Association of 14-bp insertion/deletion polymorphism of HLA-G gene with unexplained recurrent spontaneous abortion: A meta-analysis

Abstract

A meta-analysis was performed to assess the association of the human leukocyte antigen-G (HLA-G) 14 bp insertion (ins)/deletion (del) polymorphism with unexplained recurrent spontaneous abortions (URSA). The fixed or random effect pooled measure was selected based on the homogeneity test among studies that was evaluated with I(2) . Odds ratio (OR) with 95% confidence interval (CI) was used to assess the strength of the association. In this meta-analysis, 14 studies with 1464 cases and 1247 controls corresponded to the inclusion criteria were included. Significant associations between 14 bp ins/del polymorphism and risk of URSA were observed in both dominant [random effect model (REM) OR = 1.469, 95% CI = 1.127-1.914] and codominant (REM OR = 1.195, 95% CI = 1.005-1.420) models. After excluding two articles that deviated from Hardy-Weinberg equilibrium in cases and sensitivity analysis, significant associations were also observed in dominant [fixed effect model (FEM) OR = 1.224, 95% CI = 1.020-1.470] and codominant (FEM OR = 1.158, 95% CI = 1.028-1.305) models. This meta-analysis suggests that the 14 bp ins HLA-G allele is associated with increased risk of URSA. The results are of importance to the clinical practice of URSA and infertility.

Full text

Tissue Antigens ISSN 0001-2815
Association of 14-bp insertion/deletion polymorphism
of
HLA-G
gene with unexplained recurrent spontaneous
abortion: a meta-analysis
X. Wang†,W.Jiang
†& D. Zhang
Department of Public Health, Qingdao University Medical College, 266021, Qingdao, China
Key words
HLA-G
; meta-analysis; polymorphism;
unexplained recurrent spontaneous abortion
Correspondence
Wenjie Jiang
Department of Public Health
Qingdao University Medical College
Deng Zhou Street 38
266021 Qingdao
China
Tel: +86 532 82991712
Fax: +86 532 83801449
e-mail: wenjie-jiang@126.com
Received 17 September 2012; revised 22
November 2012; accepted 24 December
2012
doi: 10.1111/tan.12056
Abstract
A meta-analysis was performed to assess the association of the human leukocyte
antigen-G (HLA-G) 14 bp insertion (ins)/deletion (del) polymorphism with unex-
plained recurrent spontaneous abortions (URSA). The fixed or random effect pooled
measure was selected based on the homogeneity test among studies that was evalu-
ated with I2. Odds ratio (OR) with 95% confidence interval (CI) was used to assess
the strength of the association. In this meta-analysis, 14 studies with 1464 cases and
1247 controls corresponded to the inclusion criteria were included. Significant asso-
ciations between 14 bp ins/del polymorphism and risk of URSA were observed in
both dominant [random effect model (REM) OR =1.469, 95% CI =1.127– 1.914]
and codominant (REM OR =1.195, 95% CI =1.005–1.420) models. After excluding
two articles that deviated from Hardy–Weinberg equilibrium in cases and sensitivity
analysis, significant associations were also observed in dominant [fixed effect model
(FEM) OR =1.224, 95% CI =1.020 –1.470] and codominant (FEM OR =1.158, 95%
CI =1.028–1.305) models. This meta-analysis suggests that the 14 bp ins HLA-G
allele is associated with increased risk of URSA. The results are of importance to the
clinical practice of URSA and infertility.
Introduction
Recurrent spontaneous abortions (RSA), defined as three or
more first trimester spontaneous abortions prior to the 20th
week of pregnancy (1), is a major concern for reproductive
medicine and occurs in about 2%–5% of otherwise healthy
women (2). After ruling out identifiable causes, including
chromosomal anomalies, anatomical anomalies, endocrine
abnormalities and so forth, about 50% of RSA cases the
pathophysiology remains unexplained. These cases are,
therefore, considered as unexplained recurrent spontaneous
abortions (URSA) (3).
Research on immunepathogenesis indicates that URSA is
associated with the unbalanced maternofetal immunological
tolerance. As a non-classical major histocompatibility com-
plex class I molecule, HLA-G is expressed predominantly and
restrictedly in extravillous trophoblast cells at maternal–fetal
interface. Experiments in vitro showed that HLA-G may
contribute to maternal acceptance of the semi-allogenic
foetus, by suppressing the maternal immune system during
†These authors contributed equally to this work.
pregnancy (4–7). Among its several limited polymorphisms,
the 14 bp ins/del polymorphism at 3untranslated region
(3UTR) of HLA-G gene has been shown an important role in
post-transcriptional regulation of HLA-G molecule (8). Two
studies have found that alleles including 14 bp insertion are
related to lower level of mRNA (9, 10). With 94 healthy unre-
lated individuals, Rebmann et al. have found that individuals
carrying the HLA-G*01013 or HLA-G*0105 N allele (both
contain the 14 bp sequence) has shown significantly reduced
sHLA-G levels (11). Research on 150 normal Chinese has
found a dramatically lower sHLA-G level in plasma with
HLA-G 14 bp ins/ins genotype than that with 14 bp ins/del
and 14 bp del/del genotypes (12). Since the 14 bp ins/del
polymorphism was first reported by Harrison and colleagues
in 1993 (13), studies on the association between HLA-G
14 bp ins/del polymorphism and URSA have been performed.
However, no clear consensus has been reached. Because of
the inconsistent results from relatively small studies with
underpowered to detect the effect, this study reporting the
results of a meta-analysis including all relevant studies
is important in this specific area of human reproductive
immunology.
108 ©2013 John Wiley & Sons A/S
Tissue Antigens
, 2013, 81, 108– 115

X. Wang et al.
HLA-G
gene and unexplained recurrent spontaneous abortion
Materials and methods
Study selection
The search of all relevant available studies (up to November
2012) was conducted through the following five electronic
databases: (1) Pubmed; (2) CBM (China Biological Medical
literature database); (3) CNKI (China National Knowledge
Infrastructure); (4) VIP (Database of Chinese Scientific
and Technical Periodicals); (5) Web of Science (ISI). The
search was performed using the following keywords: ‘poly-
morphism’ or ‘mutation’; ‘HLA-G’ or ‘human leukocyte
antigen-G’, in combination with ‘URSA’, ‘unexplained recur-
rent spontaneous abortions’; ‘RSA’, ‘recurrent spontaneous
abortion’; ‘RPL’, ‘recurrent pregnancy loss’; ‘RM’, ‘recurrent
miscarriage’; ‘RFL’, ‘recurrent fetal loss’. A manual search
was also conducted for additional studies by reviewing the
bibliographies of the relevant original studies and reviews.
The inclusion criteria were as follows: (1) case–control
or cohort study to evaluate the association between the 14 bp
ins/del polymorphism and the risk of URSA; (2) numbers of
the genotype were reported in case and control groups for
case–control studies, or exposed and unexposed groups for
cohort studies, or data provided from which numbers could
be calculated; (3) subjects in each study should come from
the same period and ethnicity and (4) studies published in
English or Chinese.
Data extraction
Two investigators collected the data independently. If more
than one of the same patient population was included in
several publications, only the complete or most recent study
was included. If the two investigators disagreed about the
eligibility of an article, it was resolved by consensus with a
third reviewer.
Information was carefully extracted from all eligible pub-
lications by two investigators who reached a consensus on
all of the items. From each study, the following information
was extracted: first author, country, year of publication, ethnic
category, sample size, typing method, eligibility criteria, con-
clusion achieved in each study, distributions of genotype and
allele for both case and control groups.
Statistical analysis
Hardy–Weinberg equilibrium
Hardy–Weinberg equilibrium (HWE) for 14 bp ins/del poly-
morphism of HLA-G gene in cases and controls was tested by
chi-square analysis.
Quantitative synthesis
Pooled measure was calculated, as the inverse variance-
weighted mean of the logarithm of odds ratio (OR) with 95%
confidence interval (CI), to assess the strength of association
of the 14 bp ins/del polymorphism and the risk of URSA for
dominant (14 bp ins/ins and 14 bp ins/del vs 14 bp del/del),
recessive(14 bp ins/ins vs 14 bp ins/del and 14 bp del/del) and
codominant(14 bp ins vs 14 bp del) models, respectively.
Heterogeneity
Between-study heterogeneity was assessed by the intraclass
correlation coefficient I2statistic (ranging from 0% to 100%),
which was documented for the percentage of the observed
between-study variability due to heterogeneity rather than
chance. If substantial heterogeneity is present (I2>50%)
(14), the DerSimonian and Laird (15) random effect model
(REM) was adopted as the pooling method; otherwise, the
fixed effect model (FEM) was used as the pooling method.
Pooled ORs were in both of the FEM and REM models
and under both models the pooled ORs were presented for a
clearer comparison.
Meta-regression
Meta-regression with restricted maximum likelihood
estimation was performed to explore the potentially important
covariates: ethnicity (categorized as Asian and Europe) and
publication year that might have substantial impacts on
between-study heterogeneity.
Sensitivity analysis
Theoretical considerations and empirical evidence suggest that
specific genetic variants causally associated with common
diseases will have small effects (risk ratios mostly <2.0)
(16, 17), and considering the fact that original studies with
relatively small participants might be underpowered to detect
the effect. Thus, for sensitivity analysis, we excluded the
studies with OR >3.0 as the criteria to control the impact
of outlier values resulting from low cell counts within each
single study on the pooled effect.
Influence analysis
Influence analysis was conducted, which described how robust
the pooled estimator is for the removal of individual studies.
An individual study is suspected of excessive influence, if the
point estimate of its omitted analysis lies outside the 95% CI
of the combined analysis.
Publication bias
Small-study effect in terms of publication bias was estimated
using modified Egger’s linear regression test as proposed by
Peters et al. (18).
Software used
HWE was tested by the chi-square test using a web-based
programme (http://www.ihg.gsf.de/cgi-bin/hw/hwa1.pl). All
©2013 John Wiley & Sons A/S 109
Tissue Antigens
, 2013, 81, 108– 115

HLA-G
gene and unexplained recurrent spontaneous abortion X. Wang et al.
Records identified through database searching and other
sources (n=64)
The records excluded (n=34)
Through reading the topic and abstract
The studies included in this meta-analysis (n=14)
The records excluded (n=4)
For repetitive publication
Full-text articles on reference gene polymorphisms for
detailed evaluation (n=18)
The records excluded (n=12)
(1) Three review article
(2) One with birth weight
(3) Four with pre-eclampsia
(4) Four reported a link with mRNA
expression
Records screened for 14bp ins/del polymorphisms (n=30)
Figure 1 The flow diagram for selection
process of the study.
meta-analyses were performed with stata version 10.0 (Stata
Corporation, College Station, TX). All reported probabilities
(P-values) were two-sided, with P<0.05 considered statisti-
cally significant.
Results
Characteristics of eligible studies
The process of study selection is shown in Figure 1. Sixty-
four citations were identified in the primary literature search.
After screening the titles and abstracts, 30 articles remained
for further evaluation. After examining those articles in detail,
12 articles were excluded because they were either review
articles or irrelevant to this study. In total, 18 articles were
retrieved and reviewed by two authors. Four studies were
excluded because of repeated publication. Finally, a total of
14 studies were retrieved for the meta-analysis, including a
total of 1464 cases and 1247 controls. All 14 studies were
case–control design. General characteristics and the 14 bp
ins/del polymorphism genotype distributions in the 14 studies
are shown in Table 1.
Quantitative synthesis
The results of the pooled analysis are summarized in Table 2.
The meta-analysis showed a significant association of 14 bp
ins HLA-G allele with increased risk of URSA in dominant
(REM OR =1.469, 95% CI =1.127 – 1.914) and codominant
(REM OR =1.195, 95% CI =1.005 – 1.420) models, no sig-
nificant association was found in recessive (REM OR =1.148,
95% CI =0.821–1.607) model. All articles of the control
were in HWE and there were two articles (19, 20) of the
case deviating from HWE. After excluding those two arti-
cles, associations were not altered significantly in both domi-
nant (REM OR =1.440, 95% CI =1.085–1.911) and codom-
inant (REM OR =1.211, 95% CI =1.013–1.448) models,
and no significant association was found in recessive (REM
OR =1.254, 95% CI =0.912–1.726) model.
Sources of heterogeneity
After exclusion of articles (19, 20) deviating from HWE
in cases, strong evidence for heterogeneity (I2>50%) was
found in the dominant, recessive and codominant models.
Exploratory univariate meta-regression with the covariates of
ethnicity (categorized as Asia and Europe) and publication
year as the potential sources of between-study heterogene-
ity showed that no covariates had a significant impact on
between-study heterogeneity.
Sensitivity analysis
After exclusion of articles (19, 20) deviating from HWE in
cases and sensitivity analysis of excluding the studies (21,
32) with OR >3.0 (shown in Table 2), low heterogene-
ity (I2<50%) was found in the inherited models, and the
meta-analysis also showed a significant association of 14 bp
ins HLA-G allele with increased risk of URSA in dominant
(FEM OR =1.224, 95% CI =1.020 – 1.470) (Figure 2) and
codominant (FEM OR =1.158, 95% CI =1.028 – 1.305) mod-
els, but no significant association was found in recessive (FEM
OR =1.203, 95% CI =0.975–1.484) model.
Influence analysis
After exclusion of articles (19, 20) deviating from HWE in
cases and sensitivity analysis of excluding the studies (21,
32) with OR >3.0 (shown in Table 2), in any of dominant
(Figure 3), recessive and codominant models, no individual
study was found having excessive influence on the pooled
effect considering all the comparison groups (data not shown).
110 ©2013 John Wiley & Sons A/S
Tissue Antigens
, 2013, 81, 108– 115

X. Wang et al.
HLA-G
gene and unexplained recurrent spontaneous abortion
Table 1 Characteristics of 14 bp ins/del polymorphism genotype distributions in studies included in the meta-analysis
Case Control
Author
(references) Country Ethnicity
Publish
year
Number of
miscarriages
in case
Number of
pregnancies in
control
Typing
method
Case
number
Control
number
Genotype
(del/del)/
(ins/del)/(ins/ins)
P
a
Genotype
(del/del)/
(ins/del)/(ins/ins)
P
aResults
Xue et al. (19) China Asian 2007 ≥3≥1 PCR-SSP 24 88 6/17/1 0.018*44/38/6 0.563 14 bp ins/del genotype
Shao et al. (20) China Asian 2011 ≥2≥1 PCR-SSP 39 52 11/25/3 0.035*16/22/14 0.271 14 bp ins/del genotype
Shankarkumar
et al. (21)
India Asian 2011 ≥3≥1 PCR-SSP 50 41 25/17/8 0.102 33/7/1 0.419 14 bp ins/ins and 14 bp
ins/del genotype
Sipak-Szmigiel
et al. (22)
Poland Europe 2007 ≥3≥2 PCR-SSP 58 58 27/22/9 0.222 20/28/10 0.970 No association
Vargas et al. (23) Brazil Europe 2011 ≥3≥2 PCR-SSP 60 68 20/34/6 0.124 26/29/13 0.344 haplotypic combinations
of
HLA-G
alleles and
the 14 bp insertion
allele
Hviid et al. (24) Denmark Europe 2004 ≥3≥2 PCR-SSP 61 93 19/27/15 0.386 31/52/10 0.085 14bp ins/ins genotype
Yan et al. (25) China Asian 2006 ≥3≥2 PCR-SSP 79 109 7/37/35 0.526 19/57/33 0.508 Trend for 14 bp insertion
allele
Tripathi et al. (26) India Asian 2004 ≥3≥3 PCR-SSP 120 120 27/70/23 0.066 37/51/32 0.104 14 bp ins/del genotype
Aruna et al. (27) India Asian 2011 ≥2≥1 SBT 141 150 40/67/34 0.569 42/71/30 1.000 Trend for 14 bp del/del
genotype
Suryanarayana
et al. (28)
India Asian 2008 ≥3≥1 SBT 181 92 40/92/49 0.797 20/42/30 0.466 Association with
genotypes including
the 14 bp insertion
allele and a 1570 T-C
mutation
Zhu et al. (29) China Asian 2010 ≥3≥1 PCR-SSP 326 251 111/162/53 0.634 98/124/29 0.2756 Trend for 14 bp ins/ins
genotype. Association
only in women ≥4RM
Jassem et al. (30) Iraq Asian 2012 ≥2≥1 SNaPshot 49 48 10/17/22 0.067 13/21/14 0.388 No association
Christiansen et al.
(31)
Denmark Europe 2012 ≥1≥2 xMAP 286 78 87/148/51 0.383 32/37/9 0.730 Trend for 14 bp insertion
allele
Zhaobet al. (32) China Asian 2009 ≥2≥1 PCR-SSP 40 40 10/NA/NA NA 28/NA/NA NA 14 bp ins/ins and 14 bp
ins/del genotype
del/del, 14 bp del/del genotype; ins/del, 14 bp ins/del genotype; ins/ins, 14 bp genotype; NA, not available.
aExact
P
-value for Hardy– Weinberg equilibrium test.
bWe can only get the genotype counts for +14 bp/+14 bp ++14bp/−14 bp
vs
−14 bp/−14 bp from the original study.
*
P
<0.05.
©2013 John Wiley & Sons A/S 111
Tissue Antigens
, 2013, 81, 108– 115

HLA-G
gene and unexplained recurrent spontaneous abortion X. Wang et al.
Table 2 Pooled measures on the relations of the +14 bp/−14 bp polymorphism to URSAa
Inherited models FEM pooled OR (95% CI) REM pooled OR (95% CI)
I
2(%)
Articles
excluded (references)
All included articles Dominant 1.373 (1.157–1.629) 1.469 (1.127– 1.914) 52.2 0
Recessive 1.168 (0.951– 1.435) 1.148 (0.821 –1.607) 56.0 0
Codominant 1.178 (1.051 – 1.321) 1.195 (1.005 – 1.420 50.8 0
Exclude articles deviating
from Hardy– Weinberg
equilibrium
Dominant 1.350 (1.132–1.611) 1.440 (1.085– 1.911) 55.5 (19, 20)
Recessive 1.225 (0.994– 1.509) 1.254 (0.912 –1.726) 52.0 (19, 20)
Codominant 1.189 (1.057 – 1.338) 1.211 (1.013– 1.448) 51.5 (19, 20)
Sensitivity analysis further
excluding studies with OR
>3.0
Dominant 1.224 (1.020–1.470) 1.224 (1.020– 1.470) 0 (19, 21, 32, 20)
Recessive 1.203 (0.975– 1.484) 1.208 (0.888 –1.644) 49.9 (19, 21, 20)
Codominant 1.158 (1.028 – 1.305) 1.156 (1.004– 1.330) 23.8 (19, 21, 20)
CI, confidence interval; FEM, fixed effect model; OR, odds ratio; REM, random effect model; URSA, unexplained recurrent spontaneous abortions.
aDominant model: 14 bp ins/ins and 14 bp ins/del
vs
14 bp del/del. Recessive model: 14 bp ins/ins
vs
14 bp ins/del and14 bp del/del. Codominant
model: 14 bp ins
vs
14 bp del.
Overall (I-squared = 0.0%, p = 0.578)
Aruna,2011(27)(141/150)
author,year(ref.No)(case/control)
Yan,2006(25)(79/109)
Christiansen,2012(31)(286/78)
Hviid,2004(24)(61/93)
Zhu,2010(29)(326/251)
Tripathi,2004(26)(120/120)
Jassem,2012(30)(49/48)
Suryanarayana,2008(28)(181/92)
Vargas,2011(23)(60/68)
Sipak-Szmigiel,2007(22)(58/58)
1.22 (1.02, 1.47)
1.05 (0.63, 1.75)
OR (95% CI)
2.17 (0.87, 5.45)
1.59 (0.95, 2.67)
1.11 (0.55, 2.21)
1.24 (0.88, 1.75)
1.54 (0.86, 2.74)
1.45 (0.56, 3.72)
0.98 (0.53, 1.80)
1.24 (0.60, 2.56)
0.60 (0.29, 1.28)
100.00
12.69
3.95
12.53
6.98
28.65
10.03
3.77
9.07
6.35
5.98
Weight(%)
1.183 1 5.45
Figure 2 Forest plot of odds ratio (ORs) for 14 bp ins/del polymorphism and unexplained recurrent spontaneous abortions, in the dominant model
(14 bp ins/ins and 14bp ins/del
vs
14 bp del/del), after exclusion of articles (19, 20) deviating from Hardy–Weinberg equilibrium in cases and sensitivity
analysis of excluding the articles (21, 32) with OR >3. White diamond denotes the pooled OR. Black squares indicate the OR in each study, with
square sizes inversely proportional to the standard error of the OR. Horizontal lines represent 95% confidence intervals.
Publication bias
After exclusion of articles (19, 20) deviating from HWE in
cases and sensitivity analysis of excluding the studies (21, 32)
with OR >3.0 (shown in Table 2), no significant publication
bias was detected (data not shown) in any of the inherited
models mentioned above.
Discussion
URSA is related to aberrant maternofetal immunological tol-
erance. For more than two decades, HLA-G due to its unique
expression in extravillous trophoblast cells, has been supposed
to be an important immunoloregulation molecule. Studies
in vitro have shown that HLA-G has the capability of
inhibiting immune cell functions including natural killer cells,
cytotoxic T cells and dentritic cells (8). The 14 bp ins/del
polymorphism in exon 8 of HLA-G gene is a polymorphism
in the 3UTR of the gene, which may contribute importantly
to the regulation of HLA-G expression (33). The presence of
the 14 bp is related to the removal of 92 bp from the mature
transcript, which may result in a more stable mRNA, but
lower levels of mRNA are observed in URSA cases (10, 20).
112 ©2013 John Wiley & Sons A/S
Tissue Antigens
, 2013, 81, 108– 115

X. Wang et al.
HLA-G
gene and unexplained recurrent spontaneous abortion
0.97 1.22 1.02 1.47 1.55
Sipak-Szmigiel,2007(22)(58/58)
Vargas,2011(23)(60/68)
Hviid,2004(24)(61/93)
Yan,2006(25)(79/109)
Tripathi,2004(26)(120/120)
Aruna,2011(27)(141/150)
Suryanarayana,2008(28)(181/92)
Zhu,2010(29)(326/251)
Jassem,2012(30)(49/48)
Christiansen,2012(31)(286/78)
Lower CI Limit Estimate Upper CI Limit
Meta-analysis estimates, given named study is omitted
Figure 3 Analysis of influence of individual study on the pooled
estimate for 14 bp ins/del polymorphism and unexplained recurrent
spontaneous abortions, in the dominant model (14 bp ins/ins and 14 bp
ins/del
vs
14 bp del/del), after exclusion of articles (19, 20) deviating
from Hardy– Weinberg equilibrium in cases and sensitivity analysis of
excluding the articles (21, 32) with odds ratio (OR) >3. Open circle
indicates the pooled OR, given named study is omitted. Horizontal lines
represent the 95% confidence intervals.
The 14 bp sequence also exerts on the alternative splicing
that produces different HLA-G isoforms, and lower levels
of sHLA-G is observed in individuals with the 14 bp ins/ins
genotype (12, 32). Up to now, quite many studies have been
published that have investigated the possible associations of
14 bp ins/del polymorphism and URSA. Some have observed
a trend that more URSA women than normal fertile women
are 14 bp ins/ins genotype carriers (24, 25, 29, 31). Other
studies have found that 14 bp ins/del genotype has a higher
frequency in URSA women (19, 20, 26), and yet others have
reported an uncertain conclusion (21–23, 27, 28, 30, 32).
Because the inconsistent results from the studies men-
tioned above come from relatively small sample size, this
kind of small sample study is underpowered to detect the
effect. To elucidate the role of HLA-G polymorphism in the
outcome of pregnancy, we focused on the 14 bp ins/del poly-
morphism in the 3UTR of exon-8 of HLA-G gene in this
meta-analysis.
So for the first time, our meta-analysis summarized all the
available data on the association of the 14 bp ins/del poly-
morphism and URSA, including a total of 14 studies with
1464 cases and 1247 controls. All studies included in this
meta-analysis were case–control design, and general char-
acteristics including ethnic background, genotyping method
and source of DNA sample between case and control groups
in individual studies were comparable. Eligibility criteria for
case and control selection for each study were in coinci-
dence. All DNA samples obtained for genotyping were from
maternal peripheral blood. Therefore, data collected in this
meta-analysis were with reasonably high quality.
This meta-analysis provides a highly significant association
of the 14 bp ins allele of the 14 bp ins/del polymorphism in
HLA-G gene with increased risk of URSA. And significantly
lower sHLA-G levels or lower expression of HLA-G mRNA
were also observed in case group of some individual studies
(20, 30, 32).
In meta-analysis, between-study heterogeneity is common
in genetic association studies (34). This meta-analysis also
showed significant between-study heterogeneity in the above-
mentioned inherited models. Among the studies, an indeter-
minate number of characteristics that could be the causes of
the between-study heterogeneity, characteristics of the sample,
such as non-comparable measure of genotyping, the covariates
of ethnicity, deviation from HWE in some studies, etc. Thus,
we used meta-regression, which aimed to reduce the between-
study heterogeneity, to explore the potential important causes
of the between-study heterogeneity for studies. After exclud-
ing two articles (19, 20) that deviated from HWE in cases and
our meta-analysis did not identify any of the above-mentioned
covariates as being an important contributor to between-study
heterogeneity.
Noteworthy, we should pay attention to the outlier values
of OR that could cause significant findings by chance. The-
oretical considerations and empirical evidence suggest that
specific genetic variants causally associated with common
diseases should have small effects (OR mostly <2.0) (16,
17). Relative small sample size and possible genotyping
errors might also have an impact on the effect size. Besides,
very big effect estimates could be caused by unstable effect
estimations resulting from low cell counts within each single
study. Thus, we performed a sensitivity analysis further
excluding studies (21, 32) with OR >3.0 (shown in Table 2),
low heterogeneity (I2<50%) was found in the inherited
models, and our results also showed a significant association
of 14 bp ins allele with increased risk of URSA in both dom-
inant and codominant models, but no significance was found
in recessive model. Moreover, no significant individual study
influence on the pooled effect was observed in influence anal-
ysis. In this meta-analysis, no significant publication bias was
found.
Although our study showed positive results, there is a
limitation in this meta-analysis. For the biological specimen
was from maternal peripheral blood, however, HLA-G
is foetus origin indeed, so the genetic information from
the couple and the foetus will make the findings more
reliable.
In conclusion, more convincing evidence is required to
draw a solid conclusion on the relation between the 14 bp
ins/del polymorphism and the risk of URSA. To elucidate
the role of HLA-G in URSA, combining 14 bp ins/del
polymorphism and sHLA-G level in investigations and taking
the interaction of HLA-G and cytokines, interaction of 14 bp
ins/del polymorphism and other polymorphism of HLA-G
gene into consideration would be a better design.
©2013 John Wiley & Sons A/S 113
Tissue Antigens
, 2013, 81, 108– 115

HLA-G
gene and unexplained recurrent spontaneous abortion X. Wang et al.
Acknowledgment
No specific funding was obtained.
Conflict of Interest
The authors have declared no conflicting interests.
References
1. Jauniaux E, Farquharson RG, Christiansen OB, Exalto N.
Evidence-based guidelines for the investigation and medical
treatment of recurrent miscarriage. Hum Reprod 2006: 21:
2216–22.
2. Coulam CB, Clark DA, Beer AE et al. Current clinical options
for diagnosis and treatment of recurrent spontaneous abortion.
Clinical Guidelines Recommendation Committee for Diagnosis
and Treatment of Recurrent Spontaneous Abortion. Am J
Reprod Immunol 1997: 38: 57 – 74.
3. ACOG practice bulletin. Management of recurrent pregnancy
loss. Number 24, February 2001. (Replaces Technical Bulletin
Number 212, September 1995). American College of
Obstetricians and Gynecologists. IntJGynaecolObstet 2002:
78: 179–90.
4. Hunt J, Orr H. HLA and maternal-fetal recognition. FA S E B J
1992: 6: 2344–8.
5. Le Bouteiller P, Legrand-Abravanel F, Solier C. Soluble
HLA-G1 at the materno-foetal interface – a review. Placenta
2003: 24:S10–5.
6. LeMaoult J, Le Discorde M, Rouas-Freiss N et al. Biology and
functions of human leukocyte antigen-G in health and
sickness*. Tissue Antigens 2003: 62: 273 – 84.
7. Hunt JS, Petroff MG, McIntire RH, Ober C. HLA-G and
immune tolerance in pregnancy. FA S E B J 2005: 19: 681 – 93.
8. Dahl M, Hviid TVF. Human leucocyte antigen class Ib
molecules in pregnancy success and early pregnancy loss. Hum
Reprod Update 2012: 18: 92 – 109.
9. O’brien M, McCarthy T, Jenkins D et al. Altered HLA-G
transcription in pre-eclampsia is associated with allele specific
inheritance: possible role of the HLA-G gene in susceptibility
to the disease. Cell Mol Life Sci 2001: 58: 1943 – 9.
10. Hviid TV, Hylenius S, Rorbye C, Nielsen LG. HLA-G allelic
variants are associated with differences in the HLA-G mRNA
isoform profile and HLA-G mRNA levels. Immunogenetics
2003: 55: 63–79.
11. Rebmann V, van der Ven K, Passler M, Pfeiffer K, Krebs D,
Grosse-Wilde H. Association of soluble HLA-G plasma levels
with HLA-G alleles. Tissue Antigens 2001: 57: 15 – 21.
12. Chen XY, Yan WH, Lin A, Xu HH, Zhang JG, Wang XX. The
14 bp deletion polymorphisms in HLA-G gene play an
important role in the expression of soluble HLA-G in plasma.
Tissue Antigens 2008: 72: 335 – 41.
13. Harrison GA, Humphrey KE, Jakobsen IB, Cooper DW. A
14 bp deletion polymorphism in the HLA-G gene. Hum Mol
Genet 1993: 2: 2200.
14. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring
inconsistency in meta-analyses. BMJ 2003: 327: 557–60.
15. DerSimonian R, Laird N. Meta-analysis in clinical trials.
Control Clin Trials 1986: 7: 177–88.
16. Khoury MJ, Little J, Gwinn M, Ioannidis JP. On the synthesis
and interpretation of consistent but weak gene-disease
associations in the era of genome-wide association studies. Int
J Epidemiol 2007: 36: 439 – 45.
17. Ioannidis JP. Commentary: grading the credibility of molecular
evidence for complex diseases. Int J Epidemiol 2006: 35:
572–578; discussion 593 – 576.
18. Peters JL, Sutton AJ, Jones DR, Abrams KR, Rushton L.
Comparison of two methods to detect publication bias in
meta-analysis. JAMA 2006: 295: 676–80.
19. Xue S, Yang J, Yao F, Xu L, Fan L. Recurrent spontaneous
abortions patients have more −14 bp/+14 bp heterozygotes in
the 3UT region of the HLA-G gene in a Chinese Han
population. Tissue Antigens 2007: 69 (Suppl 1): 153 – 5.
20. Shao J, Hu D, Chen XJ, Liu Y, Wang Z. Relationship of
polymorphism of HLA-G 14 bp and the expression of HLA-G
mRNA with unexplained recurrent spontaneous abortion. Chin
J Pract Gynaecol Obstet 2011: 27: 698 –701.
21. Shankarkumar U, Shankarkumar A, Chedda Z, Ghosh K. Role
of 14-bp deletion/insertion polymorphism in exon 8 of the
HLA-G gene in recurrent spontaneous abortion patients. JHum
Reprod Sci 2011: 4: 143–6.
22. Sipak-Szmigiel O, Cybulski C, Lubinski J, Ronin-Walknowska
E. HLA-G polymorphism in a Polish population and
reproductive failure. Tissue Antigens 2008: 71: 67 – 71.
23. Vargas RG, Sarturi PR, Mattar SB et al. Association of
HLA-G alleles and 3UTR 14 bp haplotypes with recurrent
miscarriage in Brazilian couples. Hum Immunol 2011: 72:
479–85.
24. Hviid TV, Hylenius S, Lindhard A, Christiansen OB.
Association between human leukocyte antigen-G genotype and
success of in vitro fertilization and pregnancy outcome. Tissue
Antigens 2004: 64:66–9.
25. Yan WH, Lin A, Chen XJ et al. Association of the maternal
14-bp insertion polymorphism in the HLA-G gene in women
with recurrent spontaneous abortions. Tissue Antigens 2006: 68:
521–3.
26. Tripathi P, Abbas A, Naik S, Agrawal S. Role of 14-bp
deletion in the HLA-G gene in the maintenance of pregnancy.
Tissue Antigens 2004: 64: 706 – 10.
27. Aruna M, Sirisha PV, Andal Bhaskar S, Tarakeswari S,
Thangaraj K, Reddy BM. Role of 14-bp insertion/deletion
polymorphism in HLA-G among Indian women with recurrent
spontaneous abortions. Tissue Antigens 2011: 77: 131 – 5.
28. Suryanarayana V, Rao L, Kanakavalli M et al. Association
between novel HLA-G genotypes and risk of recurrent
miscarriages: a case–control study in a South Indian
population. Reprod Sci 2008: 15: 817–24.
29. Zhu Y, Huo Z, Lai J et al. Case – control study of a HLA-G
14-bp insertion-deletion polymorphism in women with
recurrent miscarriages. Scand J Immunol 2010: 71:
52–4.
30. Jassem RM, Shani WS, Loisel DA, Sharief M, Billstrand C,
Ober C. HLA-G polymorphisms and soluble HLA-G protein
levels in women with recurrent pregnancy loss from Basrah
province in Iraq. Hum Immunol 2012: 73: 811 – 7.
31. Christiansen OB, Kolte AM, Dahl M et al. Maternal
homozygocity for a 14 base pair insertion in exon 8 of the
HLA-G gene and carriage of HLA class II alleles restricting
114 ©2013 John Wiley & Sons A/S
Tissue Antigens
, 2013, 81, 108– 115

X. Wang et al.
HLA-G
gene and unexplained recurrent spontaneous abortion
HY immunity predispose to unexplained secondary recurrent
miscarriage and low birth weight in children born to these
patients. Hum Immunol 2012: 73: 699 – 705.
32. Zhao Y, Zheng M, Zhang G, Zhang Y, Hua A, He J. Serum
level of sHLA-G and HLA-G14 bp insertion / deletion
polymorphism with recurrent miscarriage. Shanxi Med J 2009:
38; 390–1.
33. Hviid TV, Rizzo R, Melchiorri L, Stignani M, Baricordi OR.
Polymorphism in the 5upstream regulatory and 3untranslated
regions of the HLA-G gene in relation to soluble HLA-G and
IL-10 expression. Hum Immunol 2006: 67: 53 – 62.
34. Munafo MR, Flint J. Meta-analysis of genetic association
studies. Trends Genet 2004: 20: 439 – 44.
©2013 John Wiley & Sons A/S 115
Tissue Antigens
, 2013, 81, 108– 115