The approach to the PCOS patient undergoing IVF Roy Homburg Barzili Medical Centre, Ashkelon, Israel and Homerton University Hospital, London Antalya, October 2011

Problems – IVF for PCOS Excessive ovarian response Low fertilization rates High number of immature oocytes Reduced cleavage rates Low implantation rates High miscarriage rates

Overcoming the problems for PCOS in IVF Diagnosis and mild stimulation Agonist vs antagonist Oral contraceptive pre-treatment GnRH agonist to trigger ovulation Metformin Freeze embryos IVM

Comparison of the long protocol and the antagonist protocols more physiologic no cyst formation no hormonal withdrawal antagonist administration gonadotropin administration multiple dose protocol Patient-friendly less gona- dotropins early pregnancy? Menses Advantages of the antagonist protocol: - no cyst formation (since there is no flare up effect) - no hormonal withdrawal symptoms (since pituitary suppression is achieved after ovarian stimulation has been started and not before) - the antagonist cannot be given in early pregnancy, since this can be excluded by a pregnancy test - less gonadotropins necessary - shorter treatment duration - more physiologic, since it can be integrated in the spontaneous menstrual cycle flare up effect pituitary suppression agonist administration gonadotropin administration long protocol (mid-luteal) longer treatment pre-treatment cycle treatment cycle

O H S S 4.1% 2.6%

Hospital admission due to OHSS RR: 0.47 ~50% less risk for hospital admission due to OHSS with GnRH antagonists Kolibianakis et al. Hum Reprod Update. 2006

GnRH antagonists are safer than agonists: an update of a Cochrane review. Al-Inany HG, Youssef MA, Aboulghar M, Broekmans F, Sterrenburg M, Smit J, Abou-Setta AM. Hum Reprod Update, May 2011

Finally…..no difference in live birth rate 2011 May 11, 2011 With new information available, authors of a Cochrane Systematic Review have revised their conclusions about the relative effectiveness of two different treatments used to help women become pregnant. They now conclude that giving women GnRH-antagonists leads to similar live-birth rates compared with GnRH agonists. Previously they had concluded that women who used antagonists tended to have lower birth-rates than those using agonists. [........] In 2006, when the researchers reached their earlier conclusion, they were only able to draw data from 27 trials. Since then more research has been published, allowing them to consider the findings of 45 randomised controlled studies that involved a total of 7,511 women. “This increased amount of data lets us get a much better idea of how well the two approaches compare,” says Dr Hesham Al-Inany, who was lead author of the research and works at Cairo University, Egypt. Dr Al-Inany led a multi-centre team, with researchers also based in the Netherlands and Canada. “The reduction in ovarian hyperstimulation combined with a comparable live-birth rate mean justifies a move away from the standard GnRH agonist to using GnRH antagonists,” says Dr Al-Inany. http://eu.wiley.com/WileyCDA/PressRelease/pressReleaseId-96357.html

F&S 2008 18.1% vs 23.6%

Starting with Gn on day 2/3 after the last OC intake - OC pretreated Starting with Gn on day 2/3 after the last OC intake Fixed GnRH antagonist regimen (long-starting SD 1) Huirne et al, 2007

Incidence of OHSS Objective: to determine OHSS incidence in 2,524 antagonist-based cycles (1801 patients). Results: fifty three patients (2%) were hospitalized because of OHSS. Conclusions: clinically significant OHSS is a limitation even in antagonist cycles. “There is more than ever an urgent need for alternative final oocyte maturation – triggering medication” F&S January 2006

Day 5 , 6 or 7 antagonist start FIXED 0.25mg/day antagonist FSH Day 5 , 6 or 7 antagonist start FIXED hcg 0.25mg/day antagonist FSH day 8/9 hcg Follicle size 14mm- start antagonist Flexible regime FSH hcg

Flexible regime GnRH agonist GnRH agonist hcg 0.25mg/day antagonist FSH Day 5 , 6 or 7 antagonist start FIXED GnRH agonist 0.25mg/day antagonist FSH GnRH agonist day 8/9 Follicle size 14mm- start antagonist Flexible regime FSH hcg

Agonist: 1932 patients, not a single case of OHSS! hCG: 84 cases in 1760 patients, 4.8% OHSS % (n) n Ovulation trigger Oocyte source Trial type Reference 0 (0/13) 31(4/13) 15 13 GnRHa hCG own RCT, high risk Babayof et al 2006 0 (0/33) 31 (10/32) 33 32 Engamnn et al 2008 0 (0/30) 17 (5/30) 30 donors RCT Acevedo et al 2006 0 (0/1046) 1.3 (13/1031) 1046 1031 Retrospective Bodri et al 2009 0 (0/20) 20 Observational, High risk Griesinger et al 2007 0 (0/152) 2 (3/150) 152 150 Humaidan et al 2009 0 (0/23) 4 (1/23) 23 Retrospective, case-controlled, high risk Engmann et al 2006 0 (0/42) 42 hCG - cancelled Retrospective case-control, high risk Manzanares et al 2009 0 (0/254) 6 (10/175) 254 175 Hernandez et al 2009 0 (0/82) 7 (5/69) 82 69 Retrospective, high risk Orvieto et al 2006 0 (0/32) 1 (1/42) Retrospective, high risk: agonist arm only Shapiro et al 2007 0 (0/44) 7 (3/44) 44 Sismanoglu et al 2009 8 (1/12) 12 GnRH, luteal rescue with hCG 1500IU Observational, high risk 0 (0/106) 8 (9/106) 106 Galindo et al 2009 0 (0/45) 15 (33) 4 45 Shahrokh et al 2010

Agonist versus HCG for oocyte triggering Youssef et al. Cochrane Review 2010

Freeze and thaw cycles Luteal phase support 0.25mg/day antagonist FSH Day 5 start FIXED GnRH agonist Freeze and thaw cycles Luteal phase support FSH hcg

Triggering of final oocyte maturation with GnRH-a or HCG: Live birth after frozen-thawed embryo replacement cycles P = 0.02 Griesinger et al., Fertil Steril 2007

Frozen-thawed cycles Manzanares et al, 2009

Luteal phase support: 1. Massive doses P +/- E2 0.25mg/day antagonist FSH Day 5 start FIXED GnRH agonist Luteal phase support: 1. Massive doses P +/- E2 2. 1500 IU hCG on day OPU (Humaidan 2009) FSH hcg

GnRH agonist vs hCG in high risk IVF patients RCT, n=66 with PCO’s Antag + GnRH trigger vs Agonist + hCG trigger OHSS – 0% vs 31% Ongoing pregnancy rates – 53% vs 48% Adequate E2 , P supplementation in luteal phase Engmann et al, 2008

Luteal phase support: 1500 IU hCG on day OPU (Humaidan 2009) 0.25mg/day antagonist FSH Day 5 start FIXED GnRH agonist Luteal phase support: 1500 IU hCG on day OPU (Humaidan 2009) No significant difference in outcome compared with hCG trigger FSH hcg

Humaidan et al, 2010 GnRH agonist + 1500 U hCG on day OPU vs hCG Table 3. Pregnancy outcome in GnRHa vs. hCG-group. GnRH agonist + 1500 U hCG on day OPU vs hCG Variable GnRHa hCG OR (95% CI) P Value Patients, n 152 150 Clinical pregnancy (%) 50/152 (33) 55/150 (37) 0.8 (0.7–0.9) .29 Ongoing pregnancy (%) 40/152 (26) 49/150 (33) 0.7 (0.6–0.8) .69 Delivery rate (%) 36/152 (24) 47/150 (31) .16 Early pregnancy loss, n (% of positive hCG) 13/63 (21) 11/66 (17) 1.3 (0.7–1.9) .36 Humaidan et al, 2010

Beyond the context of OHSS: Patient-friendly luteal phase Abdominal pain and discomfort due to enlarged ovaries. How to minimize ovarian volume post oocyte retrieval?

Clinical use of agonist trigger opinion Primarily in the context of OHSS prevention. Prevention is total. A major reason to use GnRH antagonists in ovarian stimulation of high-risk patients: to keep the option of agonist trigger if needed.

Metformin for IVF - metformin (2G/d) - placebo for 16 weeks n=73 PCOS for IVF/ICSI - metformin (2G/d) - placebo for 16 weeks No difference in any stimulation, IVF or clinical criteria. BUT in group with BMI < 28, pregnancy rates double on metformin. Kjotrod et al, 2004

Metformin in IVF Single centre, double-blind RCT Tang, Bart & Balen, 2005 Single centre, double-blind RCT 94 patients, PCOS, BMI 27.8 101 IVF/ICSI cycles, long agonist protocol Metformin (850mg bd) or placebo from start of agonist to OPU

Metformin in IVF No difference: Total dose FSH No. of oocytes Fertilisation rates Tang, Barth & Balen, 2005

Metformin for IVF in PCOS Tang et al., Hum Reprod 2005

Metformin in IVF Short term co-treatment with metformin for PCOS in IVF/ICSI : Does not improve response to stimulation Improves pregnancy rates Reduces the risk of OHSS Tang, Bart & Balen, 2005

Endometrial dysfunction Low luteal phase serum glycodelin and IGFBP-1 (Jacubowicz et al, 2001) Plasma endothelin-1 levels high in PCOS (Diamantis-Kandarakis et al, 2005) Inadequate endometrial blood flow (Orio et al, 2005) All induced by hyperinsulinemia and improved by metformin.

IVM from unstimulated PCO N=118 women, PCOS. 152 cycles OPU day 9-14 ET – 140 cycles Clinical pregnancy rate – 40% / transfer 56 livebirths and another 10 ongoing. Zhao et al, F&S, 2008

Summary –High responders IVF The GnRH antagonist protocol appears to be an attractive option for PCOS patients undergoing IVF. Ovulation triggering with GnRH-a may be a better option than cycle cancellation or prolonged coasting. The addition of metformin to the treatment protocol may be beneficial for PCOS. Pretreatment with an OCP may be beneficial.