gonadotropins. The increased number of antral follicles is also reflected by elevation of anti‐Müllerian hormone (AMH) levels in women with PCOS compared with those with normal ovaries.
Management options
All patients undergoing ovarian stimulation, whether to correct anovulation or for assisted reproduction techniques (ART), should of course have a pretreatment ultrasound scan and if polycystic ovaries are detected the dose of gonadotropin lowered (to a starting dose of no more than 50–150 IU depending upon age and other factors).
OHSS is usually associated with the presence of a large number of small to moderate sized follicles (< 14 mm diameter) rather than larger, more mature follicles. When serum estradiol concentrations are 10,000–15,000 pmol/L with 20–30 follicles, the patient is at risk of OHSS. A decision then has to be taken as to whether or not to proceed with oocyte retrieval and then potentially electively cryopreserve any embryos created or decide if it is safe to perform a fresh embryo transfer, with the inherent increased risk of late‐onset and prolonged OHSS. If the serum estradiol concentration becomes greater than 15,000 pmol/L (5,000 pg/ml) with more than 30–40 follicles, the treatment cycle should be canceled.
The long GnRH agonist protocol
Historically a “long” protocol with pituitary desensitization using a GnRH agonist was used although this has now largely been replaced by the GnRH antagonist protocols in women with PCOS. If a long protocol of GnRH agonist treatment is followed by treatment with one of the “pure” or recombinant FSH preparations, one must also be aware that the lack of LH changes the usual relationship of follicle number to circulating estradiol levels. In this situation measurement of serum estradiol concentrations underestimates follicle development. It is therefore essential that endocrine monitoring is supported by high quality ultrasound, otherwise low circulating estradiol concentrations may encourage further and inappropriate gonadotropic stimulation despite adequate follicular development. Indeed, these days, most use ultrasound as the keyway of monitoring follicular growth. Meta‐analyses of the different gonadotropin preparations have indicated no significant differences in the risk of developing OHSS [6].
For the patient with overstimulated ovaries who is approaching the time of human chorionic gonadotropin (hCG) administration several strategies to make treatment safer may be considered. The first is to administer a low dose of hCG to initiate oocyte maturation (i.e. not more than a single injection of 5,000 IU, rather than the dose of 10,000 IU which many clinics use in routine practice) and, in patients receiving GnRH agonist treatment and who therefore require luteal support, to give progesterone rather than hCG (which is virtually obsolete now as a form of luteal support).
Insulin resistance and compensatory hyperinsulinemia contribute to the pathogenesis of PCOS. A number of studies have investigated the effects of using the insulin sensitizing agents, mainly metformin, on women with PCOS undergoing IVF treatment. The use of metformin as an adjunct for IVF is associated with no significant difference in live birth rates (OR 1.39, 95% CI 0.81 to 2.40, five RCTs, 551 women, low‐quality evidence), but a significant lowering of the incidence of OHSS (OR 0.29; 95% CI 0.18 to 0.49, eight RCTs, 798 women, moderate‐quality evidence), but with a higher incidence of gastrointestinal side effects [7]. However, as stated, the GnRH agonist regimens have been superseded by the use of GnRH antagonist regimens for women with PCOS undergoing IVF.
The GnRH antagonist protocol
The GnRH antagonist cycle is now widely recognized as superior to the agonist cycle in reducing the risk of OHSS in women with polycystic ovaries [8]. With equivalent pregnancy rates, the adoption of this strategy appears unquestionable. It is also prudent to consider the option of using a GnRH agonist trigger instead of hCG to further reduce OHSS rates in those at risk.
In a meta‐analysis of studies looking at agonist versus antagonist protocols for pituitary suppression during IVF, the GnRH antagonist protocols were associated with a reduced risk of OHSS (OR 0.61, 95% CI 0.51 to 0.72; 36 RCTs; n = 7944; moderate quality evidence) whilst not compromising the live birth rate (OR 1.02, 95% CI 0.85 to 1.23; 12 RCTs; n = 2303; moderate quality evidence) [8].
The GnRH antagonist cycle also allows the use of GnRH agonist rather than hCG as the pre‐ovulatory trigger further reducing the risk of OHSS (OR 0.15, 95% CI 0.05 to 0.47; eight RCTs, 989 women; moderate‐quality evidence) [9]. However, the use of a GnRH agonist in this way may be associated a lower live birth rate due to a deficient luteal phase (OR 0.47, 95% CI 0.31 to 0.70; five RCTs, 532 women, moderate‐quality evidence) [9]. Two approaches have been suggested to overcome the luteal phase deficiency – supplementation with exogenous estrogen and progesterone in the luteal phase or rescuing the corpus luteum using low dose hCG given either alongside the GnRH agonist trigger or on the day of the oocyte retrieval. Use of modified luteal phase support appears to lead to similar live birth rates with a GnRH agonist trigger as with hCG (OR 0.84, 95% CI 0.62 to 1.14; five RCTs; n = 857) [10].
An alternative approach has been elective cryopreservation of embryos following oocyte retrieval, followed by transfer in a frozen embryo transfer cycle – the “segmentation” approach [11]. A recent multicenter trial including 1508 women found that the segmentation approach led to a higher live birth rate (49.3% vs. 42.0%; RR 1.17, 95% CI 1.05 to 1.31) with a lower risk of OHSS (1.3% vs. 7.1%, RR 0.19, 95% CI, 0.10 to 0.37), but a higher risk of pre‐eclampsia (4.4% vs. 1.4%, RR 3.12, 95% CI, 1.26 to 7.73) [12].
New concepts include the possibility of using Kisspepetin as the pre‐ovulatory trigger [13], which has generated much interest. It is also possible to use dopamine agonists, such as cabergoline, to inhibit phosphorylation of the receptor for VEGF, which has also been shown to reduce the incidence of OHSS [14]. There is the possibility to consider in vitro maturation (IVM) of oocytes collected from unstimulated or minimally stimulated ovaries, although this requires particular expertise in the clinic and laboratory, and despite some clinics demonstrating success with this approach it has not gained widespread popularity [15].
Key points
Challenge: Polycystic ovaries and PCOS in IVF treatment.
Background:
Polycystic ovarian morphology is present in 20–30% of IVF patients.
Not all women with polycystic ovaries have polycystic ovary syndrome (PCOS).
The presence of polycystic ovaries is associated with sensitive response to stimulation and an increased risk of OHSS.
Assessment:
Baseline pelvic USS provides morphological appearance of polycystic ovaries.
Baseline endocrine profile enables appropriate regimen choice.
Assessment of glucose tolerance is important if overweight.
Counsel for increased obstetric risk (gestational diabetes, preeclampsia and fetal morbidity) if overweight.
Management options:
Treatment plan aimed to minimize risk of OHSS which is a life‐threatening condition
Use low dose stimulation in a short GnRH‐antagonist protocol.
Metformin therapy may reduce the risk of OHSS in a long GnRH‐agonist protocol
Use progestogens and not hCG for luteal support
Consider a GnRH‐agonist trigger in GnRH‐antagonist protocols if there is a significant risk of OHSS
Consider segmentation, that is elective cryopreservation of all embryos for use in subsequent frozen embryo replacement cycles
Answers