ORIGINAL ARTICLE: EARLY PREGNANCY

Role of ovarian reserve markers, antim€ ullerian hormone and antral follicle count, as aneuploidy markers in ongoing pregnancies and miscarriages Maribel Grande, B.Sc., Ph.D.,a Virginia Borobio, M.D.,a Mar Bennasar, M.D., Ph.D.,a , Ph.D.,a Narcís Masoller, M.D.,a Iosifina Stergiotou, M.D., Ph.D.,a Immaculada Mercade b ~ arrubia, M.D., Ph.D., and Antoni Borrell, M.D., Ph.D.a Joana Pen a Department of Maternal-Fetal Medicine and b Department of Gynecology and Reproduction, Institute Gynecology, Obstetrics and Neonatology, Hospital Clínic Barcelona, Catalonia, Spain

Objective: To assess the role of two ovarian reserve markers, antim€ ullerian hormone (AMH) and antral follicle count (AFC), as markers of the background risk for fetal trisomy. Design: Prospective study. Setting: Tertiary referral hospital. Patient(s): Assessment was carried out either in ongoing pregnancies or miscarriages in our center. Intervention(s): AFC was assessed transvaginally during a routine (11–13 weeks) or referral scan. AMH was determined either during the first-trimester maternal serum markers assessment or in cases referred for chorionic villi sampling after the invasive procedure. Main Outcome Measure(s): AMH reference ranges were constructed according to maternal age, and AMH- and AFC-derived ovarian ages were compared among three different cytogenetic groups (normal karyotype, autosomal trisomies, and other chromosomal anomalies) in both ongoing pregnancies and miscarriages. Result(s): In autosomal trisomies, the median AFC-derived ovarian age was 3–5 years above the median maternal age. No differences were observed between AMH-derived ovarian age and maternal age. Conclusion(s): AFC-derived ovarian biologic age reflects a more precise background risk for fetal aneuploidy that is not observed for AMH-derived age. (Fertil SterilÒ 2015;-:-–-. Use your smartphone Ó2015 by American Society for Reproductive Medicine.) to scan this QR code Key Words: Antral follicle count, antim€ ullerian hormone, ovarian age, aneuploid marker, and connect to the maternal age Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/grandem-aneuploidy-markers-pregnancies-miscarriages/

O

varian aging is characterized by the age-related decline in the quantity of primordial follicles and quality of the oocytes

within (1). After peak fertility at 20 years of age, women are expected to lose three-fourths of their follicular reserve from 30 to 40 years of age,

Received August 19, 2014; revised and accepted February 18, 2015. M.G. has nothing to disclose. V.B. has nothing to disclose. M.B. has nothing to disclose. I.S. has nothing to disclose. I.M. has nothing to disclose. N.M. has nothing to disclose. J.P. has nothing to disclose. A.B. has nothing to disclose.  n Sanitaria (grant PI 11/00685). Supported by Instituto de Salud Carlos III. Fondo de Investigacio Reprint requests: Antoni Borrell, M.D., Ph.D., Department of Maternal-Fetal Medicine, Institute of Gynecology, Obstetrics, and Neonataology, Hospital Clínic Barcelona, Sabino de Arana 1, Barcelona 08028, Catalonia, Spain (E-mail: [email protected]). Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.02.022 VOL. - NO. - / - 2015

discussion forum for this article now.*

* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.

with a gradual decline in fertility thereafter until menopause (2–4). However, poor response to ovarian stimulation in young patients and the wide range of chronologic age at which menopause occurs suggest that women's reproductive potential depends not only on chronologic age but also on the ovarian reserve (5, 6). Ovarian reserve can be evaluated by endocrine and ultrasound markers, mostly by antim€ ullerian hormone (AMH) and antral follicle count (AFC) (7–9). AMH is expressed by the 1

ORIGINAL ARTICLE: EARLY PREGNANCY granulosa cells of primary to early antral follicles (10), and is inhibiting maturation of primordial follicles (11). AFC consists in counting antral follicles 2–10 mm in diameter in both ovaries with the use of vaginal sonography. Both markers reflect the age-dependent decline of ovarian function (12, 13), but it is still unclear whether they should be assessed concurrently or if one of these markers should be preferably used. It has been suggested by earlier studies that the ovarian reserve decline has not only a quantitative (decreased number of primordial follicles), but also a qualitative effect (low-quality oocytes and increased trisomy risk). Namely, it has been described that unilateral oophorectomy in female mice results in an earlier than expected rise in aneuploidy (14), and that menopause occurs at an earlier age among women with trisomic pregnancies (15). Recently, our group has reported a 3–6 years higher ovarian-derived age in women carrying trisomic pregnancies (16). It seems, therefore, that ovarian rather than chronologic age potentially reflects a more precise background risk for fetal trisomy. Because AMH and AFC are currently considered to be the best markers of ovarian reserve markers, we sought to assess their performance as fetal aneuploidy markers when determined in pregnant women at the same time of routine maternal serum markers or first-trimester scan. Although AMH levels in the first trimester may be slightly decreased as compared with those reported in healthy women (17–19), AFC has not been previously assayed in pregnant women by any other group.

MATERIALS AND METHODS Population During a 21-month period (April 2012–December 2013) pregnant women from Barcelona-West Health District assessed in our center, with either ongoing pregnancies or miscarriages, were offered to be enrolled in the study. The former group was attending for routine first-trimester scan at 11–13 weeks or were referred either for chorionic villi sampling (CVS) or for termination of pregnancy because of fetal aneuploidy. In the latter group, first-trimester pregnancy loss was diagnosed in the ultrasound or emergency department and CVS was offered for genetic counseling (20). Multiple pregnancies and pregnancies achieved by medically assisted reproductive technology (ART) were excluded from the study, because some kind of ovarian stimulation or egg donation was almost invariably carried out. Postnatal follow-up was sought in the nonkaryotyped low-risk pregnancies. Ethics approval was received by the local Research Ethics Committee, and written informed consent was obtained from each of the recruited pregnant women. Maternal and clinical data were analyzed with the use of the statistical package SPSS version 17 (Statistical Package for the Social Sciences).

Electric, GE Healthcare Austria; and Aloka Prosound a7, Aloka). A 25-minute slot was assigned to perform a routine first-trimester scan, scheduled at 11–13 weeks.

Antral Follicle Count After the routine or referral scan, AFC was assessed by means of transvaginal scan with the use of a 7.5 Mhz probe according to the technical considerations of Broekmans et al.’s guidelines (21). Considering that our study comprised pregnant women with spontaneous conception, we disregarded the clinical consideration of counting follicles during days 2–4 of the cycle, and we discarded ovaries with a corpus luteum large enough to massively compress the ovarian parenchyma. Roughly, round or oval sonolucent structures in the ovaries, not capturing color flow, were regarded as follicles, but only follicles measuring 2–10 mm in mean diameter were included in the AFC (16). A data reanalysis was carried out in case of pregnancies with a single ovary assessed, after AFC was estimated to be twofold the obtained value from the single ovary.

€ llerian Hormone Assay Antimu In women from the Barcelona-West Health District, 1 mL blood was drawn at the time of the first pregnancy sampling at 8–12 weeks, during the first-trimester maternal serum markers assessment. In referrals, maternal blood was drawn after the scan, and in cases of pregnancy loss before 8 weeks, blood was obtained in the Emergency Department. Maternal serum was stored frozen until basal AMH levels (ng/mL) were determined by means of AMH enzyme immunoassay Gen II (Instrumentation Laboratory and Beckman-Coulter). The analytical sensitivity of this ELISA was estimated to be 0.1 ng/mL. The intra-assay coefficient of variation (CV) was 4% and the interassay CV was 6%.

Chorionic Villi Sampling and Karyotyping CVS was offered to all pregnant women from the BarcelonaWest Health District with a risk R1/250 for trisomy 21, 18, or 13 at the combined test regardless of maternal age, to highrisk referrals, and to cases with pregnancy loss. CVS was performed transcervically with the use of a round-tip curved steel forceps (1.9 mm in diameter and 27 cm in length), first introduced by Rodeck (Area Medica), under continuous ultrasound guidance (22). At the laboratory, samples were inspected under the dissecting microscope to release villi from maternal material. Cytogenetic analysis was carried out after both short- and long-term cultures (20, 23), and quantitative fluorescent polymerase chain reaction (QF-PCR) was applied only to ongoing pregnancies (24). Three groups were defined for comparison according to cytogenetic results: normal karyotype, autosomal trisomies, and other chromosomal anomalies.

First-trimester Scan First-trimester scans were performed by four sonologists with the use of primarily three ultrasound machines (Acuson Antares, Siemens Medical Solutions; Voluson E6 General 2

AMH Reference Ranges Reference ranges for AMH according to maternal chronologic age were constructed by means of the lambda-mu-sigma VOL. - NO. - / - 2015

Fertility and Sterility® (LMS) method (25) including 907 ongoing pregnancies with a normal karyotype. The LMS method summarizes the changing distribution with the use of three curves representing the skewness expressed as Box-Cox power (L), median (M), and coefficient of variation (S). The final curves of percentiles are produced by these three smooth curves (25). Degrees of freedom for each curve (L, M, and S) were selected according to changes in the model deviance. Percentile values were computed with the use of LMS Chartmaker software (LMS Chartmaker Light version 2.54; Medical Research Council). A table reporting the mean and the 90% interval of AMH (5th and 95th percentiles) for each age interval was created. In addition, AMH changes were assessed across gestational age (in euploid ongoing pregnancies), and AMH medians from 10 to 14 weeks of gestation were compared with the use of Mann-Whitney nonparametric test.

AFC Reference Ranges Reference ranges for AFC were previously constructed by our group and already reported for pregnancies enrolled in the first 15 months of the present cohort from April 2012 to June 2013 (16). The present cohort was expanded with pregnancies studied during a further 6-month period (July– December 2013). Additionally, slopes of regression lines for AMH and AFC were calculated in chromosomally normal ongoing pregnancies to assess changes with gestational age during the first trimester (crown-rump length [CRL] 40–80 mm).

nancy loss. Among ongoing pregnancies, 836 were attending for routine first-trimester assessment from Barcelona-West Health District, 97 were referred for CVS, and 8 for pregnancy termination because of trisomy 21. In the whole study population, the mean maternal age was 32.7 years (range 16–46 years), and the mean CRL was 61.8 mm (range 3–89 mm). The mean CRL was 63 mm (12þ5 weeks) in routine scans, 64 mm (12þ5 weeks) in cases referred for CVS, 70 mm (13þ1 weeks) in cases referred for trisomy 21, and 22 mm (8þ6 weeks) in cases of pregnancy loss. AMH reference ranges according to chronologic maternal age were constructed with the use of 907 chromosomally normal ongoing pregnancies (after exclusion of 34 cases of aneuploidy or fetal loss). The 5th, 50th, and 95th percentiles according to maternal age are presented in Table 1 and Figure 1. Because maternal age varied from 16 to 46 years, the 50th percentile for AMH ranged from 3.7 to 0.2 ng/mL, the 5th percentile from 1.0 to 0 ng/mL, and the 95th percentile from 9.2 to 1.1 ng/mL. AFC reference ranges according to maternal age have been previously reported by our group (16). When AMH changes were assessed across gestational age, a steady nonsignificant decrease was observed from 11 to 14 weeks with median values of 2.25 ng/mL at 11 weeks,

TABLE 1 €llerian hormone (AMH) reference ranges according to Antimu woman's age constructed in a series of 907 pregnant women with a spontaneously conceived chromosomally normal singleton fetus.

Comparison of AMH- and AFC-derived Ages In pregnancies with complete AMH and AFC data, two different ovarian ages were derived from the two ovarian markers. An AMH-derived ovarian age was assigned to each pregnancy, with the use of the maternal age corresponding to the observed AMH level according to the newly constructed 50th-percentile curve. Similarly, an AFC-derived ovarian age was assigned according to medians recently published by our group (16). The median chronologic age was compared with the median AMH and AFC ovarian ages, with the use of Mann-Whitney nonparametric test for each of the three cytogenetic groups in both viable and nonviable pregnancies. In addition, bivariate correlation analysis with two-tailed test was applied to one of the cytogenetic groups, the autosomal trisomies group, among AMH level/age, AFC count/ age, and maternal age. Correlation analysis was repeated after stratification according to pregnancy outcome (ongoing/lost pregnancy).

RESULTS Among 1,604 singleton pregnancies initially recruited at the time of ultrasound assessment, 602 were excluded because of the following reasons: ART (n ¼ 104); blood sample not available (n ¼ 492); and unknown karyotype in a pregnancy loss (n ¼ 6). Thus, the study population for AMH assessment included 1,002 spontaneously conceived singleton pregnancies, namely, 941 ongoing pregnancies and 61 cases of pregVOL. - NO. - / - 2015

AMH percentile (ng/mL) Maternal age (y)

5th

50th

95th

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

1.0 0.9 0.9 0.8 0.8 0.8 0.7 0.7 0.7 0.6 0.6 0.6 0.5 0.5 0.5 0.4 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0

3.7 3.6 3.5 3.4 3.3 3.2 3.1 3.0 2.9 2.8 2.8 2.7 2.6 2.5 2.4 2.2 2.1 2.0 1.8 1.7 1.5 1.3 1.2 1.0 0.8 0.7 0.6 0.5 0.4 0.3 0.2

9.2 9.0 8.9 8.8 8.6 8.5 8.4 8.3 8.2 8.1 8.0 7.9 7.7 7.5 7.3 7.0 6.8 6.5 6.2 5.9 5.5 5.1 4.7 4.2 3.6 3.1 2.7 2.2 1.8 1.4 1.1

Grande. AMH and AFC as aneuploidy markers. Fertil Steril 2015.

3

ORIGINAL ARTICLE: EARLY PREGNANCY

FIGURE 1

TABLE 2 Chromosomal anomalies (autosomal trisomies and other anomalies) €llerian observed in 753 pregnant women having both antimu hormone (AMH) and antral follicle count (AFC) determined, in both ongoing pregnancies and pregnancy losses.

Cytogenetic group

€llerian hormone (AMH) reference ranges according to Antimu woman's age, with lines indicating the 5th, 50th, and 95th percentiles. Dots indicate the observed AMH levels in 907 ongoing pregnancies with a spontaneously conceived euploid singleton fetus. Grande. AMH and AFC as aneuploidy markers. Fertil Steril 2015.

Ongoing pregnancies (n [ 23)

Autosomal trisomies

13 trisomies 21 3 trisomies 18 2 trisomies 13 1 trisomy 12

Other chromosomal anomalies

3 monosomies X 1 sex trisomy

Pregnancy losses (n [ 36) 7 trisomies 16 5 trisomies 22 2 trisomies 7 2 trisomies 9 2 trisomies 10 2 trisomies 15 2 trisomies 18 2 trisomies 20 1 trisomy 14 1 trisomy 21 5 triploidies 4 monosomies X 1 tetrasomy 16q

Grande. AMH and AFC as aneuploidy markers. Fertil Steril 2015.

2.00 ng/mL at 12 weeks, 1.8 ng/mL at 13 weeks, and 1.00 ng/ mL at 14 weeks. The slope of the linear regression of AMH with gestational age, expressed by CRL, was 0.0081, and the regression line is plotted in Supplemental Figure 1A (available online at www.fertstert.org). The corresponding slope for AFC was 0.003 and plotted in Supplemental Figure 1B. The median AMH value in the autosomal trisomies group was 1.40 ng/mL, significantly lower than 2.00 ng/mL as obtained in the euploid group (P¼ .036). However, comparison of AMH regardless of the maternal age distribution may be misleading, because median maternal ages were significantly different among the two groups: 36 versus 33 years, respectively (P< .001). To appropriately compare AMH- and AFC-derived ovarian ages, 249 pregnancies were excluded because one (n ¼ 214; 21%) or both (n ¼ 35; 3.5%) ovaries could not be properly assessed. Thus, 753 pregnancies remained for comparison: 699 ongoing pregnancies (676 with a normal karyotype, 19 with autosomal trisomies, and 4 with other chromosomal anomalies) and 54 cases of pregnancy loss (18 with a normal karyotype, 26 with autosomal trisomies, and 10 with other chromosomal anomalies; Table 2). When chronologic maternal age was compared with the two different ovarian ages, differences were observed only for AFC in the autosomal trisomies' group, with a significant 5-year excess in AFC-derived ovarian age in aneuploidy cases (Table 3). After stratification according to pregnancy outcome, a 3-year significant excess was observed for ongoing pregnancies. In cases of pregnancy loss, the observed 5-year excess age achieved significance only after data reanalysis with 214 added pregnancies with a single ovary studied, possibly owing to the extra ten cases with chromosomal anomalies (36 vs. 41 years; P¼ .018; data not shown). No differences were observed in the euploid cases or in other chromosomal anomalies. In the autosomal trisomies' group, a moderate correlation was observed between both markers, AFC and AMH, either in 4

the overall series (r ¼ 0.57; P< .001) or after stratification into ongoing pregnancies and cases of pregnancy loss (Supplemental Table 1, available online at www.fertstert.org). In the correlation analysis between ovarian markers with maternal age, AFC demonstrated a moderate correlation in the overall series (r ¼ 0.54; P< .001) as well as after stratification into ongoing pregnancies and cases of pregnancy loss, whereas no significant correlation was observed with AMH (Supplemental Table 1).

DISCUSSION In this study, both AMH- and AFC-derived ovarian ages were calculated and compared with chronologic maternal age with the use of newly constructed AMH and recently published AFC reference ranges. The results provide evidence that during the first trimester of pregnancy, AFC, but not AMH, biologic age, is a better marker for autosomal trisomies, being higher than chronologic age in both viable and nonviable pregnancies. AMH and AFC compares favorably with other previously described ovarian reserve and response markers, such as early follicular FSH and inhibin B (26, 27). AMH reflects the agerelated decline of reproductive capacity (28) and can also identify poor and hyperresponders to ART (29). Our AMH results during the first trimester support a gradual AMH decline with chronologic age, which is in agreement with earlier studies in women with proven fertility (28), either in those having ART because of a male factor (17) or in firsttrimester pregnant women (19, 30). Although AMH has been previously determined during pregnancy and medians for the age groups reported, the present study is the first to construct AMH reference ranges according to individual maternal years of age. Because it has not been proven that AMH decreases during the first trimester, our reference ranges and cases were assessed at the same gestational period. In our series, a decline of medians, from 3.3 ng/mL VOL. - NO. - / - 2015

Fertility and Sterility®

TABLE 3 €llerian hormone (AMH)–derived age (P1), and median antral Comparison between the median maternal chronologic age and median antimu follicle count (AFC)–derived age (P2), in the three cytogenetic groups: reanalysis of data stratified by pregnancy outcome, ongoing pregnancies, and pregnancy losses. Cytogenetic group All pregnancies Normal karyotype Autosomal trisomies Other chromosomal anomalies Ongoing pregnancies Normal karyotype Autosomal trisomies Other chromosomal anomalies Pregnancy losses Normal karyotype Autosomal trisomies Other chromosomal anomalies

n

Median maternal age (y)

Median ovarian age by AM, (y)

P1 value

Median ovarian age by AFC (y)

P2 value

694 45 14

33 36 31.5

33 36.5 34.8

.321 .910 .541

30.5 41 29.5

.115 .003 .454

676 19 4

33 39 28.5

33 39 39

.208 .907 .386

30.5 42 34.8

.076 .010 .564

34 36 32

38 35.8 33

.116 .769 .940

39.3 41 27.5

.205 .068 .198

18 26 10

Grande. AMH and AFC as aneuploidy markers. Fertil Steril 2015.

at 20 years to 0.8 ng/mL at 40 years, was observed, which was very similar to previously reported results in a first-trimester pregnancy population of fertile women: 2.98 ng/mL at 20– 24.9 years and 0.81 ng/mL at R40 years (30). Although AMH levels during the first trimester of pregnancy have been reported to be similar to those in nonpregnant healthy women (18), a significant drop has been demonstrated from the first to the second trimester and from the second to the third trimester (19). The comparison with the observed decline in a presumably nonpregnant healthy cohort from 3.59 ng/mL at

Role of ovarian reserve markers, antimüllerian hormone and antral follicle count, as aneuploidy markers in ongoing pregnancies and miscarriages.

To assess the role of two ovarian reserve markers, antimüllerian hormone (AMH) and antral follicle count (AFC), as markers of the background risk for ...
626KB Sizes 0 Downloads 10 Views