Article. Laura Detti, MD, Frank D. Yelian, MD, PhD, Michael L. Kruger, MA, Michael P. Diamond, MD, Elizabeth E. Puscheck, MD

Transcription

1 Article Endometrial Thickness Dynamics and Morphologic Characteristics During Pituitary Downregulation With Antagonists in Assisted Reproductive Technology Cycles Laura Detti, MD, Frank D. Yelian, MD, PhD, Michael L. Kruger, MA, Michael P. Diamond, MD, Elizabeth E. Puscheck, MD Objective. The purpose of this study was to evaluate whether the dynamics of endometrial stripe thickness during gonadotropin-releasing hormone (GnRH) antagonist pituitary downregulation in in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) cycles are related to implantation and pregnancy outcomes. Methods. This retrospective cohort study evaluated 115 conventional IVF/ICSI cycles. All patients underwent ovarian stimulation with gonadotropins and the GnRH antagonist ganirelix acetate. The endometrial stripe was measured transvaginally daily from the day of initial GnRH antagonist administration to the day of the human chorionic gonadotropin (hcg) trigger and then transabdominally on the day of embryo transfer. We created 5 categories (0 4) of endometrial thickness variation, considering significant a daily variation of 1.5 mm. Our aim was to predict whether the endometrial thickness dynamics or morphologic characteristics were related to the duration of ovarian stimulation, duration of ganirelix use, or estradiol levels during ovarian stimulation and whether they would influence implantation and pregnancy rates. Results. No relationship was found between the duration of ovarian stimulation, duration of ganirelix use, and estradiol level (expressed as the area under the curve), and endometrial thickness dynamics or morphologic characteristics. Despite a thinner endometrial thickness in 37% of the cycles on the day of the hcg trigger compared with the beginning of GnRH antagonist stimulation, there was no correlation between endometrial dynamics and pregnancy outcomes. There was, instead, a positive relationship between a trilaminar endometrial morphologic pattern with a positive pregnancy test result, successful implantation, and ongoing pregnancy (P <.05). Conclusions. Despite a net decrease in thickness in almost 50% of cases, endometrial dynamics did not correlate with pregnancy outcomes. Conversely, a trilaminar endometrial morphologic pattern on the day of embryo transfer was positively related to pregnancy outcomes. Key words: endometrial thickness; gonadotropin-releasing hormone antagonist; in vitro fertilization; outcome; sonography. Abbreviations GnRH, gonadotropin-releasing hormone; hcg, human chorionic gonadotropin; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilization Received May 8, 2008, from the Department of Obstetrics and Gynecology, Division of Reproductive Endocrinology and Infertility, Wayne State University School of Medicine, Detroit, Michigan USA. Revision requested May 28, Revised manuscript accepted for publication June 24, Address correspondence to Laura Detti, MD, Department of Obstetrics and Gynecology, Division of Reproductive Endocrinology and Infertility, Wayne State University, 3750 Woodward Ave, Suite 200-D, Detroit, MI USA. ldetti@med.wayne.edu In in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) cycles, both the embryo quality and optimal endometrial development are deemed necessary to achieve pregnancy. Despite advancements in techniques for embryo manipulation and culture, the percentage of embryo transfers resulting in live births for fresh IVF/ICSI cycles increased only 22% in previous years: from 28% in 1996 to 34% in Predicting the probability of pregnancy by assessing the degree of endometrial development on sonography at the time of the human chorionic gonadotropin (hcg) trigger has 2008 by the American Institute of Ultrasound in Medicine J Ultrasound Med 2008; 27: /08/$3.50

2 Endometrial Thickness in Assisted Reproductive Technology Cycles been the objective of many investigators since the 1980s. 2 Unfortunately, results have been discordant, possibly because of different treatment protocols used for different populations and different etiologies of patients infertility. However, all studies seem to agree that a thin endometrium is detrimental to the implantation and development of pregnancy. 3 Some authors have thought that evaluation of the changes in endometrial growth during ovulation induction cycles for IVF/ICSI could be more helpful than a single measurement on the day of the hcg trigger. 4 9 However, all of the reports dealt with cycles downregulated by gonadotropinreleasing hormone (GnRH) agonists and measurement of the endometrium at fixed intervals during ovarian stimulation as well as the luteal phase. Except for the most recent study, 9 all found no correlation between progressive endometrial growth changes and pregnancy outcomes. All of the studies, however, agreed that the final endometrial thickness on the day of hcg administration was not related to IVF/ICSI pregnancy outcomes. In recent years, a different means to obtain pituitary downregulation during ovarian stimulation has been approved by the US Food and Drug Administration and implemented in our clinic: the GnRH antagonists ganirelix acetate and cetrorelix acetate. Results of clinical trials to date suggest that with GnRH antagonists for pituitary downregulation, much shorter treatment regimens with fewer injections and possibly less gonadotropin can achieve good clinical results However, a Cochrane review showed that ovarian stimulation cycles using GnRH antagonist downregulation during IVF/ICSI cycles were reported to have worse pregnancy outcomes when compared with GnRH agonists. 13 Interestingly, it appears that endometrial receptivity, and not embryo quality, is the limiting step in implantation because similar pregnancy rates between GnRH agonists and antagonists are achieved when the derived embryos are transferred in subsequent frozen-thawed cycles. 14 In cycles downregulated with GnRH antagonists, we recently found that the endometrial thickness on the day of the hcg trigger was not related to embryo implantation or ongoing pregnancy. 15 However, a thin endometrial stripe was related to early pregnancy loss, defined as a biochemical pregnancy or miscarriage. In this study, we sought to evaluate whether the sequential sonographic appearance of the endometrium exposed to the effect of the GnRH antagonist ganirelix influenced IVF/ICSI pregnancy outcomes. We investigated its relationship with the duration of ovarian stimulation, the serum estradiol level throughout the stimulation (expressed by the area under the curve), and the pregnancy outcomes of IVF/ICSI treatments. Materials and Methods This was a retrospective cohort study that evaluated 99 patients undergoing 115 conventional IVF/ICSI cycles between January 2004 and December Cycles were included regardless of the patients diagnoses or reproductive histories, but all needed to have daily endometrial measurements during GnRH antagonist administration. All patients underwent ovarian stimulation with gonadotropins (Gonal-f, EMD Serono, Inc, Rockland, MA; and Repronex, Ferring Pharmaceuticals, Inc, Albuquerque, New Mexico) and ganirelix acetate (Antagon, Organon, Inc, West Orange, NJ) for pituitary downregulation. Patients began receiving oral contraceptives 1 month before stimulation. Gonadotropins were administered from stimulation day 1 until the day of the hcg trigger according to a step-up protocol. The GnRH antagonist (0.25 mg daily) was added from the day when at least 1 follicle reached 14 mm in diameter and continued until hcg administration. Supplemental luteinizing hormone and follicle-stimulating hormone (Repronex) in the form of 1 ampoule per day were added to the gonadotropin regimen when GnRH antagonist administration was started. This regimen was continued until and including the day of hcg trigger administration. The hcg was administered when at least 2 follicles were greater than 18 mm in diameter. Endometrial thickness was measured trans - vaginally daily from the day of initial ganirelix administration to the day of the hcg trigger. The thickest endometrial segment (between the two interfaces of the endometrial-myometrial junction) was measured transvaginally on a frozen midplane longitudinal section of the uterus by 1592 J Ultrasound Med 2008; 27:

3 Detti et al 2-dimensional sonography. Endometrial thickness was measured again transabdominally on the day of embryo transfer. (No transvaginal probe is used at the time of embryo transfer; a speculum is introduced into the vagina, and intrauterine advancement of the catheter loaded with the embryo(s) is executed under transabdominal sonographic guidance). At that time, the endometrial pattern was also recorded as trilaminar or homogeneously solid : the first pattern indicated the sonographic appearance of 3 marked lines separated by hypoechoic endo metrium; the second indicated that the endometrium was homogeneously hyperechoic or isoechoic relative to the adjacent myometrial tissue (Figure 1). We created 5 categories (0 4) of endometrial thickness variation until the hcg trigger day, considering significant a daily variation of 1.5 mm, as indicated in Table 1. The 1.5- mm variation was based on the intraobserver and interobserver variation of 1 mm in our clinic (F.D.Y., M.P.D., and E.E.P., unpublished data, 2007). Endometrial thickness variation was also simply calculated as the net difference between the thickness on the hcg trigger day minus the thickness on the ganirelix start day (3 categories based on a minimum change of 1.5 mm: a, no change; b, increase; and c, decrease). Retrieval of the oocytes was performed 36 hours after the hcg trigger by sonographically guided transvaginal follicular aspiration. Transfer of the embryos to the uterine cavity was performed on day 3 after retrieval for all of the patients. Luteal phase support was given by daily injections of progesterone in oil (100 mg intramuscularly) until a negative pregnancy result or until 10 weeks of pregnancy. Initial pregnancy was defined by a rise in serum hcg levels between days 14 and 16 after oocyte retrieval. Biochemical pregnancy was defined as a positive serum β-hcg test result with no gestational sac identified on sonography. Miscarriage was defined after a gestational sac, an embryonal heart rate, or both were detected on sonography at 6 weeks gestation. Ongoing pregnancy was defined as pregnancy continuing beyond 28 weeks gestation. None of the cycles were cancelled. Of the 137 cycles that fulfilled our inclusion criteria, 2 cycles that ended in ectopic pregnancy and 6 that had day 5 embryo transfer were excluded from our analyses. Similarly, 14 cycles in which patients received vaginal estrogen starting on the day of oocyte retrieval to improve a thin endometrial stripe were excluded. Main outcome measures were the relationship of endometrial dynamics with the duration of ovarian stimulation, the duration of ganirelix use (days), the serum estradiol level throughout ovarian stimulation, and the pregnancy outcomes (positive pregnancy test result, biochemical pregnancy, implantation, miscarriage, and ongoing pregnancy). The χ 2 test was used to assess whether endometrial thickness dynamics or morphologic characteristics would influence pregnancy outcomes. The t test was used to assess whether the endometrial thickness on the ganirelix start day, hcg trigger day, or embryo transfer day could assess pregnancy outcomes. Analysis of variance was used to assess whether endometrial thickness dynamics had any relationship with the exposure to estradiol (expressed as the area under the curve during the ovarian stimulation) Figure 1. Endometrial pattern assessed by transabdominal sonography at the time of embryo transfer. A, Trilaminar. B, Homogeneously solid. A B J Ultrasound Med 2008; 27:

4 Endometrial Thickness in Assisted Reproductive Technology Cycles Table 1. Endometrial Thickness Dynamics During Ovarian Stimulation Until the hcg Day in the 115 Cycles Pattern Description Explanation 0 No change EMT starts low then does not change (variation <1.5 mm up or down) 1 Low-up EMT starts low then increases 2 Low-up-low EMT starts low then increases then decreases again 3 Low-lower EMT starts low then decreases 4 Low-lower-up EMT starts low then decreases then increases EMT indicates endometrial thickness. or with the estradiol peak. Spearman correlations were used to assess whether endometrial thickness development depended on the duration of ovarian stimulation or the duration of ganirelix use. The SPSS version 15.0 statistical package for Windows (SPSS Inc, Chicago, IL) was used, considering a significance level of P <.05. Results Baseline characteristics of 115 cycles (99 patients) are reported in Table 2. Endometrial thickness dynamics depicted by the 5 categories in Table 1, as well as the more basic 3 categories, did not show a significant relationship with any of the pregnancy outcomes as end points (positive pregnancy test result, biochemical pregnancy, Table 2. Characteristics of the Study Population (n = 115 Cycles, 99 Patients) Variable Mean ± SD Range Age, y 34 ± Duration of stimulation, d 11 ± GnRH antagonist start, d 7 ± Duration of GnRH antagonist, d 5 ± Total FSH dose, IU 5037 ± ,700 Estradiol on hcg day, pg/ml 2474 ± Endometrial stripe on hcg day, mm 10 ± Oocytes retrieved, n 15 ± Mature oocytes, n 12 ± Fertilized oocytes, n 9 ± Embryos transferred, n 3 ± Total positive pregnancy test results, n (%) 61/115 (53) NA Implantation rate, % a 23 NA Total miscarriages, n (%) 6/115 (5) NA Total biochemical pregnancies, n (%) 12/115 (10) NA Total ongoing pregnancies, n (%) 43/115 (37) NA FSH indicates follicle-stimulating hormone; and NA, not applicable. a Number of gestational sacs/number of embryos transferred 100. miscarriage, implantation, and ongoing pregnancy). However, 37% of the cases had either a decreasing endometrial thickness or no change on the hcg trigger day compared with the ganirelix start day. Of note, the serum estradiol level increased steadily for the average administration duration of 5 days (range, 3 7 days), with an average increase ± SD of 1713 ± 874 pg/ml (range, pg/ml) in the 115 cases. Additionally, endometrial thickness dynamics were not correlated with either the duration of ovarian stimulation or the duration of ganirelix use. The endometrial thickness dynamics did not show an association with the extent of estrogenization during ovarian stimulation, as expressed by the peak serum estradiol level or the estradiol area under the curve. Similarly, a single endometrial thickness measurement on the ganirelix start day, hcg trigger day, or embryo transfer day had no significant relationship with any of the pregnancy outcomes. Only a trilaminar endometrial pattern on the embryo transfer day had a significant positive relationship with a positive pregnancy test result (P =.002), implantation (P =.001), and ongoing pregnancy (P =.001; Figure 2). Interestingly, the trilaminar pattern had a significant inverse relationship with the duration of stimulation in days (P =.02; Figure 3). If the ovarian stimulation lasted more than 11 days, the likelihood of a trilaminar endometrial pattern was almost nil. However, it was not related to the duration of ganirelix use, with the starting day of ganirelix administration, or with the endometrial thickness measurement on the transfer day. Discussion In our cohort of IVF/ICSI cycles downregulated with GnRH antagonists, a decrease or no change in endometrial thickness was noted in 37% of the patients on the day of the hcg trigger compared with the beginning of ganirelix administration. Despite this, endometrial dynamics through the ovulation induction cycle did not provide significant prognostic information with regard to the outcomes of the cycles. Additionally, the endometrial morphologic pattern seemed to be more important than its thickness or changes in thickness in predicting the 1594 J Ultrasound Med 2008; 27:

5 Detti et al pregnancy outcomes: a trilaminar endo metrium on the day of embryo transfer was positively related to a positive pregnancy test result, implantation, and ongoing pregnancy. In all of the previous reports on progressive endometrial changes during ovarian stimulation cycles downregulated with GnRH agonists, there was a net increase in endometrial thickness. 4 9 In our cohort of cycles downregulated with GnRH antagonists, there was a decrease in endometrial thickness in almost half of the cases despite a steady daily increase in the serum estradiol level. Our outcomes find substantiation in in vitro studies in which GnRH antagonists have been shown to directly influence the development of extrapituitary tissues, including the endometrium In these, endometrial thickness was not measured, but endometrial tissue exposed to GnRH antagonists had increased apoptosis and decreased expression of endometrial growth factors. This combination of effects would lead to decreased endometrial receptivity The halflife of GnRH antagonists has been reported to be about 30 hours. 19 Therefore, a residual effect on the endometrium could still persist at the time of embryo transfer (performed on day 3 after retrieval) and maybe even later, at the time of implantation. Nonetheless, in our study we could not find a relationship between the degree of endometrial development and pregnancy outcomes. In our study, the trilaminar endometrial pattern was negatively related to the duration of ovarian stimulation, as reported in Figure 3. If the ovarian stimulation lasted more than 11 days, the occurrence of a trilaminar pattern was unlikely. Notwithstanding, the morphologic pattern did not have a relationship with the duration of ganirelix use or with its starting day. Our results confirm 2 older studies by Noyes et al 5 and Bohrer et al, 20 in which the authors evaluated women s endometrial thicknesses during comparable ovarian stimulations with gonadotropins. In the first study, 516 IVF cycles were prospectively evaluated for endometrial thickness and pattern, and the authors concluded that a trilaminar morphologic pattern and an endometrium thicker than 9 mm were more favorable for successful pregnancy outcomes. 5 Similarly, of the 175 patients evaluated in the second study, pregnancy Figure 2. Relationship of the endometrial pattern on embryo transfer day with pregnancy outcomes. Preg. indicates pregnancy; and SAB, miscarriage. *P <.05. occurred in 2.9% of women with a homogeneous endometrial pattern and 23% of those with a trilaminar pattern. 20 The authors concluded that in patients receiving gonadotropins, a homogeneous pattern is a bad prognostic sign, regardless of endometrial thickness. The characteristic histologic appearance of a trilaminar versus homogeneous endometrium has not been delineated in the human. In vitro studies performed during the implantation period in animals did not contemplate the sono- Figure 3. Relationship of the endometrial pattern on the embryo transfer day with the duration of ovarian stimulation with gonadotropins. GE indicates greater than or equal to; and LT, less than. *P <.05. J Ultrasound Med 2008; 27:

6 Endometrial Thickness in Assisted Reproductive Technology Cycles graphic appearance of the endometrium. 21 Notwithstanding, these data fail to correlate endometrial parameters with ongoing pregnancy, and the link between endometrial morphologic characteristic and function has not been elucidated yet. In conclusion, sonographic morphologic evaluation of the endometrium, albeit limited by difficulties in interpretation, appears to be the most promising evaluation technique for predicting IVF/ICSI cycle outcomes. References 1. US Department of Health and Human Services, Centers for Disease Control and Prevention Assisted Reproductive Technology (ART) Report. Atlanta, GA: Centers for Disease Control and Prevention; Fleischer AC, Herbert CM, Sacks GA, Wentz AC, Entman SS, James AE Jr. Sonography of the endometrium during conception and nonconception cycles of in vitro fertilization and embryo transfer. Fertil Steril 1986; 46: Kovacs P, Matyas S, Boda K, Kaali SG. The effect of endometrial thickness on IVF/ICSI outcome. Hum Reprod 2003; 11: Gonen Y, Casper RF, Jacobson W, Blankier J. Endometrial thickness and growth during ovarian stimulation: a possible predictor of implantation in in vitro fertilization. Fertil Steril 1989; 52: Noyes N, Liu HC, Sultan K, Schattman G, Rosenwaks Z. Endometrial thickness appears to be a significant factor in embryo implantation in in-vitro fertilization. Hum Reprod 1995; 10: Lebovitz Z, Grinin V, Rabia R, et al. Assessment of endometrial receptivity for gestation in patients undergoing in vitro fertilization, using endometrial thickness and the endometrium-myometrium relative echogenicity coefficient. Ultrasound Obstet Gynecol 1999; 14: De Geyter C, Schmitter M, De Geyter M, Nieschlag E, Holzgreve W, Schneider HP. Prospective evaluation of the ultrasound appearance of the endometrium in a cohort of 1,186 infertile women. Fertil Steril 2000; 73: Bassil S. Changes in endometrial thickness, width, length and pattern in predicting pregnancy outcome during ovarian stimulation in in vitro fertilization. Ultrasound Obstet Gynecol 2001; 18: Borm G, Mannaerts B. Treatment with the gonadotrophinreleasing hormone antagonist ganirelix in women undergoing ovarian stimulation with recombinant follicle stimulating hormone is effective, safe and convenient: results of a controlled, randomized, multicentre trial. The European Orgalutran Study Group. Hum Reprod 2000; 15: Lindheim SR, Morales AJ. GnRH antagonists followed by a decline in serum estradiol results in adverse outcomes in donor oocyte cycles. Hum Reprod 2003; 18: Al-Inany HG, Abou-Setta AM, Aboulghar M. Gonadotropinreleasing hormone antagonists for assisted conception. Cochrane Database Syst Rev 2006; 3:CD Zikopoulos K, Kolibianakis EM, Camus M, et al. Fertil Steril 2004; 81: Detti L, Ambler RD, Yelian FD, Kruger M, Diamond MP, Puscheck EE. Timing and duration of use of GnRH antagonist down-regulation for IVF/ICSI cycles have no impact on oocyte quality or pregnancy outcomes. J Assist Reprod Genet 2008; 25: Hernandez E. Embryo implantation and GnRH antagonists: embryo implantation the Rubicon for GnRH antagonists. Hum Reprod 2000; 15: Luo X, Xu J, Chegini N. Gonadotropin releasing hormone analogue (GnRHa) alters the expression and activation of Smad in human endometrial epithelial and stromal cells. Reprod Biol Endocrinol 2003; 16: Meresman GF, Bilotas MA, Lombardi E, Tesone M, Sueldo C, Barañao RI. Effect of GnRH analogues on apoptosis and release of interleukin-1beta and vascular endothelial growth factor in endometrial cell cultures from patients with endometriosis. Hum Reprod 2003; 18: Duijkers IJ, Klipping C, Willemsen WN, et al. Single and multiple dose pharmacokinetics and pharmacodynamics of the gonadotrophin-releasing hormone antagonist Cetrorelix in healthy female volunteers. Hum Reprod 1998; 13: Bohrer MK, Hock DL, Rhoads GG, Kemmann E. Sonographic assessment of endometrial pattern and thickness in patients treated with human menopausal gonadotropins. Fertil Steril 1996; 66: Bourgain C, Devroey P. Histologic and functional aspects of the endometrium in the implantatory phase. Gynecol Obstet Invest 2007; 64: McWilliams GD, Frattarelli JL. Changes in measured endometrial thickness predict in vitro fertilization success. Fertil Steril 2007; 88: Fluker M, Grifo J, Leader A, et al; North American Ganirelix Study Group. Efficacy and safety of ganirelix acetate versus leuprolide acetate in women undergoing controlled ovarian hyperstimulation. Fertil Steril 2001; 75: J Ultrasound Med 2008; 27: