Anti-Müllerian hormone testing is useful for individualization of stimulation protocols in oocyte donors

Transcription

1 Reproductive BioMedicine Online (2010) 20, ARTICLE Anti-Müllerian hormone testing is useful for individualization of stimulation protocols in oocyte donors Gary S Nakhuda *, Nataki C Douglas, Melvin H Thornton, Michael M Guarnaccia, Rogerio Lobo, Mark V Sauer Columbia University College of Physicians and Surgeons, Division of Reproductive Endocrinology and Infertility, Centre for Women s Reproductive Care, 1790 Broadway, New York, NY 10019, USA * Corresponding author. address: gsn16@columbia.edu (GS Nakhuda). Dr Nakhuda is an assistant clinical professor in obstetrics and gynecology at Columbia University, where he completed both his residency in obstetrics and gynecology and his fellowship in reproductive endocrinology and infertility. His current research interests include assessment of ovarian reserve, HIV and assisted reproduction, and clinical embryology. Abstract While the age of a donor is a fundamental factor to the success of donor IVF, no serum markers have been demonstrated to be useful in predicting variability of ovarian response in individual donors. Anti-Müllerian hormone (AMH) has been described as an accurate marker of ovarian response in patients undergoing IVF, but has not been applied to oocyte donors. AMH concentrations from 104 anonymous oocyte donors between the ages of years were studied and IVF outcome parameters compared. AMH was correlated with several parameters including the number of oocytes retrieved (r = 0.232, P = 0.024), the peak oestradiol concentrations (r = 0.235, P = 0.024) and the need to decrease gonadotrophin dose in order to avoid ovarian hyperstimulation syndrome (r = 0.274, P = 0.007). Receiver operating curve analysis was able to identify an AMH threshold that rendered about 70% sensitivity and 70% specificity for predicting the need to decrease gonadotrophin dosing. The clinical pregnancy rate was 77% per recipient and was not related to the donors AMH concentrations. For oocyte donors, measurement of AMH appears most useful for determining gonadotrophin sensitivity in order to mitigate symptoms consistent with ovarian hyperstimulation. RBMOnline ª 2009, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. KEYWORDS: anti-müllerian hormone, donor IVF, in-vitro fertilization, Müllerian inhibiting substance, oocyte donation, ovarian reserve Introduction The age of the egg donor is fundamentally important to the pregnancy success experienced by recipients of oocyte donation IVF. Numerous investigators have examined various age thresholds in identifying optimal donors, with most authorities agreeing that donors younger than 35 years are ideal (Cohen et al., 1999; Faber et al., 1997). Pregnancy /$ - see front matter ª 2009, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. doi: /j.rbmo S88 Reprinted from Vol. 20, No. 1 (2010) pp

2 AMH testing in oocyte donors 43 rates resulting from oocyte donation are consistently higher than autologous egg IVF, presumably partly because the oocytes come from healthy young women with no history of infertility. Even when oocyte yields are low, recipient pregnancy rates remain robust, ostensibly because of the high quality of the oocytes retrieved (Letterie et al., 2005). Nevertheless, even in healthy young oocyte donors, there is a great deal of variation in ovarian response to gonadotrophin stimulation. The extremes of ovarian response, both hypoand hyper-, do not appear to be predictable in the young donors using current clinical techniques or demographics. The availability of a marker that can predict the degree of ovarian response in oocyte donors prior to undertaking stimulation would be very useful in selecting the most optimal donor candidates and potentially lessen the associated untoward outcomes inherent to poor response as well as ovarian hyperstimulation syndrome (OHSS), which jeopardize the outcome of donor IVF cycles. In recent years, anti-müllerian hormone (AMH), also known as Müllerian inhibiting substance (MIS), has been described as an accurate marker of ovarian response to gonadotrophin stimulation in patients undergoing IVF. AMH is a 170 kda growth factor from the transforming growth factor b family that is produced by small antral follicles in women of reproductive age. Numerous studies have convincingly demonstrated that AMH accurately predicts poor response, positively correlates with IVF outcomes, such as the number of oocytes retrieved and peak serum oestradiol, and identifies patients at risk of OHSS (Hazout et al., 2004; Lee et al., 2008; Nakhuda et al., 2006; Seifer et al., 2002; van Rooij et al., 2002). A few studies have also reported that AMH predicts oocyte and embryo quality, as well as pregnancy and live birth rates (Ebner et al., 2006; Nelson et al., 2007; Silberstein et al., 2006). Measurement of AMH has been primarily utilized to assess ovarian reserve in patients seeking infertility treatment, but its use in the management of oocyte donors has not been elucidated. To this end, baseline AMH values were compared with donor IVF outcomes in a group of young oocyte donors in order to determine if measurement of AMH may be clinically useful in this context. Materials and methods In order to qualify to participate as anonymous oocyte donors, candidates had to be between the ages of 21 and 32 years and cleared by routine psychological, medical, reproductive and genetic risk-factor-based screening. All donors were normoovulatory, with bilateral ovaries and without evidence of polycystic appearance. Serum AMH was retrospectively measured from serum samples taken at the time of physical examination using a commercially available assay (DSL; Webster, TX, USA). Samples were collected between June 2007 and June Columbia University institutional review boards approved the investigational protocol. These samples were taken irrespective of the phase of the menstrual cycle or use of oral contraceptives. Baseline FSH and oestradiol were not measured since these parameters do not consistently correlate with ovarian reserve in young patients (Barnhart and Osheroff, 1999; Esposito et al., 2002). Antral follicle count is not reported because ultrasounds were performed by numerous physicians, thus consistency and standardization are uncertain. For reasons of consistency, all donors included in this study were started on an identical IVF protocol: down-regulation with 20 U of subcutaneous leuprolide acetate (Lupron; Abbott Laboratories, Illinois, USA) followed by gonadotrophin stimulation with an initial daily dosage of 150 U of intramuscular recombinant FSH (Gonal F; Serono Biotech, Geneva, Switzerland) plus 75 U human menopausal gonadotrophin (Repronex; Ferring Pharmaceuticals, Suffern, NY, USA). Dosage decreases were made according to physician discretion, based on serum oestradiol or follicular response; for example, if serum oestradiol exceeded 500 pg/ml on day 5 of stimulation. Similarly dosage increases were also made per physician judgment; for example if the serum oestradiol after 5 days of stimulation was less than 100 pg/ml or if fewer than five growing follicles were apparent. Decisions to cancel cycles were also made by physician discretion, but the primary reasons were the risk of impending OHSS or the probability of poor response (fewer than five growing follicles by day 8 of gonadotrophin stimulation). Final follicular maturation was induced via intramuscular administration of human chorionic gonadotrophin (Novarel; Ferring); 5000 U were prescribed if the oestradiol on the day of administration was greater than or equal to 3500 pg/ml, or 10,000 U were prescribed if the oestradiol was below 3500 pg/ml. The method of fertilization was indicated by semen analysis parameters, and embryo transfer occurred on day 3 or 5 depending on embryo quality. High-quality embryos that were not transferred were cryopreserved. Recipients were synchronized for embryo transfer using leuprolide acetate (depending on menopausal status), oral oestradiol and vaginal progesterone suppositories as previously described (Klein and Sauer, 2002). Ongoing viable pregnancies and live births were included in the clinical pregnancy rate. Statistical analysis was performed with the Statistical Package for Social Sciences (SPSS) version 16.0 for Mac (SPSS, USA). Data were evaluated as appropriate with Pearson regression coefficient, Kruskal Wallis comparison of means and receiver operator characteristics (ROC) and independent samples t-test. Results Oocyte donor characteristics and cycle parameters are summarized in Table 1. Recipient and outcome parameters are summarized in Table 2. A total of 104 cycles were initiated with unique donors, four of whom were cancelled during stimulation. One cycle was cancelled due the risk of impending hyperstimulation: a 26-year-old donor with a baseline AMH of 5.11 ng/ml with an oestradiol of 7970 pg/ ml on day 7 of stimulation. Three donors were cancelled due to a risk of unsatisfactory response, with fewer than five maturing follicles by day 8: a 24-year-old with an AMH of 1.87 ng/ml, a 23-year-old with an AMH of 2.19 ng/ml and a 32-year-old with an AMH of 0.26 ng/ml. The 100 remaining recipients underwent a total of 117 transfers (100 fresh +17 frozen). The clinical pregnancy rate Reprinted from Vol. 20, No. 1 (2010) pp S89

3 44 GS Nakhuda et al. Table 1 Characterisitic Table 2 Ooctye and donor cycle characteristics. Recipient cycle characteristics and outcomes. Characteristic/outcome Value Donor age (years) 25.5 ± 0.29 ( ) BMI (kg/m 2 ) ± 0.17 ( ) AMH (ng/ml) 3.17 ± 0.20 ( ) Peak oestradiol (pg/ml) 3824 ± ( ) Cumulative 2025 ± 62.7 ( ) gonadotrophin dose a Total oocytes 21 ± 0.92 ( ) retrieved (n) Values are mean ± SE (95% CI). AMH = anti-müllerian hormone; BMI = body mass index. a Data from Lee et al. (2008). Value Recipient age (years) 43.6 ± 0.50 ( ) ICSI 60/100 (60.0) Embryos transferred 2.16 ± 0.06 ( ) Clinical pregnancy/ 77/117 (65.8) transfer (fresh + frozen) Spontaneous abortion 9/117 (7.7) Ectopic pregnancy 1/117 (0.85) Multiple pregnancy 33/77 (42.9) Total clinical pregnancy 77/100 (77.0) rate/patient Values are number/total (percentage) or mean ± SE (95% CI). per patient was 77% when fresh and frozen transfers were included (70/100 fresh and 7/17 frozen). The median number of embryos transferred was two. Of the fresh transfers, 30 occurred on day 3 and 70 transfers occurred on day 5. Additional embryos available for cryopreservation were obtain from 67% (67/100) of the retrievals. The multiple pregnancy rate was 42.9% (33/77), of which 90.9% (30/33) were twins, while three were triplet pregnancies. Table 3 summarizes the statistical relationships of AMH and other variables. The distribution of AMH was skewed such that the mean value was 3.17 ng/ml and the median was 2.59 ng/ml; however, 70% (73/104) of AMH values were included within 1 standard deviation of the mean. The standard deviation was 2.03 ng/ml and the standard error was 0.20 ng/ml. The values ranged from 0.26 to 10.4 ng/ml. AMH was not significantly associated with donor age or body mass index (BMI). AMH was significantly correlated with the peak oestradiol (r = 0.235, P = 0.024) and the number of oocytes retrieved (r = 0.232, P = 0.024). AMH had a significant inverse correlation with the total dose of gonadotrophin used per cycle (r = 0.35, P = 0.05) and the gonadotrophin sensitivity, as calculated by the ratio of peak oestradiol to total gonadotrophin dose (r = 0.357; P = 0.045). Notably, neither donor age nor BMI was related to peak oestradiol, total number of oocytes retrieved, gonadotrophin sensitivity or AMH (data not shown). Predictably, peak oestradiol was correlated with the number of oocytes retrieved (r = 0.451, P < 0.001). AMH and outcome parameters were analysed in relation to the need to adjust daily gonadotrophin dose based on the intracycle parameters. Of the 104 initiated cycles, 56 required no dosage adjustment (Group N), 37 required a decrease in dose (Group D) and 11 required increase in dose (Group I). The Kruskal Wallis test was used for multiple comparisons because of the small sample size of Group I. Age and BMI were statistically similar between all three groups. The AMH values were significantly different between all groups: Group N, 3.04 ± 0.31; D, 3.94 ± 0.34; Group I, 1.89 ± 0.19; P = (Kruskal Wallis). The peak oestradiol concentrations and total number of oocytes retrieved were also significantly different between all groups (P < 0.001). The clinical pregnancy rate in all groups (live birth and ongoing) was statistically similar. Table 4 summarizes these relationships. No donor in this cohort suffered severe OHSS that required inpatient or outpatient treatment. However, numerous donors had surrogate markers that are considered risk factors for developing OHSS, such as a high peak oestradiol higher than 3500 pg/ml (46/100) and more than 20 oocytes retrieved (54/100). Using ROC analysis, an AMH of 2.94 ng/ml rendered approximately 70% sensitivity and 70% specificity in predicting the need to decrease gonadotrophin dose. Of the 46% (46/100) of donors with an AMH greater than the threshold determined by ROC analysis of 2.94 ng/ml, 54% (25/46) had peak oestradiol concentrations higher than 3500 pg/ml and 65% (30/46) had more than 20 oocytes retrieved. Conversely, of the 54% (54/100) of donors with an AMH concentration less than the 2.94 ng/ml, 39% (21/54) had peak oestradiol concentrations higher than 3500 pg/ml and 44% (24/54) had more than 20 oocytes retrieved. Using the cut-point of 2.94 ng/ml for AMH, comparison of high AMH versus low AMH groups showed a significant difference in the tendency for a peak oestradiol higher than 3500 pg/ml (4306 versus 3417 pg/ml (mean difference (95% CI ), P = (t-test)), but no difference in the tendency to have more than 20 oocytes retrieved (23.4 versus 20.3). Table 3 Association of AMH to donor and outcome parameters. Donor age Donor BMI Peak oestradiol Total oocytes Total gonadotrophin dose r (Pearson) P-value NS NS BMI = body mass index; NS = not statistically significant. S90 Reprinted from Vol. 20, No. 1 (2010) pp

4 AMH testing in oocyte donors 45 Neither AMH, nor any other study parameter was associated with pregnancy outcome. There was no difference in clinical pregnancy rates between ICSI (n = 60) and insemination (n = 40) groups (t-test). Of note, there were no differences in clinical pregnancy rates for day 3 compared with day 5 transfers (63.3% versus 71.0%) (t-test). However, when day 3 transfers were performed, the mean number of embryos transferred was a slightly greater when compared with day 5 [2.4 versus 2.08; P = (t-test)]. The multiple pregnancy rate was not related to day of transfer. The availability of embryos for cryopreservation was statistically associated with the total number of oocytes retrieved [r = 0.217; P = (t-test)] but not with AMH. Discussion When young women volunteer for oocyte donation, robust responses to gonadotrophin stimulation can be generally expected based on the association of youth with good ovarian reserve. However, because individual variation is common, an independent predictor of gonadotrophin response would be helpful. To this end, AMH, a marker of both poor and excessive ovarian response in infertile populations, was studied as a potentially predictive marker for performance in a cohort of oocyte donors. Unlike the general infertility population, oocyte donors comprise a relatively homogenous group, given the uniformity of age and the lack of conditions associated with infertility. Furthermore, when only eumenorrhoeic women are used as donors, ovulatory disorders, such as polycystic ovarian syndrome, that may complicate ovarian response are minimized. Despite these features, variability in gonadotrophin response, belied by the homogeneity of the population, is not uncommon. According to the data, some of the variability may be predicted by the measurement of AMH prior to stimulation. Consistent with numerous reports that have examined the association of AMH with IVF outcome parameters, in donors in the present study, AMH significantly correlated with the degree of gonadotrophin sensitivity (r = 0.357; P = 0.045), peak oestradiol concentration (r = 0.235, P = 0.024) and the number of oocytes retrieved (r = 0.232, P = 0.024). Thus, even in this young population in whom other tests of ovarian reserve may not be applicable (Esposito et al., 2002), AMH does appear to have some prognostic relevance. One of the fundamental principles in the management of egg donors is the prevention of OHSS. Because donors may be predisposed to OHSS based on their age, vigilant monitoring of developing signs and symptoms is routine and dosage decreases must be made liberally in order to reduce the incidence of this complication. In this series, 53.8% (56/ 104) of initiated cycles did not require a change of the daily gonadotrophin dose, 35.6% (37/104) required a decrease in dose based on clinical suspicion of impending hyperstimulation and 10.6% (11/104) required an increase in dose based on a low response. AMH was significantly different between these three groups: highest in the group that required a dosage reduction, lowest in those that required an increase and intermediate in the group that required no dosage change. Peak oestradiol concentration, total oocytes retrieved and gonadotrophin sensitivity also significantly varied in a similar trend (Table 4). Using the need for dosage decrease as the dependent condition and AMH as a predictive variable, an ROC curve rendered approximately 70% specificity and 70% sensitivity using an AMH threshold of 2.94 ng/ml. This cut-off of 2.94 ng/ml was used post hoc to determine if features consistent with OHSS such as peak oestradiol concentration higher than 3500 pg/ml or more than 20 oocytes retrieved could be predicted. Compared with donors with lower AMH concentrations, those with concentrations higher than 2.94 ng/ml were more likely to have a peak oestradiol concentration higher than 3500 pg/ml (P = 0.008); however, the tendency to have more than 20 oocytes retrieved was not different using this AMH threshold. At least two prior studies have reported that AMH may predict a risk for OHSS in patients undergoing IVF, although neither study included donors (Lee et al., 2008; Nakhuda et al., 2006). Because no donor required in- or out-patient treatment for OHSS, these surrogate markers may not necessarily correlate with clinical outcomes in this context. The predictability of poor response in donor IVF cycles based on AMH is questionable. In general IVF patients, especially those of advanced age, the prevalence of poor response is relatively high and AMH has demonstrated strong prognostic accuracy for poor response (Ficicioglu et al., 2006; Nakhuda et al., 2007). In the current series of donors, Table 4 Need for dosage adjustment in relation to cycle characteristics. Age BMI AMH Peak oestradiol (pg/ml) Total no. of oocytes retrieved Clinical pregnancy rate (%) a Increased ± ± ± ± ± 2.5 7/11 (63.63) dose ( ) ( ) ( ) ( ) ( ) No dose ± ± ± ± ± /56 (66.07) change ( ) ( ) ( ) ( ) ( ) Decreased ± ± ± ± ± /37 (70.27) dose ( ) ( ) ( ) ( ) ( ) P-value b NS NS <0.001 <0.001 NS Data are mean ± SE (95% CI) unless otherwise stated. BMI = body mass index; NS = not statistically significant. a Fresh pregnancy rate per initiated cycle. b Kruskal Wallis. Reprinted from Vol. 20, No. 1 (2010) pp S91

5 46 GS Nakhuda et al. only three cycles were cancelled for poor response determined by less than five maturing follicles by day 8 of stimulation, despite gonadotrophin dose increases. Two of these donors had AMH concentrations (1.87 and 2.19 ng/ml) within one standard deviation of the mean for the donor population (3.07 ng/ml ± 2.03). One donor did have a notably lower AMH (0.26 ng/ml). Because the incidence of poor response in this series is low, it is difficult to determine if AMH is predictive in this regard. However, given the normal AMH concentrations in two of the three cancelled donors, the sensitivity for poor response in this population may not be as high compared with general IVF populations. Pregnancy rates in recipients from this series were excellent, regardless of any independent variable examined, likely due to the strict age limitations in the donors. The cumulative experience with oocyte donation supports the principal that donor age is the main determinant of successful pregnancy, thus restricting the age requirement between the ages of 21 and 32 years is most likely instrumental in assuring optimal outcomes. Even donor IVF cycles with relatively low oocyte yields are known to lead to high pregnancy rates (Letterie et al., 2005), a fact which was corroborated in this series, where clinical pregnancy rate did not correlate with the number of oocytes retrieved. Thus, while AMH may help in the management of the donor cycles in terms of potentially mitigating excessive response, it does not appear to be useful in predicting pregnancy in this context. Studying AMH in the donor population provides unique insight into the value of this marker. Unlike the general infertility population, donors are more homogeneous in several important characteristics such as age, BMI and general health status. Because AMH is consistently higher in patients with PCOS (Cook et al., 2002; Fallat et al., 1997), only eumenorrhoeic donors were included in this series. Standard baseline FSH and oestradiol were not considered in this series, as these tests are known to add little value to the diagnosis of young patients (Barnhart and Osheroff, 1999; Esposito et al., 2002). AMH was sampled randomly in respect to phase of menstrual cycle and oral contraceptive status, as several investigators have demonstrated that AMH is not significantly affected by these variables (Hehenkamp et al., 2006; Somunkiran et al., 2007; Tsepelidis et al., 2007). While antral follicle count (AFC) has been shown to have a high degree of correlation with AMH (de Vet et al., 2002), this was omitted from the current data due to the inconsistency of measurement but multiple investigators. Recently, investigators demonstrated that AFC could be used for screening oocyte donors in a manner similar to the current study of AMH (Barreto Melo et al., 2009). The inclusion of a standardized AFC in this study may have provided a valuable comparison to AMH as a predictor of ovarian response in oocyte donors. In conclusion, while the measurement of AMH in oocyte donors requires further attention, it is believed that it may have a potential role in donor screening. It may be most useful for predicting the degree of gonadotrophin sensitivity, allowing individualization of dosing protocols. While all oocyte donors may be at risk for hyperstimulation based on young age, AMH may identify those who may be at a particular risk. In this population, AMH appears less useful for predicting poor response and pregnancy outcome. References Barnhart, K., Osheroff, J., We are overinterpreting the predictive value of serum follicle-stimulating hormone levels. Fertil. Steril. 72, 8 9. Barreto Melo, M.A., Garrido, N., Alvarez, C., et al., Antral follicle count (AFC) can be used in the prediction of ovarian response but cannot predict the oocyte/embryo quality or the in vitro fertilization outcome in an egg donation program. Fertil. Steril. 91, Cohen, M.A., Lindheim, S.R., Sauer, M.V., Donor age is paramount to success in oocyte donation. Hum. Reprod. 14, Cook, C.L., Siow, Y., Brenner, A.G., Fallat, M.E., Relationship between serum mullerian-inhibiting substance and other reproductive hormones in untreated women with polycystic ovary syndrome and normal women. Fertil. Steril. 77, de Vet, A., Laven, J.S., de Jong, F.H., et al., Antimullerian hormone serum levels: a putative marker for ovarian ageing. Fertil. Steril. 77, Ebner, T., Sommergruber, M., Moser, M., et al., Basal level of anti-mullerian hormone is associated with oocyte quality in stimulated cycles. Hum. Reprod. 21, Esposito, M.A., Coutifaris, C., Barnhart, K.T., A moderately elevated day 3 FSH concentration has limited predictive value, especially in younger women. Hum. Reprod. 17, Faber, B.M., Mercan, R., Hamacher, P., et al., The impact of an egg donor s age and her prior fertility on recipient pregnancy outcome. Fertil. Steril. 68, Fallat, M.E., Siow, Y., Marra, M., et al., Mullerian-inhibiting substance in follicular fluid and serum: a comparison of patients with tubal factor infertility, polycystic ovary syndrome, and endometriosis. Fertil. Steril. 67, Ficicioglu, C., Kutlu, T., Baglam, E., Bakacak, Z., Early follicular antimullerian hormone as an indicator of ovarian reserve. Fertil. Steril. 85, Hazout, A., Bouchard, P., Seifer, D.B., et al., Serum antimullerian hormone/mullerian-inhibiting substance appears to be a more discriminatory marker of assisted reproductive technology outcome than follicle-stimulating hormone, inhibin B, or oestradiol. Fertil. Steril. 82, Hehenkamp, W.J., Looman, C.W., Themmen, A.P., et al., Anti-Mullerian hormone levels in the spontaneous menstrual cycle do not show substantial fluctuation. J. Clin. Endocrinol. Metab. 91, Klein, J., Sauer, M.V., Oocyte donation. Best Pract. Res. 16, Lee, T.H., Liu, C.H., Huang, C.C., et al., Serum anti- Mullerian hormone and oestradiol levels as predictors of ovarian hyperstimulation syndrome in assisted reproduction technology cycles. Hum. Reprod. 23, Letterie, G., Marshall, L., Angle, M., The relationship of clinical response, oocyte number, and success in oocyte donor cycles. J. Assist. Reprod. Genet. 22, Nakhuda, G.S., Chu, M.C., Wang, J.G., et al., Elevated serum mullerian-inhibiting substance may be a marker for ovarian hyperstimulation syndrome in normal women undergoing in vitro fertilization. Fertil. Steril. 85, Nakhuda, G.S., Sauer, M.V., Wang, J.G., et al., Mullerian inhibiting substance is an accurate marker of ovarian response in women of advanced reproductive age undergoing IVF. Reprod. BioMed. Online 14, Nelson, S.M., Yates, R.W., Fleming, R., Serum anti-mullerian hormone and FSH: prediction of live birth and extremes of response in stimulated cycles implications for individualization of therapy. Hum. Reprod. 22, S92 Reprinted from Vol. 20, No. 1 (2010) pp

6 AMH testing in oocyte donors 47 Seifer, D.B., MacLaughlin, D.T., Christian, B.P., et al., Early follicular serum mullerian-inhibiting substance levels are associated with ovarian response during assisted reproductive technology cycles. Fertil. Steril. 77, Silberstein, T., MacLaughlin, D.T., Shai, I., et al., Mullerian inhibiting substance levels at the time of HCG administration in IVF cycles predict both ovarian reserve and embryo morphology. Hum. Reprod. 21, Somunkiran, A., Yavuz, T., Yucel, O., Ozdemir, I., Anti- Mullerian hormone levels during hormonal contraception in women with polycystic ovary syndrome. Eur. J. Obstet. Gynecol. Reprod. Biol. 134, Tsepelidis, S., Devreker, F., Demeestere, I., et al., Stable serum levels of anti-mullerian hormone during the menstrual cycle: a prospective study in normo-ovulatory women. Hum. Reprod. 22, van Rooij, I.A., Broekmans, F.J., te Velde, E.R., et al., Serum anti-mullerian hormone levels: a novel measure of ovarian reserve. Hum. Reprod. 17, Declaration: The authors report no financial or commercial conflicts of interest. Received 3 November 2008; refereed 3 December 2008; accepted 24 September Reprinted from Vol. 20, No. 1 (2010) pp S93