Hum. Reprod. Advance Access originally published online on September 5, 2007
Human Reproduction 2007 22(11):2829-2833; doi:10.1093/humrep/dem264
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Strong evidence that skewed X-chromosome inactivation is not associated with recurrent pregnancy loss: an incident paired case–control study
1 EA 3878 (GETBO), Department of Internal Medicine and Chest Diseases, Brest University Hospital, La Cavale Blanche Hospital, 29609 Brest, France 2 Division of Gynaecology, Brest University Hospital, Morvan Hospital, 29220 Brest, France 3 INSERM U613, EFS, 29220 Brest, France
4 Correspondence address. Tel: +33-298-347-336; Fax: +33-298-347-944; E-mail: elisabeth.pasquier{at}chu-brest.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
BACKGROUND: Previous studies have reported conflicting results regarding recurrent pregnancy loss and skewed X-chromosome inactivation. Hence, we sought an association by carrying out a specifically designed incident paired case–control study with required statistical power.
METHODS: Design incident 1:3 matched case–control study, from 2003 to 2007. Setting: University Hospital of Brest. Patients: Women, from the Brittany area, consecutively referred for at least two unexplained consecutive spontaneous abortions. Controls: Women from the same geographic area, with no history of pregnancy loss and at least one normal pregnancy, recruited using electoral lists and then paired with cases, with respect to age, to within 1 year. Intervention: Assessment of skewed X-chromosome inactivation. Statistical analysis: Comparison of the ratio of >90% skewed X-chromosome inactivation by conditional logistic regression.
RESULTS: Five hundred and forty-three controls (mean age: 34.3 years) were paired within 1 year to 200 cases. The cases (mean age: 33.6 years) had experienced between 2 and 14 consecutive losses (median 3). The rate of >90% skewed X-chromosome inactivation was not statistically different (P = 0.33, odds ratio: 0.58, 95% confidence interval: 0.19–1.77) between cases and paired controls, 2.27% versus 4.1%, respectively.
CONCLUSIONS: We conclude that there is no association between skewed X-chromosome inactivation and recurrent pregnancy loss, defined as two or more unexplained consecutive spontaneous abortions.
Key words: paired case–control study/recurrent spontaneous abortion/skewed X-chromosome inactivation
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Recurrent spontaneous abortion (RSA), defined as three or more consecutive pregnancy losses before 20–22 weeks of gestation (Stirrat, 1990
; Christiansen et al., 2006), affects one to two percent of couples trying to have children (Brigham et al., 1999). A growing tendency is to use other criteria to define RSA, such as at least two consecutive or three or more non-consecutive losses (Campana et al., 1986; Gilchrist et al., 1991; Lanasa et al. 1999; Sangha et al., 1999; Lanasa et al., 2001; Robinson et al., 2001; Uehara et al., 2001). Nevertheless, the clinical relevance of such criteria needs to be assessed.
Standard anatomical, cytogenetic, hormonal and immunological investigations fail to reveal any apparent cause in about 50% of the cases (Stirrat, 1990; Hatasaka, 1994). Prothrombotic biological features, especially Factor V Leiden, Prothrombin G20 210A mutation and protein S deficiency, are implicated (Rey et al., 2003). However, they do not account for a large number of cases. Consequently, new markers are sought in an attempt to regroup patients with respect to the mechanism of their pregnancy loss. The aim is to design therapeutic trials applied to well-selected patients. Thus, circulating procoagulant microparticles (Laude et al., 2001; Carp et al., 2004) and soluble CD 146 (Pasquier et al., 2005) have been measured in order to explore prothrombotic and vascular dysfunctions, whereas other authors have drawn attention to a genetic hypothesis by studying the patterns of X-chromosome inactivation.
X-chromosome inactivation is a physiological event consisting of a methylation sensitive transcription inactivation of one of the two X-chromosomes in each somatic cell of healthy human females. This inactivation most often occurs randomly between the maternal and paternal X-chromosomes. All women are therefore mosaic, with about one half of their cells expressing the paternally derived X, the others expressing the maternal X-chromosome. However, some women preferentially express the X-chromosome of just one parent in most of their cells. This is known as skewed X-chromosome inactivation (Migeon, 1998). A weak but significant positive correlation was observed between age and the degree of X-skewing over the whole age range. Skewing values become higher at older ages, from 30 to 55 years according to the studies of Hatakeyama et al. (2004) and Knudsen et al. (2007).
Four previous reports (Lanasa et al., 1999, 2001; Sangha et al., 1999; Uehara et al., 2001) have described an increased skewed X-chromosome inactivation in women with idiopathic RSA compared with controls. The authors drew a parallel between some X-linked syndromes, where female carriers exhibit highly skewed X-chromosome inactivation, and where mutations are lethal to hemizygous male offspring who inherit them (Devriendt et al., 1997). Consequently, it was first hypothesized that some women who experienced spontaneous abortions were carrying hidden X-linked mutations which underlie both skewed X-chromosome inactivation and RSA of the male fetus (Lanasa et al., 1999). Another mechanism has been proposed: a reduction in the number of ovarian follicles by restriction in the precursor pool size during development or by oocyte atresia due to X mutations (Robinson et al., 2001).
We found these first preliminary studies (Lanasa et al.; 1999, 2001; Sangha et al.; 1999; Uehara et al., 2001) attractive, despite some shortcomings in the method used (either the design of the study, the size of the case and control samples, the difference in age between the compared groups or the type of controls). We therefore initiated in February 2003, an incident paired case–control study specially designed to settle the question of an association between skewed X-chromosome inactivation and unexplained RSA.
Thus, we consecutively enrolled a large number of cases and, above all, we focused on the recruitment and selection of age-matched control women.
On the basis of the availability of a well-defined methylation sensitive assay and on the high informativity rate among the Caucasian population (heterozygosity), the human androgen receptor locus (AR gene, HUMARA, Xq11–12) was used to differentiate between the active and inactive X-chromosome as well as between its maternal or paternal origin, respectively (Allen et al., 1992).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Study design and setting
The study was set up to compare >90% skewed X-chromosome inactivation between women referred for recurrent pregnancy loss, defined as two or more unexplained consecutive spontaneous abortions, and age-matched control women.
This was a 1:3 matched incident case–control study which took place from February 2003 to March 2007, at the University Hospital of Brest, France. The women gave their informed consent. The study was approved by the local ethics committee, the CPP of Brest University Hospital. The participants were seen once, in regard to this study, by one of the investigators for a medical review using a standard questionnaire and venous blood sampling.
Cases
All the women were from the Brittany area, aged between 18 and 45 years, and consecutively referred for a history of unexplained recurrent pregnancy loss, by the obstetricians of the area, whether in private and/or public practice. Unexplained recurrent pregnancy loss was defined as two or more unexplained consecutive pregnancy losses at/or before 21 weeks of gestation.
Exclusion criteria consisted of: a maternal or paternal carrier of a structural chromosomal rearrangement, maternal antiphospholipid antibodies or any anatomical abnormality possibly responsible for pregnancy loss. A standard evaluation comprised maternal testing for lupus anticoagulant, determination of immunoglobulin G and M against cardiolipin and ß-2 glycoprotein-1, hysterosalpingogram or a hysteroscopy, maternal and paternal cytogenetic analysis.
Controls
The controls were recruited during the same period, among women aged between 18 and 45 years from the same geographic area, using electoral lists. They were sent a letter with exhaustive information on the study. If a control wished to participate, she contacted us by phone to plan a medical review with an investigator. Women were potentially eligible if they had given birth to at least one living child. Non-inclusion criteria included pregnancy loss and preclinical spontaneous abortions.
Samples
Peripheral blood was collected in EDTA. DNA was extracted from blood lymphocytes by a standard salting out procedure and concentrations were normalized by picogreen fluorescence measurement (Miller et al., 1988).
Quantitative X-chromosome inactivation assay
The X-chromosome inactivation assay was performed at the CAG repeat of the AR locus by a methylation-specific enzymatic digestion with HpaII enzyme as described by Allen et al. (1992). The PCR surrounding the CAG repeat was carried out with 100 ng of digested and undigested DNA, with a fluorescent (FAM labelled) forward primer and then amplicons were electrophoresed on a Megabace 500 (GE Healthcare) automated sequencer. Peak heights were analysed using Genetic Profiler software (GE Healthcare) and normalized in relation to the undigested amplification according to Lanasa et al. (2001) in order to estimate the degree of inactivation considered as skewed (positive) at >90% and random (negative) at 
90%. All borderline results above 80% of preferential inactivation were repeated at least twice in an independent experiment and averaged. Homozygous genotypes, i.e. the same CAG repeat in both alleles, were uninformative and thus these samples were discarded and classified as non-informative for this assay.
Number of patients and controls required
The sample size was estimated on the basis of the data published in 2002 at the time of the design of this study (Table 1).
|
The expected mean age of our controls was between 30 and 35 years. Thus, we estimated at 5.5% the rate of skewed X-chromosome inactivation (defined as X-inactivation skewing >90%) in controls according to the data of Uehara et al. (2001)
and Sangha et al. (1999). In fact, in these two previous reports, the mean age of the controls was as in our study.
In other respects, as we used the same method for studying X-inactivation as Lanasa et al. (2001), we assumed that about 15% of women will not have an informative X-inactivation pattern due to homozygous genotype.
We planned to match one case with three controls with respect to age within 1 year. Thus, about 200 cases and 600 controls were required when setting the alpha risk at 5%, the odds ratio at 2.5 and the β risk at 20%.
Statistical analysis
The cases were paired with controls, to within 1 year of age, before any biological measurement was undertaken. Then, to compare the ratio of skewed X-chromosome inactivation between patients and controls, a conditional logistic regression was performed using SAS software (SAS v9, SAS Institute Inc, Cary, NC, USA). P < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
From February 2003 to March 2007, 200 cases were consecutively enrolled.
During the same period, 28 047 invitations for potential participation as a control subject were mailed. Twenty-three percentage (6453 letters) were returned because of wrong address. Finally, 600 controls were included after a medical review by an investigator.
Only 543 controls could be paired for age within 1 year to the 200 cases. Ages were normally distributed, the mean being 33.6 [95% confidence interval (CI): 32.8–34.2] and 34.3 (95% CI: 33.8–34.6) years for cases and controls, respectively.
Cases had experienced between 2 and 14 consecutive spontaneous abortions with a median of three spontaneous abortions and a mean gestation time of 7 weeks. Among these 200 women, 13 (6.5%) had also suffered a late pregnancy loss after 21 weeks of gestation. Controls had a median of two normal pregnancies.
Cases and controls were assessed simultaneously for their X-chromosome inactivation pattern. Twenty-four cases (12%) and 44 controls (8.1%) were non-informative regarding the X-chromosome inactivation assay (P = 0.13).
Finally, after excluding these 24 non-informative cases and their matched controls, we conducted the analysis on the remaining 176 cases and on their 440 informative controls, matched for age. Four cases (2.27%) and 18 controls (4.1%) had a skewed X-chromosome inactivation (P = 0.33, odds ratio: 0.58, 95% CI: 0.19–1.77, using conditional logistic regression analysis) (Fig. 1, Table 2).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
We did not find any association between recurrent pregnancy loss and skewed X-chromosome inactivation.
Nine studies, whose main objective was to seek a link between skewed X-chromosome inactivation and idiopathic RSA, have been published to date (Lanasa et al., 1999, 2001; Sangha et al., 1999; Uehara et al., 2001; Beever et al., 2003; Sullivan et al., 2003; Kim et al., 2004; Bagislar et al., 2006; Hogge et al., 2007) (Table 1). Whereas the first four report an association (Lanasa et al., 1999; Sangha et al., 1999; Lanasa et al., 2001; Uehara et al., 2001), the others, except one (Bagislar et al., 2006), do not. It must be pointed out that some of these studies include patients and/or controls of previous studies. The second study of Lanasa et al. and Hogge et al. includes cases and controls from prior studies. Likewise, the patients of Sangha et al. (1999) were also included in the study of Beever et al. (2003). Moreover, conclusions are sometimes conflicting, Lanasa et al. (1999, 2001) versus Hogge et al. (2007) and Sangha et al. (1999) versus Beever et al. (2003).
The discrepancies between the results of all previous studies may be partly related to differences in methodology. Four major points must be stressed: the type of control (age matched or not, with at least one normal pregnancy, from the same geographic area, enrolled simultaneously with the cases, and specifically or not for this particular study), the characteristics of the cases (the number of spontaneous abortions, consecutive or not), the design (retrospective versus incident case-control study), and the statistical power which partly depends on the size of the case and control samples. Thus, Hogge et al. and Lanasa et al. have hypothesized that their conflicting results could be related to initially small samples of cases and, above all, to the wide difference in age (36 versus 25 years) between patients and their first control group (Hogge et al., 2007).
In other respects, the controls were not usually prospectively enrolled. They were either healthy anonymous donors or recruited from one or more other studies which were sometimes designed for other purposes. This constitutes a well-known bias, which may result in misleading conclusions.
In this incident, paired-age case–control study, we did not find any association between recurrent pregnancy loss and skewed X-chromosome inactivation, although it was specifically designed to detect a ‘true link’.
One of the strong points of our study lies in the highly time-consuming recruitment of 600 controls from the electoral lists, with the aim of conducting a real bias-free incident case–control study as much as possible. Thus, the enrolment of cases and controls was carried out in parallel in the same geographic area at the same period. The controls had on average two healthy children and had not experienced any pregnancy loss. The major confounder for skewed X-chromosome inactivation is age. This was avoided by matching the cases and controls according to age within 1 year. The assessment of the degree of skewing was performed simultaneously for the cases and controls by the most commonly used method of measurement. The rate of non-informative patients and controls was not statistically different, 12% versus 8.1%, and was close to the lower rates of non-informative skewing observed in previous studies. We chose the cut-off value of skewed X-chromosome inactivation at 90%, like most, but not all the prior studies. Using a less stringent cut-off of 90%, rather than 95%, could theoretically impact the results. However, the choice of a less stringent cut-off might be less clinically relevant in identifying ‘a true link’.
The criterion of at least two consecutive spontaneous abortions could have had a negative impact on the results. In fact, when a 15% rate of clinically detected pregnancy losses is taken into account, the risk of three random consecutive spontaneous abortions (0.33%) is clearly below the observed prevalence of three consecutive spontaneous abortions RSA (1–2%). Nevertheless, women are often referred for care after two losses rather than three. Hence, the criteria used in RSA studies have increasingly been changed into either two consecutive (Uehara et al., 2001; Kim et al., 2004) or three non-consecutive losses (Bagislar et al., 2006) and even two non-consecutive losses (Lanasa et al., 1999, 2001; Hogge et al., 2007).
The latest study published on RSA and X-chromosome inactivation (Hogge et al., 2007) is noteworthy due to its recruitment of 357 women with at least two spontaneous abortions. They did not find any skewing differences when taking into account the criterion of at least two versus three spontaneous abortions.
In other respects, despite the large enrolment of age-matched women, our study could still be considered insufficient in detecting a small but significant difference in skewing. Nevertheless, we point out that there is a higher rate of skewing in controls compared with cases (4.1% versus 2.27%) and that the upper limit of the 95% CI of the odds ratio computed by conditional logistic regression is under 2.
To conclude, we agree with Hogge et al. (2007), who also did not demonstrate an association between RSA and skewed X-chromosome inactivation. Therefore, presently there is no clinical indication to test for skewed X-chromosome inactivation in unexplained RSA.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
The study was supported by a grant (PHRC) from the French Health Office. We wish to thank Gynobs 29 and the obstetricians in Brittany who referred their patients.
|
Footnotes |
|---|
E.P. designed the study, performed the clinical part of the research and wrote the paper. C.B. performed the clinical part of the research and wrote the paper. L.De S.M. analysed data. M.T.Le M., S.R., Y.L., M.C. and D.M. performed the clinical part of the research. C.Le M. and C.F. carried out the biological part of the research.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAcknowledgementsReferences |
|---|
Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM, Belmont JW. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X chromosome inactivation. Am J Hum Genet (1992) 51:1229–1239.[Web of Science][Medline]
Bagislar S, Ustuner I, Cengiz B, Soylemez F, Akyerli CB, Ceylaner S, Ceylaner G, Acar A, Ozcelik T. Extremely skewed X-chromosome inactivation patterns in women with recurrent spontaneous abortion. Aust N Z J Obstet Gynaecol (2006) 46:384–387.[CrossRef][Web of Science][Medline]
Beever CL, Stephenson MD, Penaherrera MS, Jiang RH, Kalousek DK, Hayden M, Field L, Brown CJ, Robinson WP. Skewed X-chromosome inactivation is associated with trisomy in women ascertained on the basis of recurrent spontaneous abortion or chromosomally abnormal pregnancies. Am J Hum Genet (2003) 72:399–407.[CrossRef][Web of Science][Medline]
Brigham SA, Conlon C, Farquharson RG. A longitudinal study of pregnancy outcome following idiopathic recurrent miscarriage. Hum Reprod (1999) 14:2868–2871.
Campana M, Serra A, Neri G. Role of chromosome aberrations in recurrent abortion: a study of 269 balanced translocations. Am J Med Genet (1986) 24:341–356.[CrossRef][Web of Science][Medline]
Carp H, Dardik R, Lubetsky A, Salomon O, Eskaraev R, Rosenthal E, Inbal A. Prevalence of circulating procoagulant microparticles in women with recurrent miscarriage: a case-controlled study. Hum Reprod (2004) 19:191–195.
Christiansen OB, Nielsen HS, Kolte A, Pedersen AT. Research methodology and epidemiology of relevance in recurrent pregnancy loss. Semin Reprod Med (2006) 24:5–16.[CrossRef][Web of Science][Medline]
Devriendt K, Matthijs G, Legius E, Schollen E, Blockmans D, van Geet C, Degreef H, Cassiman JJ, Fryns JP. Skewed X-chromosome inactivation in female carriers of dyskeratosis congenita. Am J Hum Genet (1997) 60:581–587.[Web of Science][Medline]
Gilchrist DM, Livingston JE, Hurlburt JA, Wilson RD. Recurrent spontaneous pregnancy loss. Investigation and reproductive follow-up. J Reprod Med (1991) 36:184–188.[Web of Science][Medline]
Hatakeyama C, Anderson CL, Beever CL, Penaherreraa MS, Brown CJ, Robinsona WP. The dynamics of X-inactivation skewing as women age. Clin Genet (2004) 66:327–332.[CrossRef][Web of Science][Medline]
Hatasaka HH. Recurrent miscarriage: epidemiologic factors, definitions, and incidence. Clin Obstet Gynecol (1994) 37:625–634.[CrossRef][Web of Science][Medline]
Hogge WA, Prosen TL, Lanasa MC, Heather AH, Matthew FR. Recurrent spontaneous abortion and skewed X-inactivation: is there an association? Am J Obstet Gynecol (2007) 196:384–386.[Medline]
Kim JW, Park SY, Kim YM, Kim JM, Han JY, Ryu HM. X-chromosome inactivation patterns in Korean women with idiopathic recurrent spontaneous abortion. J Korean Med Sci (2004) 19:258–262.[Web of Science][Medline]
Knudsen GP, Pedersen J, Klingenberg O, Lygren I, Ørstavik KH. Increased skewing of X chromosome inactivation with age in both blood and buccal cells. Cytogenet Genome Res (2007) 116:24–28.[CrossRef][Web of Science][Medline]
Lanasa MC, Hogge WA, Kubik C, Blancato J, Hoffman EP. Highly skewed X-chromosome inactivation is associated with idiopathic recurrent spontaneous abortion. Am J Hum Genet (1999) 65:252–254.[CrossRef][Web of Science][Medline]
Lanasa MC, Hogge WA, Kubik CJ, Ness RB, Harger J, Nagel T, Prosen T, Markovic N, Hoffman EP. A novel X chromosome-linked genetic cause of recurrent spontaneous abortion. Am J Obstet Gynecol (2001) 185:563–568.[CrossRef][Web of Science][Medline]
Laude I, Rongieres-Bertrand C, Boyer-Neumann C, Wolf M, Mairovitz V, Hugel B, Freyssinet JM, Frydman R, Meyer D, Eschwège V. Circulating procoagulant microparticles in women with unexplained pregnancy loss: a new insight. Thromb Haemost (2001) 85:18–21.[Web of Science][Medline]
Migeon BR. Non-random X chromosome inactivation in mammalian cells. Cytogenet Cell Genet (1998) 80:142–148.[CrossRef][Web of Science][Medline]
Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res (1988) 16:1215.
Pasquier E, Bardin N, De Saint Martin L, Le Martelot MT, Bohec C, Roche S, Mottier D, Dignat-George F. The first assessment of soluble CD146 in women with unexplained pregnancy loss. A new insight? Thromb Haemost (2005) 94:1280–1284.[Web of Science][Medline]
Rey E, Kahn SR, David M, Shrier I. Thrombophilic disorders and fetal loss: a meta-analysis. Lancet (2003) 361:901–908.[CrossRef][Web of Science][Medline]
Robinson WP, Beever C, Brown CJ, Stephenson MD. Skewed X inactivation and recurrent spontaneous abortion. Semin Reprod Med (2001) 19:175–181.[CrossRef][Web of Science][Medline]
Sangha KK, Stephenson MD, Brown CJ, Robinson WP. Extremely skewed X-chromosome inactivation is increased in women with recurrent spontaneous abortion. Am J Hum Genet (1999) 65:913–917.[CrossRef][Web of Science][Medline]
Stirrat GM. Recurrent miscarriage. Lancet (1990) 336:673–675.[CrossRef][Web of Science][Medline]
Sullivan AE, Lewis T, Stephenson M, Odem R, Schreiber J, Ober C, Branch DW. Pregnancy outcome in recurrent miscarriage patients with skewed X chromosome inactivation. Obstet Gynecol (2003) 101:1236–1242.[CrossRef][Web of Science][Medline]
Uehara S, Hashiyada M, Sato K, Sato Y, Fujimori K, Okamura K. Preferential X-chromosome inactivation in women with idiopathic recurrent pregnancy loss. Fertil Steril (2001) 76:908–914.[CrossRef][Web of Science][Medline]
Submitted on May 3, 2007; resubmitted on July 13, 2007; accepted on July 24, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||