SIEP publications are protected by copyright and all rights are reserved.  SIEP publications may not be reproduced in any form or by any means without written permission from the copyright owner. This includes the posting of electronic files on the Internet, transferring electronic files to other persons, distributing printed output, and photocopying. Requests for authorization to make photocopies should be directed to:   SIEP, 1697 Lark Lane, Cherry Hill, NJ 08003-3157
Tel: (856) 429 2699, Fax:  (856) 429 7414, Email: barnea@earlypregnancy.org.

---

EARLY PREGNANCY: Biology and Medicine Editor-in-Chief: Eytan R. Barnea MD, FACOG

April 2000
Volume IV, Number 2
ISSN: 1537-6583
Pages: 144-153

Shared HLA Antigens In Couples With 5 Or More, Compared To 3 Or 4 Consecutive Recurrent Miscarriages

Short title: Shared HLA Antigens & Recurrent Miscarriage

Key words: HLA, consecutive recurrent miscarriages, anti paternal antibodies

Correspondence: Dr. S. Orgad, Division of Transplantation Immunology; Tissue typing lab, Sheba Medical Center, Tel Hashomer 5621, Israel, Telephone: 972-3-5302829,
Fax: 972-3-5345964

Abbreviations: HLA - Human Leucocyte Antigen, CRM - consecutive recurrent miscarriages, APA - anti paternal antibodies, XM - crossmatch

Objective
Couples with consecutive recurrent miscarriages (CRM) are not generally believed to share more HLA antigens with their spouses than expected by chance. This paper attempted to determine the situation in patients with five or more miscarriages.

Methods
The number of shared HLA class I and class II antigens, HLA phenotype, and the ability to mount an antibody response in a large cohort of 425 couples with CRM was assessed, according to whether the patient had three or four, or five or more miscarriages.

Results
There was no significant difference in the frequency of shared HLA antigens in women with five or more miscarriages when compared to three or more miscarriages or control patients. No specific HLA antigen or phenotype was associated with CRM in male or female partners of either group. The number of shared antigens did not influence ability to develop anti paternal antibodies (APA). Moreover, HLA antigen sharing had no influence on the subsequent pregnancy after paternal leucocyte immunization.

Conclusion
Class I and Class II HLA antigens are not diagnostic for immunologically mediated abortion, do not predict the ability to mount an antibody response, or the outcome of a subsequent pregnancy.

Introduction

The maternal immunologic responses to pregnancy, both cellular and humoral, are not harmful and may even benefit the embryo by protecting it from harmful maternal immune reactions, (Ahron 1971, Nymand et al, 1971). Excess sharing of HLA antigens between spouses has been considered by some to be a mechanism leading to maternal hyporesponsiveness to paternal antigens encountered in pregnancy and therefore subsequent miscarriage (Beer et al, 1981). It has been reported (Komlos et al, 1977, Beer et al, 1981) that couples with consecutive recurrent spontaneous miscarriages share a higher proportion of HLA antigens than would be expected by chance. However, subsequent reports have contradictory. Some have confirmed excess HLA antigen sharing between spouses in recurrently aborting couples (Taylor et al, 1981), while others have not found excess sharing of HLA antigens (Muller Eckhardt et al, 1984, Oksenberg et al, 1984, Caudle et al, 1985, Pennesi et al, 1998). The whole subject has been fully reviewed elsewhere (Ober 1988).

The relevant response to protect pregnancy from immunologically mediated miscarriage has not been fully elucidated but may include a cytokine response leading to a cellular response, and might involve an antibody response. It has been suggested that hyporesponsiveness to the developing pregnancy may be reflected by a lower incidence of anti paternal antibodies (APA) in couples with recurrent miscarriage (Mowbray et al 1983). The concept of excess sharing of HLA antigens between spouses and subsequent immunological hyporesponsiveness has had clinical consequences and led to immunomodulation being used to overcome the hyporresponsiveness and improve the subsequent live birth rate. Immunomodulation has either been active with paternal leukocytes, or passive with intravenous immunoglobulin infusion. The initial reports on paternal leukocyte immunization used the sharing of HLA antigens between spouses as a criterion for immunization (Beer et al 1981, Taylor and Faulk, 1981). It is still controversial whether immunomodulation improves the subsequent live birth rate. It is also controversial to what extent the number of shared HLA antigens and antipaternal antibodies (APA) influence the outcome of the subsequent pregnancy.

We considered that the controversy about the influence of shared antigens may be due to previous studies assessing a small study group, and by their assessing all couples with three or more miscarriages. We therefore assessed a more homogeneous group with five or more miscarriages, and compared the antigen sharing to patients with three or four miscarriages, and both of these groups to a control group of patients.

Patients
The study comprised of 425 couples who attended the recurrent miscarriage clinic of the Sheba Medical Center at Tel Hashomer between 1987 and 1995. One hundred and twenty two couples had 5-11 previous miscarriages and 303 couples had 3 or 4 miscarriages. These patients were compared to 104 control couples who had HLA antigens assessed for paternity determination. Patients with artificial abortions prior to or between the spontaneous miscarriages were excluded from the study. Each patient was assessed for other known causes of miscarriage and were included in the study only after the following presumptive etiological factors were found to be within normal limits: -

1. Karyotype of both parents
2. Glucose tolerance test
3. Toxoplasmosis serology
4. Hysterosalpingogram to exclude anatomical malformations, intrauterine adhesions and cervical incompetence
5. Thyroid function
6. Serum prolactin
7. Normal luteal phase of at least 12 days and plasma progesterone above 8 ng/100 cc
8. ANF, using rat liver as substrate and fluoresceinated rabbit anti-human IgG
9. Anticardiolipin antibody by Elisa and Lupus anticoagulant, according to the Kaolin clotting time (KCT)

Both the 122 patients with 5-11 previous consecutive recurrent miscarriages, and the 303 couples with 3 or 4 miscarriages elected either to be immunized or to enter a control group. The patients who declined immunization did so either due to the advice of their own physician or fear of potential side effects of experimental treatment. This control group was therefore random and matched but not double blind. In total, 49 couples in the >5 CRM group and 81 couples in the 3-4 CRM group formed the control non-immunized group .

Immunization Procedure
Our immunization protocol has been fully described elsewhere (Carp et al, 1990). As there is no reliable laboratory test at present to diagnose immunologically mediated miscarriages, treatment was offered to all women with three or more unexplained pregnancy losses if none of the causes of miscarriage listed above were present. All patients were informed of the possible risks and benefits of immunization, and all signed an informed consent form. Patients and spouses undergoing immunization were screened for Hepatitis B antigen and antibodies to Syphilis, Hepatitis C, CMV and HIV. If any of these were positive, immunization was not performed and the patient was excluded from the study.

Laboratory Testing
The presence of anti paternal antibodies (APA) was detected by a crossmatch (XM) between maternal undiluted fresh serum with paternal peripheral lymphocytes (Mittal et al, 1968). Rabbit serum was used as the source of complement. The reaction was scored as positive when the test serum killed 40% or more paternal cells.

HLA typing was performed by the NIH standard micro-lymphocytotoxicity technique. One hundred forty-four alloantisera were used for HLA class I antigens and 70 alloantisera were used to type for DR antigens. Alloantisera were procured locally or obtained by collaboration with colleagues and from international HLA workshops serum sets. Peripheral blood lymphocytes were isolated according to Boyum’s (1968) method. HLA DR typing was performed on nylon wool T cell depleted, B cell enriched lymphocytes by extended incubation micro-lymphocytotoxicity testing (Terasaki et al, 1978). The control group for the HLA study comprised of 104 couples who had been tissue typed for paternity determination, or parents typed to determine suitability for donating bone marrow to their children.

Statistics
The results were evaluated by 2x2 tables and c ² with the Pearson’s correlation.

Results

The HLA antigen frequencies were similar in all the groups studied. (Data not shown). No specific HLA antigen or HLA phenotype was found to be associated with consecutive recurrent abortion in either the male or a female partner. There was no significant difference in the number of shared HLA antigens between both groups of patients with recurrent miscarriage and the control group (Fig 1). There was no excess HLA antigen sharing in 65 of the 122 couples with 5 or more miscarriages (53%). This was a similar incidence to the 176 of 303 couples with 3 or 4 miscarriages (58%), and to the 47 of 104 control patients (43.3%). One HLA Class I antigen was shared in 40 (33%) patients with >5 CRM, 99 (33%) patients with 3-4 CRM and 35 (34%) of control patients. Two Class I antigens were shared by 16 (13%) of patients with >5 CRM, 26 (9%) patients with 3 or 4 miscarriages and 20 (19%) control patients. The numbers of couples sharing 3 Class I antigens were (0.6%), 2(0.8%), and 2(2%), in the >5 CRM, 3-4 CRM and control groups respectively. No couple in any group shared 4 HLA Class I antigens. These differences were not statistically significant.

There was no excess sharing of HLA Class II antigens in 57 (48%) couples with >5 CRM, 167 (57%) couples with 3-4 CRM, and 10 (77%), control patients. The number of couples with one shared Class II antigen was 52 (46%), 112 (38%) and 3 (23%) in patients with 5 or more, 3-4 miscarriages and control patients respectively. Two Class II antigens were shared in 7 (6%), 14(5%), and 0(0%), in >5 CRM, 3-4 CRM and in control patients respectively (Fig 2). These differences were not statistically significant.

Two hundred and sixty seven women were immunized with paternal leukocytes, 66 in the >5 CRM group and 201 in the 3-4 CRM group. (Table 1). Very few immunized women failed to develop antipaternal antibodies: In the >5 CRM group the number of women who seroconverted to APA positive was 81% for Class I antigens and 90% for Class II antigens. A similar result was seen in the 3-4 CRM group, (88% for HLA Class I antigens and 87.7% for Class II.)

The ability to mount an immune response to paternal antigens was not significantly different between women sharing 1, 2, or 3 HLA antigens and those with no shared antigens. This rule held true for both Class I and Class II antigens and for both CRM groups.

Discussion

This study summarizes the results of a large cohort of 425 women with 3-11 spontaneous recurrent miscarriages (CRM). These 425 patients were compared to 104 control couples who were investigated for paternity studies. The recurrently miscarrying women were either immunized with paternal leucocytes or were assigned to control group when no immunization was carried out.

The classification into three or four CRM and five or more CRM groups is unusual. We considered that the usual criterion of three or more miscarriages is too heterogeneous a group. Three miscarriages can occur by chance, as the number of miscarriages increases, it becomes less likely that the cause is a random event (Christiansen 1996). Therefore, it is conceivable that patients with five or more miscarriages might have excess antigen sharing, but the greater number of patients with three miscarriages might have skewed the findings in previous studies. It has even been suggested that the beneficial effect of immunopotentiation may be greater in patients with a higher number of miscarriages (Recurrent Miscarriage Immunotherapy Trialists Group, 1994, Daya and Gunby, 1994, Carp et al, 1997). Hence, we regarded it to be more valid to assess a more homogeneous group with a poor prognosis, therefore patients with five or more miscarriages were selected to assess HLA antigens. However, even in this highly selected group there was no excess HLA antigen sharing between spouses.

Our results showed no excess antigen sharing in either CRM group when compared to the control group. The frequencies of antigen sharing are similar to those described for Northern European populations (Terasaki et al, 1978). Terasaki et al, (1978), have reported that 46% of patients shared no HLA antigens, 42% shared one Class I antigen; 10.8% share 2 antigens and that only 0.9% share 3 Class I antigens. These results do not concur with those that have reported (Komlos et al, 1977, Beer et al, 1981) that a higher proportion of couples with consecutive recurrent miscarriages share HLA antigens with their spouses. There are several possible explanations for this discrepancy.

Early studies such as Komlos et al, (1977), only tested for the 10 HLA-A Antigens and 14 HLA-B Antigens which were known at the time. Since the publication of this early study, much progress has been made in the understanding of the HLA system. It is much more polymorphic than originally thought. Many antigens previously thought to be one antigen have been subsequently split into two or more antigenic specificities. E:g. HLA-A9 is to-day known to be two antigens A23 and A24, A10 has been split and is known to-day as A25, A26, A34 and A66 etc. Therefore it is not surprising that the actual level of sharing of HLA antigens is much lower than that published previously. Moreover, with the recent refinement of HLA typing by DNA sequencing, it has become apparent that the polymorphism of the human MHC is even greater than was suspected after the "splitting" of certain antigens. Therefore, the so-called "sharing" of HLA antigens by serological means is most probably not true sharing. Moreover, most published studies have assessed a relatively small number of couples (between 12-60). Therefore there may have been some selection bias in their populations. In this larger study of 425 couples no excess antigen sharing was observed, either of HLA Class I and or Class II, between both CRM groups and the 104 couples in the control group that were assessed for paternity.

Indeed, 78% of couples with 5 or more CRM and 93% of couples with 3-4 CRM who shared two HLA antigens (class I and or class II) were able to elicit anti paternal cytotoxic antibodies. These antibodies have been reported to correlate with the outcome of the subsequent pregnancy, (Carp et al, 1990, Recurrent Miscarriage Immunotherapy Trialists Group, 1994). Therefore it was interesting to see if the antigen sharing might affect the ability to seroconvert to APA positive. We found no difference in the ability of women to seroconvert after immunization, whether they shared HLA Class I and Class II antigens or not. Indeed, 78-93% of couples who shared two HLA antigens (class I and or class II) were able to elicit anti paternal cytotoxic antibodies. The number of women who did not respond to immunization was similar in the >5 and 3-4 CRM groups, and the number of shared antigens played no role in the response to immunization.

The issue of consanguinity was not assessed in this study. It is conceivable that in cases of consanguinity with subsequent excess HLA sharing, either genetic or immunological mechanisms could lead to spontaneous abortion or an increased incidence of malformations.

The reports on antigen sharing in couples with recurrent miscarriage led to several studies assessing the role of HLA antigens and their influence on the outcome of pregnancy (Beer et al, 1981, Taylor and Faulk, 1981), and led to immunopotentiation being used to prevent subsequent miscarriages. The initial reports on paternal leucocyte immunization used the sharing of HLA antigens between spouses as a criterion for immunization (Beer et al, 1981, Taylor and Faulk 1981). This excess sharing was considered to be responsible for the hyporesponsiveness to paternal antigens leading to miscarriage. Although this treatment is still controversial as to its efficacy, it seems to day that its effect is not related to HLA antigen sharing. Recently, it has been suggested that sharing of allele C4B*Q0 of the complement system, may induce complement dependant damage to the extra embryonic tissues. This factor has been reported to be elevated in women with CRM (Pennesi et al, 1998). However, there may be other mechanisms leading to miscarriage which have not yet been elucidated.

For personal use. Only reproduce with permission from SIEP.

References

Ahron, S. (1971). HLA Antibodies: Influence on the Human Fetus. Tissue Antigens 1, 129-136

Beer, A. E., Quebbman, J. F., Ayers, J. W. T., Haines, R. F. (1981). Major histocompatibility complex antigens, maternal and paternal immune responses and chronic habitual miscarriage. Am. J. Obstet. Gynecol. 141, 987-999

Boyum, A. (1968). Separation of leukocytes from blood and bone marrow: Scand. J. Clin. Lab. Invest. (Suppl 97) 21, 77-81

Carp, H. J. A., Toder, V, Gazit, E., Orgad, S., Mashiach, S., Serr, D.M., Nebel, L. (1990). Immunization by Paternal Leucocytes for Prevention of Primary Habitual Miscarriage: Results of a Matched Controlled Trial. J. Gynaecol. & Obstet. Invest. 29, 16-21

Carp, H. J. A., Torchinsky, A., Portuguese, S., Gazit, E., Mashiach, S., Toder, V., and the Recurrent Miscarriage Immunotherapy Trialists Group. (1997). Paternal Leucocyte Immunization After Five Or More Miscarriages. Hum. Reprod. 12, 250-255

Caudle, M. R., Rote, N. S., Scott, J. R., De Witte, C., Barney, M. F. (1985). Histocompatibility in couples with recurrent spontaneous miscarriages and normal fertility. Fertil. Steril. 39, 793-798

Christiansen, O.B. (1996). A fresh look at the causes and treatments of recurrent miscarriage, especially its immunological aspects. Human Reprod. Update 2, 271-293

Daya, S., Gunby, J. (1994). The Effectiveness of Allogeneic Leucocyte Immunization in Unexplained Primary Recurrent Spontaneous Abortion. Am. J. Reprod. Immunol. 32, 294-302

Komlos, L., Zamir, R., Joshuah, H., Halbrecht, I. (1977). HLA antigens in couples with repeated miscarriages. Clin. Immunol. Immunopathol. 7, 330-335

Mittal, K. K., Mickey, T. R., Singall, D. P., Terasaki, P. I. (1968). Serotyping for Transplantation XVIII. Refinement of microdroplet cytotoxicity test. Transplantation 6, 913-916

Mowbray, J. F., Gibbings, C. R., Sidgwick, A. S., Ruskiewicz, M., Beard, R. W. (1983). Effects of transfusions in women with recurrent spontaneous abortion. Transpl. Proc. 10, 896-899

Muller-Eckhardt, G., Mallmann, P., Neppert, J., Lattermann, A. (1994) Immunogenic and Serological Investigations in Non pregnant and pregnant women with a history of recurrent Spontaneous Miscarriages. J. Reprod. Immunol. 27, 95-109

Nymand, G., Heron, I., Jensen, K. G., Lundsgaard, A. (1971). Occurrence of cytotoxic antibodoes during pregnancy. Vox. Sang. 21, 21-29

Ober, C (1998). Immunogenetics ’98 HLA and pregnancy: The paradox of fetal allograft. Am. J. Hum. Genet. 62, 1-5

Oksenberg, J. R., Persitz, E., Amar, A. (1984). Maternal-paternal histocompatibility: Lack of association with habitual miscarriages. Fertil. Steril. 42, 389-395

Pennesi, G., Brioli, G., Lulli, P., Mariani, B, Morellini, M,. Nicotra, M., Trabace, S. (1998). HLA and complement factors, allele sharing in Italian couples with recurrent spontaneous abortions. Human Immunol. 59, 382-386

Recurrent Miscarriage Immunotherapy Trialists Group. (1994). Worldwide Collaborative Observational Study And Metaanalysis On Allogenic Leucocyte Immunotherapy For Recurrent Spontaneous Abortion Am. J. Reprod. Immunol. 32, 55-72

Taylor, C., Faulk, W. P. (1981). Prevention of recurrent miscarriage with leucocyte transfusions. Lancet ii, 68-69

Terasaki, P.I., Bernoco, D., Park, M. S., Ozturk, G., Iwaki, Y. (1978). Microdroplet testing for HLA -A,B,C and D antigens. Am. J. Clin. Pathol. 69, 103-105 1978

Figure 1

Number of shared HLA Class I antigens

Sharing of Class I HLA Antigens. No HLA antigen sharing was observed in 65/122 (53%) of patients with >5 CRM, 176/303 (58%) of patients with 3-4 CRM or 47/ 104 (43.3%) of control patients. One HLA Class I antigen was shared in 40 (33%), 99 (33%) 35 (34%) in the >5 CRM, 3-4 CRM and control groups respectively. Two HLA Class I antigens were shared by 16 (13%), 26 (9%) and 20 (19%) in the >5 CRM, 3-4 CRM and control groups respectively. There was sharing of 3 Class I antigens in (0.6%), 2(0.8%), and 2(2%), of >5 CRM, 3-4 CRM and control groups respectively. No couple in any group shared 4 HLA Class I antigens. These differences were not statistically significant.

Figure 2

Number  of Shared HLA Class II antigens

Sharing HLA class II antigens. There was no sharing of HLA Class II antigens in 57 (48%), 167 (57%) and 10 (77%), in the >5 CRM, 3-4 CRM and control groups respectively. One HLA Class II antigen was shared by 52 (46%), 112 (38%) and 3 (23%) %), in the >5 CRM, 3-4 CRM and control groups respectively. Two HLA Class II antigens were shared in 7 (6%), 14(5%), and 0(0%), in the >5 CRM, 3-4 CRM and in control patients respectively. These differences were not statistically significant.

Influence of Shared HLA Antigens on the ability of women to elicit an immune response to paternal cells after Immunization

APA = anti paternal antibodies
Numbers in parentheses denote percentages
The differences between the groups are not statistically significant