1 Introduction

Since the early days of in vitro fertilisation (IVF), the results of IVF treatment have much improved with a 32.8% live birth rate being reported for women aged under 35 years in the United Kingdom in the year 2012 [1]. The paradigm shift from natural unifollicular IVF treatment cycles to multifollicular stimulated IVF treatment cycles has been an important contributing factor to this improvement, largely enabled by the availability of ovulation induction drugs. It has led to the evolution of the concept of superovulation whereby the ovaries are stimulated to produce high numbers of good quality oocytes that will compensate in part for the deficiencies in IVF and cleavage, and facilitate a yield of good numbers of high quality embryos available for transfer, thereby increasing the probability of pregnancy. Ovarian stimulation is now an essential part of IVF with 98.3% of IVF in the United Kingdom being stimulated cycles in 2013 [1].

The use of superovulation regimens led to the introduction of controlled ovarian stimulation (COS) in order to achieve better cycle control by the avoidance of a premature luteinising hormone (LH) surge. A premature LH surge leads to high cycle cancellation and poor pregnancy rates as a result of either premature ovulation or inappropriate luteinisation before oocyte retrieval. In conjunction with the drugs that cause multifollicular stimulation of the ovaries, pituitary suppression with gonadotrophin releasing hormone (GnRH) analogues which eliminate endogenous gonadotrophin interference caused by exogenous superovulation regimens and timed ovulation trigger control all events in the process of COS.

An understanding of the physiology of ovulation is important to comprehend the use of stimulation regimens in IVF treatment. During a normal menstrual cycle following the involution of the corpus luteum and the consequent fall in oestrogen production, follicle stimulating hormone (FSH) levels rise during the luteo-follicular transition [2]. This rise in FSH stimulates the recruitment of a cohort of follicles. Further development of these follicles during the follicular phase is dependent on continued stimulation by gonadotrophins. According to the concept of an FSH ‘threshold’ postulated by Brown in 1978, FSH concentrations need to exceed a certain level for follicular development to proceed [3]. When the FSH ‘threshold’ is surpassed in a normal cycle, it stimulates the growth of a cohort of small antral follicles and ensures further preovulatory follicular development [4]. The duration of this period in which the threshold is exceeded (FSH ‘window’) is limited in the normal cycle as there follows a decrease in FSH during the early-mid follicular phase [2]. Extension of the ‘window’ period leads to multiple follicular development. In a normal menstrual cycle, a single follicle continues to develop despite falling levels of FSH due to an increased sensitivity to the hormone [5], but the remaining follicles undergo atresia in response to the declining levels of FSH. Ovulation is triggered by a sharp and transient surge in LH as a result of the positive feedback from oestrogen produced by the dominant growing preovulatory follicle.

Thus the duration of elevated FSH plays an important role in determining the number of follicles that will undergo further development. In superovulation regimens, a supra-physiological dose of FSH is used to recruit multiple ovarian follicles with higher ‘threshold’ FSH requirements. Exogenous administration of FSH stimulates the granulosa cells of the ovarian follicles and induces multiple follicular growth.

Ovarian stimulation is important as it enhances the oocyte yield and the number of oocytes retrieved is an important prognostic variable for IVF success [6]. The aim of COS is to optimise the number of oocytes with individualised stimulation regimens by fine tuning the different stages of COS to achieve maximum efficacy and safety [7].

The components of COS are pituitary suppression, ovarian stimulation and ovulation triggering preparing for oocyte retrieval. The various events of COS in relation to the hypothalamic pituitary ovarian axis and the commonly used drugs at each stage are represented schematically in Figure 17.1.

Fig. 17.1 Stages of COS

2 Pituitary Suppression Regimens

The commonly employed COS regimens in IVF are the long GnRH agonist regimen, the short GnRH agonist regimen, the GnRH antagonist regimen and modifications of these.

2.1 Long GnRH Agonist Regimen

With the GnRH agonist long regimen, pituitary down-regulation with GnRH agonist is commenced in the follicular phase or more commonly in the mid luteal phase of the menstrual cycle. Menstruation usually follows within two weeks of starting GnRH agonist. Pituitary down-regulation is confirmed by transvaginal ultrasound scan demonstrating quiescent ovaries with follicles of size ≤ 10 mm diameter and endometrium ≤ 5 mm in thickness. On confirmation of down-regulation, ovarian stimulation is commenced with gonadotrophin injections. Ultrasound monitoring is performed during ovarian stimulation to assess follicular recruitment and growth. Gonadotrophin and GnRH agonist administration are continued until ovulation trigger. Human chorionic gonadotrophin (hCG) is administered subcutaneously when the criteria for triggering final oocyte maturation is met. Figure 17.2a is a schematic representation of events in the long GnRH agonist regimen.

Fig. 17.2a Long GnRH agonist regimen

A modification of the long agonist regimen is the long stop regimen where the GnRH agonist is stopped on commencement of ovarian stimulation. The use of this regimen is based on the rationale that prolonged pituitary down-regulation with continuous GnRH agonist administration during ovarian stimulation with gonadotrophins suppresses ovarian response. However, current evidence does not favour the use of the long stop regimen over the standard long GnRH agonist regimen [8].

2.2 Short GnRH Agonist Regimen

With the short GnRH agonist regimen, treatment is started in the early follicular phase (days 1 to 3 of the menstrual cycle) after a transvaginal ultrasound scan confirming quiescent ovaries and a thin endometrium ≤ 5 mm in thickness. Gonadotrophin injections are commenced one day following the start of the GnRH agonist. Ultrasound monitoring is performed during ovarian stimulation to assess follicular recruitment and growth. Gonadotrophin and GnRH agonist administration are continued until ovulation trigger. hCG is administered subcutaneously when the criteria for triggering ovulation is met. Figure 17.2b is a schematic representation of events in the short GnRH agonist regimen.

A modification of the short agonist regimen is the ultra-short regimen where the GnRH agonist is stopped following commencement of ovarian stimulation with gonadotrophins.

Fig. 17.2b Short GnRH agonist regimen

2.3 GnRH Antagonist Regimen

With the GnRH antagonist regimen, ovarian stimulation with gonadotrophin injections is commenced in the early follicular phase of the menstrual cycle after a transvaginal ultrasound scan confirming quiescent ovaries and a thin endometrium (≤ 5 mm). The GnRH antagonist is commenced on day 6 of stimulation (fixed administration) or when the leading follicle is ≥14 mm (flexible administration). Both the gonadotrophin and the GnRH antagonist are continued until the day of ovulation triggering. Figure 17.2c is a schematic representation of events in the GnRH antagonist regimen. In an antagonist cycle the trigger is either hCG as above or GnRH agonist.

Fig. 17.2c GnRH antagonist regimen

3 Gonadotrophin Dose and Type

It is imperative to use the right gonadotrophin dose to optimise the number of oocytes retrieved, maximise live birth rates following IVF and at the same time minimise risks such as ovarian hyperstimulation syndrome (OHSS) and cycle cancellation. When exogenous gonadotrophin is administered, the number of mature follicles recruited largely depends upon the number of follicles attaining FSH sensitivity. Hence administration of a high gonadotrophin dose may induce excessive ovarian response consequently leading to a high risk of OHSS. On the other hand, administration of an inappropriately low gonadotrophin dose may lead to the growth of a low number of follicles resulting in an ‘iatrogenic’ poor response.

The successful therapeutic use of urinary gonadotrophins started with the first generation product human menopausal gonadotrophin (hMG) or menotropin, which contained 75 IU of FSH and 75 IU of LH in each standard ampoule. This was followed in the early 1980s by the development of urofollitropin, the second generation product from which the LH activity had been reduced to 0.1 IU/ 75 IU FSH. Subsequently, the third generation product, highly purified urofollitropin (Metrodin HP^®) with practically no residual LH activity was developed in the early 1990s. Due to its enhanced purity with very small amount of protein, Metrodin HP^® could be administered subcutaneously which was an advantage over the previous generations which had to be administered intramuscularly. The more recent fourth generation gonadotrophin is produced in vitro through recombinant deoxy ribo nucleic acid (DNA) technology, by genetically engineered Chinese hamster ovary cells. This is recombinant human FSH (r-FSH or follitropin) which is free of LH and contains less than 1% of contaminant proteins. There are two preparations of r-FSH that are commercially available for clinical use, follitropin-α and follitropin-β, both of which have the advantage of subcutaneous administration. Over the years there have been numerous randomised controlled trials (RCTs) comparing urinary gonadotrophins versus recombinant FSH for COS. Current evidence suggests that the two gonadotrophin preparations are comparable in IVF outcomes [9].

4 Ovulation Trigger

Following recruitment and growth of follicles to the mature stage resulting from ovarian stimulation, the next step is maturation of oocytes facilitated by the ovulation trigger in COS regimens. The LH surge that induces germinal vesicle breakdown and ovulation in a natural menstrual cycle is not suppressed in stimulated multifollicular cycles necessitating artificial triggering of ovulation. hCG which is naturally produced by the human placenta and excreted in large quantities in the urine of pregnant women bears a close molecular resemblance to LH and has a similar effect on the LH receptor. hCG is used because of this molecular resemblance and also due to its longer serum half-life (36 hours) compared to the short serum half-life of LH (108–148 minutes) [10], which avoids the inconvenience of repeated administrations. Administration of hCG results in luteinisation of the granulosa cells, progesterone biosynthesis, resumption of meiosis, oocyte maturation and subsequent follicular rupture 36–40 hours later. It is administered after the stimulated development of mature preovulatory follicles in order to induce maturation, but oocyte retrieval is undertaken before ovulation. The usual criteria for the administration of hCG is the presence of ≥ 3 follicles of ≥ 18 mm in diameter. The preparations of hCG that are available for clinical use are the urinary and recombinant forms and are comparable for IVF outcomes [11]. The usual dose of hCG for final ovulation triggering is between 5000 IU and 10,000 IU as a single dose.

The GnRH agonist trigger has been proposed as an alternative to the hCG trigger by virtue of inducing a rise in endogenous LH and FSH due to its initial flare effect. The specific mode of action of the antagonist by competitive blockade of the pituitary receptors and a shorter half-life means that the pituitary remains responsive to the GnRH agonist thus enabling its use for triggering ovulation. The mechanism of action of the GnRH agonist in causing down-regulation and desensitisation of the pituitary receptors precludes the use of agonist trigger in those cycles. Use of the GnRH agonist trigger significantly lowers the incidence of OHSS compared to the hCG trigger [12].

5 Individualised COS

The main objective of individualisation of treatment in IVF is to offer every single woman the best treatment tailored to her own unique characteristics, thus maximising the chances of pregnancy and eliminating the iatrogenic and avoidable risks resulting from ovarian stimulation. It is therefore important to categorise women based on their predicted response in order to individualise COS regimens. Women can be identified as having a poor response, normal response or a hyper-response based on individual characteristics and ovarian reserve tests (ORTs). Among the various ORTs including basal FSH, basal oestradiol, inhibin B, antral follicle count (AFC) and anti-mullerian hormone (AMH), AFC and AMH have the highest accuracy for the prediction of either a poor or an excessive response following ovarian stimulation [13]. Results from individual patient data (IPD) meta-analyses of patient characteristics and ORTs demonstrated age as being the most important among patient characteristics for the prediction of poor or excessive response and AFC or AMH as having the highest predictive accuracy among ORTs [14,15]. The cut-off levels of AFC and AMH for prediction of poor response is an AFC of < 5 to <7 and AMH of <0.5 ng/ml to <1.1 ng/ml [16]. The cut-off levels for AFC and AMH for the prediction of hyper-response is an AFC of >14 to >16 [17,18] and AMH of 3.5 ng/ml to 3.9 ng/ml [19,20]. Individualisation involves tailoring the different stages of COS to suit each woman.

With regards to the optimal gonadotrophin dosage an RCT, comparing a gonadotrophin dose of 225 IU daily versus 150 IU daily in women aged 23–41 years undergoing IVF demonstrated the number of oocytes to be significantly higher with 225 IU daily compared to 150 IU daily [21]. This study excluded women with basal FSH > 10 IU/L, polycystic ovary syndrome (PCOS), previous poor response and previous OHSS. Another RCT comparing a daily gonadotrophin dose of 225 IU versus 300 IU daily among women predicted as normal responders based on a total AFC of 8–21 showed no significant difference in the number of oocytes retrieved between the two doses [22]. This evidence would therefore suggest that the ideal gonadotrophin dose for women predicted as normal responders is 225 IU daily. An RCT comparing gonadotrophin doses of 300 IU versus 375 IU versus 450 IU daily among women predicted as poor responders based on a total AFC of <12 showed no significant difference in the number of oocytes retrieved nor live birth rates between the three arms suggesting an unlikely benefit with gonadotrophin doses > 300 IU daily [23]. Women with PCOS and those predicted to have a hyper-response should be stimulated with a lower gonadotrophin dose of ≤ 150 IU daily to avoid excessive stimulation. Excessive response (>20 oocytes) is associated with a decrease in live birth rate in a fresh embryo transfer IVF cycle [6] in addition to a higher incidence of OHSS.

Updated evidence has demonstrated comparable pregnancy rates with the GnRH antagonist and GnRH agonist regimens in unselected women in addition to a lower risk of OHSS with the antagonist regimen [24]. Between the long and the short GnRH agonist regimens, the long agonist regimen has better outcomes in terms of the number of oocytes retrieved and pregnancy rates compared to the short agonist regimen [8]. The GnRH antagonist and long GnRH agonist regimens are therefore suitable options for pituitary suppression. An RCT comparing the long GnRH agonist regimen versus short GnRH agonist regimen versus GnRH antagonist regimen in women with a previous poor ovarian response demonstrated the long agonist and antagonist regimens to be suitable for these women with regards to the number of oocytes retrieved [25]. A recent meta-analysis of studies comparing GnRH antagonist versus GnRH agonist regimes in women with PCOS showed comparable pregnancy rates between the two groups and a significantly lower incidence in severe OHSS in the GnRH antagonist group [26]. An added advantage of the use of GnRH antagonist based protocols is the use of GnRH agonist as a substitute for hCG in triggering of final oocyte maturation, potentially eliminating the risk of OHSS [12]. The Cochrane review comparing the GnRH agonist versus the hCG trigger however, demonstrated significantly reduced live birth rates in fresh autologous cycles with the use of the GnRH agonist trigger although there was no reduction in live birth rates in oocyte donor/recipient cycles [12]. Following initial use of the GnRH agonist trigger the need to modify the standard luteal support to obtain reliable reproductive outcomes was soon recognised. Study groups have since endeavoured to fine tune the luteal phase support in IVF cycles using the GnRH agonist trigger to optimise clinical outcomes. Recent suggestions and developments in overcoming luteal insufficiency which occurs in such cycles are use of (i) a ‘dual trigger’ [27], (ii) low dose hCG supplementation [28], (iii) intensive luteal oestradiol and progesterone supplementation [29], (iv).rec-LH supplementation [30] and (v) luteal GnRH agonist administration [31]. Alternatively, all the available embryos are frozen for transfer in a subsequent cycle.