Effect of HCG-day serum progesterone and oestradiol concentrations on pregnancy outcomes in GnRH agonist cycles

Transcription

1 Reproductive BioMedicine Online (2012) 24, ARTICLE Effect of HCG-day serum progesterone and oestradiol concentrations on pregnancy outcomes in GnRH agonist cycles Ze Wu a,b, Rong Li a, Yanping Ma b, Bo Deng b, Xiaomei Zhang b,c, Yushi Meng c, Xinna Chen a, Ping Liu a, Jie Qiao a, * a Department of Obstetrics and Gynecology, Reproductive Medical Centre, Peking University Third Hospital, No. 49 North Huayuan Road, Haidian District, Beijing , People s Republic of China; b Department of Reproduction and Genetics, Reproductive Medical Centre, The First People s Hospital of Yunnan Province, No. 157 Jin Bi Road, Kunming , People s Republic of China; c Reproductive Medical Centre of The Second Hospital Affiliated Kunming Medical College, No. 374 Dian Mian Road, Kunming , People s Republic of China * Corresponding author. addresses: wuzes@hotmail.com (Z Wu), roseli001@sina.com (R Li), jie.qiao@263.net (J Qiao). Qiao Jie is chief physician and head at the Department of Obstetrics and Gynecology of Peking University Third Hospital and is now the president-elect for reproductive medical branch of the Chinese Medical Association. She attended Queen Mary Hospital, the University of Hong Kong as a visiting scholar in and Stanford University Medical Center as a postdoctoral research scholar in She specializes in treatment, education and research on female fertility preservation and improvement and the diagnosis and treatment for infertility and reproductive endocrine diseases, focusing on polycystic ovary syndrome. She is treasurer of the Asia Pacific Society for Reproductive Medicine, board member of the AEPCOS Society and Asian Journal of Andrology and editorial member for Reproductive Biology and Endocrinology and Seminars in Reproductive Medicine. Abstract This study analysed the relationship between serum progesterone/oestradiol concentrations and IVF pregnancy outcomes in gonadotrophin-releasing hormone agonist protocols. A total of 2921 infertile women undergoing IVF were assigned to four groups according to serum progesterone and oestradiol concentrations on the day of human chorionic gonadotrophin (HCG) administration: group 1 (control) progesterone < 3.34 nmol/l and oestradiol < 19,124 pmol/l; group 2 (high oestradiol); group 3 (high progesterone); group 4 (high progesterone and high oestradiol). Compared with group 1, group 4 had lower clinical pregnancy and live birth rates as well as the highest ectopic pregnancy rate (29.15% versus 45.91%; 18.67% versus 34.34%; 18.10% versus 5.82%; P < 0.05). Group 3 had lower clinical pregnancy and live birth rates per embryo-transfer cycle (29.78% versus 45.91%; 20.28% versus 34.34%, respectively; P < 0.05). Clinical pregnancy rates were similar in frozen thawed embryo transfers (FET) among the four groups. In conclusion, elevated progesterone was detrimental to live birth rates. High serum oestradiol concentration on HCG day did not affect the IVF pregnancy outcome. In combination with the elevated progesterone, high oestradiol concentrations had a potential negative effect. For these patients, FET should be suggested to improve the pregnancy outcomes. RBMOnline ª 2012, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. KEYWORDS: IVF, oestradiol, pregnancy outcomes, progesterone /$ - see front matter ª 2012, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. doi: /j.rbmo

2 512 Z Wu et al. Introduction During IVF cycles, the endometrium and embryo are exposed to supra-physiological concentrations of oestradiol and progesterone during ovarian stimulation, which could influence pregnancy outcomes. However, an elevated serum progesterone concentration on the day of human chorionic gonadotrophin (HCG) administration has previously been reported to adversely affect IVF/embryo-transfer pregnancy outcomes (Bosch et al., 2003; Ozcakir et al., 2004; Yu et al., 2010). A more recent publication by the current study group (Li et al., 2008) has also shown that high serum progesterone on the day of HCG administration had detrimental effects on the pregnancy rate of IVF/embryo transfer after using a gonadotrophin-releasing hormone (GnRH) agonist/ recombinant FSH (225 IU/day) short protocol. However, other studies have not found that probability of pregnancy decreased significantly when serum progesterone was above a threshold concentration on the day of HCG administration (Doldi et al., 1999; Martinez et al., 2004; Urman et al., 1999). Hofmann et al. (1993a, 1996) observed no significant differences in pregnancy rates in patients undergoing IVF/embryo transfer with high or low progesterone concentrations on the day of HCG administration and in patients who received oocytes from women with high or low progesterone concentrations. Furthermore, a negative association between the probability of pregnancy and oestradiol concentrations on the day of HCG administration has been reported (Arslan et al., 2007; Simon et al., 1995; Yu Ng et al., 2000). However, the unfavourable effect of elevated oestradiol has not been found in other studies (Chen et al., 2003; Kyrou et al., 2009; Papanikolaou et al., 2009). There is still controversy surrounding this issue due to absence of a large-scale study. The effects of elevated progesterone and oestradiol on the day of HCG administration on pregnancy outcomes is a controversial topic. However, the researches on these effects are scarce. Previous studies on the relationship between sex hormones and pregnancy outcomes are limited to elevated progesterone or oestradiol concentrations separately, not in combination. Some studies have mentioned that elevated progesterone concentrations often accompany elevated oestradiol concentrations (Azem et al., 2008; Kiliçdag et al., 2010; Li et al., 2008), but, as far as is known, there is no reported research analysing the effects of combined oestradiol and progesterone concentrations on pregnancy outcomes. Whether the elevation of both progesterone and oestradiol on the day of HCG administration could influence pregnancy outcomes remains unknown. Accordingly, the role of the elevation of both progesterone and oestradiol required further exploration. The aim of the present study was to analyse the relationship between these hormones and pregnancy outcomes in GnRH agonist protocols. Materials and methods Patients and study design A multicentre, large sample retrospective study was performed. A total of 2921 infertile women undergoing IVF/intracytoplasmic sperm injection (ICSI) treatment with a GnRH agonist protocol from April 2008 to April 2009 were enrolled in this retrospective study. These women received treatment at one of the following hospitals: the Reproductive Medical Centre of Peking University Third Hospital, the Department of Reproductive and Genetics of the First People s Hospital of Yunnan Province and the Reproductive Medical Centre of The No. 2 Hospital Affiliated Kunming Medical College. The inclusion criteria included the following: women <38 years of age, no previously diagnosed adnexal masses, basal serum FSH on day 2 < 10 IU/l, 4 retrieved oocytes, a day-3 embryo transfer, two embryos transferred when patient s age <35 years and a maximum of three embryos transferred in the fresh and frozen thawed embryo transfers (FET). Only the first stimulation cycle for each woman was considered. In this retrospective study, poor responders, with <5 oocytes harvested, were excluded. There were no other special exclusion criteria used to assign patients as short protocol and long protocols were widely used in the study centre s IVF/ovarian stimulation clinical practice. Patients assigned to either short protocol or long protocols were included; other options were ignored. Clinical indications for IVF/ICSI included tubal factors (53.1%), endometriosis (2.3%), polycystic ovary syndrome (3.2%), sperm abnormalities (22.7%), unexplained infertility (6.5%) and mixed factors (12.2%). Pregnancy rates in subsequent FET cycles were evaluated in 970 patients who did not become pregnant in the fresh embryo-transfer cycles. All patients signed an informed consent form before receiving IVF treatment. Protocols for ovarian stimulation There were two GnRH agonist protocols for ovarian stimulation. The details for the GnRH agonist short protocol (1970 cycles) have been published previously (Li et al., 2008). Briefly, GnRH agonist was administered from day 2 of menstruation (0.1 mg Diphereline; Beaufour Ipsen, France). Ovarian stimulation started on day 3 with IU recombinant FSH/HMG (75 IU Gonal-F; Serono, Switzerland; human menopausal gonadotrophin (HMG; 75 IU; Livzon Pharmaceutical, China). On day 2, prior to administering GnRH agonist, the serum FSH, oestradiol and progesterone concentrations were measured and the recombinant FSH/HMG dose was adjusted according to the ovarian response. Patients were evaluated by ultrasonography and serum oestradiol, LH and progesterone concentrations from day 8 until the day of the HCG administration. Specified ultrasound technicians monitored the ovarian response by vaginal ultrasonography. HCG (5000 IU per ampoule, Profasi; Serono) ,000 IU was given i.m. when the two leading follicles reached a mean diameter of 18 mm. Transvaginal ultrasound-guided oocyte retrieval was scheduled h after HCG administration. For the GnRH agonist long protocol (951 cycles), ovarian stimulations were carried out with recombinant FSH/HMG after pituitary down-regulation with GnRH agonist. The pituitary suppression was achieved with either daily 0.1 mg triptorelin (Diphereline) from the mid-luteal phase of the cycle preceding the treatment cycle, and after ovarian suppression was achieved, the dose of triptorelin was reduced to 0.05 mg until the day of HCG administration; or a

3 Serum progesterone and oestradiol on the day of HCG and pregnancy outcomes mg dose of triptorelin (3.75 mg per ampoule Diphereline) was administrated on day 21 of the cycle once. Pituitary down-regulation was confirmed by both transvaginal ultrasonography and serum oestradiol concentration on the second day of the treatment cycle. If there were no cysts 2 cm and the oestradiol was <180 pmol/l, gonadotrophin stimulation with IU recombinant FSH/HMG was started, consistent with the short protocol. The standard IVF procedures have been described elsewhere (Zhang et al., 1990). Two or three embryos were transferred on day 3, according to the Code of Practice for Assisted Reproductive Technology formulated by the Ministry of Health of the People s Republic of China: the number of embryos transferred per cycle must be fewer than three. Embryo transfers were performed by specified doctors with the same standard embryo transfer protocol in each medical centre. Surplus embryos of good quality were frozen for the subsequent transfer of embryos if the woman did not become pregnant after the fresh cycle. The standard embryo freezing procedures have been described elsewhere (Zheng et al., 2008). Luteal supplementation The luteal phase was supported by a 2000 IU HCG injection on the day of embryo transfer and repeated three times every 2 days. When the serum oestradiol concentration on the day of HCG administration was >15,000 pmol/l or the number of retrieved oocytes was >15, HCG injections were discontinued and 60 mg progesterone injection were given i.m. daily (20 mg per ampoule; Shanghai General Pharmaceutical, China) instead, from the day of embryo transfer for 14 days. To assess treatment outcomes, serum bhcg was measured 14 days after transfer. An increase in serum bhcg > 30 IU/l indicated pregnancy; a clinical pregnancy was defined by the ultrasonic observation of fetal cardiac activity 30 days after the embryo transfer. Hormonal measurements A hormonal assessment was performed at the initiation of stimulation, on day 8 and on the day of the HCG administration. Additional blood samples were taken before HCG administration, as necessary. Serum LH, FSH, oestradiol, progesterone and HCG samples were tested using the Immulite 1000 assay based on chemiluminescence (DPC, USA). The preparation, set up, dilutions and adjustment, assay and quality control procedures were performed according to the manufacturer s instructions. For all measurements, the inter-assay coefficient of variation was less than 10% and the intra-assay variation was less than 15%. Definition of progesterone/oestradiol elevation and groupings According to Li et al. (2008), receiver operating characteristic (ROC) curve analysis was used to identify the most efficient cut-off point for serum progesterone concentration to discriminate between successful and unsuccessful clinical pregnancy outcome. A serum progesterone concentration 3.34 nmol/l (1.05 ng/ml) was the threshold set for an elevated progesterone concentration and predicted a poor pregnancy outcome. Oestradiol elevation was defined when serum oestradiol concentration exceeded the 75th percentile (19,124 pmol/l; pg/ml), according to a previous study (Bahceci et al., 2006). The conversion of progesterone concentration was performed as progesterone (nmol/l)/ 3.18 = ng/ml. The conversion of oestradiol concentration was performed as oestradiol (pmol/l)/3.67 = pg/ml. Patients were assigned to one of four groups according to their serum progesterone and oestradiol concentrations on the day of HCG administration as follows: group 1 (control), progesterone <3.34 nmol/l and oestradiol <19,124 pmol/l; group 2 (high oestradiol): progesterone <3.34 nmol/l and oestradiol 19,124 pmol/l; group 3 (high progesterone), progesterone 3.34 nmol/l and oestradiol <19,124 pmol/l; group 4 (high progesterone and high oestradiol), progesterone 3.34 nmol/l and oestradiol 19,124 pmol/l. The clinical data for the four groups were analysed. Definition of clinical outcomes A clinical pregnancy was defined by the presence of one or more gestational sacs, including ectopic pregnancy or the demonstration of gestational production in the uterine evacuate. The clinical pregnancy rate was the number of clinical pregnancies per embryo-transfer cycle. The implantation rate was the proportion of embryos transferred that resulted in an intrauterine gestational sac. The live birth rate was the number of live births per embryo-transfer cycle. Early miscarriage was defined as the cessation or lack of detection of cardiac activity in the gestational sac or the inability to detect a previously defined gestational sac during the first 12 weeks of pregnancy. Late miscarriage was defined as pregnancy loss between gestation weeks 12 and 28. Miscarriage rate was calculated among women who were identified as having a clinical pregnancy. Statistical analysis Values were expressed as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) with multiple comparisons (Tukey s HSD) for continuous data and chi-squared test for categorical data were used for data analysis. Correlation was assessed by the Pearson method. The statistical analysis was performed with the Statistical Package for Social Sciences version 11.0 (SPSS, Chicago, IL, USA). A P-value of < 0.05 was considered to be statistically significant. Results A total of 2921 GnRH agonist cycles were included in this study: 1970 GnRH agonist short protocol cycles and 951 GnRH agonist long protocol cycles. Of those, 1863 were conventional IVF cycles and 1058 were ICSI cycles. In all, 411 cycles were cancelled due to risk of ovarian hyperstimulation syndrome (411/2921, 14.1%). Embryo transfer was carried out in 2510 cycles and there were 964 clinical pregnancies. In the 964 clinical pregnancies, there were 82 ectopic pregnancies (8.5%, 82/964), three heterotopic pregnancies (0.31%), 140 miscarriages, which included 121 early miscarriages (86.43%) and 19 late miscarriages (13.57%),

4 514 Z Wu et al. and 674 live birth cycles (69.92%); 63 cases were lost to follow-up (6.54%). There were five neonatal malformations in the live births: one child had trisomy 21 syndrome, one had a lower lip hemangioma, one had a patent ductus arteriosis, one had no left ear canal and one had two deformed left thumbs. Three of the malformations were from IVF pregnancies and two were from ICSI, which suggests ICSI did not increase the neonatal malformation rate. The area under the serum progesterone concentration ROC curve was (95% confidence interval (CI) ) (Figure 1). The serum progesterone cut-off value was 3.34 nmol/l. This value had 65% sensitivity, 52% specificity, 70.36% positive predictive value and 45.67% negative predictive value for not achieving a successful pregnancy. To assess the predictive value of the serum oestradiol concentration on the day of HCG administration of a successful pregnancy, a ROC curve analysis was also performed (Figure 1). The area under the curve (AUC) was (95% CI ), which was not adopted due to its low sensitivity (58.1%) and specificity (49.2%). Percentile analysis was performed to define elevation of oestradiol on the day of the HCG administration. An elevated oestradiol concentration was defined as an oestradiol concentration higher than the 75th percentile on the day of HCG administration (19,124 pmol/ml). Clinical features are presented in Table 1. Basal FSH, endometrial thickness on the day of HCG administration and the number of transferred embryos were not different between the groups. The gonadotrophin dosages were statistically different between groups, and were the lowest in group 2 ( ± IU) and the highest in group 3 ( ± IU) (one-way ANOVA P < 0.01). Significantly more oocytes were retrieved in groups 2 and 4 compared with groups 1 and 3 (group 2: ± 6.49; c l Figure 1 Area under the receiver operating characteristic curve for serum hormone measurements. For serum progesterone concentration (HCGP), the area under the curve was (95% CI ). For serum oestradiol concentration (HGCE2), the area under the curve was (95% CI ). group 4: ± 7.18; versus group 1: ± 5.84; group 3: ± 5.85; one-way ANOVA P < 0.001). More oocytes were fertilized and more high-quality embryos were seen in groups 2 and 4. However, the fertilization rate and rate of high-quality embryos did not reach statistical significance between the groups (Table 1). The clinical pregnancy rates in groups 3 and 4 were significantly lower than in group 3 (29.78%) and group 4 (29.15%) than group 1 (45.91%) or group 2 (44.44%; P < 0.05; Table 2). A significant difference existed between group 1 and groups 3 and 4 (P < 0.05), as well as between group 2 and groups 3 and 4 (P < 0.05). Of note, the ectopic pregnancy rate in group 4 was significantly higher than in the other three groups (chi-squared test P < 0.05). Implantation rates in groups 3 and 4 were also significantly lower than in groups 1 or 2 (P < 0.05; Table 2). The live birth rate was the lowest in group 4. A significant difference existed between group 4 (18.67%) and groups 1 and 2 (34.34% and 32.58%; P < 0.05), but no difference was seen between groups 4 and 3 (20.28%). A subgroup of 970 patients had one FET because of a failed transfer, early pregnancy loss or cancellation in a fresh IVF/embryo-transfer cycle. In FET cycles, the clinical pregnancy rate was similar between the four groups (Table 2). Discussion As far as is known, this was the first study of elevated serum progesterone and oestradiol on the day of HCG administration and IVF outcomes in GnRH agonist protocols. Previous studies have only concentrated on the relationship of single elevated progesterone or oestradiol concentrations on the day of HCG administration and the IVF/embryo-transfer pregnancy outcomes. The database included 2921 infertile patients from three teaching hospitals in China. The results indicated that the elevation of both progesterone and oestradiol concentrations was detrimental to pregnancy outcomes because the transfer of embryos with both high progesterone and high oestradiol serum concentrations on the day of HCG administration resulted in lower clinical pregnancy and live birth rates and a significantly higher ectopic pregnancy rate. According to the findings, a single elevated progesterone concentration could adversely affect the clinical outcome of IVF. In the case of oestradiol, the results showed no association between a single elevated oestradiol concentration and pregnancy outcomes. With regard to progesterone concentrations, the data showed that an increase in progesterone on the day of HCG administration impairs pregnancy, implantation and live birth rates, which was in agreement with a study of 1045 GnRH agonist cycles by Kiliçdag et al. (2010) and further confirmed the current study centre s previous data (Li et al., 2008), in which 251 infertile patients undergoing IVF/embryo transfer with the uniform GnRH agonist down-regulation and stimulation were prospectively studied. All the cycles were grouped according to serum progesterone concentration on the day of HCG administration (<3.97 nmol/l or 3.97 nmol/l). The pregnancy rate was significantly lower (25.9 versus 48.75%; P < 0.001) in the elevated progesterone group. If the serum progesterone

5 Serum progesterone and oestradiol on the day of HCG and pregnancy outcomes 515 Table 1 Clinical characteristics. Group 1 Group 2 Group 3 Group 4 P-value Patients 1375 (47.07) 194 (6.64) 816 (27.94) 536 (18.35) Long protocol cycles 559 (58.78) 32 (3.36) 284 (29.86) 76 (7.99) <0.01 a,b,c,d,e Short protocol cycles 816 (41.42) 162 (8.22) 532 (27.01) 460 (23.35) Age (years) ± ± ± ± 3.41 <0.05 b,e Basal FSH (IU/l) 6.90 ± ± ± ± 1.64 NS Gonadotrophin dosage ± ± ± ± <0.01 a,b,c,d,e (IU) Endometrial thickness ± ± ± ± 1.89 NS (mm) Oestradiol on HCG day ± ± ± ± <0.01 a,b,c,d,e,f (pmol/l) Progesterone on HCG day 1.99 ± ± ± ± 3.75 <0.001 b,c,d,e,f (nmol/l) LH on HCG day (miu/ml) 1.73 ± ± ± ± 2.39 <0.05 a,b,c,d,e,f Progesterone/oestradiol 0.29 ± ± ± ± a,b,c,d,e,f on HCG day Mature oocytes retrieved ± ± ± ± 7.18 <0.001 a,b,c,d Fertilized oocytes (2PN) 7.50 ± ± ± ± 5.85 <0.001 a,b,c,d 2PN rate (%) ± ± ± ± NS Good-quality embryos 4.86 ± ± ± ± 5.54 <0.001 a,b,c,d Good-quality embryo ± ± ± ± NS rate (%) Cancelled cycle rate (%) 140/1375 (10.18) 59/194 (30.41) 74/816 (9.07) 138/536 (25.75) <0.05 b,d,e,f Transferred embryos 2.22 ± ± ± ± 0.44 NS Values are n (%), n/total (%) or mean ± SD. Group 1 = normal (oestradiol < 19,124 pmol/l, progesterone < 3.34 nmol/l); group 2 = high oestradiol (oestradiol 19,124 pmol/l, progesterone < 3.34 nmol/l); group 3 = high progesterone (progesterone 3.34 nmol/l, oestradiol < 19,124 pmol/l); group 4 = high oestradiol + high progesterone (progesterone 3.34 nmol/l, oestradiol 19,124 pmol/l). ANOVA and chi-squared test. HCG = human chorionic gonadotrophin; NS = not significant; 2PN = 2 pronuclei. a Significant difference between groups 1 and 2 (P < 0.05). b Significant difference between groups 2 and 3 (P < 0.05). c Significant difference between groups 3 and 4 (P < 0.05). d Significant difference between groups 1 and 4 (P < 0.05). e Significant difference between groups 1 and 3 (P < 0.05). f Significant difference between groups 2 and 4 (P < 0.05). was >6.0 nmol/l, the pregnancy outcome was less successful. Many studies have described an adverse relationship between elevated progesterone concentrations and IVF pregnancy outcomes (Fanchin et al., 1993; Schoolcraft et al., 1991; Yovel et al., 1995). Although the precise mechanism is unclear, the effect could be attributed to a deleterious effect on either the embryo quality or endometrial receptivity. In this study, the fertilization rate, the rate of high-quality embryos and number of embryos transferred were not significantly different among the four groups, making the proposed theory of decreased embryo quality less likely. A subgroup analysis was performed in FET cycles and the pregnancy rate was similar among all groups, supporting the theory of reduced endometrial receptivity with high progesterone concentrations in the fresh embryo-transfer cycle. This finding was consistent with Li et al. (2008).In a recent study by Papanikolaou et al. (2009), 628 infertile women undergoing a GnRH antagonist/recombinant FSH treatment also demonstrated that even a modest raise in progesterone during the follicular phase had detrimental effects on the implantation potential of a high-quality cleavage-stage embryo. Despite routine suppression of gonadotrophins by GnRH agonist, a premature elevation of plasma progesterone had been reported in ovarian stimulation cycles, not only in flare protocols (short protocols), but also in long protocols. The frequency of elevated serum progesterone concentrations varies, to as high as 35% (5 35%) of stimulated cycles in individuals treated with GnRH agonists (Edelstein et al., 1990; Silverberg et al., 1991). The cause of the premature elevation of progesterone in GnRH agonist cycles remains unknown. Many researchers in the past have adopted the term premature luteinization to describe patients with a progesterone elevation on the day of HCG administration for the final oocyte maturation (Venetis et al., 2007). Most studies have arbitrarily used an absolute progesterone concentration on the day of HCG administration as an indicator of premature luteinization, and the cut-off concentration differs from 0.9 to 3.0 ng/ml (Bosch et al., 2003; Martinez et al., 2004; Zhong et al., 2002). A meta-analysis on the correlation of progesterone elevation and IVF pregnancy outcomes by Venetis et al. (2007) showed that the ROC curve was adopted only in three studies to determine the threshold of progesterone elevation: they were 0.9, 0.9 and 1.2 ng/ml respectively, while the others

6 516 Z Wu et al. Table 2 Pregnancy outcomes. Group 1 Group 2 Group 3 Group 4 P-value Transfer cycles 1235 (49.20) 135 (5.38) 742 (29.56) 398 (15.86) Long protocol cycles 483 (61.22) 18 (2.28) 245 (31.05) 43 (5.45) <0.01 a,b,c,d,e Short protocol cycles 752 (43.70) 117 (6.80) 497 (28.88) 355 (20.63) Clinical pregnancy rate 567/1235 (45.91) 60/135 (44.44) 221/742 (29.78) 116/398 (29.15) <0.05 b,d,e,f Implantation rate 750/2737 (27.40) 80/300 (26.67) 282/1673 (16.86) 155/878 (17.65) <0.05 b,d,e,f Ectopic pregnancy rate 33/567 (5.82) 7/60 (11.67) 21/221 (9.50) 21/116 (18.10) <0.05 c,d Heterotopic pregnancy rate 2/567 (0.35) 0/60 (0) 0/221 (0) 1/116 (0.86) NS Lost in follow up 41/567 (7.23) 3/60 (5.00) 12/221 (5.43) 7/116 (6.03) NS Early miscarriage rate 67/526 (12.74) 7/57 (12.28) 35/209 (16.75) 12/109 (11.01) NS Late miscarriage rate 14/526 (2.66) 0 3/209 (1.44) 2/109 (1.83) NS Total miscarriage rate 81/526 (15.40) 7/57 (12.28) 38/209 (18.18) 14/109 (12.84) NS Live birth rate/et 410/1194 (34.34) 43/132 (32.58) 148/730 (20.28) 73/391 (18.67) <0.05 b,d,e,f Neonatal malformation rate 3/526 (0.57) 1/57 (1.75) 0/209 (0) 1/109 (0.92) NS FET clinical pregnancy rate 84/187 (47.51) 46/91 (50.55) 110/235 (46.81) 118/242 (48.76) NS Values are n (%), n/total (%) or mean ± SD. Group 1 = normal (oestradiol < 19,124 pmol/l, progesterone < 3.34 nmol/l); group 2 = high oestradiol (oestradiol 19,124 pmol/l, progesterone < 3.34 nmol/l); group 3 = high progesterone (progesterone 3.34 nmol/l, oestradiol < 19,124 pmol/l); group 4 = high oestradiol + high progesterone (progesterone 3.34 nmol/l, oestradiol 19,124 pmol/l). Chi-squared test. ET = embryo transfer; FET = frozen thawed embryo transfer; NS = not significant. a Significant difference between groups 1 and 2 (P < 0.05). b Significant difference between groups 2 and 3 (P < 0.05). c Significant difference between groups 3 and 4 (P < 0.05). d Significant difference between groups 1 and 4 (P < 0.05). e Significant difference between groups 1 and 3 (P < 0.05). f Significant difference between groups 2 and 4 (P < 0.05). set the threshold arbitrarily. Shulman et al. (1996) showed, in 786 cycles of IVF cycles induced by long protocol, that 261 (33.2%) had plasma progesterone concentrations >0.9 ng/ml on the day of HCG administration (223 (28.37%) with plasma progesterone concentrations 1 2 ng/ml and 38 (4.83%) with concentrations >2 ng/ml) and the result suggested that early plasma progesterone rise had a negative impact on endometrial receptivity but not on egg and embryo quality. A more recent study with long protocol (Bosch et al., 2010) showed elevated serum progesterone concentrations on the day of HCG administration was associated with reduced ongoing pregnancy rates. In particular, serum progesterone concentrations of 1.5 ng/ml was associated with lower ongoing pregnancy rates following IVF/ICSI cycles irrespective of the GnRH analogue used for pituitary down-regulation. Younis et al. (1998, 2001) proposed a progesterone/oestradiol ratio >1 as the definition of premature lutenization and demonstrated that it was associated with low ovarian reserve and poor pregnancy outcomes. The progesterone/oestradiol >1 definition of premature lutenization could be used to differentiate between progesterone secretion from immature follicles as a product of poor ovarian reserve and a physiological progesterone secretion from multiple healthy mature follicles in women with good ovarian reserve (Kiliçdag et al., 2010). In a recent study by Lai et al. (2009), premature lutenization was defined as a progesterone/oestradiol ratio 1.2 in 139 infertile women with a normal ovarian reserve; the results suggested that an increased serum progesterone/oestradiol ratio on the day of HCG administration has a poor predictive value for IVF outcomes in infertile women with normal ovarian reserve treated with a long GnRH agonist protocol. Lee et al. (2009) demonstrated that using a progesterone/oestradiol ratio on the day HCG administration is theoretically possible, but the low sensitivity and positive predictive value made the use of progesterone/oestradiol clinically unfeasible. Premature lutenization, as defined by the progesterone/oestradiol ratio, was more prevalent in poor ovarian responders (Elnashae, 2010). In the present study, poor responders were excluded and only patients with 4 retrieved oocytes were examined. The proportion of progesterone/oestradiol >1 was only 3.9% (114/2921), therefore this study did not adopt the progesterone/oestradiol >1 definition of premature lutenization. The ROC curve was used to identify the threshold of elevated progesterone, and a serum progesterone 3.34 nmol/l (1.05 ng/ml) was selected to be the cut-off value. The incidence of progesterone elevation was 46.29%, which was higher than the 36.7% reported by the study centre s previous work (Li et al., 2008). In both the high progesterone and high oestradiol groups, the elevated progesterone might be attributed to an excess number of follicles, each one producing a normal amount of progesterone for the late follicular phase. Thus, the excess of proliferating granulosa cells leads to an increased in progesterone production independently of LH exposure (Li et al., 2008; Kiliçdag et al., 2010). We noticed that the number of retrieved oocytes in the high progesterone only group was similar with the normal group, but the total FSH dose was significantly higher in the high progesterone only group. This finding is consistent with Filicori et al. (2002), who found a strong positive correlation between the administered FSH dose and follicular progesterone concentrations. This might explain the progesterone elevation in the high progesterone

7 Serum progesterone and oestradiol on the day of HCG and pregnancy outcomes 517 group. The high incidence of progesterone elevation in this large sample study might also be attributed to the lack of uniform GnRH agonist down-regulation and stimulation protocols, as different gonadotrophin-dosage forms contained LH. In the case of oestradiol, the results showed no association between single oestradiol concentrations and achievement of pregnancy, which is in agreement with Kyrou et al. (2009) and Yu Ng et al. (2000). Although the percentile curves presented a more objective tool, there are still no uniformly accepted oestradiol percentiles. Kyrou et al. (2009) used 25th and 75th percentiles to divide the patients into three groups according to oestradiol concentrations on the day of HCG administration (<1142, , >2446 pg/ml). Their results showed that in patients with oestradiol concentrations higher than the 75th percentile (oestradiol concentration >2446 pg/ml), the pregnancy rates remained the same as compared with the medium and lower percentile group, although the embryo quality was better than the two other groups. In Ulug et al. (2006), 5273 ovarian stimulation and ICSI cycles were segregated according to serum oestradiol concentrations percentiles (<24th, 25th, 74th, and >75th), the result showed that the subclinical pregnancy rate was not influenced by the concentrations of oestradiol during ovarian stimulation. The present study used percentile analysis to divide the patients according to oestradiol concentration on the day of HCG administration. The results showed that the implantation rates, pregnancy rates, early miscarriage rates and live birth rates in cycles with a high oestradiol concentration on the day of HCG administration (>75th percentile, oestradiol 19,124 pmol/l ( pg/ml) were not significantly different when compared with the normal group. This finding indicated that a high oestradiol concentration on the day of HCG administration did not adversely affect endometrial receptivity in fresh embryo-transfer cycles. In a similar study, Chen et al. (2007) evaluated 1196 GnRH agonist cycles grouped by peak oestradiol percentile distribution into three groups: normal responders between the 25 75th percentiles (oestradiol pg/ml, n = 595 cycles); moderate-high responders between the 75 90th percentiles (oestradiol pg/ml, n = 180 cycles); and high responders > 90th percentiles (oestradiol > 4128 pg/ml, n = 119 cycles). The high-response group showed decreasing trends in implantation and pregnancy rates compared with normal responders, but only the difference in implantation rates was statistically significant. The oestradiol cut-off concentration (19,124 pmol/l) in the present study, which was higher than those reported (Chen et al., 2007; Kyrou et al., 2009), may be explained by the exclusion of poor responders. In a meta-analysis by Kosmas et al. (2004), two studies showed a positive association and two studies showed a negative association of oestradiol and pregnancy rates, while the majority of studies showed no association (Chen et al., 2003; Dor et al., 1992; Papageorgiou et al., 2002; Sharara and McClamrock, 1999). The inconsistent results could partially be explained by the retrospective design of the studies, their small sample sizes and their methodologies, such as the inclusion of more than one cycle for analysis of the same patients in more than three studies (Kosmas et al., 2004). In this retrospective study, poor responders were excluded and GnRH agonist short protocol and long protocol were included. When patients undergoing the short or long protocol were analysed separately, there were no differences in the proportions of short protocol and long protocol in each group, and the same trends were found in that elevated progesterone was detrimental to live birth rates, transfers with high progesterone and oestradiol on the day of HCG resulted in not only low clinical pregnancy and live birth rates, but also a significantly higher ectopic pregnancy rate (data not shown). In order to avoid patient differences and the basic treatment differences among groups, large multicentre samples were included. The purpose of the present study was to analyse the relationship between progesterone or/and oestradiol and pregnancy outcomes in GnRH agonist protocols. So, combined GnRH agonist short protocols and long protocols were included in the final analysis. The results showed that patients in the elevated progesterone and oestradiol group on the day of HCG administration had lower clinical pregnancy and live birth rates. Also, the ectopic pregnancy rate was significantly higher in this group. Because only a high oestradiol concentration had no effect on pregnancy outcome, elevated progesterone may be the major negative factor on pregnancy outcomes. When combined with premature elevation of progesterone, high oestradiol concentration may have a potential detrimental effect on pregnancy outcome. However, the data on the risk factors for developing an ectopic pregnancy after IVF are still inconsistent. The incidence of ectopic pregnancy after IVF generally has been reported to be from 2.1 to 8.6% of all clinical pregnancies (Clayton et al., 2006; Nazari et al., 1993). Theoretically, differences between natural conception and conception via assisted reproductive technology might affect the risk of ectopic pregnancy. The different hormonal milieu, the reproductive health characteristics of infertile women, technical issues of the IVF procedures (such as zygote intra-fallopian transfer, assisted hatching) and estimated embryo implantation potential were possible risk factors (Chang and Suh, 2010). In the present study, the proportion of patients with a history of ectopic pregnancy and known tubal factors in group 4 (both high progesterone and high oestradiol) were not significantly different from the normal group (9.1 versus 8.6% and 54.5% versus 45.7%, respectively). More oocytes were retrieved in group 4, but the cleavage rate, high-quality embryo rate and transferred embryo rate were similar between group 4 and normal group. Furthermore, the embryos were transferred by specified doctors using the same standard embryo-transfer protocol on day 3. This suggests that embryo quality was not the major factor for ectopic pregnancy and that abnormally elevated oestradiol and/or progesterone concentrations or the imbalance of their ratio may interrupt the implantation environment, relating to increasing ectopic pregnancy rates. Higher hormone concentrations might affect tubal peristalsis and the egg or zygote transport. Progesterone was shown to contribute to uterine quiescence (Fanchin et al., 2001). Increasing uterine contractility might allow for decreased implantation within the uterine cavity and favour migration of the embryos into the Fallopian tubes.

8 518 Z Wu et al. In IVF/embryo-transfer cycles, supra-physiological progesterone concentrations could exceed the concentration of normal conception and may result in more uterine relaxation during IVF/embryo transfer. Also, high oestradiol concentrations in IVF cycles affected tubal peristalsis through the control of tubal smooth muscle contractility and ciliary activity (Fernandez et al., 1991). Paltieli et al. (2000) found that, after the addition of progesterone to the culture medium, higher concentrations of progesterone caused a significant decrease in the ciliary beat frequency, oestradiol induced a slight increase in frequency, and Metrodin, Pergonal or LH did not affect ciliary motility, which suggested that higher concentrations of progesterone caused ciliary dysfunction and subsequently might be a cause of ectopic pregnancy. However, Pyrgiotis et al. (1994) did not demonstrate a difference in oestradiol concentrations on the day of HCG administration between IVF/embryo-transfer patients with and without ectopic pregnancies. The possible relationship between the risk of ectopic pregnancy and the oestradiol and progesterone concentrations at the time of embryo transfer should be further investigated. Ovarian stimulation is known to advance endometrial maturation, and progesterone might hasten the closure of the implantation window (Hofmann et al., 1993b). A recent study by Papanikolaou et al. (2009) showed that progesterone rise on the day of HCG administration impaired pregnancy outcome in day-3 single-embryo transfers, whilst it had no effect on day-5 single-blastocyst transfers. It was proposed that high follicular progesterone concentrations greatly advance the endometrium, and therefore the placement of a day-3 embryo in an asynchronous endometrium (earlier than a natural pregnancy) resulted in the failure to establish an embryo endometrium cross-dialogue and a failed implantation. The negative impact of premature luteinization on pregnancy rates with blastocyst transfer suggests that the endometrium has already significantly recovered from the violation induced from the supra-physiological steroid concentrations on the fifth luteal day. Thus, a proposed strategy for the cases with progesterone elevation on the day of HCG administration was the selection for embryo transfer on day 5. In the current study, a subgroup analysis was performed in FET cycles and the pregnancy rate was similar among the groups. This finding further supported the theory of reduced endometrial receptivity with high progesterone or accompanying high oestradiol concentration on the day of HCG administration in fresh embryo-transfer cycles and suggested the transfer of frozen thawed embryos in such cycles. In conclusion, progesterone elevation on the day of HCG administration in GnRH agonist protocols was detrimental to IVF pregnancy outcomes by reducing endometrial receptivity. No association was found between elevated oestradiol concentrations alone and successful pregnancies. However, in combination with the premature elevation of progesterone, high oestradiol concentrations had a negative effect on pregnancy outcome, as the transfer of embryos with high progesterone and oestradiol concentrations on the day of HCG administration resulted in low clinical pregnancy and live birth rates and a significantly higher ectopic pregnancy rate. For patients with high progesterone concentrations or high progesterone concentrations with accompanying high oestradiol concentrations on the day of HCG administration, embryos should be frozen and transferred in a later cycle to achieve a successful pregnancy outcome. Acknowledgements This study was supported by the National Science Fund for Distinguished Young Scholars ( ) and National Basic Research Program of China, 973 Program (2007CB511901). References Arslan, M., Bocca, S., rslan, E.O., Duran, H.E., Stadtmauer, L., Oehninger, S., Cumulative exposure to high estradiol levels during the follicular phase of IVF cycles negatively affects implantation. J. Assist. Reprod. Genet. 24, Azem, F., Lessen, J.B., Malcov, M., Ben-Yosef, D., Almog, B., Amit, A., Does high serum progesterone level on the day of human chorionic gonadotrophin administration affect pregnancy rate after intracytoplasmic sperm injection and embryo transfer? Gynecol. Endocrinol. 24, Bahceci, M., Ulug, U., Erden, H.F., Mesut, A., Jozwiak, E.A., Elevated oestradiol concentrations are not associated with increased first trimester miscarriage rates of singleton gestations conceived by assisted conception treatment. RBM online 12, Bosch, E., Valencia, K., Escudero, E., Crespo, J., Simón, C., Remohí, J., Pellicer, A., Premature luteinization during gonadotropin-releasing hormone antagonist cycles and its relationship with in vitro fertilization outcome. Fertil. Steril. 80, Bosch, E., Labarta, E., Crespo, J., Simo n, C., Remohı, J., Jenkins, J., Pellicer, A., Circulating progesterone levels and ongoing pregnancy rates in controlled ovarian stimulation cycles for in vitro fertilization: analysis of over 4000 cycles. Hum. Reprod. 25, Chang, H.J., Suh, C.S., Ectopic pregnancy after assisted reproductive technology: what are the risk factors? Curr. Opin. Obstet. Gyn. 22, Chen, C.H., Zhang, X.Q., Barnes, R., Confino, E., Milad, M., Puscheck, E., Kazer, R.R., Relationship between peak serum estradiol levels and treatment outcome in in vitro fertilization cycles after embryo transfer on day 3 or day 5. Fertil. Steril. 80, Chen, Q.J., Sun, X.X., Li, L., Gao, X.H., Wu, Y., Gemzell-Danielsson, K., Cheng, L.N., Effects of ovarian high response on implantation and pregnancy outcome during controlled ovarian hyperstimulation (with GnRH agonist and rfsh). Acta Obstet. Gyn. Scan. 86, Clayton, H.B., Schieve, L.A., Peterson, H.B., Jamieson, D.J., Reynolds, M.A., Wright, V.C., Ectopic pregnancy risk with assisted reproductive technology procedures. Obstet. Gynecol. 107, Doldi, N., Marsiglio, E., Destefani, A., Gessi, A., Merati, G., Ferrari, A., Elevated serum progesterone on the day of HCG administration in IVF is associated with a higher pregnancy rate in polycystic ovarian syndrome. Hum. Reprod. 14, Dor, J., Seidman, D.S., Ben-Shlomo, I., Levran, D., Karasik, A., Mashiach, S., The prognostic importance of the number of oocytes retrieved and estradiol levels in poor and normal responders in in vitro fertilization (IVF) treatment. J. Assist. Reprod. Genet. 9, Edelstein, M.C., Seltman, H.J., Cox, B.J., Robinson, S.M., Shaw, R.A., Muasher, S.J., Progesterone levels on the day of human chorionic gonadotropin administration in cycles with gonadotropin-releasing hormone agonist suppression are not predictive of pregnancy outcome. Fertil. Steril. 54,

9 Serum progesterone and oestradiol on the day of HCG and pregnancy outcomes 519 Elnashae, A.M., Progesterone rise on the day of HCG administration (premature luteinization) in IVF: An overdue update. J. Assisted Reprod. Genet. 27, Fanchin, R., de Ziegler, D., Taieb, J., Hazout, A., Frydman, R., Premature elevation of plasma progesterone alters pregnancy rates of in vitro fertilization and embryo transfer. Fertil. Steril. 59, Filicori, M., Cognigni, G.E., Pocognoli, P., Tabarelli, C., Spettoli, D., Taraborrelli, S., Ciampaglia, W., Modulation of folliculogenesis and steroidogenesis in women by graded menotrophin administration. Hum. Reprod. 17, Fernandez, H., Coste, J., Job-Spira, N., Controlled ovarian hyperstimulation as a risk factor for ectopic pregnancy. Obstet. Gynecol. 78, Fanchin, R., Righini, C., de Ziegler, D., Olivennes, F., Ledée, N., Frydman, R., Effects of vaginal progesterone administration on uterine contractility at the time of embryo transfer. Fertil. Steril. 75, Hofmann, G.E., Bentzien, F., Bergh, P.A., Garrisi, G.J., Williams, M.C., Guzman, I., Navot, D., 1993a. Premature luteinization in controlled ovarian hyperstimulation has no adverse effect on oocyte and embryo quality. Fertil. Steril. 60, Hofmann, G.E., Bergh, P.A., Guzman, I., Masuku, S., Navot, D., 1993b. Premature luteinization is not eliminated by pituitary desensitization with leuprolide acetate in women undergoing gonadotrophin stimulation who demonstrated premature luteinization in a prior gonadotrophin-only cycle. Hum. Reprod. 8, Hofmann, G.E., Khoury, J., Johnson, C.A., Thie, J., Scott Jr., R.T., Premature luteinization during controlled ovarian hyperstimulation for in vitro fertilization embryo transfer has no impact on pregnancy outcome. Fertil. Steril. 66, Kiliçdag, E.B., Haydardedeoglu, B., Cok, T., Hacivelioglu, S.O., Bagis, T., Premature progesterone elevation impairs implantation and live birth rates in GnRH-agonist IVF/ICSI cycles. Arch. Gynecol. Obstet. 81, Kosmas, I., Kolibianakis, E., Devroey, P., Association of estradiol levels on the day of hcg administration and pregnancy achievement in IVF: a systematic review. Hum. Reprod. 19, Kyrou, D., Popovic-Todorovic, B., Fatemi, H.M., Bourgain, C., Haentjens, P., Van Landuyt, L., Devroey, P., Does the estradiol level on the day of human chorionic gonadotrophin administration have an impact on pregnancy rates in patients treated with rec-fsh/gnrh antagonist? Hum. Reprod. 24, Lai, T.S., Lee, F.K., Lin, T.K., Horng, S.G., Chen, S.C., Chen, Y.H., Wang, P.C., An increased serum progesterone-to-estradiol ratio on the day of human chorionic gonadotropin administration does not have a negative impact on clinical pregnancy rate in women with normal ovarian reserve treated with a long gonadotropin releasing hormone agonist protocol. Fertil. Steril. 92, Lee, F.K., Lai, T.S., Lin, T.K., Horng, S.G., Chen, S.C., Relationship of progesterone/estradiol ratio on day of hcg administration and pregnancy outcomes in high responders undergoing in vitro fertilization. Fertil. Steril. 92, Li, R., Qiao, J., Wang, L., Zhen, X., Lu, Y., Serum progesterone concentration on day of HCG administration and IVF outcome. RBM Online 6, Martinez, F., Coroleu, B., Clua, E., Tur, R., Buxaderas, R., Parera, N., Barri, P.N., Balasch, J., Serum progesteron concentrations on the day of HCG administration cannot predict pregnancy in assisted reproduction cycles. RBM Online 8, Nazari, A., Askari, H.A., Check, J.H., O Shaughnessy, A., Embryo transfer technique as a cause of ectopic pregnancy in in vitro fertilization. Fertil. Steril. 60, Ozcakir, H.T., Levi, R., Tavmergen, E., Göker, E.N., Premature luteinization defined as progesterone estradiol ratio > 1 on hcg administration day seems to adversely affect clinical outcome in long gonadotropin-releasing hormone agonist cycles. J. Obstet. Gynaecol. Res. 30, Paltieli, Y., Eibschitz, I., Ziskind, G., Ohel, G., Silbermann, M., Weichselbaum, A., High progesterone levels and ciliary dysfunction a possible cause of ectopic pregnancy. J. Assist. Reprod. Genet. 17, Papageorgiou, T., Guibert, J., Goffinet, F., Patrat, C., Fulla, Y., Janssens, Y., Zorn, J.R., Percentile curves of serum estradiol levels during controlled ocarina stimulation in 905 cycles stimulated with recombinant FSH show that high estradiol is not detrimental to IVF outcome. Hum. Reprod. 17, Papanikolaou, E.G., Kolibianakis, E.M., Pozzobon, C., Tank, P., Tournaye, H., Bourgain, C., Van Steirteghem, A., Devroey, P., Progesterone rise on the day of human chorionic gonadotropin administration impairs pregnancy outcome in day 3 single-embryo transfer, while it has no effect on day 5 single blastocyst transfer. Fertil. Steril. 91, Pyrgiotis, E., Sultan, K.M., Neal, G.S., Liu, H.C., Grifo, J.A., Rosenwaks, Z., Ectopic pregnancies after in vitro fertilization and embryo transfer. J. Assist. Reprod. Genet. 11, Schoolcraft, W., Sinton, E., Schlenker, T., Huynh, D., Hamilton, F., Meldrum, D.R., Lower pregnancy rate with premature luteinization during pituitary suppression with leuprolide acetate. Fertil. Steril. 55, Sharara, F.I., McClamrock, H.D., Ratio of oestradiol concentration on the day of human chorionic gonadotrophin administration to mid-luteal oestradiol concentration is predictive of in-vitro fertilization outcome. Hum. Reprod. 14, Shulman, A., Ghetler, Y., Beyth, Y., Ben-Nun, I., The significance of an early (premature) rise of plasma progesterone in in vitro fertilization cycles induced by a long protocol of gonadotropin releasing hormone analogue and human menopausal gonadotropins. J. Assist. Reprod. Genet. 13 (3), Silverberg, K.M., Burns, W.N., Olive, D.L., Riehl, R.M., Schenken, R.S., Serum progesterone levels predict success of in vitro fertilization/embryo transfer in patients stimulated with leuprolide acetate and human menopausal gonadotropins. J. Clin. Endocrinol. Metab. 73, Simon, C., Cano, F., Valbuena, D., Remohi, J., Pellicer, A., Clinical evidence for a detrimental effect on uterine receptivity of high serum E2 concentrations in high and normal responder patients. Hum. Reprod. 10, Ulug, U., Tosun, S., Jozwiak, E.A., Mesut, A., Sismanoglu, A., Bahceci, M., Subclinical pregnancy losses among women undergoing in-vitro fertilization with ICSI. J. Assist. Reprod. Genet. 23, Urman, B., Alatas, C., Aksoy, S., Mercan, R., Isiklar, A., Balaban, B., Elevated serum progesterone level on the day of human chorionic gonadotropin administration does not adversely affect implantation rates after intracytoplasmic sperm injection and embryo transfer. Fertil. Steril. 72, Venetis, C.A., Kolibianakis, E.M., Papanikolaou, E., Bontis, J., Devroey, P., Tarlatzis, B.C., Is progesterone elevation on the day of human chorionic gonadotrophin administration associated with the probability of pregnancy in in vitro fertilization? A systematic review and meta-analysis. Hum. Reprod. Update 13, Younis, J.S., Haddad, S., Matilsky, M., Ben-Ami, M., Premature luteinization: could it be an early manifestation of low ovarian reserve? Fertil. Steril. 69, Younis, J.S., Matilsky, M., Radin, O., Ben-Ami, M., Increased progesterone/estradiol ratio in the late follicular phase could be related to low ovarian reserve in in vitro fertilization-embryo