ANXA5 haplotype with recurrent pregnancy loss (RPL) in PCOS patients

M E TAB O LI S M CL I NI CA L A N D EX P ER IM EN T AL 6 2 (2 0 1 3) 1 05 7 –1 0 60

Available online at www.sciencedirect.com

Metabolism www.metabolismjournal.com

Independent association of the M2/ANXA5 haplotype with recurrent pregnancy loss (RPL) in PCOS patients Nina Rogenhofer a , Laura Engels a , Nadja Bogdanova b , Frank Tüttelmann b , Christian J. Thaler a , Arseni Markoff c,⁎ a

Division of Gynecological Endocrinology and Reproductive Medicine, Department of Gynecology and Obstetrics, Klinikum der Ludwig-Maximilians-Universität, Munich, Campus Grosshadern, D-81377 Munich, Germany b Insititute of Human Genetics, UKM and WWU Muenster, D-48149 Muenster, Germany c Institute of Medical Biochemistry, ZMBE and IZKF, WWU Muenster, Von Esmarch Str. 56, D-48149 Muenster, Germany

A R T I C LE I N FO Article history:

AB S T R A C T Objective. The aim of this study was to analyze the contribution of the M2 haplotype of

Received 20 July 2012

ANXA5 gene, previously identified as a risk factor for RPL and thrombophilia related

Accepted 13 February 2013

pregnancy complications, to repeated miscarriage observed in PCOS patients.

Keywords:

controls were genotyped for M2/ANXA5.

Patients/Methods. 100 PCOS patients, 500 fertile women and 533 random population Annexin A5

Results. M2 haplotype carriers faced a 3.4 fold elevated RPL risk (odds ratio 5.3, 95%

Recurrent miscarriage

confidence interval 3–9.2) compared to female fertile controls and 2.1 (odds ratio 2.6, 95%

Risk factor

confidence interval 1.6–4.3) compared to population controls. The relative population risks

Polycystic ovarian syndrome

in subgroups of PCOS patients with primary and secondary RPL were 2.3 (odds ratio 2.5, 95% confidence interval 1.2–5) and 3.3 (odds ratio 3.6, 95% confidence interval 1.5–8.4) respectively. As compared to the fertile women group, the relative risks equaled 4 (odds ratio 5, 95% confidence interval 2.3–10.8) and 6 (odds ratio 7.2, 95% confidence interval 3–17.7). Estimated relative risks for M2 carriers among PCOS RPL patients matched the values previously obtained for repeated miscarriage populations. The essential phenotypes, clinically defining PCOS, associated neither with RPL in their diagnostically relevant combinations, nor with M2 carriage as RPL risk factor in the PCOS RPL subgroups. Conclusions. M2/ANXA5 seems an independent RPL risk factor in PCOS patients that progressively correlates with the number of first trimester pregnancies. From our pilot study in PCOS women it appears relevant to offer M2/ANXA5 diagnostic analysis to such patients with RPL complications, to possibly guide proper therapeutic decisions. © 2013 Elsevier Inc. All rights reserved.

1.

Introduction

Polycystic ovarian syndrome (PCOS) is one of the most common endocrine disorders in women of childbearing age and a frequent cause of infertility and pregnancy loss. This condition affects about 7% of women worldwide, but its prevalence might

be as high as 15% when applying the Rotterdam criteria [1]. Clinical phenotypes of the syndrome include disturbed menstrual cycle, traits of hyperandrogenism and obesity. Infertility among PCOS women is a logical consequence of the oligoovulatory/anovulatory state and, together with recurrent pregnancy loss (RPL), a frequent complication [2,3]. At the third ESHRE/

Abbreviations: ANXA5, annexin A5 gene or protein; AF, allele frequency; DNA, deoxyribonucleic acid; M2/ANXA5, M2 haplotype of ANXA5; PCOS, polycystic ovarian syndrome; RPL, recurrent pregnancy loss. ⁎ Corresponding author. Tel.: +49 2518352126; fax: + 49 2518356748. E-mail address: [email protected] (A. Markoff). 0026-0495/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.metabol.2013.02.005

1058

M ET AB O LI S M CL I NI CA L A N D EX PE R IM EN T AL 6 2 (2 0 1 3) 1 05 7–1 06 0

ASRM PCOS consensus workshop it was concluded that women with PCOS desiring a pregnancy may be at an increased risk for adverse pregnancy outcomes, which can be exacerbated by obesity and/or insulin resistance [4]. A recent proteomic analysis study identified annexin A5 as a protein overexpressed in the omental adipose tissue of PCOS patients [5]. On the other hand, reduction of ANXA5 abundance in placenta has been noted for patients with antiphospholipid syndrome, APS [6]; preeclampsia, PE [7,8]; fetal growth restriction, FGR [8,9]; and non-pregnant women with previous RPL [10]. A promoter haplotype of ANXA5 identified in 2007 was shown to drive down gene expression [11,12]. This haplotype designated M2 has been associated with RPL in populations of Central Europe [11,13] and recently in the Japanese population [14]. Very recently, a murine model of ANXA5 loss of function has demonstrated fetal loss, FGR and placental thromboses in homo- and heterozygous animals [15]. In the current study we aimed to evaluate the contribution of M2/ANXA5 as risk factor to the RPL observed in PCOS patients.

2.

Materials and methods

2.1.

Study populations

The present study complied with the ethical guidelines of the institutions involved, as it was approved by the Universities’ Review Boards and informed consent was obtained from all subjects examined. 100 PCOS patients were consecutively recruited between June 2011 and February 2012 in the Division of Gynecological Endocrinology and Reproductive Medicine for a prospective study. PCOS was diagnosed according to the Rotterdam criteria [1]. The study sample (Table 1) was divided in the following subgroups [16]: 1) patients with no recorded

pregnancies, n = 27; 2) patients with ≥ 2 consecutive primary RPLs, n = 45; and 3) patients with ≥ 2 consecutive secondary RPLs, n = 28. Patients with RPL (subgroup 2 and 3) had been prescreened negative for additional potential causes of their RPL as previously described [17]. DNA was extracted from peripheral blood, using the QIAmp DNA blood mini kit (Qiagen, Hilden, Germany) and stored in 100 μl aliquots at 4 °C for further analyses. Previously recruited control groups consisted of a fertile women cohort (n = 500) from the registry of the Institute of Human Genetics, UKM Muenster, and a population control sample drafted from the PopGen biobank at UKSH Kiel (n = 533) [11].

2.2.

Genotyping and statistical analysis

Genotyping of extracted DNA and population statistics were performed as previously described [11]. Differences of clinical phenotypes distribution and of M2 carriage among patient subgroups were assessed using the two-tailed Fisher's exact test. Statistical significance was interpreted at odds ratios of 2 (± 0.25) and above, characteristic of the RPL populations genotyped so far [11,13,14]. The significance of the analysis was interpreted at p < 0.05.

3.

Results

The M2/ANXA5 haplotype [11] and the additional SNP rs1050606 [14] were analyzed as possible recurrent miscarriage factors in this cohort. The haplotype M2 was detected in 32 PCOS patients in heterozygous condition, whereas the SNP rs1050606 G/T was not found to associate with any allelic variant of the haplotype. The G/T allele ratios observed in controls and in the patient sample equaled 1.19 and 1.06 accordingly.

Table 1 – Relevant clinical data of PCOS patients.

Age years mean (min–max) Pregancies n mean (min–max) Deliveries n mean (min–max) Miscarriages n mean (min–max) Early RPL ≤12 weeks of gestation Late RPL 12–20 weeks of gestation Early and late RPL Primary RPL b Secondary RPL c Number of miscarriages 2 3 4 ≥5

subgroup 1 a (n = 27)

subgroup 2 (n = 45)

subgroup 3 (n = 28)

p-value

29 (21–38) 0 (0) 0 (0) 0 (0) 0 0 0 0 0

30 (23–40) 3 (2–6) 0 (0) 3 (2–6) 39 (87%) 1 (2%) 5 (11%) 45 0

33 (22–41) 4 (3–11) 1 (1–2) 3 (2–10) 24 (85%) 1 (4%) 3 (11%) 0 28

n.s. n.s. n.s. n.s.

20 (44%) 14 (31%) 7 (16%) 4 (9%)

10 (36%) 11 (39%) 3 (11%) 4 (14%)

0 0 0 0

(n: number, % percentage, RPL: recurrent pregnancy loss, n.s.: non significant). a Subgroup 1, no pregnancies; subgroup 2, primary RPL; subgroup 3, secondary RPL. b Primary RPL: no infant born either dead or alive after the 20th completed week of gestation and/or weighing more than 500 g before the series of miscarriages [16]. c Secondary RPL: at least one infant born either dead or alive after the 20th completed week of gestation or weighing more than 500 g before the series of abortions [16].

1059

M E TAB O LI S M CL I NI CA L A N D EX P ER IM EN T AL 6 2 (2 0 1 3) 1 05 7 –1 0 60

Table 2 – Genotype distributions of ANXA5 gene promoter haplotypes in the 3 subgroups of PCOS patients and in the whole group. subgroup 1 a

Genotype

Observed N/N N/M1 M1/M1 N/M2, M1/M2 b M2/M2 Total

17 (63%) 3 (11%) 0 (0%) 7 (26%) 0 (0%) 27

Expected 17.8 2.5 0.1 6.2 0.4 27

subgroup 2 a

subgroup 3

PCOS group

Observed

Observed

Expected

Expected

Observed

Expected

21 (46.7%) 10 (22.2%) 0 (0%) 14 (31.1%) 0 (0) 45

14 (50%) 3 (10.7%) 0 (0%) 11 (39.3%) 0 (0%) 28

15.7 2.3 0 9 1 28

24.1 7.4 0.5 12 1 45

52 (52%) 16 (16%) 0 (0%) 32 (32%) 0 (0%) 100

57.7 12.2 0.6 27 2.5 100

(% percentage). Subgroup 1, no pregnancies; subgroup 2, primary RPL; subgroup 3, secondary RPL. b Genotype M1/M2 was not observed in the PCOS patients sample. a

The association with M2 carriage in the PCOS group was confirmed by comparing the 100 PCOS patients with fertile and population controls [11]. M2 carriers appeared to face a 3.4 times higher risk of RPL than non-carriers (odds ratio 5.3, 95% confidence interval 3–9.2, p = 0.000) in the comparison with fertile controls. Similar results were obtained when comparing to population controls. For M2 carriage the odds ratio equals 2.6 with 95% confidence interval of 1.6– 4.3, p = 0.000 and calculated relative (population) risk is 2.1 for this constellation. Subgroup analysis on the patient sample was performed to verify and expand the elevated risk values obtained so far (Table 2). Patients of the first subgroup had a relative population risk of 1.9 (odds ratio 1.9, 95% confidence interval 0.7–5, p = 0.234). The relative RPL risk calculated in comparison with fertile controls was 3.5 (odds ratio 3.9, 95% confidence interval 1.4–10.5, p = 0.006). For the second subgroup RPL population and fertile controls risks were 2.3 (odds ratio 2.5, 95% confidence interval 1.2–5, p = 0.012) and 4 (odds ratio 5, 95% confidence interval 2.3–10.8, p = 0.000), accordingly. For the third subgroup RPL population and fertile controls risks were 3.3 (odds ratio 3.6, 95% confidence interval 1.5–8.4, p = 0.003) and 6 (odds ratio 7.2, 95% confidence interval 3–17.7, p = 0.000) accordingly. The crosstab analysis shows progressive correlation of M2 carriage with the number of miscarriages in the last two patient subgroups (r = 0.92). No combination of clinical features (PCOS phenotypes [1]) significantly distinguished PCOS only vs PCOS/RPL patients. The distribution of M2 carriage as RPL risk factor stratified according to the diagnostic qualifiers for PCOS also did not significantly differ between subgroup 1 and subgroups 2 + 3.

4.

Discussion

Miscarriages in the first trimester have previously been reported to be rather frequent in PCOS women from several studies [3,18,19]. The relative abundance of the M2 haplotype was evident in the patient group, compared to both, fertile and population controls, but no allele of rs1050606 G/T was more prevalent. Minor allele frequency (MAF) of 0.485 obtained for this SNP is

rather close to 0.457 (population controls), and to 0.46 (unrelated individuals of European descent, Centre d’Etudes du Polymorphisme Humaine). In contrast, the reported MAF in the Japanese population is 0.85 [14]. These differences led to the conclusion that the proposed risk role of rs1050606 should be population specific. Altogether PCOS patients carrying M2 seem to face an elevated RPL risk of 3 compared to fertile women, but the population relative risk appears closer to 2, very similar to results previously obtained for RPL patients with 2 or more miscarriages [11,13]. The risk stratification analysis confirmed this finding (Table 2) and demonstrated a trend for M2 carriage to correlate with the mean number of pregnancies in the last 2 subgroups. The abundance of M2 in the subgroup with no pregnancies corresponded to an elevated RPL risk of almost 2 with the general population. However, the 95% confidence interval for the M2 carriers’ odds ratio is rather wide (0.7–5) here, indicating that such conclusion would not be formally justified, despite the perfect Hardy–Weinberg equilibrium for ANXA5 genotypes. Along these lines, we propose that the observed deviation of Hardy–Weinberg equilibrium for the last 2 subgroups, due to the excess of M2 heterozygotes, reflects a positive ascertainment bias. Taken together, these findings imply that 1) M2 may act as independent RPL risk factor in PCOS patients; and that 2) the RPL risk of M2 carriers is not likely to be exacerbated by the PCOS condition. Despite the relatively small number of M2 carriers in the various qualifying categories there is no statistically discernible tendency that distinguished PCOS only from PCOS/RPL patients in terms of phenotype classifiers or M2 distributions within them. The conclusion is that 3) there is no detectable association trend of characteristic PCOS phenotypes with RPL; and reconfirms that M2 carriage demonstrates as independent risk factor, possibly not associated with any PCOS characteristic clinical features. Therapeutic intervention with anticoagulants has been demonstrated to be efficient for PCOS patients with repeated idiopathic miscarriages [20], probably because of the involvement of thrombophilic factor(s) in such cohorts. From this initial study it would appear that M2 in ANXA5 presents as independent risk factor for RPL among PCOS women, since the extensive patients selection protocol applied excluded other more trivial and known causes for

1060

M ET AB O LI S M CL I NI CA L A N D EX PE R IM EN T AL 6 2 (2 0 1 3) 1 05 7–1 06 0

repeated miscarriages. Clearly this finding is interesting for improved clinical work up of PCOS cases with RPL, in the light of possibly successful anticoagulants intervention noted above. These results would need replication from other PCOS cohorts, where enhanced patient numbers among the relevant clinical criteria divisions would render adequate statistical power to answer possible association of M2/ANXA5 with any of the qualifying features characterizing this complex condition.

[5]

[6]

[7]

Author contributions

[8]

NR, LE, NB, FT, CJT and AM participated in the design and conduct of the study and in writing the manuscript.

[9]

[10]

Funding This research received no specific grant from any funding source. The authors acknowledge the continuous support of IZKF Muenster.

[11]

[12]

Acknowledgments The authors are grateful to Ursula Antkowiak for technical assistance with this study.

[13]

[14]

Conflict of interest

[15]

The authors declare no competing interests.

[16]

REFERENCES [17] [1] Rotterdam ESHRE⁄ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004;81:19–25. [2] Simpson JL. Genetic factors in common disorders of female infertility. Reprod Med Rev 2001;8:173–202. [3] Glueck CJ, Wang P, Bornovali S, et al. Polycystic ovary syndrome, the G1691A factor V Leiden mutation, and plasminogen activator inhibitor activity: associations with recurrent pregnancy loss. Metabolism 2003;52:1627–32. [4] Amsterdam ESHRE/ASRM-Sponsored 3rd PCOS Consensus Workshop Group. Consensus on women's health aspects of

[18]

[19] [20]

polycystic ovary syndrome (PCOS). Hum Reprod 2012;27: 14–24. Cortón M, Botella-Carretero JI, López JA, et al. Proteomic analysis of human omental adipose tissue in the polycystic ovary syndrome using two-dimensional difference gel electrophoresis and mass spectrometry. Hum Reprod 2008;23: 651–61. Rand JH, Wu XX, Guller S, et al. Reduction of annexin-V (placental anticoagulant protein-I) on placental villi of women with antiphospholipid antibodies and recurrent spontaneous abortion. Am J Obstet Gynecol 1994;171:1566–72. Shu F, Sugimura M, Kanayama N, et al. Immunohistochemical study of annexin V expression in placentae of preeclampsia. Gynecol Obstet Invest 2000;49:17–23. Chinni E, Tiscia GL, Colaizzo D, et al. Annexin V expression in human placenta is influenced by the carriership of the common haplotype M2. Fertil Steril 2009;91:940–2. Sifakis S, Soufla G, Koukoura O, et al. Decreased annexin A5 mRNA placental expression in pregnancies complicated by fetal growth restriction. Thromb Res 2010;125:326–31. Rand JH, Arslan AA, Wu XX, et al. Reduction of circulating annexin A5 levels and resistance to annexin A5 anticoagulant activity in women with recurrent spontaneous pregnancy losses. Am J Obstet Gynecol 2006;194:182–8. Bogdanova N, Horst J, Chlystun M, et al. A common haplotype of the annexin A5 (ANXA5) gene promoter is associated with recurrent pregnancy loss. Hum Mol Genet 2007;16:573–8. Markoff A, Gerdes S, Feldner S, et al. Reduced allele specific annexin A5 mRNA levels in placentas carrying the M2/ANXA5 allele. Placenta 2010;31:937–40. Tiscia G, Colaizzo D, Chinni E, et al. Haplotype M2 in the annexin A5 (ANXA5) gene and the occurrence of obstetric complications. Thromb Haemost 2009;102:309–13. Miyamura H, Nishizawa H, Ota S, et al. Polymorphisms in the annexin A5 gene promoter in Japanese women with recurrent pregnancy loss. Mol Hum Reprod 2011;17:447–52. Ueki H, Mizushina T, Laoharatchatathanin T, et al. Loss of maternal annexin A5 increases the likelihood of placental platelet thrombosis and foetal loss. Sci Rep 2012;2:827. Rowe P, Comhaire F, Hargreave T, et al. WHO manual for the standardized investigation and diagnosis of the infertile couple. World Health Organization: Cambridge University Press; 1993. Rogenhofer N, Engels L, Bogdanova N, et al. Paternal and maternal carriage of the annexin A5 M2 haplotype are equal risk factors for recurrent pregnancy loss: a pilot study. Fertil Steril 2012;98:383–8. Glueck CJ, Phillips H, Cameron D, et al. Continuing metformin throughout pregnancy in women with polycystic ovary syndrome appears to safely reduce first-trimester spontaneous abortion: a pilot study. Fertil Steril 2001;75: 46–52. Homburg R. Pregnancy complications in PCOS. Best Pract Res Clin Endocrinol Metab 2006;20:281–92. Ramidi G, Khan N, Glueck CJ, et al. Enoxaparin–metformin and enoxaparin alone may safely reduce pregnancy loss. Transl Res 2009;153:33–43.