In Vitro Fertilization for Polycystic Ovarian Syndrome

JESSICA R. ZOLTON, DO,* and SAIOA TORREALDAY, MD† *Program in Reproductive Endocrinology and Gynecology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health; and †Walter Reed National Military Medical Center, Bethesda, Maryland

Abstract: In vitro fertilization is indicated for infertile treatment for infertility. Guidelines indi- women with polycystic ovarian syndrome (PCOS) after cate that IVF should be offered after failed unsuccessful treatment with ovulation induction agents or in women deemed high-risk of multiple gestations ovulation induction with oral agents or 1 who are ideal candidates for single embryo transfers. gonadotropin treatment. However, due to PCOS patients are at increased risk of ovarian hyper- the risk of twins and higher order multi- stimulation syndrome; therefore, attention should be ples, which is more commonly seen when made in the choice of in vitro fertilization treatment gonadotropin medications are utilized, protocol, dose of gonadotropin utilized, and regimen to achieve final oocyte maturation. Adopting these strat- IVF may be considered after failed ovula- egies in addition to close monitoring may significantly tion induction with clomiphene citrate or reduce the ovarian hyperstimulation syndrome risk. letrozole.2 In addition, PCOS patients are Future developments may improve pregnancy out- ideal candidates for consideration of elec- comes and decrease complications in PCOS women tive single embryo transfer to mitigate the undergoing fertility treatment. Key words: infertility, in vitro fertilization, polycystic risk of multiple pregnancies while under- ovarian syndrome, ovarian hyperstimulation syn- going IVF. drome, in vitro maturation PCOS patients undergoing IVF, without additional infertility confounders, are gen- In vitro fertilization (IVF) for the treat- erally good prognosis patients to achieve a ment of anovulatory polycystic ovarian live birth. During IVF stimulation, the large syndrome (PCOS) is considered third-line number of antral follicles seen within this population commonly results in a high Correspondence: Jessica R. Zolton, DO, Program in number of oocytes retrieved, the develop- Reproductive Endocrinology and Gynecology, Eunice ment of supernumerary embryos, and Kennedy Shriver National Institute of Child Health and consequently, excess embryos for cryopre- Human Development, National Institutes of Health; 10 Center Drive, Building 10, Room 8N248B, MSC 1840, servation. IVF cycles undertaken in the Bethesda, MD. E-mail: [email protected] PCOS population are characterized by The authors declare that they have nothing to disclose. slightly longer stimulation (1.2 d longer),

CLINICAL OBSTETRICS AND GYNECOLOGY / VOLUME 64 / NUMBER 1 / MARCH 2021

www.clinicalobgyn.com | 39 Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved. 40 Zolton and Torrealday higher number of developing of follicles, intractable nausea and vomiting, hypovo- higher estradiol levels, and retrieval of a lemia, and hypercoagulability. Electrolyte higher number of cumulus-oocyte com- disturbances resulting in hyperkalemia, hy- plexes (COC) [2.9 more COC; 95% con- ponatremia, and abnormal hepatic function fidence interval (CI) = 2.2, 3.6)] than in are also commonly present.7 Although women without PCOS.3 PCOS patients severe cases of OHSS are rare, maternal may require relatively high doses of gona- deaths have been reported.8 Therefore, it is dotropins at the start of stimulation due to important that infertility specialists identify their underlying hyperandrogenism, which women who are at risk of OHSS before the contributes to altered steroidogenesis and initiation of treatment and adopt strategies impaired follicular development.4 Increas- to prevent OHSS. ing gonadotropin dosage should be cautioned and closely monitored as these patients are also prone to overrespond to IVF Protocols stimulation, increasing the risk of develop- Treatment strategies employed during IVF ing ovarian hyperstimulation syndrome aim to achieve “controlled” growth and (OHSS). OHSS is an iatrogenic complica- development of a cohort of ovarian fol- tion that results from the stimulation of licles that is available and responsive to multiple corpus lutei following exogenous exogenous gonadotropin stimulus, while human chorionic gonadotropin (hCG) ad- concomitantly suppressing the hypothala- ministration. OHSS occurs in up to 3% to mic-pituitary-ovarian axis. In the absence 6%5 of all IVF cycles and results due to of a suppressive strategy, escalating estra- excess production of vascular endothelial diol levels that arise with the development growth factor (VEGF) and cytokines which of ovarian follicles can initiate an endog- cause arteriolar vasodilation and increased enous luteinizing hormone (LH) surge capillary permeability.6 Womenwithan with subsequent ovulation. During the antral follicle count > 24 and an anti- early days of IVF, suppression of the Mullerian hormone ≥ 3.5 ng/mL have been hypothalamic-pituitary-ovarian axis was identified to be at increased risk of OHSS; extensively done utilizing gonadotropin- bothparametersarecommoninPCOS releasing hormone (GnRH) agonist in patients. During the IVF cycle, high estra- efforts to prevent the premature luteiniza- diol values > 3500 pg and ≥ 24 oocytes tion during controlled ovarian hyperstimu- retrieved are also associated with an in- lation. GnRH agonists bind to the creased risk of OHSS.7 Early-onset OHSS, pituitary GnRH receptors and stimulate which occurs within 10 days of oocyte the release of the endogenous gonadotro- retrieval, is due to exposure to hCG when pins, follicle-stimulating hormone (FSH) used as a trigger for oocyte maturation. and LH. This sudden increase in FSH and Late-onset OHSS ( ≥ 10 d from oocyte LH secretion following receptor binding is retrieval) is caused by the secretion of oftenreferredtoasthe“flare-effect.” After hCG by the developing embryo after em- continuous stimulation with a GnRH ago- bryo transfer. OHSS symptoms range from nist, desensitization and downregulation mild, with patients reporting abdominal of GnRH receptors occurs resulting in distention, mild dyspnea, and nausea and FSH and LH suppression. vomiting, to moderate, with laboratory There are 2 primary GnRH agonist evidence of hemoconcentration and ultra- protocols, the long luteal and the short sound evidence of ascites, to severe cases GnRH agonist protocols, that are com- requiring hospitalization. In severe cases, monly used in IVF cycles. In the long patients may require fluid resuscitation and luteal GnRH agonist protocol, the GnRH anticoagulation to treat oliguria/anuria, agonist is typically initiated in the luteal www.clinicalobgyn.com Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved. IVF for PCOS 41 phase of the preceding cycle and exoge- protocol). GnRH antagonist protocols nous gonadotropins are not started until permit the use of either hCG or GnRH after downregulation has occurred. In agonist trigger for final oocyte matura- contrast, in the short GnRH agonist prot- tion, whereas, in a GnRH agonist cycle ocol, GnRH agonist is administered at hCG trigger is the only option due to the the beginning of the menstruation cycle depletion of pituitary LH stores. The concomitantly with the start of exogenous disadvantage of exogenous hCG is that gonadotropins. The short GnRH agonist it has a longer half-life then LH (24 h vs. protocol is often reserved for poor res- 60 min) and the duration of the hCG ponders due to the advantage of having surge is also significantly longer than the endogenous gonadotropin “flare- LH (24 to 36 h vs. 6 d).12 As a result, effect” and to prevent the excessive pitui- hCG stimulates the corpus luteum for a tary suppression often seen in long GnRH prolonged period of time which promotes agonist cycles. Since GnRH agonist VEGF production and thus increases the cycles exhaust the pituitary of its gonado- risk of OHSS.6 tropin stores, the use of hCG as a surro- For patients that receive an hCG trigger gate LH surge is a necessity. hCG, which and are deemed high risk for OHSS, shares structural similarities to LH and recommendation to delay transfer and therefore can bind to the LH receptor, subsequently cryopreserve all embryos is induces luteinization of granulosa cells, encouraged. Therefore, adopting a GnRH achieves final oocyte maturation, and antagonist protocol with GnRH agonist allows for resumption of meiosis. trigger is more beneficial for PCOS pa- In PCOS patients who are generally tients. In a randomized controlled trial “good responders” to gonadotropins, the (RCT) of 220 patients with PCOS, GnRH use of the short GnRH agonist protocol antagonist protocols were associated with with subsequent “flare-effect” can result in a decreased risk in OHSS in comparison to an increased risk of OHSS. One study the GnRH agonist protocol (40% vs. 60%, reported that GnRH agonist protocols with difference −20%, 95% CI = −7.1 to hCG trigger for oocyte maturation were −31.9%, P < 0.01).13 hCG trigger was used associated with a 20% incidence of OHSS in in both groups which explains the high PCOS patients compared with a 7% inci- incidence of OHSS overall.13 In a meta- dence in non-PCOS groups (P = 0.53).9 analysis including 9 RCTs and 1294 With the introduction of GnRH ago- women, GnRH antagonist protocols were nists for oocyte maturation in 199010 and associated with a decrease risk of OHSS in GnRH antagonist protocols in 1995,11 comparison to GnRH agonists protocols clinical management of IVF for PCOS [odds ratio (OR) = 0.53; 95% CI = 0.3, patients have evolved. PCOS patients are 0.95].14 Oocyte yield and ongoing preg- ideal candidates for an antagonist proto- nancy and live birth rates were not signifi- col because there is no initial release of cantly different among the 2 protocols.14 gonadotropins at the start of treatment, as The use of a GnRH agonist for oocyte evident with the short GnRH agonist maturation almost entirely eliminates the protocol. In a GnRH antagonist proto- risk of OHSS, with only a small number of col, the gonadotropin therapy is started publishedcasesintheliterature.15 following menstruation. The GnRH an- For all patients who receive a GnRH tagonist is initiated either around day 5 of agonist for oocyte maturation, there is gonadotropin stimulation (fixed protocol) concern for insufficient luteal phase sup- or when a lead follicle of 12 to 13 mm is port. A RCT found a significant reduction visible on ultrasound and estradiol levels in clinical pregnancy rate [36% vs. 6% range from 200 to 400 pg/mL (flexible (P = 0.002)] and elevated risk of pregnancy

www.clinicalobgyn.com Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved. 42 Zolton and Torrealday loss [4% vs. 79% (P = 0.005)] in patients RR = 0.67; 95% CI = 0.54, 0.80; P ≤ 0.001) receiving a GnRH agonist trigger versus and OHSS (1.3% vs. 7.1%; RR = 0.19; 95% hCG trigger.16 The surge following GnRH CI = 0.1, 0.37; P < 0.001) compared with agonist is shortened, as described previ- womenwithPCOSwhohadafresh ously, and serum levels of estradiol and transfer.19 Overall, IVF practice has shifted progesterone are lower after GnRH ago- to an increase in FET cycles in all patient nist trigger when compared with a hCG groups and it is anticipated that this trend trigger. Therefore, additional steps are will continue as studies consistently demon- necessary to ensure that appropriate luteal strate an increase in live birth in freeze-all hormonal support is attained to ad- protocols. equately support a pregnancy. Although an optimal protocol has not been estab- lished, methods to compensate for the Miscarriage Risk After IVF hormonal deficit include a microdose of Studies suggest that PCOS patients may hCG at oocyte retrieval and the addition be at an increased risk of miscarriage due of progesterone and estrogen supplemen- to insulin resistance, hyperandrogenism, tation.17 For women at risk of OHSS, a and increased body mass index (BMI).20 microdose of hCG on day of oocyte These conditions usually coexist and in- retrieval may not be appropriate and dependently are associated with negative should be cautioned. Supplementation pregnancy outcomes. Studies have sug- with progesterone and estradiol in the gested that pregnancy loss in PCOS pa- absence of a hCG trigger has been asso- tients is due to impaired endometrial ciated with a higher risk of pregnancy receptivity21 while others have suggested loss.16 An alternative approach is to adopt that embryonal aneuploidy is increased.22 a freeze-all strategy in patients who under- Alterations in endometrial receptivity go an antagonist cycle utilizing a GnRH have been suggested due to impaired agonist trigger. insulin and glucose signaling and over- As PCOS patients represent a high res- expression of androgen receptors in ponder group, serum levels of estradiol and the endometrium23 resulting in dysregu- progesterone are usually markedly elevated lation of the cell signaling necessary in comparison to other patients undergoing for implantation.20 Studies have also re- controlled ovarian hyperstimulation. There ported that embryonal aneuploidy is in- is evidence that elevated progesterone levels creased in PCOS patients compared with on the day of hCG trigger are negatively a non-PCOS patient population (61.3% in associated with live birth rates in fresh PCOS patients vs. 47.8% in the control cycles. However, employing a freeze-all group).22 PCOS diagnosis was identified strategy restores live-birth rates comparable to be an independent risk factor for to cycles without progesterone elevation, aneuploidy after adjustments for BMI, suggesting that the endometrial environ- maternal age, embryo quality, and other ment is negatively affected by the supra- infertility diagnoses. Although the mech- physiological hormonal levels.18 Further anism for an increased aneuploidy rate in evidence supports the use of frozen embryo PCOS patients has not been elucidated, transfer (FET) cycles in all patients with studies suggest that altered steroidogene- PCOS regardless of clinical characteristics. sis and impaired oocyte metabolism in Employing a freeze-all strategy in PCOS this patient population may lead to DNA patients improved live-birth rate [49% vs. instability.24 More research in this area is 42%; relative risk (RR) = 1.17; 95% CI = needed to determine if PCOS patients 1.05, 1.31; P = 0.004], and lowered the have a higher inherent risk of creating risk of pregnancy loss (22% vs. 32.7%; aneuploid embryos. www.clinicalobgyn.com Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved. IVF for PCOS 43

Adjunctive Treatment 40 women).32 Furthermore, there was no Metformin, an oral biguanide, is tradi- difference in live birth rate, miscarriage tionally prescribed to PCOS patients due rate, number of oocytes collection, or cycle to the high incidence of insulin resistance cancellation between the metformin and in this population. Metformin decreases placebo groups. No set protocols for start- systemic insulin levels and acts at the level ing metformin exist; doses ranged from of the ovary to decrease androgen pro- metformin 500 mg 2 to 3 times daily to duction through reduced CYP17 activity 850 mg twice daily starting 16 weeks be- in PCOS women.25 A double-blind RCT fore IVF to the day of trigger. found that while metformin did not increase pregnancy rates among nonobese PCOS women undergoing IVF, it did Dopamine Agonists result in an overall increase in clinical Dopamine receptor (D2) agonists are pregnancy and live birth rates as a result routinely prescribed for patients with of an increase in spontaneous conceptions hyperprolactinemia undergoing infertility after randomization.26 Metformin treat- treatment but more recently has also been ment may be most beneficial for women utilized to prevent OHSS. Cabergoline, a with increased BMI. A RCT, which in- dopamine agonist, has been found to cluded PCOS women with a mean BMI of decrease vascular permeability in women 27 kg/m2 who either received metformin at risk for OHSS. It has also been dem- or placebo, demonstrated that the pa- onstrated in human granulosa cell culture tients who received metformin had a that cabergoline decreases the production higher pregnancy and live birth rates of VEGF.33 Treatment is usually initiated compared with the those who were as- on the day of hCG trigger in patients who signed to placebo.27 are identified to be at increased risk for Metformin has also been proposed as OHSS. A RCT including 182 women an agent to decrease the incidence of identified as being high risk for OHSS OHSS in PCOS patients.28 In vitro, lower were randomized to receive either placebo insulin levels reduces the production of or 1 of 3 assigned doses of the dopamine VEGF, a key player in the pathogenesis of agonist starting on the day of hCG OHSS.29 Thus, metformin has also been trigger. The study reported a reduced risk used an adjunct to IVF cycles in efforts to of moderate/severe OHSS (OR = 0.28; decrease the risk of OHSS. A RCT includ- 95% CI = 0.09, 0.81; P = 0.019) with dop- ing 120 patients found that metformin amine agonist treatment.34 The incidence significantly reduced the incidence of of moderate/severe OHSS in the study OHSS and cycle cancellation in GnRH was significant higher in the placebo arm agonist cycles.30 Peak estradiol levels were [23% (12/53)] than in the dopamine ago- lower in women treated with metformin, nist arm [11% (14/129)].34 Higher doses of likely a result of decreased insulin receptor the medication were most efficacious in signaling and, subsequently, lowering reducing OHSS, however, the side effect androgen precursors available for aromat- profile consisting of nausea, vomiting, ization.31 A Cochrane meta-analysis pub- and dizziness was also more pronounced. lished in 2014 including 9 RCTS and 816 A systematic review including 7 studies women found that PCOS patients who and 858 women also reported a decrease took metformin had a lower incidence of in OHSS in the dopamine agonist coh- OHSS in GnRH agonist protocols (OR = ort compared with the no treatment 0.29; 95% CI = 0.16, 0.51, 758 women) but group (RR = 0.38; 95% CI = 0.29, 0.51; found no difference for GnRH antagonist P < 0.00001) without impacting clinical protocols (OR = 0.3; 95% CI = 0.03, 3.15, pregnancy rates.35 Of note, the dose of

www.clinicalobgyn.com Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved. 44 Zolton and Torrealday cabergoline and timing of treatment var- more research, IVM will be adopted as an ied across the studies.35 Although there is additional treatment for infertility. The evidence in support of dopamine agonists process involves retrieval of immature oo- for the prevention of OHSS, few studies cytes (arrested in prophase I) and IVM to have been performed exclusively in PCOS the metaphase II stage. Since IVM proto- patients. One study suggested that women cols either involve mild ovarian stimulation with PCOS treated with cabergoline may or avoid stimulation of the ovary entirely, have a lower risk of cycle cancellation due the process is less expensive and requires to OHSS, although the overall low num- less monitoring.39 hCG trigger during IVM ber of OHSS cancellations in the study retrieval is usually performed when the lead limits the interpretation.36 Overall, stud- follicle is only 14 mm in size, which is ies have demonstrated consistent evidence earlier than conventional IVF protocols that dopamine agonists may be a modal- that administer the trigger when the lead ity to decrease the risk of OHSS. follicle reaches at least 18 mm. There is no specific IVM protocol, but COCs are typicallyculturedinmediawithFSH, Coasting hCG, and growth hormone for 20 hours. During ovarian hyperstimulation, ultra- Intracytoplasmic sperm injection is usually sound surveillance of ovarian response in performedtofacilitatefertilization.Aret- conjunction with serum estradiol values rospective review found that live birth rates assist in the identification of patients who in fresh cycles after IVM were suboptimal are at risk of developing complications. to IVF (18.8% vs. 31%, P = 0.021) Early practice was to simply cease gona- although live birth rates in FET cycles were dotropin stimulation and “coast” until similar.40 The total number of suitable estradiol levels plateaued, or the rate blastocysts were lower in the IVM group of follicular growth slowed. A 2017 compared with the IVF cohort (2.5 ± 2.1 Cochrane meta-analysis reported that vs. 3.9 ± 3.4, P ≤ 0.001).40 There was no coasting was associated with a reduction difference in biochemical pregnancy, clin- in OHSS (OR = 0.11; 95% CI = 0.05, ical pregnancy, or miscarriage rates be- 0.24) and a lower number of oocytes tween the 2 treatment groups in fresh or retrieved.37 There were no differences in frozen cycles. Furthermore, cumulative live birth or miscarriage between the 2 pregnancy rate was lower and fewer blas- groups. The evidence was considered low tocysts were available for cryopreservation. quality due to incomplete data and small A retrospective case-cohort study reported sample size. Other studies have similarly a significant decrease in live birth rate (16.5 reported a reduction in oocyte yield and vs. 44.3, P < 0.0001) between IVM and decrease implantation rate when coasting IVF/intracytoplasmic sperm injection.41 was applied.38 As the evidence for coast- Although miscarriage rates were not differ- ing is weak, other interventions should ent, IVM pregnancies had a lower implan- be pursued to prevent the incidence of tation rate (12.9% vs. 25.6%, P < 0.0001) OHSS. potentially due to impaired endometrial receptivity.41 This may be due to embryo/ endometrial asynchrony, as hCG trigger In Vitro Maturation (IVM) for oocyte maturation is done at a consid- IVM may be considered a future option for erably earlier time in the menstrual cycle.40 PCOS women who want to avoid the risk The lower implantation rate may also be a of OHSS entirely. Although IVM is cur- failure of oocyte cytoplasmic maturation, a rently considered experimental, it is ex- process which modern technology still pected that with increased experience and cannot assess. www.clinicalobgyn.com Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved. IVF for PCOS 45

The safety of IVM has also been ques- and safety of patients, with complete tioned. Increased chromosomal abnormal- elimination of OHSS, is the goal of all ities have been noted in oocytes that health care providers and is well within required prolonged culture time. However, reach. IVM embryos created from oocytes that were matured for <48 hours showed chromosomal abnormalities that were comparable to IVF References 42 embryos. Animal studies have also raised a 1. Teede HJ, Misso ML, Costello MF, et al. Recom- concern for the potential of IVM to alter the mendations from the international evidence-based genome permanently via epigenetic changes, guideline for the assessment and management of polycystic ovary syndrome. Fertil Steril. 2018;110: thus increasing the incidence of imprinting – 43 364 379. disorders. However, a study reporting on 2. Zolton JR, Lindner PG, Terry N, et al. Gonado- long-term outcomes (mean follow-up of 7.5 tropins versus oral ovarian stimulation agents for y) of 184 children born from IVM found no unexplained infertility: a systematic review and difference in birth weight, congenital anoma- meta-analysis. Fertil Steril. 2020;113:417–425. lies, or growth development compared with 3. Heijnen EM, Eijkemans MJ, Hughes EG, et al. A 44 meta-analysis of outcomes of conventional IVF in 366 children conceived via IVF. As more women with polycystic ovary syndrome. Hum studies are published reporting the long-term Reprod Update. 2006;12:13–21. health of infants born from IVM, IVM will 4. De Vos M, Pareyn S, Drakopoulos P, et al. likely be adopted as a routine assisted repro- Cumulative live birth rates after IVF in patients ductive technique available to patients. Cur- with polycystic ovaries: phenotype matters. Reprod Biomed Online. 2018;37:163–171. rently, however, while IVM may be an 5. Delvigne A. Symposium: update on prediction appealing strategy for younger women with and management of OHSS. Epidemiology of PCOS because it eliminates the risk for OHSS. Reprod Biomed Online. 2009;19:8–13. developing OHSS, enthusiasm for this option 6. Wang TH, Horng SG, Chang CL, et al. Human is currently somewhat dampened by the lesser chorionic gonadotropin-induced ovarian hyperstimulation syndrome is associated with up-regulation of success rates. vascular endothelial growth factor. J Clin Endocrinol Metab. 2002;87:3300–3308. 7. Practice Committee of the American Society for Conclusions Reproductive Medicine. Prevention and treatment The goal of IVF is to achieve a singleton of moderate and severe ovarian hyperstimulation syndrome: a guideline. Fertil Steril. 2016;106: pregnancy while minimizing the risks of 1634–1647. OHSS and multiple pregnancies. Manage- 8. Kupka MS, D’Hooghe T, Ferraretti AP, et al. ment of PCOS patients requires a detailed Assisted reproductive technology in Europe, clinical assessment and risk stratification 2011: results generated from European registers and the choice of therapy needs to be by ESHRE. Hum Reprod. 2016;31:233–248. 9. Nardo LG, Gelbaya TA, Wilkinson H, et al. individualized for each patient. Serious Circulating basal anti-Mullerian hormone levels attention should be made in the choice of as predictor of ovarian response in women under- IVF treatment protocol, dose of gonado- going ovarian stimulation for in vitro fertilization. tropin utilized, and regimen selected to Fertil Steril. 2009;92:1586–1593. achieve final oocyte maturation. Close 10. Gonen Y, Balakier H, Powell W, et al. Use of gonadotropin-releasing hormone agonist to trig- monitoring of ovarian and hormonal ger follicular maturation for in vitro fertilization. response as well as preempting the risk J Clin Endocrinol Metab. 1990;71:918–922. for OHSS development is critical. As 11. Olivennes F, Fanchin R, Bouchard P, et al. Scheduled assisted reproduction technologies contin- administration of a gonadotrophin-releasing hormone ue to improve, further development on antagonist (Cetrorelix) on day 8 of in-vitro fertilization cycles: a pilot study. Hum Reprod. 1995;10: ways to mitigate OHSS and improve 1382–1386. outcomes in PCOS patients will likely 12. Humaidan P, Kol S, Papanikolaou EG. GnRH continue to occur. The long-term health agonist for triggering of final oocyte maturation:

time for a change of practice? Hum Reprod 26. Kjotrod SB, Carlsen SM, Rasmussen PE, et al. Update. 2011;17:510–524. Use of metformin before and during assisted 13. Lainas TG, Sfontouris IA, Zorzovilis IZ, et al. reproductive technology in non-obese young in- Flexible GnRH antagonist protocol versus fertile women with polycystic ovary syndrome: a GnRH agonist long protocol in patients with prospective, randomized, double-blind, multi- polycystic ovary syndrome treated for IVF: a centre study. Hum Reprod. 2011;26:2045–2053. prospective randomised controlled trial (RCT). 27. Tang T, Glanville J, Orsi N, et al. The use Hum Reprod. 2010;25:683–689. of metformin for women with PCOS under- 14. Lambalk CB, Banga FR, Huirne JA, et al. GnRH going IVF treatment. Hum Reprod. 2006;21: antagonist versus long agonist protocols in IVF: a 1416–1425. systematic review and meta-analysis accounting 28. Costello MF, Misso ML, Balen A, et al. Evidence for patient type. Hum Reprod Update. 2017;23: summaries and recommendations from the inter- 560–579. national evidence-based guideline for the assess- 15. Devroey P, Polyzos NP, Blockeel C. An OHSS- ment and management of polycystic ovary Free Clinic by segmentation of IVF treatment. syndrome: assessment and treatment of infertility. Hum Reprod. 2011;26:2593–2597. Hum Reprod Open. 2019;2019:hoy021. 16. Humaidan P, Bredkjaer HE, Bungum L, et al. 29. Stanek MB, Borman SM, Molskness TA, et al. GnRH agonist (buserelin) or hCG for ovulation Insulin and insulin-like growth factor stimulation induction in GnRH antagonist IVF/ICSI cycles: a of vascular endothelial growth factor production prospective randomized study. Hum Reprod.2005; by luteinized granulosa cells: comparison between 20:1213–1220. polycystic ovarian syndrome (PCOS) and non- 17. Benadiva C, Engmann L. Luteal phase support PCOS women. J Clin Endocrinol Metab. 2007;92: after gonadotropin-releasing hormone agonist 2726–2733. triggering: does it still matter? Fertil Steril. 2018; 30. Palomba S, Falbo A, Carrillo L, et al. Metformin 109:763–767. reduces risk of ovarian hyperstimulation syn- 18. Healy MW, Patounakis G, Connell MT, et al. drome in patients with polycystic ovary syndrome Does a frozen embryo transfer ameliorate the during gonadotropin-stimulated in vitro fertiliza- effect of elevated progesterone seen in fresh trans- tion cycles: a randomized, controlled trial. Fertil fer cycles? Fertil Steril. 2016;105:93–99. Steril. 2011;96:1384–1390. 19. Chen ZJ, Shi Y, Sun Y, et al. Fresh versus frozen 31. Mahamed RR, Maganhin CC, Sasso GRS, et al. embryos for infertility in the polycystic ovary Metformin improves ovarian follicle dynamics by syndrome. N Engl J Med. 2016;375:523–533. reducing theca cell proliferation and CYP-17 20. Schulte MM, Tsai JH, Moley KH. Obesity and expression in an androgenized rat model. J PCOS: the effect of metabolic derangements on Ovarian Res. 2018;11:18. endometrial receptivity at the time of implanta- 32. Tso LO, Costello MF, Albuquerque LE, et al. tion. Reprod Sci. 2015;22:6–14. Metformin treatment before and during IVF or 21. Wang Q, Luo L, Lei Q, et al. Low aneuploidy rate ICSI in women with polycystic ovary syndrome. in early pregnancy loss abortuses from patients Cochrane Database Syst Rev. 2014:CD006105. with polycystic ovary syndrome. Reprod Biomed 33. Ferrero H, Garcia-Pascual CM, Gaytan M, et al. Online. 2016;33:85–92. Dopamine receptor 2 activation inhibits ovarian 22. Li Y, Wang L, Xu J, et al. Higher chromosomal vascular endothelial growth factor secretion in an aberration rate in miscarried conceptus from ovarian hyperstimulation syndrome (OHSS) ani- polycystic ovary syndrome women undergoing mal model: implications for treatment of OHSS assisted reproductive treatment. Fertil Steril.2019; with dopamine receptor 2 agonists. Fertil Steril. 111:936–943. 2014;102:1468–1476. 23. Apparao KB, Lovely LP, Gui Y, et al. Elevated 34. Busso C, Fernandez-Sanchez M, Garcia-Velasco endometrial androgen receptor expression in JA, et al. The non-ergot derived dopamine agonist women with polycystic ovarian syndrome. Biol quinagolide in prevention of early ovarian hyper- Reprod. 2002;66:297–304. stimulation syndrome in IVF patients: a random- 24. Snider AP, Wood JR. Obesity induces ovarian ized, double-blind, placebo-controlled trial. Hum inflammation and reduces oocyte quality. Repro- Reprod. 2010;25:995–1004. duction. 2019;158:R79–R90. 35. Leitao VM, Moroni RM, Seko LM, et al. Ca- 25. Diamanti-Kandarakis E, Christakou CD, Kan- bergoline for the prevention of ovarian hyper- daraki E, et al. Metformin: an old medication stimulation syndrome: systematic review and of new fashion: evolving new molecular mecha- meta-analysis of randomized controlled trials. nisms and clinical implications in polycystic Fertil Steril. 2014;101:664–675. ovary syndrome. Eur J Endocrinol. 2010;162: 36. Papaleo E, Doldi N, De Santis L, et al. Cabergoline 193–212. influences ovarian stimulation in hyperprolactinaemic www.clinicalobgyn.com Copyright r 2020 Wolters Kluwer Health, Inc. All rights reserved. IVF for PCOS 47

patients with polycystic ovary syndrome. Hum Re- and cumulative cycle outcomes. Hum Reprod. prod. 2001;16:2263–2266. 2015;30:88–96. 37. D’Angelo A, Amso NN, Hassan R. Coasting 41. Gremeau AS, Andreadis N, Fatum M, et al. In (withholding gonadotrophins) for preventing vitro maturation or in vitro fertilization for ovarian hyperstimulation syndrome. Cochrane women with polycystic ovaries? A case-control Database Syst Rev. 2017;5:CD002811. study of 194 treatment cycles. Fertil Steril. 2012; 38. Nardo LG, Cheema P, Gelbaya TA, et al. The 98:355–360. optimal length of ‘coasting protocol’ in women at 42. Zhang XY, Ata B, Son WY, et al. Chromosome risk of ovarian hyperstimulation syndrome under- abnormality rates in human embryos obtained going in vitro fertilization. Hum Fertil. 2006;9: from in-vitro maturation and IVF treatment 175–180. cycles. Reprod Biomed Online. 2010;21:552–559. 39. Paulson RJ, Fauser B, Vuong LTN, et al. Can we 43. Kerjean A, Couvert P, Heams T, et al. In modify assisted reproductive technology practice vitro follicular growth affects oocyte imprinting to broaden reproductive care access? Fertil Steril. establishment in mice. Eur J Hum Genet. 2003;11: 2016;105:1138–1143. 493–496. 40. Walls ML, Hunter T, Ryan JP, et al. In vitro 44. Yu EJ, Yoon TK, Lee WS, et al. Obstetrical, maturation as an alternative to standard in vitro neonatal, and long-term outcomes of children fertilization for patients diagnosed with polycystic conceived from in vitro matured oocytes. Fertil ovaries: a comparative analysis of fresh, frozen Steril. 2019;112:691–699.