Maternal C-reactive protein levels are raised at 4 weeks gestation

G.P. Sacks1,2, L. Seyani1, S. Lavery1 and G. Trew1

1 Department of Reproductive Medicine, Hammersmith Hospital, Du Cane Road, London W12 0HS, UK

2 To whom correspondence should be addressed. e-mail: g.sacks{at}imperial.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The aim of this study was to determine whether there is evidence of a systemic maternal inflammatory response in very early pregnancy. METHODS: Successive women receiving treatment by IVF or ICSI had serum C-reactive protein (CRP) levels measured on the day of their pregnancy blood test at 4 weeks gestation (14 days post-egg collection). Women with positive {beta}HCG levels had ongoing pregnancies confirmed by serial transvaginal ultrasound scans up to 8 weeks gestation. RESULTS: Pregnant women (n = 40) were significantly younger (mean age 34 years) than women who failed to become pregnant (n = 95, mean age 37 years, P < 0.001), received significantly lower treatment doses of recombinant FSH (2000 versus 2400 IU, P < 0.05) and had significantly more eggs collected (11 versus 8, P < 0.01). There were no significant differences in body mass index, parity, a history of smoking, endometriosis or polycystic ovaries, pre-treatment CRP levels and white cell counts, peak serum estradiol levels and numbers of embryos transferred. Pregnant women had significantly higher CRP levels (median 3.68 mg/l) than those who were not pregnant (median 1.495 mg/l, P < 0.0001), a difference that persisted after excluding potential confounding variables. Six pregnant women with ovarian hyperstimulation syndrome had higher CRP levels than those who did not (P < 0.01). CONCLUSIONS: This well-controlled study is the first to demonstrate that maternal CRP levels are raised as early as 4 weeks gestation and thus that the maternal inflammatory response is established during the earliest phases of implantation. It is hypothesized that an abnormal response (either exaggerated or absent) could cause some cases of miscarriage.

Key words: C-reactive protein/early pregnancy/inflammatory response/IVF/miscarriage


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In normal pregnancy, there is evidence of widespread activation of innate (as opposed to specific) elements of the maternal immune system, and these changes constitute the maternal inflammatory response (Sacks et al., 1999Go; Veenstra van Nieuwenhoven et al., 2003Go). Most evidence for such a response is from the third trimester, and investigators have postulated that an abnormal inflammatory response causes pregnancy complications such as pre-eclampsia (Redman et al., 1998Go) or preterm labour (Hvilsom et al., 2002Go). The stimulating factor or factors for the response are still unknown, as is the gestation at which the response is first generated.

There is reason to suspect that an inflammatory response is present early in the first trimester. From as early as 6 weeks gestation, there are increased numbers of circulating monocytes (Smarason et al., 1986Go), while in first trimester decidua there is a massive influx of pro-inflammatory macrophages and natural killer (NK) cells (Heikkinen et al., 2003Go). There is considerable interest in the roles these cells might have in limiting implantation on the one hand (Loke and King, 2000Go; Petroff et al., 2002Go), and encouraging trophoblast growth and appropriate maternal physiological changes on the other (Sacks et al., 1999Go; Egawa et al., 2002Go; Monzon-Bordonaba et al., 2002Go; Heikkinen et al., 2003Go). The interaction between the invading trophoblast and maternal inflammatory cells is a key stage of early pregnancy and may be a factor in some cases of miscarriage.

It is not possible to assess the maternal–fetal interface in ongoing or threatened pregnancies directly. Hence the interest in circulating factors such as cytokines (Ogasawara et al., 2000Go) or NK cells (Emmer et al, 2000Go; Ntrivalas et al, 2001Go) that may provide surrogate measures of the maternal–fetal interface. One of the most important markers of an inflammatory response is C-reactive protein (CRP), first discovered in 1930 as an acute phase protein that reacted with the C polysaccharide of Pneumococcus bacteria, but now established as an inflammatory marker in a wide range of conditions (Kluft and de Maat, 2002Go). CRP levels can rise many-fold in the presence of infection or trauma, but very small increases above baseline are also useful predictors of cardiovascular disease and are believed to indicate low-grade inflammation in atherosclerotic plaques (Ablij and Meinders, 2002Go). CRP is produced by hepatocytes primarily in response to interleukin (IL)-6, and is equally distributed in the vascular compartment (Ablij and Meinders, 2002Go). Numerous functions have been described, including the ability to recognize foreign pathogens and damaged cells of the host and to initiate their elimination by activation of the classical complement pathway and phagocytosis (Ablij and Meinders, 2002Go; Kluft and deMaat, 2002Go).

In pregnancy, CRP levels are raised in women with pre-eclampsia (Teran et al., 2001Go), in women with ruptured membranes complicated by chorioamnionitis (Yoon et al., 1996Go) and in women who develop preterm labour (Hvilsom et al., 2002Go). In the first trimester, raised CRP levels have been reported (Rebelo et al., 1995Go), and more recently it was shown that women with higher CRP levels at 9–13 weeks are more likely to develop gestational diabetes mellitus (Wolf et al., 2003Go) and pre-eclampsia (Wolf et al., 2001Go).

The aim of this study was to determine whether circulating CRP levels are raised at 4 weeks gestation, during the earliest phases of implantation. A busy IVF unit provided an ideal setting for recruiting women at such an early time in their pregnancies. In view of the likely confounding factors associated with a stimulated IVF cycle, women who failed to become pregnant after similar treatment formed the control group.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Subjects
All women in this study were being treated by IVF or ICSI at the IVF unit in Hammersmith Hospital. Ethical approval was obtained from the Hammersmith Hospital Research Ethics Committee. Consecutive women starting IVF/ICSI cycles were enrolled, but only those who completed a full cycle and had a pregnancy test were included.

In this study, the pregnant group of women comprised those with viable ongoing pregnancies only. Thus, they had raised serum levels of the {beta} subunit of HCG (>2 IU/ml) 14 days after egg collection (i.e. 4 weeks gestation), followed by the detection of a fetal heart beat on a transvaginal ultrasound scan performed in the IVF unit 2 weeks later. All those women also had a normal further follow-up scan 2 weeks after that. In practice, it was found that all those women had {beta}HCG levels >50 IU/ml. Women with pregnancy loss at any time during follow-up were excluded from this study. The non-pregnant group comprised women with {beta}HCG values <2 IU/ml at 4 weeks gestation.

IVF/ICSI cycles
Two standard stimulation protocols were used. Briefly, in a GnRH agonist protocol, on day 2 of their cycles, women began self-administered injections of buserelin acetate (Shire Pharmaceuticals, Andover, UK) for 2 weeks before starting ovarian stimulation with injections of recombinant FSH (Puregon; Organon, Cambridge, UK) as well. And, in a GnRH antagonist protocol, women began with FSH injections first and added the GnRH antagonist ganirelix (Orgalutran; Organon) when the leading follicle reached 14 mm in size. In both protocols, there was close monitoring by regular blood tests for serum estradiol and LH, and transvaginal ultrasound. When at least three follicles were >17 mm in size, women were instructed to stop both injections and have a late night injection of 10 000 IU of HCG (Organon) to simulate the LH surge of a natural cycle. After 36 h, they came to the unit for transvaginal ultrasound-guided egg collection under light sedation. All women were given a 7-day course of doxycycline after this. Eggs were fertilized in vitro by conventional IVF or single sperm injection (ICSI), and day 2/3 embryos were placed back in the uterine cavity by passing a fine transcervical catheter under ultrasound guidance. Women were given a single injection of progesterone on the day of embryo transfer, and then started daily 400 mg progesterone suppositories (Cyclogest 400) (Shire Pharmaceuticals) for 2 weeks until their pregnancy blood test.

Data collection
Blood was collected by nurses in the IVF unit into standard SST gel tubes (Becton Dickinson, Plymouth, UK). Assays for serum estradiol, FSH and LH (Abbott Laboratories, IL), CRP (Olympus Diagnostics, County Clare, Ireland) and for full blood counts (Cell-Dyn 4000, Abbott Laboratories) were performed as routine clinical tests by hospital laboratory staff. For this study, blood was taken from cycle day 2 (before starting any injections) and the day of the pregnancy test. The standard CRP assay was unreliable at levels below 2 mg/l and, since many samples were found to be below or around that level, serum samples from the day 14 pregnancy test were assayed in addition by a ‘high sensitivity’ CRP assay kit (Olympus Diagnostica).

Statistics
Data were analysed on GraphPad Prism software (San Diego, CA). The groups (pregnant and non-pregnant) were compared by two-tailed t-tests or, for non-parametric data, Mann–Witney U-tests. Changes in CRP levels between the start and end of treatment cycles were compared by the Wilcoxon signed rank test. The associations between CRP levels and potential confounding variables were tested with the Spearman correlation coefficient. In all tests, significance was assumed when P < 0.05.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
All women who completed an IVF/ICSI cycle agreed to participate in this study. Consecutive cases of women who were pregnant with viable ongoing pregnancies (n = 40) and not pregnant (n = 95) after treatment were included. There were no omissions or exclusions from these groups. Women who became pregnant were significantly younger (mean ± SD, 34.13 ± 3.80 years) than those who did not (37.03 ± 4.29 years, P < 0.001). There were no significant differences between pregnant and non-pregnant women in terms of body mass index (BMI), parity, smoking history, hepatitis and human immunodeficiency virus (HIV) screening test results (Table I). The most common cause of infertility in both groups was male factor, although this was more common in the pregnant group (57.5 versus 34.8%). Pregnant women were less likely to have tubal disease (22.5 versus 31.5%) and more likely to have polycystic ovaries (20 versus 12%). The numbers of women with endometriosis and unexplained infertility were similar in both groups (Table I). White blood cell counts taken before starting treatment were not significantly different between pregnant and non-pregnant women, both means being in the normal range for a non-inflammatory state (<11 x 109/l)


View this table:
[in this window]
[in a new window]
 
Table I. Pre-treatment characteristics of pregnant and non-pregnant groups of women
 
Women who became pregnant were more likely to have had ICSI (50 versus 33.7%) and the agonist protocol (52.5 versus 32.6%) (Table II), and they received significantly less recombinant FSH (median 2000 IU, range 750–5200) than those who did not (median 2400 IU, range 800–5025, P < 0.05). However, there were no significant differences in the number of days of stimulation, endometrial thickness or peak serum estradiol (Table II). All women in this study required sedation and local anaesthetic only for egg collection. Significantly more eggs were collected from women who became pregnant (median 11, range 2–25) than those who did not (median 8, range 1–20, P < 0.01).


View this table:
[in this window]
[in a new window]
 
Table II. Stimulation cycle profiles of pregnant and non-pregnant groups of women
 
In all cases, one, two or three embryos were replaced. There was no significant difference between pregnant (one woman had one transferred, 36 had two, and three had three) and non-pregnant groups (13, 73 and nine) in the number of embryos transferred.

Amongst the pregnant women, there were 11 twin pregnancies and 29 singletons, and six women who became pregnant developed moderate ovarian hyperstimulation syndrome (OHSS) requiring hospital admission for fluid monitoring and thromboprophylaxis. Women who developed OHSS had significantly more eggs collected (median 17.5, range 10–25) compared with pregnant women who did not have OHSS (median 10, range 2–23, P < 0.01). When the women with OHSS were excluded from the pregnant group, there was no significant difference in the number of eggs collected compared with the non-pregnant group (P = 0.0727).

Median serum CRP levels measured before the start of treatment were below the level of detection of the assay (2 mg/l) in both groups of women, and were not significantly different (Table III). In both pregnant and non-pregnant groups, there was a small but significant rise in CRP at the day 14 blood test after the treatment cycle (Table III). However, the median CRP level in the pregnant group was significantly higher than in the non-pregnant group (P < 0.0001). The day 14 serum samples were also assessed with the high sensitivity CRP assay. This confirmed the highly significant difference between pregnant and non-pregnant groups (Table III). There was a significant correlation between the standard and high sensitivity CRP assays (Spearman coefficient r = 0.7433, P < 0.0001). Pregnant women had a mean serum {beta}HCG of 159.5 IU/l (range 65–370 IU/l) and there was no significant correlation between serum CRP and {beta}HCG levels measured at the day 14 blood test.


View this table:
[in this window]
[in a new window]
 
Table III. Serum CRP levels (mg/l)
 
Women with OHSS had significantly raised CRP levels compared with pregnant women without a diagnosis of OHSS (Table IV). However, even excluding those with OHSS, pregnant women (n = 34) had significantly higher CRP levels (median 2.5) than the non-pregnant group (P < 0.0001). Women with twin pregnancies (n = 11) had a median CRP level of 4 compared with singletons (n = 29) with 3, but this difference was not statistically significant. There was no significant difference between agonist and antagonist treatment cycles, and both demonstrated raised CRP levels in pregnancy (Table IV). Similarly, women whose pregnancies resulted from both IVF and ICSI had raised CRP levels, although women who had ICSI (n = 20) had significantly higher levels than those who had IVF (n = 20, P < 0.05). Finally, women with polycystic ovaries (PCO) did not have significantly raised CRP levels compared with those without PCO, and excluding women with PCO did not affect the difference in CRP between pregnant and non-pregnant women (Table IV).


View this table:
[in this window]
[in a new window]
 
Table IV. Subgroup analysis of day 14 serum CRP levels (mg/l)
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study provides evidence of a mild systemic maternal inflammatory response from as early as 4 weeks gestation. This is the earliest time in pregnancy that such a response has been reported. Although all women in this study had been treated by IVF/ICSI, those who were pregnant had significantly higher CRP levels than those who were not.

The obvious difficulty in investigating pregnancy at such an early stage of gestation was overcome in an IVF unit where women were closely monitored before and after their pregnancy tests. This does of course mean that a particular subgroup of pregnant women were being studied, but potential bias such as cause of infertility or of the treatment cycle itself could be controlled for by direct comparison with women who failed to fall pregnant after treatment. However, the obvious main drawback of this study remains the possibility of confounding factors. The numbers involved do not enable a comprehensive exclusion of bias, although many possible factors can be excluded or are unlikely to have a significant affect. Thus, in this study, women who became pregnant were significantly younger than those who did not, consistent with the well-known powerful effect of age on IVF success rates. Baseline CRP levels are associated with increasing age (Woodward et al., 2003Go), although in this study higher CRP levels were found in the younger group who were pregnant.

Higher CRP levels are associated with increasing BMI (Visser et al., 1999Go). However, a major criterion in selecting patients for IVF is a normal BMI and, hence, in this study, BMIs in both groups were closely matched and in the normal range. Similarly, the presence of PCO as a cause of infertility has a well-known association with BMI, and there were more women with PCO in the pregnant group. However, although numbers were small, women with PCO did not have raised CRP levels in this study, and excluding women with PCO did not alter the finding that CRP was raised in pregnancy.

Couples who became pregnant were more likely to have male factor than tubal factor causes of infertility, and hence more were treated by ICSI than by conventional IVF. It might have been hypothesized that tubal factors and associated chronic pelvic infection would cause higher CRP levels but, as with age, the CRP data are in the opposite direction; pregnant women have less tubal disease but higher CRP levels.

Endometriosis is another possible background factor that could affect CRP levels, but numbers were too small to analyse statistically. Most women undergoing treatment at Hammersmith Hospital have a diagnostic laparoscopy as part of their fertility investigations before IVF treatment, and it is therefore unlikely that significant cases have been missed. There were similar numbers of women with unexplained infertility in each group. Also, there was no significant difference between pregnant and non-pregnant groups in their smoking history, hepatitis and HIV screening, and parity. Thus, pre-treatment, there was no obvious pro-inflammatory factor present in one group rather than the other, and this was confirmed by pre-treatment CRP levels and white cell counts which were not significantly different.

Women who became pregnant received significantly less stimulating FSH than those who did not. This most probably represented the age difference between the groups as older women have less responsive ovaries and are given higher starting doses of FSH. There is no reason to suspect that less FSH stimulation would result in higher day 14 CRP levels and, moreover, the peak serum estradiol levels (used as a marker of ovarian stimulation) were not significantly different.

An important possible confounding variable was the number of eggs collected which was significantly higher in women who became pregnant. More eggs collected means more ovarian trauma and may itself result in a small rise in CRP. However, firstly, egg collection occurred 14 days before the CRP blood test. None of the women had complications following the procedure, and all were well when they came in for their pregnancy and CRP blood test. CRP is a very sensitive inflammatory marker and it is highly unlikely that, even if there were higher levels immediately after egg collection in the pregnant group, they would carry on for 2 weeks. Secondly, higher numbers of eggs are well known to be associated with OHSS. This association was confirmed in this study. Additionally, when women with OHSS were excluded from the analysis, there was no significant difference in the number of eggs collected in the pregnant and non-pregnant women. However, CRP levels were still significantly raised in pregnant women without OHSS.

The only difference between ICSI and conventional IVF is the method of fertilization in vitro, and it was surprising to find that women pregnant after ICSI had higher CRP levels than those after IVF. There is no reason to suspect that male factor infertility (requiring ICSI) would result in higher CRP levels per se. It may be hypothesized that the piercing of the zona pellucida for ICSI enhances hatching and this might lead to earlier implantation and thus a more advanced maternal inflammatory response.

Six pregnant women developed OHSS requiring otherwise uneventful hospital admission. Their CRP levels were significantly raised compared with pregnant women without OHSS, a finding which has not, as far as we are aware, been reported before. This is consistent with previous reports of raised white blood cell counts and the ‘inflammatory hypothesis’ for the development of OHSS (Delvigne and Rozenberg 2002Go).

Thus this study provides evidence for a maternal inflammatory response with raised CRP levels as early as 4 weeks gestation. Although the increase is small, there is a precedent in the ability of CRP to act as a marker for low-grade chronic inflammation in cardiovascular disease (Ablij and Meinders, 2002Go). The increase was detected using a standard hospital CRP assay, and it was confirmed with a high sensitivity assay. This finding is consistent with previously reported raised CRP levels later in the first trimester (Rebelo et al., 1995Go) and third trimester (Teran et al., 2001Go), and therefore extends the presence of a systemic maternal inflammatory response—which is well described in the third trimester (Sacks et al., 1999Go)—to the earliest phases of implantation.

It is not known what causes CRP levels to rise so early in pregnancy. The primary stimulant for CRP production is IL-6 (Volanakis, 2001Go) which is produced by macrophages and activated monocytes (Diehl and Rincon, 2002Go). Macrophages are, along with NK cells, the main leukocyte cell type found in the decidua in early pregnancy (Quenby et al., 1999Go). They are undoubtedly a key component of the interaction between trophoblast and the maternal immune system, but are extremely difficult to study. There is, however, increasing recognition that they are functional (Sacks et al., 1999Go; Petroff et al., 2002Go; Heikkinen et al., 2003Go) and hence are likely to be a source of IL-6 and the cause of raised CRP levels.

Numerous factors have been hypothesized to stimulate macrophages/monocytes in pregnancy (Sacks et al., 1999Go). In very early pregnancy, two possible candidates are corticotrophin-releasing hormone, which is produced by preimplantation embryos and stimulates an inflammatory response (L.Linton, personal communication), and {beta}HCG itself. As one of the first products of conception detectable in the maternal circulation, {beta}HCG has been shown to stimulate monocytes in vitro (Kosaka et al., 2002Go), although in this study there was no correlation between {beta}HCG and CRP levels.

If the inflammatory response that we have detected in the earliest phases of pregnancy is indeed an important feature of all ongoing pregnancies, there are wide-reaching consequences. First, it provides a possible mechanism for the recently reported association between the use of non-steroidal anti-inflammatory drugs and miscarriage (Li et al., 2003Go), and thus strengthens the argument against their use in early pregnancy. Secondly, an inflammatory marker such as CRP may be a useful early barometer of the success of implantation. There is already evidence that an exaggerated CRP response in the late first trimester may predispose to pre-eclampsia (Wolf et al., 2001Go) for which poor placentation is a powerful predisposing factor. We hypothesize that an abnormal CRP response (absent or exaggerated) may indicate that there is an imbalance in the establishment of the maternal–fetal interface and that such responses are also associated with some cases of miscarriage. Thus such a marker may be a useful predictor, or even a screening tool to select cases in which some form of immunotherapy may be of benefit. We currently are extending this work to include non-IVF pregnancies and to assess CRP levels in pregnancy failure.


    Acknowledgements
 
We would like to thank Kirsten Marks, Loren Sanders and Lauren Wallis for their invaluable help in performing the blood sample assays, and Many Donaldson for the high sensitivity CRP assay.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Ablij HC and Meinders AE (2002) C-reactive protein: history and revival. Eur J Intern Med 13,412–422.[CrossRef][Medline]

Delvigne A and Rozenberg S (2002) Systemic review of data concerning etiopathology of ovarian hyperstimulation syndrome. Int J Fertil Womens Med 47,211–226.[Medline]

Diehl S and Rincon M (2002) The two faces of IL-6 on Th1/Th2 differentiation. Mol Immunol 39,531–536.[CrossRef][Medline]

Egawa H, Fujiwara H, Hirano T, Nakayama T, Higuchi T, Tatsumi K, Mori T and Fujii S (2002) Peripheral blood mononuclear cells in early pregnancy promote invasion of human choriocarcinoma cell line, BeWo cells. Hum Reprod 17,473–480.[Abstract/Free Full Text]

Emmer PM, Nelen WLDM, Steegers EAP, Hendriks JC, Veerhoek M and Joosten I (2000) Peripheral natural killer cytotoxicity and CD56+CD16+ cells increase during early pregnancy in women with a history of recurrent spontaneous abortion. Hum Reprod 15,1163–1169.[Abstract/Free Full Text]

Heikkinen J, Mottonen M, Komi J, Alanen A and Lassila O (2003) Phenotypic characterization of human decidual macrophages. Clin Exp Immunol 131,498–505.[CrossRef][Medline]

Hvilsom GB, Thorsen P, Jeune B and Bakketeig LS (2002) C-reactive protein: a serological marker for preterm delivery? Acta Obstet Gynecol Scand 81,424–429.[CrossRef][Medline]

Kluft C and de Maat MPM (2002) Sensitive markers of inflammation make it possible to study the chronic process: the rise of interest in low levels of C-reactive protein. Vasc Pharmacol 39,99–104.[CrossRef]

Kosaka K, Fujiwara H, Tatsumi K, Yoshioka S, Sato Y, Egawa H, Higuchi T, Nakayama T, Ueda M, Maeda M et al. (2002) Human chorionic gonadotrophin (HCG) activates monocytes to produce interleukin-8 via a different pathway from luteinizing hormone/HCG receptor system. J Clin Endocrinol Metab 87,5199–5208.[Abstract/Free Full Text]

Li D-K, Liu L and Odouli R (2003) Exposure to non-steroidal anti-inflammatory drugs during pregnancy and risk of miscarriage: population based cohort study. Br Med J 327,368–372.[Abstract/Free Full Text]

Loke YW and King A (2000) Decidual natural-killer-cell interaction with trophoblast: cytolysis or cytokine production? Biochem Soc Trans 28,196–198.[Medline]

Monzon-Bordonaba F, Vadillo-Ortega F and Feinberg RF (2002) Modulation of trophoblast function by tumour necrosis factor-{alpha}: a role in pregnancy establishment and maintenance? Am J Obstet Gynecol 187,1574–1580.[CrossRef][Medline]

Ntrivalas EI, Kwak-Kim JYH, Gilman-Sachs A, Chung-Bang H, Ng SC, Beaman KD, Mantouvalos HP and Beer AE (2001) Status of peripheral blood natural killer cells in women with recurrent spontaneous abortions and infertility of unknown aetiology. Hum Reprod 16,855–861.[Abstract/Free Full Text]

Ogasawara MS, Aoki K, Aoyama T, Katano K, Iinuma Y, Azaki Y and Suzumori K (2000) Elevation of transforming growth factor-beta1 is associated with recurrent miscarriage. J Clin Immunol 20,453–457.[CrossRef][Medline]

Petroff, MG, Sedlmayr P, Azzola D and Hunt JS (2002) Decidual macrophages are potentially susceptible to inhibition by class Ia and class Ib HLA molecules. J Reprod Immunol 56,3–17.[CrossRef][Medline]

Quenby S, Bates M, Doig T, Brewster J, Lewis-Jones DI, Johnson PM and Vince G (1999) Pre-implantation endometrial leukocytes in women with recurrent miscarriage. Hum Reprod 14,2386–2391.[Abstract/Free Full Text]

Rebelo I, Carvalho-Guerra F, Pereira-Leite L and Quintanilha A (1995) Lactoferrin as a sensitive blood marker of neutrophil activation in normal pregnancies. Eur J Obstet Gynecol Reprod Biol 62,189–194.[CrossRef][Medline]

Redman CWG, Sacks GP and Sargent IL (1998) Preeclampsia: an excessive maternal inflammatory response to pregnancy. Am J Obstet Gynecol 180,499–506.

Sacks GP, Sargent IL and Redman CWG (1999) An innate view of human pregnancy. Immunol Today 20,114–118.[CrossRef][Medline]

Smarason AK, Gunnarsson A, Alfredsson JH and Valdimarsson H (1986) Monocytosis and monocytic infiltration of deciduas in early pregnancy. J Clin Lab Immunol 21,1–5.[Medline]

Teran E, Escudero C, Moya W, Flores M, Vallance P and Lopez-Jaramillo P (2001) Elevated C-reactive protein and pro-inflammatory cytokines in Andean women with pre-eclampsia. Int J Gynaecol Obstet 75,243–249.[CrossRef][Medline]

VeenstravanNieuwenhoven AL, Heineman MJ and Faas MM (2003). The immunology of successful pregnancy. Hum Reprod Update 9,347–357.[Abstract/Free Full Text]

Visser M, Bouter LM, McQuillan GM, Wener MH and Harris TB (1999) Elevated C-reactive protein levels in overweight and obese adults. J Am Med Assoc 282,2131–2135.[Abstract/Free Full Text]

Volanakis JE (2001) Human C-reactive protein: expression, structure, and function. Mol Immunol 38,189–197.[CrossRef][Medline]

Wolf M, Kettyle E, Sandler L, Ecker JL, Roberts J and Thadhani R (2001) Obesity and preeclampsia: the potential role of inflammation. Obstet Gynecol 98,757–762.[Abstract/Free Full Text]

Wolf M, Sandler L, Hsu K, Vossen-Smirnakis K, Ecker JL and Thadhani R (2003) First trimester C-reactive protein and subsequent gestational diabetes. Diabetes Care 26,819–824.[Abstract/Free Full Text]

Woodward M, Rumley A, Lowe GD and Tunstall-Pedoe H (2003) C-reactive protein: associations with haematological variables, cardiovascular risk factors and prevalent cardiovascular disease. Br J Haematol 122,135–141.[CrossRef][Medline]

Yoon BH, Jun JK, Park KH, Syn HC, Gomez R and Romero R (1996) Serum C-reactive protein, white blood cell count, and amniotic fluid white blood cell count in women with preterm premature rupture of membranes. Obstet Gynecol 88,1034–1040.[Abstract/Free Full Text]

Submitted on August 11, 2003 resubmitted on November 13, 03 ; accepted on December 12, 2003