1 Department of Obstetrics and Gynaecology and 2 Department of Clinical Immunology, Aalborg Hospital, Aalborg and 3 Fertility Clinic 4071, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
4 To whom correspondence should be addressed at: Rolighedsvej 6, DK-9400 Norresundby, Denmark. e-mail: ckruse{at}dadlnet.dk
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: abortion/cytokines/L-selectin/recall antigens/recurrent miscarriage
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
All previous studies on cytokine production by peripheral blood mononuclear cells (PBMC) from RM patients have been based on a single blood sampling from each patient, taken before or at the time of miscarriage. Serial measurements in pregnant women with RM have not yet been published, but are relevant in order to clarify whether changes in the cytokine profile during pregnancy has a larger prognostic value than a single measurement. Besides, serial measurements are important for the evaluation of reproducibility of the test, and might also provide more information about the optimal time for testing in pregnancy.
A possible alternative to monitoring cytokine production may be to measure the distribution of Th1/Th2 cells using a flow cytometric analysis of the lymphocyte surface membrane markers, as two studies have indicated that Th1 and Th2 type CD4+ memory cells can be discriminated by their expression of the homing receptor CD62L (L-selectin). CD62L+ and CD62 expression was found to correspond with the cytokine profiles and functional capacities of Th2 and Th1 cells respectively (Kanegane et al., 1996; Mitra et al., 1999
). In a human pregnancy study, a reduced expression of CD62L was found in leukocytes from women with pre-eclampsia compared with normal pregnant women (Sacks et al., 1998
), whereas CD62L has not been investigated previously in studies of RM patients.
As another expression of the modulation of the maternal immune response in pregnancy, a single study (Bermas and Hill, 1997) has demonstrated loss of the normal lymphoproliferative response to microbial recall antigens (e.g. tetanus and influenza) in the majority of normal pregnant women, whereas in a significantly higher percentage of RM women the response during pregnancy was similar to the non-pregnant response. Among these RM patients, more women had a repeat miscarriage compared with RM patients who had lost their proliferative response. As yet, no further studies of this subject have been published.
The primary aim of the present study was to investigate cytokine production by PBMC, the proliferative responses of PBMC to recall antigens, the distribution of CD4+CD45RO+ CD62+/CD62 PBMC in pregnant RM patients and in normal pregnant women, and to carry out serial measurements of the same parameters in RM patients.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The control group comprised 15 pregnant women admitted to the authors clinic for legal abortion before 12 weeks gestation. None of the controls had miscarriages or stillbirths in their history. The median age was 29 (range 1945) years, and median parity was 1 (range 02). Intrauterine viable pregnancy was confirmed by ultrasound examination.
Information about reproductive history, time of last tetanus vaccination and history of herpes simplex virus (HSV) infections was obtained from all patients and controls.
Blood samples (10 ml in EDTA and 40 ml in heparin) were taken from all patients in pregnancy weeks 56, 78, 911 and 1314, or until the time of miscarriage. Similar blood samples were collected from all controls in pregnancy weeks 78 before legal abortion was carried out. PBMC were isolated from heparinized blood using standard laboratory methods and stored in liquid nitrogen for subsequent analysis.
-hCG and serum progesterone levels were measured in all blood samples using standard methods.
PBMC cultures
The cryopreserved PBMC were thawed, the concentration of viable cells was determined by trypan blue exclusion, and the cells were resuspended in a culture medium (RPMI-1640 supplemented with 1% penicillin/streptomycin and 10% fetal bovine serum) at a concentration of 2x106 viable cells per ml for the cultures stimulated with phytohaemagglutinin (PHA) or with microbial antigens. For stimulation with allogeneic cells the concentration was adjusted to 1x106 viable cells per ml.
Lymphocyte proliferative responses
PBMC were cultured in flat-bottomed wells with 50 µl cell suspension per well supplemented with 50 µl culture medium alone, or stimulated with 50 µl of the antigens tetanus toxoid, 1:10000 (Statens Serum Institut, Denmark), Candida albicans, 1:10 (Department of Microbiology, Aalborg Hospital, Denmark) or HSV, 1:10 (Behring, UK), 50 µl of the mitogen PHA, 33 mg/ml (Bie & Berntsen, Denmark) or a pool of irradiated allogeneic mononuclear cells (75 µl, concentration 1x106/ml). The stimulations with PHA and allogeneic cells were carried out to document that the proliferative potentials against unspecific stimuli were similar in the various subgroups. All analyses were performed in triplicate. After incubation at 37°C and 5% CO2 for 3 days (PHA) or 5 days (microbial antigens and allogeneic lymphocytes), the cultures were pulsed with 1 µCi [3H]-thymidine, harvested 24 h later, and scintigraphed using a -particle liquid scintillation counter (Minaxi 4430; Packard Instruments). The stimulation index was calculated by dividing the number of counts per minute (c.p.m.) of the stimulated cultures by the c.p.m. of the background response for each sample (PBMC suspension in medium alone).
Cytokine measurement
For PHA stimulation, 1x106 PBMC were cultured with 2 ml PHA (33 mg/ml) per well. For stimulation with allogeneic cells, 1x106 PBMC were cultured with 1x106 cells from a pool of allogeneic mononuclear cells obtained from a number of healthy donors. After 24 h, the supernatants were harvested for measurement of IL-2 and IL-4, and after 72 h for measurement of IL-10, IFN- and TNF-
. The supernatants were stored at 80°C in 500 µl test tubes. The time intervals to harvesting of supernatants were decided on basis of a previous study (McHugh et al., 1996
), supplied by a series of initial experiments conducted in the authors laboratory which showed that IL-2 and IL-4 concentrations in the supernatants peaked after 24 h, whereas IL-10, IFN-
and TNF-
concentrations peaked after 72 h of culture.
Cytokine production was measured using an enzyme-linked immunoabsorbent assay (ELISA). For each cytokine, a commercial kit (DuoSet; R&D Systems, UK) was used according to the manufacturers instructions. A standard curve was constructed for each cytokine set-up. The limits of detection were: 15.6 pg/ml (IL-2), 1.7 pg/ml (IL-4), 132 pg/ml (IL-10), 125 pg/ml (IFN-) and 125 pg/ml (TNF-
) in ELISA kits used for PHA-stimulated cultures, and 7.8 pg/ml (IL-2), 1.7 pg/ml (IL-4), 33 pg/ml (IL-10), 7.8 pg/ml (IFN-
) and 7.8 pg/ml (TNF-
) in kits used for allogeneic-stimulated cultures.
Measurement of surface antigen expression
The following monoclonal antibodies were used: fluorescein isothiocyanate (FITC)-conjugated anti-CD62L (DAKO), peridinin-chlorophyll-protein (PerCP)-conjugated anti-CD4 and phycoerythrin (PE)-conjugated anti-CD45RO (Becton Dickinson). Samples (100 µl) of EDTA blood were incubated with 10 µl negative control or 15 µl anti-CD4/CD45R0/CD62L for 30 min at 4°C in five Falcon tubes. Lysing solution (2 ml) (Becton Dickinson) was added and tubes were incubated for 10 min at room temperature in the dark. After centrifugation, the cells were washed in phosphate-buffered saline, suspended in sheet fluid (Becton Dickinson) with 1.35% formaldehyde, and stored at 4°C. Acquisition was performed within 24 h, using a FACScan 80122 flow cytometer (Becton Dickinson). Based on the expression of CD45RO and CD62L on CD4+ cells, three CD4+ T-cell populations were isolated using three-colour fluorescence and the relative distribution of CD45ROCD62L+ (naive), CD45RO+CD62L+ and CD45RO+CD62L CD4+ cells was calculated. A typical example of the flow cytometric analysis is depicted in Figure 1.
|
Protocol
In the protocol, the primary study outcomes were defined as cytokine production, proliferative responses and CD62L expression of lymphocytes for RM patients and controls. The secondary outcomes were the development of the same parameters during pregnancy in RM patients and the relationship to pregnancy outcome.
Ethical aspects
Informed consent was obtained from all patients and controls, and the study protocol was approved by the local ethics committee. The study was considered in agreement with Helsinki declaration II regarding research in humans.
Statistical analysis
Median values of cytokine production by PBMC, the proliferative responses of PBMC and the frequency of investigated lymphocyte surface markers were compared between patients and controls using the MannWhitney test. Intra-individual variability of measurements was compared with inter-individual variability using Friedmans test. The SPSS 10.0 program was used for these analyses.
The magnitude and development of the measured parameters during the first-trimester pregnancy in subsets of patients were analysed both in a purely linear model and in a log-linear model. The goodness-of-fit of the latter model was best, as the residuals were more normally distributed and the determination coefficient (Ra2) was higher. It was therefore assumed that each of the investigated parameters changes in a logarithmic-linear manner depending on the time of gestation. If rij indicates the jth measured parameter for the ith patient group (e.g. birth or miscarriage), ij represents random and identical normally distributed variation with mean zero, t indicates time of gestation, ai the slope coefficient and bij the intercept of the regression curve for the ith patient group it is assumed that:
log rij = µij + ij where µij = ait + bij
For each of the investigated parameters, the regression coefficients were estimated by the method of least squares. For each parameter, the log-linear models achieved for the two patient groups were compared using the ordinary F-test to decide whether the slope coefficients and intercept coefficients were significantly different. A P-value < 0.05 was considered statistically significant.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
After PHA stimulation, the supernatants of all cultures contained detectable amounts of all investigated cytokines, whereas after stimulation with allogeneic lymphocytes IL-2 and IL-4 were undetectable in all cultures.
Cytokine production, proliferative responses of PBMC cultures and distribution of CD45RO+CD62L and CD45RO+CD62L+ CD4+ cells in blood samples from all patients in pregnancy weeks 56 were compared with the samples from controls. Similar comparisons between patients with successful pregnancies and those who miscarried showed no significant differences (data not shown).
The median values of cytokines and Th1/Th2 cytokine ratios from supernatants of cultures of PBMC stimulated by PHA and allogeneic lymphocytes in patients and controls are listed in Table II. No significant differences were found. The Th1/Th2 cytokine ratios in subsets of RM patients positive or negative for ACA, and in patients with serum progesterone values above or below the median value, are listed in Table III. The cytokine ratios for the PHA-stimulated PBMC were generally lower in patients with high progesterone values, and IFN-/IL-10 and TNF-
/IL-10 ratios were lower in patients positive for ACA, though the difference was not statistically significant.
|
|
|
|
In the 12 RM patients who had at least two PBMC samples taken during pregnancy, a significantly smaller intra-individual than inter-individual variation was found for measurements of all five cytokines from PHA-stimulated PBMC (P < 0.05), in measurements of IL-10 from allogeneic-stimulated PBMC (P < 0.01), and in the frequencies of CD4+CD45RO+CD62L+ and CD4+CD45RO+CD62L PBMC (P < 0.001). For measurements of proliferative responses, the difference was significant only for HSV-stimulated PBMC (P < 0.05).
In the log-linear regression analysis the slope coefficients for serial measurements of cytokines from PHA-stimulated PBMC were close to, and not significantly different from, zero; this indicated that there was no trend to either increase or decrease with progressing gestation. No difference was found between IvIg-treated and untreated patients (Table VI).
|
With regard to the TNF- production by PBMC, the intercept coefficients were significantly different between patients who miscarried and those with successful pregnancies (P < 0.05), indicating that TNF-
levels were generally higher in miscarrying patients. The non-transformed curves for the TNF-
measurements are shown in Figure 2.
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The present control group comprised patients admitted for legal abortion. As none of the these women had experienced a previous pregnancy loss, and viable intrauterine pregnancy was confirmed in all pregnancies, the potential risk of miscarriage in this groupif pregnancy had proceededcould be considered as <5% (Brigham et al., 1999).
In the comparisons of cytokine production, a higher TNF- level was found in miscarrying patients than in those with successful pregnancies (Table VI; Figure 1), but no other differences were detected. For all cytokines measured in PHA-stimulated PBMC, the intra-individual variability was significantly lower than the inter-individual, and no decreasing or increasing trend was found in the log-linear regression analysis. It is considered that these serial measurements made a relevant contribution to the investigation of the role of cytokines in pregnancy. For all diagnostic tests it is crucial that results are reproducible within an individual or in a pregnancy. The results of the present study showed that measurements were mainly consistent in each patient, implying that the timing of tests during the first trimester of pregnancy is not crucial for the results if the test is conducted before miscarriage. In contrast, blood samples taken at the time of miscarriage might reflect changes in cytokine profile that are secondary to the pregnancy failure, but not its cause.
In the present study, no significant difference was found in cytokine production by PBMC between patients and controls (Table II). This finding was at variance with other previous studies (Marzi et al., 1996; Piccinni et al., 1998
; Makhseed et al., 2001
), in which a Th2-biased cytokine profile associated with successful pregnancy was found compared with a Th1-biased profile in patients with pregnancy loss. However, other studies have reported results which contradict the Th1/Th2 hypothesis (Vassiliadis et al., 1998
; Bates et al., 2002
), and this suggests that the interplay between cytokines in pregnancy is more complex than was first assumed. Unfortunately, no firm conclusions as to the importance of cytokine profiles in pregnancy can be drawn from the present results as the statistical power was limited by the small number of patients involved.
Herein, it was decided not to exclude the three patients who were positive for ACA, as the association between pregnancy loss and autoantibodies may be secondary to a predominant Th1 cytokine response towards the trophoblast (Gleicher, 2002), where autoantibodies develop as epiphenomena. Alternatively, patients with autoantibodies might have an antibody-biased (i.e. Th2-biased) T-cell profile. Although the inclusion of few patients in the present study prevented firm conclusions from being drawn, it is possible that the exclusion of autoantibody-positive patients might have introduced bias as these patients had the most pronounced Th1 response. Moreover, their inclusion in the present study would not reduce any differences seen between patients and controls in terms of cytokine production.
Previous studies have reported that progesterone, through stimulation of immunomodulatory proteins, might stimulate the production of Th2 cytokines at the feto-maternal interface (Szekeres-Bartho and Wegmann, 1996; Piccinni et al., 2000
). The present finding of generally lower Th1/Th2 cytokine ratios in RM patients with high serum progesterone levels might reflect an influence of serum progesterone on cytokine production. This has not been reported previously in clinical studies, but should be confirmed in larger investigations.
When studying cytokine profiles in pregnancy, it must be remembered that monitoring cytokine production by PBMC in a simple in-vitro set-up only partly reflects the complex in-vivo scenario at the feto-maternal interface, where T lymphocytes are in close contact with other regulatory lymphocytes, natural killer (NK) cells and many interacting cytokines (Chaouat et al., 2002). The cell populations of interest in pregnancy might be absent in (or disappear from) the peripheral blood and localize at the feto-maternal interface. Investigations of cytokine production in decidual lymphocytes (Piccinni et al., 1998
), or in subsets of lymphocytes such as
T-cells (Nagaeva et al., 2002
), might produce more specific results. Other studies have focused on a possible role of NK cells in pregnancy loss, as changes in NK cell subsets have been found in endometrial or peripheral NK cells (Coulam and Beaman, 1995
; Lachapelle et al., 1996
) from women with RM. Likewise, experiments conducted in mice have shown decidual NK cells to be important for normal reproduction (Guimond et al., 1998
).
During the second phase of the study, investigations were made into the proliferative responses of PBMC to stimulation with three microbial antigens, to PHA, and to allogeneic lymphocytes. Significantly higher median stimulation indices were found in patients compared with controls in response to HSV and tetanus antigens (Table IV). As there were no significant differences between patients and controls in response to PHA and allogeneic lymphocytes, the unspecific proliferative potential in samples from patients and controls appears to be similar. The present results were in agreement with those of a previous study (Bermas and Hill, 1997), which showed a diminished responsiveness to microbial recall antigens in normal pregnant women compared with RM patients. These findings may be another expression of the modulation of the immune system which might be present in normal pregnant women, but absent in those with RM. Further studies are needed to confirm these findings, however.
The final phase of the study incorporated a flow cyto metric analysis to measure the proportions of PBMC expressing the surface antigens CD4+CD45RO+CD62L+ and CD4+CD45RO+CD62L. Two previous studies showed that Th1-type and Th2-type memory CD4+ T-cells are characterized by the expression of CD45RO+CD62L and CD45RO+CD62+ respectively (Kanegane et al., 1996; Mitra et al., 1999
). This method has not been used previously in studies of RM, but may offer a more simple and more precise means to distinguish a Th1-dominated from a Th2-dominated response than by measuring cytokine production. The present results were surprising, as a significantly lower CD62L/CD62L+ ratio was found in patients compared with controls (Table V), and this was in direct opposition to the expected finding based on the Th1/Th2 hypothesis. However, the role of CD45RO+CD62L+ as a marker for Th2-type CD4+ cells in humans has not yet been established, and other studies in humans and mice have indicated different roles (Sacks et al., 1998
; Van Wely et al., 1999
; Ahmadzadeh et al., 2001
). Hence, further studies are required to clarify the possible role of CD62L expression in pregnancy failure.
In conclusion, the present results emphasize the need for improved methods to evaluate the maternal cellular immune responses relevant to pregnancy outcome. Although several studies have provided support for a role for Th1 cytokines in the pathogenesis of RM, others have emphasized the need for assays which have greater sensitivity and specificity in order to identify RM patients at risk of new miscarriage, and which are valuable in clinical practice. Future investigations might focus on intracellular cytokines (Kwak-Kim et al., 2003), genetic polymorphisms in the cytokine genes, cytokine receptor concentrations (Austgulen et al., 1992
), or the possible role of relevant surface membrane markers. The importance of the maternal cellular response towards microbial recall antigens should also be further explored.
![]() |
Acknowledgements |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Austgulen, R., Liabakk, N.B., Brockhaus, M. and Espevik, T. (1992) Soluble TNF receptors in amniotic fluid and urine from pregnant women. J. Reprod. Immunol., 22, 105116.[CrossRef][ISI][Medline]
Bates, M.D., Quenby, S., Takakuwa, K., Johnson, P.M. and Vince, G.M. (2002) Aberrant cytokine production by peripheral blood mononuclear cells in recurrent miscarriage? Hum. Reprod., 17, 24392444.
Bermas, B.L. and Hill, J.A. (1997) Proliferative responses to recall antigens are associated with pregnancy outcome in women with a history of recurrent spontaneous abortion. J. Clin. Invest., 100, 13301334.
Brigham, S.A., Conlon, C. and Faquharson, R.G. (1999) A longitudinal study of pregnancy outcome following idiopathic recurrent miscarriage. Hum. Reprod., 14, 28682871.
Chaouat, G., Zourbas, S., Ostojic, S., Lappree-Delage, G., Dubanchet, S., Ledee, N. and Martal, J. (2002) A brief review of recent data on some cytokine expressions at the maternal-fetal interface which might challenge the classical Th1/Th2 dichotomy. J. Reprod. Immunol., 53, 241256.[CrossRef][ISI][Medline]
Christiansen, O.B. (1996) A fresh look at the causes and treatments of recurrent miscarriage, especially its immunological aspects. Hum. Reprod. Update, 2, 271293.[Abstract]
Christiansen, O.B., Pedersen, B., Rosgaard, A. and Husth, M. (2002) A randomized, double-blind, placebo-controlled trial of intravenous immunoglobulin in the prevention of recurrent miscarriage: evidence for a therapeutic effect in women with secondary recurrent miscarriage. Hum. Reprod., 17, 809816.
Coulam, C.B. and Beaman, K.D. (1995) Reciprocal alteration in circulating TJ6+ CD19+ and TJ6+ CD56+ leukocytes in early pregnancy predicts success or miscarriage. Am. J. Reprod. Immunol., 34, 219224.[ISI][Medline]
Gleicher, N. (2002) Some thoughts on the Reproductive Autoimmune Failure Syndrome (RAFS) and the Th-1 versus Th-2 immune responses. Am. J. Reprod. Immunol., 48, 252254.[CrossRef][ISI][Medline]
Guimond, M.J., Wang, B. and Croy, B.A. (1998) Engraftment of bone marrow from severe combined immunodeficient (SCID) mice reverses the reproductive deficits in natural killer-deficient Tg epsilon 26 mice. J. Exp. Med., 187, 217223.
Hill, J.A. (1995) T-helper 1-type immunity to trophoblast: evidence for a new immunological mechanism for recurrent abortion in women. Hum. Reprod., 10, 114120.[ISI][Medline]
Hill, J.A., Polgar, K. and Anderson, D.J. (1995) T-helper 1-type immunity to trophoblast in women with recurrent spontaneous abortion. JAMA, 273, 19331936[Abstract]
Kanegane, H., Kasahara, Y., Niida, Y., Yachie, A., Sugii, S., Takatsu, K., Taniguchi, N. and Miyawaki, T. (1996) Expression of L-selectin (CD62L) discriminates Th1- and Th2-like cytokine-producing memory CD4+ T cells. Immunology, 87, 186190.[ISI][Medline]
Kwak-Kim, J.Y., Chung-Bang, H.S., Ng, S.C., Ntrivalas, E.I., Mangubat, C., Beaman, K.D., Beer, A.E. and Gilman-Sachs, A. (2003) Increased T helper 1 cytokine responses by circulating T cells are present in women with recurrent pregnancy losses and in infertile women with multiple implantation failures after IVF. Hum. Reprod., 18, 767773.
Lachapelle, M.H., Miron, P., Hemmings, R. and Roy, D.C. (1996) Endometrial T, B, and NK cells in patients with recurrent spontaneous abortion. Altered profile and pregnancy outcome. J. Immunol., 156, 40274034.[Abstract]
Makhseed, M., Raghupathy, R., Azizieh, F., Omu, A., Al-Shamali, E. and Ashkanani, L. (2001) Th1 and Th2 cytokine profiles in recurrent aborters with successful pregnancy and with subsequent abortions. Hum. Reprod., 16, 22192226.
Marzi, M., Vigano, A., Trabattoni, D., Villa, M.L., Salvaggio, A., Clerici, E. and Clerici, C. (1996) Characterization of type 1 and type 2 cytokine production profile in physiologic and pathologic human pregnancy. Clin. Exp. Immunol., 106, 127133.[ISI][Medline]
McHugh, S., Deighton, J., Rifkin, I. and Ewan, P. (1996) Kinetics and functional implications of Th1 and Th2 cytokine production following activation of peripheral blood mononuclear cells in primary culture. Eur. J. Immunol., 26, 12601265.[ISI][Medline]
Mitra, D.L., De Rosa, S.C., Luke, A., Balamurugan, A., Khaitan, B.K., Tung, J., Mehra, N.K., Terr, A.I., OGarra A., Herzenberg, L.A. et al. (1999) Differential representation of memory T cell subsets are characteristic of polarized immunity in leprosy and atopic diseases. Int. Immunol., 11, 18011810.
Nagaeva, O., Jonsson, L. and Mincheva-Nilsson, L. (2002) Dominant IL-10 and TGF- mRNA expression in
T cells of human early pregnancy decidua suggests immunoregulatory potential. Am. J. Reprod. Immunol., 48, 917.[CrossRef][ISI][Medline]
Ogasawara, M., Aoki, K., Okada, S. and Suzumori, K. (2000) Embryonic karyotype of abortuses in relation to the number of previous miscarriages. Fertil. Steril., 73, 300304.[CrossRef][ISI][Medline]
Piccinni, M.-P., Belloni, L., Livi, C., Maggi, E., Scarselli, G. and Romagnani, S. (1998) Defective production of both leukemia inhibitory factor and type 2 T-helper cytokines by decidual T cells in unexplained recurrent abortions. Nature Med., 4, 10201024.[CrossRef][ISI][Medline]
Piccinni, M.-P., Scaletti, C., Maggi, E. and Romagnani, S. (2000) Role of hormone-controlled Th1- and Th2-type cytokines in successful pregnancy. J. Neuroimmunol, 109, 3033.[CrossRef][ISI][Medline]
Raghupathy, R. (1997) Th1-type immunity is incompatible with successful pregnancy. Immunol. Today, 18, 478482.[CrossRef][ISI][Medline]
Raghupathy, R., Makhseed, M., Azizieh, F., Omu, A., Gupta, M. and Farhat, R. (2000) Cytokine production by maternal lymphocytes during normal pregnancy and in unexplained recurrent spontaneous abortion. Hum. Reprod., 15, 713718.
Sacks, G.P., Studena, K., Sargent, I.L. and Redman, C.W.G. (1998) Normal pregnancy and preeclampsia both produce inflammatory changes in peripheral blood leukocytes akin to those of sepsis. Am. J. Obstet. Gynecol., 179, 8086.[ISI][Medline]
Szekeres-Bartho, J. and Wegmann, T.G. (1996) A progesterone-dependent immunomodulatory protein alters the Th1/Th2 balance. J. Reprod. Immunol., 31, 8195.[CrossRef][ISI][Medline]
Van Wely, C.A., Beverly, P.C.L., Brett, S.A., Britten, C.J. and Tite, J.P. (1999) Expression of L-Selectin on Th1 cells is regulated by IL-12. J. Immunol., 163, 12141221.
Vassiliadis, S., Ranella, A., Papadimitriou, L., Makrygiannakis, A. and Athanassakis, I. (1998) Serum levels of pro- and anti-inflammatory cytokines in non-pregnant women, during pregnancy, labour and abortion. Mediators Inflamm., 7, 6972.[CrossRef][ISI][Medline]
Submitted on January 15, 2003; resubmitted on June 5, 2003; accepted on July 22, 2003.