guideline. Fertility: assessment and treatment for people with fertility problems. 2004; 60–2. RCOG Press. ISBN 1-900364-97-2.


van Wely M, Westergaard LG, van der Veen F, Bossuyt PMM. Effectiveness of human menopausal gonadotrophin versus recombinant follicle-stimulating hormone for controlled ovarian hyperstimulation in assisted reproductive cycles: a meta-analysis. Fertil Steril. 2003;80:1086–93.PubMedCrossRef


van Wely M, Kwan I, Burt AL, Thomas J, Vail A, van der Veen F, Al-Inany HG. Recombinant versus urinary gonadotrphin for ovarian stimulation in assisted reproductive technology cycles. Hum Reprod Update. 2012;18:111.PubMedCrossRef


Harlin J, Aanesen A, Csemiczky G, et al. Delivery rates following IVF treatment, using two recombinant FSH preparations for ovarian stimulation. Hum Reprod. 2002;17:304–9.PubMedCrossRef


Fleming R, Lloyd F, Herbert M, et al. Effects of profound suppression of luteinizing hormone during ovarian stimulation on follicular activity, oocyte and embryo function in cycles stimulated with purified follicle stimulating hormone. Hum Reprod. 1998;13:1788–92.PubMedCrossRef


Daya S, Gunby J, Hughes EG, et al. Follicle-stimulating hormone versus human menopausal gonadotrophin for in-vitro fertilization cycles: a mets-analysis. Fertil Steril. 1997;67:889–99.CrossRef


Humaidan P, Bungum L, Bungum M, et al. Ovarian response and pregnancy outcome related to mid-follicular LH levels in women undergoing assisted reproduction with GnRH agonist down-regulation and recombinant FSH stimulation. Hum Reprod. 2002;17:2016–21.PubMedCrossRef


Mochtar MH, van der Veen F, Ziech M, van Wely M, Musters A. Recombinant LH for controlled ovarian hyperstimulation in assisted reproductive cycles. Cochrane Database Syst Rev. 2007;(2):CD005070.


The European Recombinant LH Study Group. Human recombinant luteinizing hormone is as effective as, but safer than, urinary human chorionic gonadotrophin in inducing final follicular maturation and ovulation in in-vitro fertilization procedures: results of a multicenter double-blind. J Clin Endocrinol Metab. 2001;86:2607–18.CrossRef


Daya S. Gonadotropin releasing hormone agonist protocols for pituitary desensitization in in-vitro fertilization and gamete intra-fallopian transfer cycles. Cochrane Library, Cochrane Collaboration. 2000;(2). Oxford: Update Software.


Pellicer A, Simon C, Miro F, et al. Ovarian response and outcome of in-vitro fertilization in patients treated with gonadotrophin-releasing hormone analogues in different phases of the menstrual cycle. Hum Reprod. 1989;4:285–9.PubMed


Maheshwari A, Caserta D, Siristatidis CS, Bhattacharya S. Gonadotrophin-releasing hormone agonist protocols for pituitary suppression in assisted reproduction. Cochrane Database Syst Rev. 2011;(8):CD006919.


Oliviennes F, Diedrich K, Frydman R, Felberbaum RE, Howles CM, Cetrotide Multiple Dose International Study Group. Safety and efficacy of a 3 mg dose of the GnRH antagonist cetrorelix in preventing premature LH surges: report of two large multicentre, multinational, phase IIIb clinical experiences. Reprod Biomed Online. 2003;6:432–8.CrossRef


Escudero E, Bosch E, Crespo J, Simon C, Remohi J, Pellicer A. Comparison of two different starting multiple dose gonadotropin-releasing hormone antagonist protocols in a selected group of in vitro fertilization-embryo transfer patients. Fertil Steril. 2004;81:562–6.PubMedCrossRef


Al-Inany HG, Youssef MA, Aboulgar M, et al. Gonadotrophin-releasing hormone antagonists for assisted reproductive technology. Cochrane Database Syst Rev. 2011;(1):CD008046.

If you find an error or have any questions, please email us at admin@doctorlib.info. Thank you!