Early follicular antimüllerian hormone as an indicator of ovarian reserve

Early follicular antimüllerian hormone as an indicator of ovarian reserve Cem Fiçiciogˇlu, M.D., Tayfun Kutlu, M.D., Elif Baglam, M.D., and Zeynep Bakacak, M.D. IVF Clinic, Zeynep Kamil Women’s and Children’s Hospital, Istanbul, Turkey

Objective: To determine the predictive value of antimüllerian hormone (AMH) as a marker for ovarian reserve and to compare its value with the markers currently being used. Design: Prospective analysis. Setting: In vitro fertilization (IVF) clinic of a tertiary medical center. Patient(s): Fifty women undergoing assisted reproduction cycles. Intervention(s): None. Main Outcome Measure(s): Comparison of day-3 serum AMH levels among women with less than five retrieved oocytes and five or more oocytes. Antral follicle count, mature oocyte count, age, basal follicle-stimulating hormone (FSH), estradiol (E2), maximum serum E2 levels, and pregnancy success were also compared. Result(s): The mean serum AMH levels of patients with more than five retrieved oocytes were found to be higher (0.67 ⫾ 0.41 vs. 0.15 ⫾ 0.11 pg/mL). Mature oocyte counts, antral follicle counts, and maximum E2 levels were found to be statistically significantly different in the two groups despite similar ages and levels of basal FSH and E2. Although the receiver operator characteristics analysis revealed that the most sensitive and specific indicator of ovarian reserve is the level of AMH, it does not indicate pregnancy success as well when 0.25 pg/mL is taken as a cut-off value. Conclusion(s): These data demonstrate an association between early follicular serum AMH and ovarian response, but no association with pregnancy success. (Fertil Steril威 2006;85:592– 6. ©2006 by American Society for Reproductive Medicine.) Key Words: Serum antimüllerian hormone (AMH), ovarian reserve

Careful evaluation of patients and proper treatment with right techniques are essential for successful outcome of assisted reproduction. To obtain satisfactory results, it is necessary to assess ovarian reserve before planning treatment. Contemporary markers for ovarian reserve include age, basal follicle-stimulating hormone (FSH) levels, and estradiol (E2) levels (1, 2). Antral follicle count, serum inhibin B levels, ovarian volume, and vascular resistance have also been studied as markers of ovarian reserve (3–7). Provocative and dynamic tests such as a gonadotropinreleasing hormone (GnRH) agonist test and clomiphene citrate test have also been introduced recently, and more effective parameters are being sought (8, 9). Antimüllerian hormone (AMH) is a dimeric glycoprotein made up of two monomers attached to each other by disulfide bonds. The 72-kd molecule belongs to the transforming growth factor-B superfamily, which acts on tissue growth and differentiation (10). Sertoli cells in the male produce AMH, which induces the degeneration of the müllerian ducts and provides the normal formation of the male genital system. Sertoli cells’ secretion of AMH continues for a lifetime, but the significance of AMH in adult male is not known. In Received May 16, 2005; revised and accepted September 6, 2005. Reprint requests: Tayfun Kutlu, M.D., Zeynep Kamil Women’s and Children’s Hospital, Selimiye Kavak Iskele Cad. 16/7 34668 Uskudar, Istanbul, Turkey (FAX: 90-216-3377629; E-mail: [email protected]).

592

females, the granulosa cells of the ovary express AMH postnatally. Serum AMH levels are lower in females compared with males. After puberty and onset of menstrual cycles, serum AMH level decreases progressively until it becomes undetectable at around menopause (10 –13). Secreted from preantral and early antral follicles, AMH regulates ovarian activity and follicular steroidogenesis. Animal studies have revealed that not only does AMH decrease aromatase activity of FSH-stimulated granulosa cells, but it also decreases the number of leuteinizing hormone (LH) receptors, and regulates testosterone production in theca cells (14). The regulation of oocyte function is through endocrine and paracrine factors. Although the basic hormones causing oocyte growth are FSH and LH, receptors for gonadotropins could not be found on oocytes. The action of gonadotropins on oocytes is probably through mediators like epithelial growth factor (EGF), vascular endothelial growth factor (VEGF), insulinlike growth factor 2 (IGF2). The intrafollicular androgen to estrogen ratio also acts on oocyte function, and AMH plays a major role in the regulation of this ratio. In case of high androgen/estrogen levels, follicles degenerate, and when it is low, germinal vesicles rupture (15–17). The objective of our study was to demonstrate the value of AMH as an indicator of ovarian reserve in comparison with other parameters of ovarian reserve.

Fertility and Sterility姞 Vol. 85, No. 3, March 2006 Copyright ©2006 American Society for Reproductive Medicine, Published by Elsevier Inc.

0015-0282/06/$32.00 doi:10.1016/j.fertnstert.2005.09.019

MATERIALS AND METHODS This prospective study included a total of 50 women (age range: 24 to 39 years) attending the in vitro fertilization/ intracytoplasmic sperm injection (IVF/ICSI) program of Zeynep Kamil Women’s and Children’s Hospital IVF Department for the first time between January 2003 and December 2004. All of the participants in this research were asked for their consent. Our hospital has no institutional review board, so we conducted the study under the surveillance of the clinical director; none of the authors of this study had a conflict of interest. The patients’ histories and former treatments were noted. Patients underwent a general physical and gynecologic examination, spermiogram, ultrasonography on cycle day 3, and hysterosalpingography. Serologic markers (HBsAg, anti-HCV, anti-HIV, rubella-IgG, Toxo-IgG), cervicovaginal smear, cervical culture, blood and Rh type, and complete blood count were assessed. On day 3 of the menstrual cycle, each woman underwent blood sampling by venipuncture for measurement of serum levels of FSH, LH, E2, and AMH at approximately 09:00 AM. Later in the morning, ovarian ultrasound scanning was performed to evaluate the number and sizes of antral follicles and the size of ovary. On day 3 of the cycle, 5 mL of the blood drawn; the blood was centrifuged at 3500 cycles/minute for 10 minutes, and the sera stored in 1.5-mL polypropylene tubes at ⫺80°C. All stored sera were submitted to AMH determination with ACTIVE MIS/AMH ELISA DSL 10-14400 kit of Diagnostic System Laboratories (Weber, TX).

patients with other infertility causes. In two patients, cycles were canceled for economic reasons. The transfer of the embryos was performed 48 to 72 hours after the procedure. Before transfer, the embryos were evaluated microscopically to ensure six to eight cells in each embryo of grades 1 to 2. A maximum of three embryos was transferred. On the 12th to 14th days of the transfer, a serum ␤-hCG test was performed to confirm pregnancy. Statistical analysis was performed using SPSS (version 11.5; SPSS, Inc., Chicago, IL). Data were analyzed using student’s - test, Pearson chi-square, and Fisher exact tests. Predictivity of the parameters was compared with correlation analysis and receiver operator characteristics (ROC) curves. P⬍.05 was considered statistically significant. RESULTS The study included 50 patients in the IVF program. Two patients dropped out of the study for economic reasons. Four patients’ AMH levels could not be determined for technical reasons. Of the remaining 44 couples, the causes of infertility were male factor (25 couples), tubal factor (seven couples), and idiopathic (12 couples). The ages of the female patients were between 24 and 39 years. The mean infertility period was 9 ⫾ 3.9 years, and the mean body mass index of female patients was in the normal range (see Table 1). Ovarian reserve was assessed by basal hormone levels on cycle day 3, changes of these hormone levels on the day hCG was administered, and antral follicle count on ultrasound (see Table 1).

Long-protocol GnRH analogue was used for menstrual suppression. On cycle day 21, 0.5 mg of subcutaneous leuprolide acetate was begun; E2 levels less than 50 pg/mL on cycle day 2 and endometrial thickness less than 4 mm indicated down-regulation. Treatment with recombinant FSH, 150 –300 IU/day, was started. Patients with serum E2 levels less than 100 pg/mL and follicle count less than 3 on cycle day 5 were submitted to step-up protocol. Step-up or step-down protocols were decided according to subsequent ultrasound controls. In cases of less than two growing follicles or a serum E2 level of less than 300 pg/mL on cycle day 14, treatment cycle was canceled.

Treatment was started with a mean dose of 244 ⫾ 59 IU/day of recombinant FSH (150 –300 IU); a total dose of 2410 ⫾ 978 IU/day (1050 –5400 IU) was reached. A mean period of 9 ⫾ 2 days (5–14) was spent on the treatment. The results of the study were evaluated according to the number of growing follicles, oocytes, and mature oocytes retrieved (see Table 1).

After more than three follicles larger than 18 mm were observed, 10,000 IU of hCG was administered intramuscularly, and 36 hours later follicles were aspirated under general anesthesia. Mature oocytes were retrieved from follicular fluid and placed in cleavage (G-1 TM version 3; Vitrolife, Goteborg, Sweden) and blastocyst development mediums (G-2 TM version 3; Vitrolife).

The study group was divided into two subgroups according to the number of oocytes retrieved. Patients with an oocyte count of five or more were considered good responders and less than five as poor responders. All parameters were compared for these two groups. Characteristics such as age, infertility period, and body mass index were similar between good and poor responders.

The study group was divided into two subgroups according to the number of oocytes retrieved. Patients with an oocyte count of five or more were considered good responders, and patients with less than five as poor responders.

Good responders had a mean beginning dose of 249 ⫾ 58 IU/day, and a total dose of 2363 ⫾ 969 IU was reached in 8 ⫾ 2 days. Poor responders had a mean starting dose of 231 ⫾ 62 IU/day, and a total dose of 2550 ⫾ 1041 IU was reached in a mean period of 9 ⫾ 1 days. These parameters were statistically alike. Parameters such as basal serum FSH

The 25 patients being treated for male factor infertility received ICSI; and IVF ⫹ ICSI was provided for the 23 Fertility and Sterility姞

Fifteen couples ended in pregnancy. Two of them continued as twin pregnancies, and eight as singletons. Three pregnancies were aborted, and two were biochemical pregnancies.

593

TABLE 1 Demographic characteristics, markers of ovarian reserve, and ovarian response to recombinant FSH treatment in the study group.

Age (y) Infertility period (y) Body mass index Day-3 FSH (mIu/mL) Day-3 E2 (pg/ml) Day-3 LH (mIu/mL) Day-3 AMH (pg/mL) Antral follicle count hCG day E2 (pg/mL) Follicle count Retrieved oocyte count Mature oocyte count

Study group (n ⴝ 44)

Poor responders (n ⴝ 11)

Good responders (n ⴝ 33)

P value

31.22 ⫾ 4.02 9.0 ⫾ 3.94 24.69 ⫾ 3.55 7.49 ⫾ 2.56 33.97 ⫾ 13.98 6.16 ⫾ 5.98 0.54 ⫾ 0.42 9.70 ⫾ 4.08 2251.75 ⫾ 1032.12 10.27 ⫾ 3.75 8.90 ⫾ 4.37 7.56 ⫾ 4.33

32.63 ⫾ 3.58 12.0 ⫾ 4.17 24.68 ⫾ 5.14 7.35 ⫾ 3.49 36.40 ⫾ 22.86 8.44 ⫾ 11.17 0.15 ⫾ 0.11 6.90 ⫾ 3.41 1252.54 ⫾ 605.29 6.09 ⫾ 2.42 3.54 ⫾ 1.21 2.6 ⫾ 1.2

30.75 ⫾ 4.10 8.0 ⫾ 3.37 24.69 ⫾ 2.94 7.53 ⫾ 2.23 33.16 ⫾ 9.83 5.40 ⫾ 2.58 0.67 ⫾ 0.41 10.63 ⫾ 3.89 2584.81 ⫾ 927.02 11.66 ⫾ 3.01 10.69 ⫾ 3.47 9.21 ⫾ 3.68

NS NS NS NS NS NS ⬍.001 ⬍.01 ⬍.001 ⬍.001 ⬍.001 ⬍.001

Note: Values are mean ⫾ standard deviation. NS ⫽ Not statistially significant; AMH ⫽ antimüllerian hormone. Fiçiciogˇlu. AMH as an ovarian reserve marker. Fertil Steril 2006.

and E2 levels were also not statistically different. Meanwhile, the difference between serum AMH levels, maximum E2 levels, antral follicle count, hCG day follicle counts, and mature oocyte counts were statistically significant (see Table 1). The main parameter of the study, AMH, was found to be considerably higher in good responders; mean level was 0.67 ⫾ 0.41 pg/mL. The AMH level was 0.15 ⫾ 0.11 pg/mL in poor responders (P⬍.001) (see Fig. 1).

FIGURE 1 The AMH levels in good and poor responder groups.

When we evaluated the relationship of retrieved oocyte counts with the parameters included, we found that only basal AMH levels and the number of antral follicles were statistically correlated. The number of oocytes retrieved was not correlated with age, basal FSH, or E2 levels (Table 2). No statistically significant correlation between age and basal FSH, E2, AMH levels, the number of antral follicles, oocytes retrieved, or mature oocytes could be found. Day-3 FSH levels were statistically significantly correlated only with day-3 E2 levels (r ⫽ 0.31 ; P⬍.05). Day-3 AMH levels were correlated with antral follicle count (r ⫽ 0.47 ; P⬍.01), oocyte count (r ⫽ 0.56 ; P⬍.001), and mature oocyte count (r ⫽ 0.53 ; P⬍.001). Finally, to identify the best parameter and threshold values that would yield the optimal mix of false-positive and false-

TABLE 2 Correlation coefficients between the number of oocytes collected and the parameters investigated. Variable Age Day-3 FSH Day-3 E2 Day-3 AMH Antral follicle count

r

P

⫺0.203 ⫺0.022 ⫺0.080 0.564 0.304

NS NS NS ⬍.001 ⬍.05

Note: NS ⫽ Not statistically significant; AMH ⫽ antimüllerian hormone. Fiçiciogˇlu. AMH as an ovarian reserve marker. Fertil Steril 2006.

594

Fiçiciogˇlu et al.

AMH as an ovarian reserve marker

Fiçiciogˇlu. AMH as an ovarian reserve marker. Fertil Steril 2006.

Vol. 85, No. 3, March 2006

FIGURE 2 ROC curves for AMH, age, FSH, E2, and number of the antral follicles. Reference line (yellow); antral follicles (aqua); AMH (violet); E2 (blue); FSH (green); and age (red).

clinically similar day-3 serum FSH and E2 levels in patients of all age groups. Baseline FSH, LH, and E2 levels are good predictors of ovarian reserve (1, 2). As noted by other studies, Jurema et al. (18) found that levels of baseline FSH and E2 but not LH were statistically significantly lower in cycles resulting in a normal ovarian response as well as cycles resulting in clinical pregnancy. It was also stated that a stricter FSH value (8 IU/mL) would have been required for the group of patients with poor prognosis (1, 2, 18). In our study, good and poor responder groups had similar mean baseline FSH levels (lower than 8 IU/mL), which gave us an opportunity to underline the importance of AMH and the relation between AMH and the number of retrieved oocytes. Bancsi et al. (19) found that antral follicle count is the best single indicator for ovarian response. The combination of antral follicle count with basal FSH and inhibin B levels improved prediction. Our earlier study found serum inhibin B to be the determinant with the highest sensitivity among age, body mass index, basal FSH, E2, and antral follicle count (4). Van Rooij et al. (12) combined antral follicle count with AMH and inhibin B to provide for better prediction. Fanchin et al. (20) found that AMH levels correlated more robustly with the number of early antral follicles than did inhibin B, E2, FSH, or LH on cycle day 3.

Fiçiciogˇlu. AMH as an ovarian reserve marker. Fertil Steril 2006.

negative fractions for the prediction of the number of oocytes to collect, we used logistic regression and ROC curves analyses for age, day-3 FSH, E2, AMH, and the number of the antral follicles. Among all the parameters for ROC curves, day-3 AMH level had the highest sensitivity and specificity values, and the curve referring to AMH was placed on the top (area under the curve [AUC]: 0.92), followed by antral follicles (AUC: 0.78) and age (AUC: 0.63) (Fig. 2). When 0.25 pg/mL was selected as the AMH cut-off value, with a likelihood ratio ⫽ 1, the sensitivity was 90.9% and specificity was 90.9%; the positive predictive value was 96.8% and negative predictive value 76.9%. Good responders comprised 96% of patients with AMH values more than 0.25 pg/mL (n ⫽ 30); only 23% of patients with an AMH value less than 0.25 pg/mL (n ⫽ 3) were good responders. This difference was statistically significant (P⬍.001). Among patients with AMH levels more than 0.25 pg/mL, 13 (41%) became pregnant; two (15%) pregnancies occurred in the other group; this difference was not statistically significant (P⬎.05), possibly due to the low numbers of women in both groups. DISCUSSION Our data demonstrate an association between early follicular phase serum AMH and number of retrieved oocytes despite Fertility and Sterility姞

Expression of AMH begins in the third trimester of gestation, long before it can be detected in serum. As a result, it is thought to play an important role in early follicular development. Once menstrual cycles begin, serum AMH levels fluctuate slightly from baseline due to the stimulation of a small cohort of follicles and loss of AMH production from corpus luteum (14, 21). Our results demonstrated an association among AMH and antral follicles and retrieved oocyte count, so serum AMH levels may reflect the size of antral follicle pool; it may thus provide a marker associated with the anticipated number of oocytes to be retrieved after controlled ovarian stimulation. Early follicular phase AMH levels result from the pool of follicles that began producing it in utero, and this production is independent of the gonadotropin-dependent indicators of ovarian reserve; this makes AMH unique in providing a perspective not available by using current serum markers and ultrasound (13). Similar to previous reports on ovarian reserve, we found that basal antral follicle count is correlated but weakly with the number of retrieved oocytes during assisted reproduction cycles (19, 22). In their 289-patient series, Frattarelli et al. (22) attempted to find a threshold basal antral follicle count but found no absolute value. However, patients with fewer than five follicles were found to require additional counseling. Although antral follicle count was not 100% accurate in predicting success or failure, it correlated with current measures of ovarian reserve. Their results were similar to ours except that they found basal FSH levels were related to antral follicle count. We did not attempt to find a threshold basal antral follicle count, but we did observe that patients with fewer than five retrieved oocytes had lower day 3 AMH 595

levels, fewer antral follicles, and lower hCG day E2 levels. Thus, basal antral follicle count and basal AMH levels are good tools for use in counseling patients (11–13, 19, 20). However, accurate antral follicle counts depend on the clinician’s experience and the technical properties of the ultrasound used. By contrast, AMH levels are obtained by objective measurements performed in laboratory medium and thus are free of interobserver variability and personal comments. If the cut-off level was set at 0.25 pg/mL, we found that AMH was the best indicator of ovarian reserve with a high sensitivity and specificity. Levels of AMH would predict the number of oocytes with a positive predictive rate of 96%, although it had little value for predicting pregnancy. We believe that the latter can be attributed to other factors in human fertility such as male and endometrial factors. Thus, our data indicate that early follicular AMH levels can offer a contribution to the serum markers currently being used for evaluating ovarian reserve, but AMH levels will not reveal the possibility of pregnancy. Acknowledgments: The authors thank Gelis¸im Laboratories, Istanbul, Turkey, for laboratory studies.

8.

9.

10.

11.

12.

13.

14.

15. 16.

REFERENCES 1. Bukulmez O, Arici A. Assessment of ovarian reserve. Curr Opin Obstet Gynecol 2004;16:231–7. 2. Kwee J, Elting MW, Schats R, Bezemer PD, Lambalk CB, Schoemaker J. Comparison of endocrine tests with respect to their predictive value on the outcome of ovarian hyperstimulation in IVF treatment: results of a prospective randomized study. Hum Reprod 2003;18:1422–7. 3. Hendriks DJ, Mol BW, Bancsi LF, Te Velde ER, Broekmans FJ. Antral follicle count in the prediction of poor ovarian response and pregnancy after in vitro fertilization: a meta-analysis and comparison with basal follicle-stimulating hormone level. Fertil Steril 2005;83:291–301. 4. Fiçiciogˇlu C, Kutlu T, Demirbas¸ogˇlu S, Mulayim B. The role of inhibin B as a basal determinant of ovarian reserve. Gynecol Endocrinol 2003;17:287–93. 5. Broekmans FJ, Faddy M, te Velde ER. Ovarian reserve and reproductive age may be determined from measurement of ovarian volume by transvaginal sonography. Hum Reprod 2005;20:1114 –5. 6. Wallace WH, Kelsey TW. Ovarian reserve and reproductive age may be determined from measurement of ovarian volume by transvaginal sonography. Hum Reprod 2004;19:1612–7. 7. Jarvela IY, Sladkevicius P, Kelly S, Ojha K, Campbell S, Nargund G. Quantification of ovarian power Doppler signal with three-dimensional

596

Fiçiciogˇlu et al.

AMH as an ovarian reserve marker

17.

18.

19.

20.

21.

22.

ultrasonography to predict response during in vitro fertilization. Obstet Gynecol 2003;102:816 –22. Hendriks DJ, Broekmans FJ, Bancsi LF, Looman CW, de Jong FH, te Velde ER. Single and repeated GnRH agonist stimulation tests compared with basal markers of ovarian reserve in the prediction of outcome in IVF. J Assist Reprod Genet 2005;22:65–73. Jain T, Soules MR, Collins JA. Comparison of basal follicle-stimulating hormone versus the clomiphene citrate challenge test for ovarian reserve screening. Fertil Steril 2004;82:180 –5. Lee MM, Donahoe PK, Hasegawa T, Silverman B, Crist GB, Best S, et al. Müllerian inhibiting substance in humans: normal levels from infancy to adulthood. J Clin Endocrinol Metab 1996;81:571– 6. Laven JS, Mulders AG, Visser JA, Themmen AP, De Jong FH, Fauser BCJM. Antimüllerian hormone serum concentrations in normoovulatory and anovulatory women of reproductive age. J Clin Endocrinol Metab 2004;89:318 –23. van Rooij IA, Broekmans FJ, te Velde ER, Bancsi LF, de Jong FH, Themmen AP. Serum antimüllerian hormone levels: a novel measure of ovarian reserve. Hum Reprod 2002;17:3065–71. Seifer DB, MacLaughlin DT, Christian BP, Feng B, Shelden RM. Early follicular serum müllerian inhibiting substance levels are associated with ovarian response during assisted reproductive technology cycles. Fertil Steril 2002;77:468 –71. Cook CL, Siow Y, Taylor S, Fallat ME. Serum müllerian inhibiting substance levels during normal menstrual cycles. Fertil Steril 2000;73: 859 – 61. Filicori M, Flamicnic R. The role of luteinizing hormone in folliculogenesis and ovulation induction. Fertil Steril 1999;71:405–14. Teixeria J, Maheswaran S, Donahoe TK. Müllerian inhibiting substance: an instructive developmental hormone, diagnostic and possible therapeutic applications. Endocrin Rev 2001;22:657–74. Gruijters MJ, Visser JA, Durlinger AT, Themmen AF. Antimüllerian hormone and its role in ovarian function. Mol Cell Endocrinol 2003; 15:85–90. Jurema MW, Bracero NJ, Garcia JE. Fine tuning cycle day 3 hormonal assessment of ovarian reserve improves in vitro fertilization outcome in gonadotropin-releasing hormone antagonist cycles. Fertil Steril 2003; 80:1156 – 61. Bancsi LF, Broekmans FJ, Eijkemans MJ, de Jong FH, Habbema JD, te Velde ER. Predictors of poor ovarian response in in vitro fertilization: a prospective study comparing basal markers of ovarian reserve. Fertil Steril 2002;77:328 –36. Fanchin R, Schonauer LM, Righini C, Guibourdenche J, Frydman R, Taieb J. Serum anti-Müllerian hormone is more strongly related to ovarian follicular status than serum inhibin B, estradiol, FSH and LH on day 3. Hum Reprod 2003;18:323–7. de Vet A, Laven JS, de Jong FH, Themmen AP, Fauser BC. Antimüllerian hormone serum levels: a putative marker for ovarian aging. Fertil Steril 2002;77:357– 62. Frattarelli JL, Levi AJ, Miller BT, Segars JH. A prospective assessment of the predictive value of basal antral follicles in in vitro fertilization. Fertil Steril 2003;80:350 –5.

Vol. 85, No. 3, March 2006