Citation Christiansen OB, Steffensen R, Nielsen HS. Anti-HY-Responses in Pregnancy Disorders. Am J Reprod Immunol 2011; 66 (Suppl. 1): 93–100
Cellular and humoral immune responses against male-specific minor histocompatibility (HY) antigens are important in the pathogenesis of graft-versus-host reactions and can be detected in women who have previously given birth to a boy. However, the importance of these responses for pregnancy outcome is unclear.
Method of study
Review of the current knowledge about the impact of anti-HY immunity on pregnancy outcome in terms of risk of miscarriage, placental abruption and low birth weight.
Women with secondary recurrent miscarriage (RM) more often have given birth to a boy compared with a girl prior to a series of miscarriages (P < 0.0001) and a firstborn boy seems significantly to reduce the chance of a subsequent live in these patients (P = 0.0003). Human leukocyte antigen (HLA) class II alleles known to restrict CD4 T-cell-mediated anti-HY responses were investigated among 358 patients with secondary RM and 203 of their children born prior to the miscarriages. The chance of a subsequent live birth in secondary RM patients with firstborn boys compared with firstborn girls was significantly lower in women with HY-restricting HLA class II alleles [OR: 0.17 (0.1–0.4), P = 0.0001]. In patients without these alleles, the chance of live birth was similar in those with firstborn boys and girls, respectively. In early pregnancy, both antibodies against HLA and HY antigens were found with increased prevalence in secondary patients with RM and in particular in those with a firstborn boy compared with controls (P = 0.005). The presence of these antibodies was associated with a low subsequent live birth rate, and the presence of HY antibodies was associated with a low male/female ratio (12% boys versus 88% girls) in subsequent live births (P = 0.03). Register-based population studies suggested that births of boys also are associated with subsequent adverse obstetric outcomes also in the background population.
Findings strongly indicate that aberrant maternal immune reactions against fetal HY antigens playing a role in secondary RM and other pregnancy complications. We propose pathogenetic pathways for these conditions that in our view best explain the findings.
HY antigens are encoded by genes on the Y-chromosome and are ubiquitously expressed in male cells including fetal and trophoblast cells.1 Studies using human leukocyte antigen (HLA)-class-I- and HLA-class-II-restricted CD8+ and CD4+ T-cell clones have identified a series of peptides derived from HY proteins that are recognized by these clones.2 In the non-physiologic situation of stem cell transplantation (SCT), male recipients of parous female HLA-matched SC grafts are more prone to develop graft-versus-host disease (GvHD) than recipients of either nulliparous female grafts or male grafts suggesting that previous pregnancy of the female SC donor can enhance GvHD after SCT.3,4 In SCT, anti-HY antibodies develop significantly more often in male recipients with female SC donors than in those with male donors, and the presence of the antibodies correlates with chronic GvHD.5
In the physiologic situation of pregnancy, maternal cellular immune recognition of the HY antigens is described in women following pregnancies with boys.6,7 Furthermore, anti-HY antibodies are detected in approximately 30% of women, but are rare in healthy men.8 These immune responses were presumably caused by exposure to male cells through pregnancy.
In general, cellular and humoral anti-HY immunity in pregnancy must be considered well tolerated by the fetus because half of all newborns are healthy boys. However, a series of studies primarily from our group have provided evidence that in a minority of women, anti-HY immunity may cause complications in early and late pregnancy.
We here provide a review of the current knowledge of the impact of anti-HY immunity on pregnancy complications with special focus on recurrent miscarriage (RM), severe placental abruptions and low birth weight. This knowledge is based on epidemiologic studies, HLA investigations and measurements of anti-HY and anti-HLA antibodies.
The relevant epidemiologic studies are based on cohorts of patients with unexplained RM admitted to two tertiary RM centers in addition to register and cohort studies of women in the background population.
Secondary RM Studies
In the two cohort studies of patients with RM,9,10 secondary RM was defined as at least one pregnancy of minimum 22–24 weeks gestation prior to a series of a minimum of three consecutive miscarriages before the 22nd gestational week. The condition was considered to be unexplained if the women had normal uterine anatomy, their menstrual cycles were regular, they were negative for the lupus anticoagulant and the couples had normal karyotypes.
Table I shows the frequency of patients with secondary RM with only firstborn boys or only firstborn girls, respectively, in the two studies.9,10 In both studies, there was an significantly increased frequency of patients with firstborn boys compared with the expected 51.8% in the background population, and the difference between the frequency of patients with firstborn boys and girls, respectively, was very similar in the two studies, approximately 20%.
|Study||Firstborn boys only (%)||Firstborn girls only (%)||P|
|Nielsen et al.9||184 (60.3)||121 (39.7)||0.0001|
|Ooi et al.10||45 (64.3)||25 (35.7)||<0.05|
Table II shows data of outcome of the first pregnancy after referral in patients with secondary RM according to the sex of the firstborn child(ren) in the two studies. The subsequent miscarriage rate was 22 and 13% higher in the patients with firstborn boys compared with firstborn girls, respectively, although the difference was only significant in the former study owing to low numbers in the latter. In the Danish cohort, the male/female ratio in live births subsequent to secondary RM was 0.76 compared with 1.06 in the second birth among normal Danish women (P = 0.02).11 Previously, a large study of the anatomic sex ratio of 662 miscarried singleton embryos and fetuses found that the sex ratio was 1.30 (299 boys/230 girls) among miscarried fetuses with normal anatomy, whereas it was 0.92 (59 boys/64 girls) among malformed miscarried fetuses.12
|Study||Firstborn boys only||Firstborn girls only||P|
|Birth (%)||Miscarriage (%)||Birth (%)||Miscarriage (%)|
|Nielsen et al.9||78 (56.1)||61 (43.9)||79 (78.2)||22 (21.8)||0.0004|
|Ooi et al.10||17 (60.7)||11 (39.3)||14 (73.7)||5 (26.3)||NS|
|Combined studies||95 (56.9)||72 (43.1)||93 (77.5)||27 (22.5)||0.0003|
Register Studies of Normal Women
By far, the largest studies were based on data from the Danish Birth registry.13 A total of 499,731 primipara women who gave birth to a singleton child from January 1980 to December 1998 were identified, and information about subsequent births was obtained. It was found that one or two preceding boys, respectively, reduced the mean birth weight of later-born boys by 29 g (P = 0.0001) and 38 g (P = 0.0001) and later-born girls by 17 g (P = 0.0001) and 21 g (P = 0.0001) compared with later-born siblings with no preceding boys.13 Similar findings have been reported from the Norwegian Birth Registry, where it was an unexpected finding in a study14 with a different aim and in other smaller register and cohort studies.15,16
The risk of stillbirth in subsequent children according to the sex of the firstborn child was calculated in another study based on the Danish birth cohort.17 The risk of stillbirth was increased 12% after delivery of boys compared with girls corresponding to a relative risk = 1.12, (95% CI 1.02–1.23).
HLA studies relating to the role of anti-HY immunity for pregnancy complications have only been undertaken by the Danish research group.
Studies of HY-Restricting HLA in Women With Secondary RM and Their Children
Patients with unexplained secondary RM and (in case of more than one previous child) preceding children of same sex were the focus of these studies (n = 358). Full HLA typing was performed in 358 women with secondary RM and 203 firstborn living child(ren).18 HLA-A, HLA-B, HLA-DRB1, HLA-DRB3 and HLA-DQB1 typing was performed by up-to-date DNA techniques. The term ‘HY-restricting HLA’ is used to describe those HLA alleles reported to date that functionally present HY peptides. These include the following HLA class I alleles: HLA-A*01, HLA-A*02, HLA-B*07, HLA-B*08, HLA-B*52, HLA-B*60 and HLA class II alleles: HLA-DRB1*15, -DQB1*0501/2, -DRB3*03.
HY-Restricting HLA Allele Frequencies in Patients With Secondary RM
The allele frequencies of HLA class I and class II HY-restricting alleles were analyzed in secondary RM patients with firstborn boys and compared with those with firstborn girls.18 No difference in the frequency of these alleles was found in relation to sex of firstborn child. (P = 0.47).
Outcome of Next Pregnancy in Patients With Secondary RM
In patients with secondary RM, four different analyses (a–d) of the impact of HY-restricting HLA class I and II alleles on pregnancy prognosis in the first pregnancy after referral were undertaken.18 The impact of the HY-restricting HLA class I and II alleles on subsequent pregnancy outcome was tested separately in logistic regression analyses with live birth in the first pregnancy following referral as the outcome variable. The analysis was adjusted for number of previous miscarriages.
(a) Maternal carriage of HY-restricting HLA class I alleles and subsequent miscarriage rate
Carriage of HY-restricting HLA class I alleles did not impact the pregnancy prognosis when comparing patients with secondary RM and a firstborn boy with those with a firstborn girl.
(b) Maternal carriage of HY-restricting HLA class II alleles and subsequent miscarriage rate
Among the 286 patients with secondary RM who conceived after referral, maternal carriage of HY-restricting HLA class II alleles significantly reduced the chance of a live birth in those with a firstborn boy compared with those with a firstborn girl, odds ratio (OR): 0.17 (0.08–0.4), P = 0.0001. Among the patients without HY-restricting HLA class II alleles, the chance of a live birth was similar in those with firstborn boys (66%) and those with firstborn girls (69%), P = 0.76. In patients with boys prior to the miscarriages, maternal carriage of HY-restricting HLA class II alleles reduced the OR for a live birth to 0.46 (0.2–0.9), P = 0.02 while carriage of two alleles further reduced the OR to 0.21 (0.06–0.7), P = 0.02 compared with patients with firstborn boys not carrying any HY-restricting HLA class II alleles. Thus, the presence of maternal HY-restricting HLA class II molecule significantly reduced the chance of a live birth in a dose–response manner among secondary RM patients with firstborn boys.
(c) Firstborn child carriage of HY-restricting HLA alleles and subsequent miscarriage rate
The influence of the HY-restricting HLA class II of the firstborn child on their mother’s subsequent chance of a live birth was analyzed by comparing firstborn boys with firstborn girls. Of 203 HLA-typed children born prior to the miscarriages, the mothers of 178 became pregnant after referral. HY-restricting HLA class II of the firstborn child significantly reduced the chance of a live birth in patients with firstborn boys compared with girls [OR: 0.28 (0.1–0.7), P = 0.007], but limiting the analysis to those 24 children where only the firstborn boy but not the mother carried an HY-restricting HLA class II allele, and no reduction in chance of a live birth was found [OR = 1.94 (0.3–15), P = 0.52]. Thus, the presence of the HY-restricting HLA class II alleles among firstborn boys seems not to influence the outcome of the subsequent pregnancy.
(d) Maternal HY-restricting HLA class II alleles in secondary RM and subsequent perinatal outcome
Maternal carriage of HY-restricting HLA class II alleles was associated with obstetric complications in 213 surviving pregnancies after the secondary RM diagnosis. The mean birth weight was 381 g lower (P = 0.006) and the risk of stillbirth, preeclampsia and placental abruption increased (P = 0.05) in secondary RM patients with HY-restricting HLA class II alleles and a firstborn boy compared with a firstborn girl. These parameters did not differ according to sex of the first child in patients without HY-restricting HLA class II alleles.
HY-Restricting HLA Class II Alleles in Patients With Placental Abruptions
In the Danish population of patients with secondary RM or recurrent late pregnancy losses, eight patients were identified who had experienced at least two severe placental abruptions.19 They had had a total of 22 placental abruptions in 18 of which the fetus died. Fifteen (68%) of the placental abruptions had involved male fetuses. Seven of the patients had a firstborn boy and six of these were positive for HY-restricting HLA class II alleles. Haplotypes with HY-restricting class II alleles comprised 64% of the seven patients’ HLA haplotypes compared with only 28% among 37 controls with no pregnancy losses (P = 0.009).
Presence of HLA antibodies in maternal blood is found in approximately one-third of normal successful pregnancies.20–23 Patients with primary RM have a prevalence of HLA antibodies of maximum 10%;24,25 patients with secondary RM on the other hand may have a high prevalence of HLA antibodies.26 Steinborn et al.27 found an association between an increased prevalence of HLA antibodies and placental abruption. A Danish study28 investigated HLA antibodies in patients with secondary RM and controls. HLA class I and/or class II antibody responses were significantly more frequent in secondary RM patients with a boy prior to the series of miscarriages (62%) compared with secondary RM patients with a firstborn girl (29%), P = 0.03 and compared with women with primary RM (23%), and healthy female controls (25%) P = 0.001. Among the 60 (87%) pregnant patients who were HLA-antibody positive in early pregnancy, 41% had a live birth compared with 76% of those with no HLA antibodies (Table III). The adjusted OR for a live birth in patients with HLA antibodies was reduced to approximately one-third compared with patients without these antibodies, OR = 0.29 (0.1–0.8), P = 0.02.
|Antibody status||Pregnancy outcome||P-value||Sex of subsequent children||P-value|
|Miscarriage (%)||Live birth (%)||Girl (%)||Boy (%)|
|HLA-antibody neg||8 (24)||25 (76)||0.006|
|HLA-antibody pos||17 (59)||12 (41)|
|HY-antibody neg||17 (39)||27 (61)||0.26||15 (56)||12 (44)||0.03|
|HY-antibody pos||17 (52)||16 (48)||14 (88)||2 (12)|
The presence of IgG antibodies to one or more of five recombinant HY antigens has been shown to correlate with chronic GvHD in male recipients of stem cells from female donors.5 In a Danish study,29 antibodies against these five HY antigens were investigated in patients with RM and in controls who mainly comprised women who had given birth to two or more boys. The frequency of HY-specific antibody positivity was significantly higher in patients with secondary RM: 39 (46%) compared with female controls: seven (19%) (P = 0.004), and patients with primary RM: one (8%), (P = 0.01). HY-specific antibodies were detected in 33 (49%) of secondary RM patients with a boy and in six (38%) of the women with a girl prior to the miscarriages (P = 0.33). The impact of HY antibodies was analyzed in the 77 patients with RM who were early pregnant at the time of serum sample; 43 (56%) of these pregnancies ended with a live birth. As shown in Table III, live birth rates tended to be lower in HY-antibody-positive (48%) compared with HY-antibody-negative patients (61%) (P = 0.26). Only two (12%) of the children delivered by HY-antibody-positive patients were boys, which was significantly lower than the 12 (44%) boys delivered by HY-antibody-negative patients (P = 0.03) and the 51% boys among newborns in the background population (P = 0.002).
Maternal acceptance of the fetal semi-allograft is an immunologic paradox. We here report the current evidence for the impact of anti-HY immunity on adverse pregnancy outcome with a special focus on patients with secondary RM.
Epidemiologic studies from two centers showed that firstborn boys predispose to secondary RM and affect the prognosis of subsequent pregnancies negatively with regard to the chance of live birth and perinatal complications. Whereas the male/female sex ratio is elevated before secondary RM, it is significantly decreased after secondary RM suggesting that male embryos are preferentially miscarried.
The influence of the presence of the known HY-presenting HLA class I and class II alleles in both the mother and the child born prior to the miscarriages was investigated. The maternal carriage of HY-restricting HLA class II alleles significantly decreased the chance of live birth in patients with secondary RM and firstborn boys compared with those with firstborn girls. The chance of a live birth decreased in a dose–response manner with increasing number of maternal HY-restricting HLA class II alleles in patients with firstborn boys. No impact of maternal HY-restricting HLA class II alleles on the chance of live birth was observed in secondary RM patients with firstborn girls.
The observation that HLA class II but not class I HY-restricting alleles are associated with adverse pregnancy outcomes might reflect the participation of CD4+ T cells providing help for the CD8+ cytotoxic T cells in their response against minor histocompatibility (HY) antigens. Indeed, presence of CD4+ T cells with anti-recipient activity rather than CD8+ T cells was earlier reported in association with GvHD.30–32
The association of pregnancy outcome in secondary RM and maternal (but not fetal) carriage of HY-restricting HLA class II alleles implies that the HY antigens are presented to the immune system via the indirect pathway (maternal HLA alleles present fetal HY antigens). Accepting the hypothesis that there is an association between HLA class II alleles and pregnancy complications relating to male fetuses, it is relevant to look at the presence of specific antibodies in patients with secondary RM because CD4+ Th cell recognition of HY epitopes is HLA class II restricted and Th cells activate B cells to secrete antibodies.
We found, as expected, that the frequency of anti-HLA antibodies was increased in women with secondary RM compared with primary RM, but surprisingly the frequency in secondary RM was significantly higher than in controls who had given birth at least twice and it was also significantly higher in secondary RM patients with a firstborn boy relative to a firstborn girl. Finally, the presence of HLA antibodies decreased the chance of live birth with 35% in the next pregnancy in all patients with RM (P = 0.006) (Table III).
The frequency of HY-specific IgG antibodies was also significantly increased in patients with secondary RM compared with control females who previously had given birth to boys and compared with patients with primary RM. Furthermore, the presence of these antibodies in early pregnancy seemed to be correlated with low male/female sex ratio at birth but no significantly increased clinical miscarriage rate in patients with RM (Table III). This provides support for a direct, early (preclinical) and male-specific embryotoxic response. Such impact of HY antisera has been shown in studies aiming at non-invasive techniques for sex selection of pre-implantation cattle embryos to increase profitability of dairy and beef cattle production.33 Between 80 and 87% of murine and bovine male embryos that are cultured in high-titer rat HY antisera, at the morula stage, stop their development in contrast to female embryos.33,34
In general, alloantibodies (e.g. against paternal HLA) have never convincingly been found to play any pathologic role for outcome in early pregnancy and are so far thought to be a result of previous ongoing pregnancies.35,36 The mechanisms behind the association between future miscarriage rate and the presence of HLA antibodies in the Danish study are unknown. Presence of HLA antibodies may be the direct cause of the increased miscarriage frequency or it may an epiphenomenon that reflects that immunologic recognition has taken place because of prior transfers of fetal cells to the maternal immune circulation. Microchimerism associated with previous obstetric complications may play an important role in the pathogenesis of the miscarriages in patients with secondary RM, and this microchimerism may secondary be associated with an increased prevalence of HLA and HY antibodies.
We can only speculate about the underlying afferent immunologic mechanisms of our epidemiologic finding that maternal carriage of HY-restricting HLA class II alleles significantly reduces the chance of a live birth in secondary RM patients with a firstborn boy. Anti-HY immunization may have taken place during the first ongoing pregnancy with a boy. Maternal immune recognition of HY antigens can be demonstrated in women following pregnancies with boys.6,7 These responses must in general be well tolerated but HY immunization may also have harmful effects as indicated in SCT.37–40 In late pregnancy, apoptotic syncytiotrophoblast debris is normally shed in large quantities (several grams per day) from the placenta,41 and after being processed, peptides derived from HLA, HY and other minor HY antigens are presented to CD4+ and CD8+ T lymphocytes42 in local lymph nodes. This presentation usually takes place under non-inflammatory conditions; resulting in T lymphocytes being tolerized against fetally derived peptides.43 In a significant proportion of ongoing pregnancies with boys prior to secondary RM, complications such as stillbirth, placental abruption or fetal growth retardation occur, which are associated with increased production of inflammatory cytokines systemically or locally in uterus.44–46 Accordingly, the circumstances for sensitization of the adaptive immune system against fetal or trophoblast antigens are present in patients with secondary RM; this risk may be further increased in those patients carrying the HY-restricting HLA class II alleles.
The anti-HY-related efferent pathophysiologic mechanisms for the maternal non-acceptance of the fetus in the patients with secondary RM need to be further elucidated. Anti-HY and anti-HLA antibodies may play a direct role or may only be epiphenomenons. Subnormal levels of and functionally deficient T regulatory (Treg) cells may play a role in RM.47 It is in theory possible that cellular anti-HY responses through production of Th1 cytokines such as TNF-α could be one of the mechanisms reducing Tregs with specificity for non-sex-specific trophoblast antigens thereby increasing the risk of miscarriage of both male and female fetuses.
The results of epidemiologic, immunogenetic and immunologic studies undertaken so far points toward an important role for anti-HY immunity in the pathogenesis of many types of early and late adverse pregnancy outcomes. Further research is needed to clarify the pathophysiologic mechanisms behind and the possible prevention of these pregnancy complications.
Ole Bjarne Christiansen has received a speaker's fee of €500 from Leo Pharma, Denmark, for giving a lecture about recurrent miscarriage for Danish gynaecologists.