OBJECTIVE: To estimate whether elevations of complement C3a early in pregnancy are predictive of the subsequent development of adverse pregnancy outcomes.
METHODS: A plasma sample was obtained from each enrolled pregnant woman before 20 weeks of gestation. The cohort (n=1,002) was evaluated for the development of adverse pregnancy outcomes defined as hypertensive diseases of pregnancy (gestational hypertension or preeclampsia), preterm birth (before 37 weeks of gestation), premature rupture of the membranes, pregnancy loss (during the embryonic and fetal period), intrauterine growth restriction, and the composite outcome of any adverse outcome.
RESULTS: One or more adverse pregnancy outcomes occurred in 211 (21%) of the cohort. The mean levels (ng/mL) of C3a in early pregnancy were significantly (P=<.001) higher among women with one or more adverse outcomes (858±435) compared with women with an uncomplicated pregnancy (741±407). Adjusted for parity and prepregnancy body mass index, women with levels of C3a in the upper quartile in early pregnancy were three times more likely to have an adverse outcome later in pregnancy compared with women in the lowest quartile (95% confidence interval, 1.8–4.8; P<.001). The link between early elevated C3a levels and adverse pregnancy outcomes was driven primarily by individual significant (P<.05) associations of C3a with hypertensive diseases of pregnancy, preterm birth, and premature rupture of the membranes.
CONCLUSION: Elevated C3a as early as the first trimester of pregnancy is an independent predictive factor for adverse pregnancy outcomes, suggesting that complement-related inflammatory events in pregnancy contribute to the subsequent development of poor outcomes at later stages of pregnancy.
LEVEL OF EVIDENCE: II
Elevated levels of the complement activation fragment C3a in early pregnancy are associated with subsequent adverse pregnancy outcomes.
From the Department of Obstetrics and Gynecology, University of Colorado, School of Medicine, Aurora, Colorado.
Supported by Eunice Kennedy Shriver National Institute of Child Health and Human Development K23 HD049684 (A.M.L.), American Heart Association Award 0865481G (A.M.L.), Center for Women's Health Research and the List Family Foundation at University of Colorado Denver (A.M.L.), Newborn Hope Colorado (A.M.L.), National Institutes of Health AR49772 (J.E.S.), AR38889 (J.E.S.), and National Institutes of Health RO1 AI 55007 (V.M.H.).
The authors thank the research staff attached to the University of Colorado, Department of Obstetrics and Gynecology, Colorado Baby Blanket Research Program for help with participant recruitment and follow-up, sample collection, data collection, and data entry; the staff at the Complement Laboratory at National Jewish Health for performing the complement assays; and the database staff at National Jewish Health for their attention to the complement study database.
The complement assays were conducted at National Jewish Health, Denver, Colorado.
Presented at the XXIII International Complement Workshop, August 1–5, 2010, New York, New York.
Corresponding author: Anne M. Lynch, MB, BCh, BAO, MSPH, Department of Obstetrics and Gynecology, University of Colorado Denver School of Medicine, Mail Stop B198, 12631 East 17th Avenue, Aurora, CO 80045; e-mail: Anne.Lynch@ucdenver.edu.
Financial Disclosure The authors did not report any potential conflicts of interest.
The complement system has functions critical to the innate immune response. Named for its “complementary” activity with antibodies to destroy bacteria, the complement system is a series of more than 30 proteins (soluble and membrane-bound) that provides one of the first lines of defense against pathogens. The complement system has three main functions: it defends the host against infections, it bridges innate and adaptive immunity, and it disposes of immune complexes, apoptotic bodies, and the products of inflammatory or traumatic injury.1,2 Complement has three initiating mechanisms known as the classical, lectin, and alternative pathways. The pathways and the triggers for activation of the individual pathways are described in Figure 1.1,3–6 The biologic functions of the complement system are mediated through the production of activation fragments. For example, the activation fragment C3a is an anaphylatoxin. Activation of the complement system therefore carries benefits for the host. This is especially true in early pregnancy, which is a time of intense tissue turnover, and indeed complement activation both reflects and contributes to the heightened inflammatory state seen in normal pregnancy.7
In recent times, attention has focused on the importance of the contribution of dysregulation of the complement system to adverse pregnancy outcomes.8,9 Dysregulation can present as either excessive activation or inadequate regulation of this complex system. Studies of the murine model of the antiphospholipid antibody syndrome in the early part of this decade demonstrated that excessive complement activation is the primary effector mechanism for intrauterine loss and intrauterine growth restriction (IUGR). Inhibition of the complement cascade blocked the antiphospholipid antibody-related fetal loss and IUGR.10,11 Our group translated these results into a study in human pregnancy and found that women with elevated levels of the alternative complement pathway Bb (Fig. 1) in early pregnancy were at risk of developing hypertensive diseases of pregnancy and spontaneous preterm birth later in pregnancy.12,13 The last decade has also seen groundbreaking perinatal research on the contribution of mutations in complement regulatory proteins to adverse outcomes,5 including pregnancy-associated hemolytic uremic syndrome14 and the variant of preeclampsia characterized by hemolysis, elevated liver enzymes, and low platelet count and known as HELLP syndrome.15,16 All of this recent novel research underscores the important contribution of this understudied area of innate immunity to adverse pregnancy outcomes.
The purpose of the current study was to expand our previous research to another activation fragment namely, C3a (Fig. 1). The aim of the current study was to estimate the role of C3a in adverse pregnancy outcomes to include hypertensive diseases of pregnancy, preterm birth, premature rupture of the membranes, pregnancy loss, and IUGR.
MATERIALS AND METHODS
This prospective study (June 2005 to June 2008) was approved by the Colorado Multiple Institutional Review Board. Details of the study have been described in previous publications.12,13,17 In brief, women were recruited from the University of Colorado Hospital prenatal clinics and two affiliated sites. Women were referred to the study's research assistant by the prenatal intake nurse if they were in the first half of pregnancy. Informed consent was obtained and additional plasma (for complement activation fragments) obtained before 20 weeks of gestation with the routine prenatal laboratory values. Data were gathered on the maternal medical and obstetric history. The women were followed throughout pregnancy. At or shortly after delivery, outcome data were collected and the gestational age at blood draw (recruitment visit) was assigned based on the best overall obstetric estimate incorporating assessment at the first visit and in the great majority of the cohort on early ultrasound examination (more than 90% had an ultrasonography at less than 15 weeks of gestation). Women with an uncertain timing of a pregnancy outcome were classified as having “uncertain dates.”
We excluded from the complete cohort (n=1,287) women with: repeat deliveries during the study period (n=6), women who presented after the second half of pregnancy (n=10), multiple gestations (n=47), loss to follow-up (n=49), induced abortions, congenital or chromosomal anomalies (n=20), and chronic medical disease (cardiovascular disease, insulin-dependent diabetes mellitus, autoimmune disease; n=113). Women with a missing plasma sample (resulting from a deviation from study protocol at the initial blood draw, n=40) were also removed from the analysis. After exclusions, 1,002 women remained in the analytic data set.
The main outcome of the study was adverse pregnancy outcome defined as the composite outcome of one or more of the following pregnancy complications: gestational hypertension, preeclampsia, pregnancy loss, IUGR, premature rupture of the membranes, and spontaneous and medically indicated preterm birth (less than 37 weeks of gestation). The definition of gestational hypertension was a systolic blood pressure greater than 140 mm Hg or a diastolic blood pressure greater than 90 mm Hg on two or more occasions at least 6 hours apart after 20 weeks of gestation in a woman known to have been normotensive before pregnancy and before 20 weeks of gestation.18 Preeclampsia was defined as 1) gestational hypertension with proteinuria (300 mg+ per 24-hour period) or at least 1+ on dipstick; or 2) in the absence of proteinuria gestational hypertension with cerebral symptoms, epigastric or right upper quadrant pain with nausea or vomiting or thrombocytopenia and an abnormal liver function test.18 Hypertensive disease of pregnancy was defined as the development of either preeclampsia or gestational hypertension in pregnancy. Pregnancy loss was categorized as occurring in the embryonic (less than 10 weeks of gestation) or fetal period (greater than 10 weeks of gestation)19,20 Intrauterine growth restriction was defined as a birth weight less than the tenth percentile21 for gestational age using the Colorado Intrauterine Growth Charts.22 Preterm rupture of the membranes was defined as membrane rupture before the onset of labor.23 Preterm birth less than 37 completed weeks of gestation was categorized as: 1) spontaneous preterm birth resulting from preterm premature rupture of the membranes or spontaneous labor; and 2) medically indicated preterm birth defined as a preterm birth that resulted from a medically indicated delivery for maternal or fetal indications. The gestational age for the lower boundary of preterm birth is controversial. For this analysis, we used 20 weeks of gestation as the lower boundary.24 The outcomes data described previously were collected by the perinatal database staff and entered into the complement study data set housed in the Division of Biostatistics and Bioinformatics at National Jewish Health. A secondary review of each completed maternal record was conducted by one investigator (A.M.L.) before the participant's sample was assayed.
The complement activation fragment, C3a, was the primary exposure examined. Maternal risk factors included in the analysis were: age, race or ethnicity (non-Hispanic white, Hispanic white, African American, Asian, and other), parity (nulliparous compared with multiparous), cigarette smoking at conception (yes compared with no), the maternal prepregnant body mass index (BMI, calculated as weight (kg)/[height (m)]2) (25 to 30 [overweight], higher than 30 [obese], compared with less than 25), calculated from the maternal prepregnant weight and height, and the gestational age at blood draw (less than compared with greater than 12 weeks of gestation).
Each plasma sample was centrifuged and processed within an average time of 11 minutes, aliquoted, and placed in a freezer at −80°C. At the time of the assay, an aliquot of plasma was thawed. The Pharmingen OptEIATM enzyme-linked immunosorbent assay was used to measure C3a. The inter- and intraassay coefficients of variation were 12.1% and 5.5%, respectively. The individuals performing all study assays were blinded to the participant's pregnancy outcome.
The data were analyzed in SAS 9.2. Initially univariable analysis was used. Measures of the complement activation fragments were examined as continuous and categorical (divided into quartiles) variables. Associations for categorical variables were tested using the chi square test (P<.05). Means for continuous variables are reported in the tables, but nonparametric methods (analysis of variance [Kruskal-Wallis test]) were used to test for differences in medians among groups. In some parts of the analysis, we used a contrast statement to test for linear trend within the analysis of variance structure. We performed multivariable logistic regression analysis to estimate the adjusted odds ratio (OR) of C3a for the outcome adjusted for covariates identified as being significantly associated with the outcome in the univariable stage of the analysis. The adjusted ORs of C3a examined as a continuous variable was calculated for every 1-standard deviation change in levels of the biomarker. In the final stage of the analysis, we looked at the levels of C3a in cases of the individual outcomes compared with those of participants in the control group. Deliveries at less than 20 weeks of gestation were removed from this stage of the analysis to comply with the definitions of preterm birth and hypertensive disease of pregnancy.
As seen in Table 1, 21% of this diverse cohort had an adverse pregnancy outcome. Eighty-three (8.5%) women in the cohort had a preterm birth and 8.5% developed gestational hypertension or preeclampsia. A small proportion of the women in the cohort had a pregnancy loss or IUGR. The prevalence of reported cigarette smoking at conception was low at 6%. In the relatively young cohort of women, just more than one third were either obese or overweight before they became pregnant.
Concentrations of C3a were significantly elevated in women with an adverse pregnancy outcome compared with women who had an uncomplicated pregnancy (Table 2). Examined as a categorical variable, a significant association was found between C3a levels in the upper most, third, and second quartiles for an adverse outcome compared with the lower quartile. Regarding other maternal risk factors, a significantly higher proportion of adverse outcomes was observed in nulliparous women. Adverse outcomes were also associated with increasing levels of a maternal prepregnant BMI (Table 2).
Using logistic regression analysis, we determined adjusted for BMI and parity that for every standard deviation increase in concentrations of C3a, the OR for an adverse outcome was 1.3 (confidence interval, 1.1–1.5; P=.001). Examined as a categorical variable and adjusted for BMI and parity (Table 3), concentrations of C3a in the top three quartiles were significantly associated with adverse outcomes with adjusted ORs of 2.0, 2.7, and 3.0, respectively (model 1). We explored the data further (Table 3) by stratifying the data by time of gestation of blood draw (before or after 12 weeks of gestation; models 2 and 3) by parity (models 4 and 5) and BMI (models 6 and 7). Stratified by gestational age at blood draw, the adjusted OR was significant in the two upper quartiles among women with a blood draw between 4 and 12 weeks of gestation. Levels of C3a in quartile 2 were not significantly associated with the outcome. The relationship was significant across all quartiles among women who had a blood draw between 12 and 20 weeks of gestation. Stratified by parity (models 4 and 5), the adjusted OR of C3a among nulliparous women for an adverse outcome rose significantly from just more than 3 in the second and third quartiles to 4.1 in the upper quartile. The adjusted ORs for an adverse pregnancy outcome were all elevated (significant in the upper two quartiles) among multiparous women but were slightly attenuated compared with nulliparous women. With obese women removed from the cohort, the adjusted OR of quartiles of C3a remained highly significant for a poor pregnancy outcome (model 6). Similarly, among the smaller group of obese women (model 7), the risk was increased across all categories of C3a. This relationship was significant for levels of C3a in the second and upper quartiles.
In the next stage of the analysis, we separated out the outcomes into their individual components to estimate the contribution of elevated levels of C3a to each pregnancy complication (Table 4). Significantly higher levels of C3a were found in women who developed any hypertensive disease of pregnancy (P<.001) or gestational hypertension (P<.001). However, although the mean C3a levels were higher among women who developed preeclampsia compared with women who remained normotensive during pregnancy, this relationship did not reach statistical significance (P=.2). There were significantly higher levels of the activation fragment in early pregnancy among women who subsequently had a spontaneous preterm birth or medically indicated preterm birth compared with women who had a term delivery. We also found a significant relationship with premature rupture of the membranes. Thirty-one of the 48 cases of spontaneous preterm birth were associated with preterm premature rupture of the membranes. Eleven cases of preterm premature rupture of the membranes occurred at 20 to 34 weeks of gestation and 20 were between 34 and 37 weeks of gestation. We found the highest levels of C3a among the group of women who delivered between 20 and 34 weeks of gestation and lowest levels among the term premature ruptures of the membranes. Importantly, we found a significant linear trend in levels of C3a across gestational age categories of premature rupture of the membranes (Table 4). We found no significant difference between levels of C3a in women with a spontaneous (n=13) or medically indicated preterm birth (n=6) at less than 34 weeks of gestation compared with term deliveries. Although levels of C3a were elevated among women with pregnancy losses in the fetal period compared with women who had a loss in the embryonic period or women who had no loss (Table 4), this difference was not statistically significant. Among the pregnancy losses, we found 12 women who had an uncertain gestational age at the time of the loss. We repeated the analysis with these cases (75% were losses during the embryonic period) removed. The mean±standard deviation levels of C3a among women with an embryonic (n=4), fetal loss (n=15), or no loss (n=971) were 681±118, 833±232, and 765±420, respectively (P=.8). Similarly, higher levels of C3a were found in women who had a pregnancy complicated by IUGR, but this relationship was not statistically significant.
In this analysis we found that the activation fragment C3a measured at a single point in early pregnancy is significantly elevated in women who have an adverse event in later pregnancy compared with women who had an uncomplicated pregnancy. Importantly, we demonstrated that significantly higher levels of C3a are associated with poor pregnancy outcome even among women with a sample drawn in the first trimester. The results were especially striking among nulliparous women; adjusted for BMI, nulliparous women with a C3a level in the upper quartile were 4.1 times significantly more likely to have an adverse pregnancy outcome compared with nulliparous women with levels of C3a in the lower quartile, a finding perhaps related to immune maladaptation25 in early pregnancy Overall, our study suggests an important role for complement-mediated inflammation in early pregnancy in the subsequent development of adverse pregnancy outcomes.
We cautiously approached the use of a composite outcome in the main part of the analysis for the reasons outlined recently by Cordoba et al.26 We believe that we were justified in using the composite outcome approach because of results of our previous analysis in which we demonstrated higher complement levels in early pregnancy in a subset of women destined to have several adverse pregnancy outcomes.12,13,17 We were also meticulous in our definitions of outcomes throughout the analysis and listed results of all the components of the composite outcome with P values to present clearly the outcomes having the most effect on the relationship of the exposure with the composite outcome (Table 4).26 In the breakout analysis into the individual components, we found a significant relationship between C3a concentrations in early pregnancy with the development of hypertensive disease later in pregnancy, in line with the findings of other authors.27,28 This association was driven by a strong relationship of C3a with gestational hypertension (Table 4). These findings are aligned with the results of our previous research that demonstrated a relationship between another complement activation fragment, Bb, with preeclampsia.12 In this current analysis, the relationship of C3a with preeclampsia was not so strong but in the same direction, perhaps secondary to the low number of cases in the outcome (Table 4). Another possibility is that the source of the activation fragments may be different. As previously described, Bb is primarily associated with alternative complement pathway activation,12 whereas C3a can arise from any of the complement pathways (Fig. 1). We speculate that there may be stronger links with alternative complement pathway activation in women with preeclampsia. This observation is consistent with dysregulation of the alternative pathway of complement activation recently described in HELLP syndrome.16
We found a significant relationship between elevated levels of C3a in early pregnancy with spontaneous preterm birth at less than 37 weeks of gestation. This is an interesting observation because infection, inflammation, or both is the only risk factor shown consistently to have a strong causal link with preterm birth.29 Given the important connection between the complement system and inflammation, it is remarkable that complement activation as it relates to preterm birth has only received the attention of a few authors.13,30–34 It is of great interest that we found significantly higher levels of C3a in early pregnancy in women who subsequently had an early rupture of the fetal membranes at any gestation in later pregnancy compared with women who had no premature rupture of the membranes. When we examined mean levels of C3a in women with premature rupture of the membranes across these categories of gestational age, we found a dose–response with the highest concentrations of C3a among women who had premature rupture of the membranes between 20 and 34 weeks of gestation and the lowest levels among women with a term premature rupture of the membranes (Table 4). This significant linear trend is in agreement with strong evidence that intrauterine infection or inflammation contributes significantly to spontaneous preterm birth, especially spontaneous preterm birth that occurs less than 34 weeks of gestation.35,36 These results are also in agreement with the results of our previous research in which we found a significant relationship between elevated levels of the complement activation fragment Bb in early pregnancy and spontaneous preterm birth less than 34 weeks of gestation.13 In the current analysis, we also found C3a to be associated with medically indicated preterm birth, an observation that needs to be examined in a larger cohort. Higher levels of C3a were observed in women who have a pregnancy loss in the fetal period compared with women who had an embryonic loss or with women who had no loss. This observation also needs to be pursued in a larger cohort to specifically address the role of complement activation in pregnancy loss after the first trimester.
Notwithstanding these important results, there are limitations to the study. The main issue is sample size with low number of cases of outcomes when examined individually. This mainly pertains to the outcomes preeclampsia and intrauterine loss and IUGR, discussed previously. With larger numbers, and specifically for preeclampsia, we would have had the opportunity to conduct a stratification analysis by severity of the disease and gestational age at onset of the disease. Women were also undoubtedly missed from this cohort with an intrauterine fetal death as the incidence of this outcome was lower than anticipated.37 We suggest that the death was not captured in this cohort because the event occurred at home before prenatal care was initiated. Our exclusion criteria contributed to the low number of neonates with IUGR. We also acknowledge that the Lubchenco growth curves are limited by their age and need reevaluation. However, currently the Lubchenco curves, developed in Denver, are the only growth curves available for our high altitude. The analysis is also limited by the large range of gestational age at blood draw (between 4 and 20 weeks of gestation). A narrower range of gestational age within data points would be beneficial in estimating the correlation of complement markers with one another and in estimating the more predictive biomarker. We also acknowledge that it will be important in the future to study the role of complement activation at data points beyond the first half of pregnancy. A larger data set would also facilitate exploration of the interaction of these activation fragments with other covariates, specifically maternal obesity (Table 3).17 Despite these limitations, the data presented provide additional support of the significant relationship between complement-mediated inflammation in early pregnancy with the subsequent development of adverse pregnancy outcomes.
The study results relating to preterm premature rupture of the membranes were especially important and speak to the fact that an early inflammatory episode was part of the natural history of an event that occurred later in pregnancy in a subset of women. The disappointing results of the many antibiotic intervention trials in women at risk for infection-related preterm birth38 underscore the fact that it may be important to further explore inflammation39,40 in early pregnancy as a trigger for preterm birth41,42 and indeed other adverse pregnancy outcomes.43,44 We await results from the ongoing, multicentered, Eunice Kennedy Shriver National Institute of Child Health and Human Development–sponsored Effects of Aspirin in Gestation and Reproduction randomized control trial in which low-dose aspirin has been given during preconception and throughout pregnancy in women with a history of intrauterine loss.45 The results of this study may shed light on the effect of this agent with antiinflammatory effects on the incidence of not only live birth, but also other adverse pregnancy outcomes. The results of this study build on the results of our previous research. We have now demonstrated that elevated levels of two complement activation fragments in early pregnancy are linked with the subsequent development of adverse pregnancy outcomes. The trigger for the upregulation of the complement system remains to be determined but may possibly be linked with inflammatory events in the placenta9–11 or in extraplacental tissues (eg, adipose tissue12,17). The link between complement activation and adverse pregnancy outcomes demonstrated in our research identifies a new potential marker to predict patients at risk for poor pregnancy outcomes.
1. Walport MJ. Complement. First of two parts. N Engl J Med 2001;344:1058–66.
2. Sjöberg AP, Trouw LA, Blom AM. Complement activation and inhibition: a delicate balance. Trends Immunol 2009;30:83–90.
3. Holers VM. The spectrum of complement alternative pathway-mediated diseases. Immunol Rev 2008;223:300–16.
4. Markiewski MM, Lambris JD. The role of complement in inflammatory diseases from behind the scenes into the spotlight. Am J Pathol 2007;171:715–27.
5. Pickering MC, Cook HT. Translational mini-review series on complement factor H: renal diseases associated with complement factor H: novel insights from humans and animals. Clin Exp Immunol 2008;151:210–30.
6. Ricklin D, Hajishengallis G, Yang K, Lambris JD. Complement: a key system for immune surveillance and homeostasis. Nat Immunol 2010;11:785–97.
7. Holmes CH, Simpson KL. Complement and pregnancy: new insights into the immunobiology of the fetomaternal relationship. Baillieres Clin Obstet Gynaecol 1992;6:439–60.
8. Lynch AM, Salmon JE. Dysregulated complement activation as a common pathway of injury in preeclampsia and other pregnancy complications. Placenta 2010;31:561–7.
9. Caucheteux SM, Kanellopoulos-Langevin C, Ojcius DM. At the innate frontiers between mother and fetus: linking abortion with complement activation. Immunity 2003;18:169–72.
10. Holers VM, Girardi G, Mo L, Guthridge JM, Molina H, Pierangeli SS, et al. Complement C3 activation is required for antiphospholipid antibody-induced fetal loss. J Exp Med 2002;195:211–20.
11. Salmon JE, Girardi G, Holers VM. Complement activation as a mediator of antiphospholipid antibody induced pregnancy loss and thrombosis. Ann Rheum Dis 2002;61(suppl 2): ii46–50.
12. Lynch AM, Murphy JR, Byers T, Gibbs RS, Neville MC, Giclas PC, et al. Alternative complement pathway activation fragment Bb in early pregnancy as a predictor of preeclampsia. Am J Obstet Gynecol 2008;198:385.e1–9.
13. Lynch AM, Gibbs RS, Murphy JR, Byers T, Neville MC, Giclas PC, et al. Complement activation fragment Bb in early pregnancy and spontaneous preterm birth. Am J Obstet Gynecol 2008;199:354.e1–8.
14. Fakhouri F, Roumenina L, Provot F, Sallee M, Caillard S, Couzi L, et al. Pregnancy-associated hemolytic uremic syndrome revisited in the era of complement gene mutations. J Am Soc Nephrol 2010;21:859–67.
15. Fang CJ, Fremeaux-Bacchi V, Liszewski MK, Pianetti G, Noris M, Goodship TH, et al. Membrane cofactor protein mutations in atypical hemolytic uremic syndrome (aHUS), fatal Stx-HUS, C3 glomerulonephritis, and the HELLP syndrome. Blood 2008;111:624–32.
16. Fang CJ, Richards A, Liszewski MK, Kavanagh D, Atkinson JP. Advances in understanding of pathogenesis of aHUS and HELLP. Br J Haematol 2008;143:336–48.
17. Lynch AM, Murphy JR, Gibbs RS, Levine RJ, Giclas PC, Salmon JE, et al. The interrelationship of complement-activation fragments and angiogenesis-related factors in early pregnancy and their association with pre-eclampsia. BJOG 2010;117:456–62.
18. Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005;365:785–99.
19. Goldstein SR. Embryonic death in early pregnancy: a new look at the first trimester. Obstet Gynecol 1994;84:294–7.
20. Oshiro BT, Silver RM, Scott JR, Yu H, Branch DW. Antiphospholipid antibodies and fetal death. Obstet Gynecol 1996;87:489–93.
21. Resnick R, Creasy R. Intrauterine growth restriction. In: Creasy R, Resnik R, Iams J, Lockwood C, Moore T, editors. Creasy and Resnick's maternal–fetal medicine: principles and practice. Philadelphia (PA): WB Saunders; 2009. p. 635–50.
22. Lubchenco LO, Hansman C, Dressler M, Boyd E. Intrauterine growth as estimated from liveborn birth-weight data at 24 to 42 weeks of gestation. Pediatrics 1963;32:793–800.
23. Mercer BM. Preterm premature rupture of the membranes: diagnosis and management. Clin Perinatol 2004;31:765–82, vi.
24. Iams JD, Romero R, Creasy RK. Preterm labor and birth. In: Creasy RK, Resnik R, Iams JD, Lockwood CJ, Moore TR, editors. Creasy and Resnick's maternal–fetal medicine: principles and practice. 6th ed. Philadelphia (PA): WB Saunders; 2009. p. 545.
25. Dekker G, Robillard PY. Pre-eclampsia: is the immune maladaptation hypothesis still standing? An epidemiological update. J Reprod Immunol 2007;76:8–16.
26. Cordoba G, Schwartz L, Woloshin S, Bae H, Gotzsche PC. Definition, reporting, and interpretation of composite outcomes in clinical trials: systematic review. BMJ 2010;341:c3920.
27. Derzsy Z, Prohaszka Z, Rigo J Jr, Fust G, Molvarec A. Activation of the complement system in normal pregnancy and preeclampsia. Mol Immunol 2010;47:1500–6.
28. Haeger M, Bengtson A, Karlsson K, Heideman M. Complement activation and anaphylatoxin (C3a and C5a) formation in preeclampsia and by amniotic fluid. Obstet Gynecol 1989;73:551–6.
29. Romero R, Espinoza J, Goncalves LF, Kusanovic JP, Friel LA, Nien JK. Inflammation in preterm and term labour and delivery. Semin Fetal Neonatal Med 2006;11:317–26.
30. Soto E, Romero R, Richani K, Espinoza J, Nien JK, Chaiworapongsa T, et al. Anaphylatoxins in preterm and term labor. J Perinat Med 2005;33:306–13.
31. Soto E, Romero R, Richani K, Yoon BH, Chaiworapongsa T, Vaisbuch E, et al. Evidence for complement activation in the amniotic fluid of women with spontaneous preterm labor and intra-amniotic infection. J Matern Fetal Neonatal Med 2009;22:983–92.
32. Vaisbuch E, Romero R, Erez O, Mazaki-Tovi S, Kusanovic JP, Soto E, et al. Activation of the alternative pathway of complement is a feature of pre-term parturition but not of spontaneous labor at term. Am J Reprod Immunol 2010;63:318–30.
33. Swierzko AS, Atkinson AP, Cedzynski M, Macdonald SL, Szala A, Domzalska-Popadiuk I, et al. Two factors of the lectin pathway of complement, l-ficolin and mannan-binding lectin, and their associations with prematurity, low birthweight and infections in a large cohort of Polish neonates. Mol Immunol 2009;46:551–8.
34. Elimian A, Figueroa R, Canterino J, Verma U, Aguero-Rosenfeld M, Tejani N. Amniotic fluid complement C3 as a marker of intra-amniotic infection. Obstet Gynecol 1998;92:72–6.
35. Watts DH, Krohn MA, Hillier SL, Eschenbach DA. The association of occult amniotic fluid infection with gestational age and neonatal outcome among women in preterm labor. Obstet Gynecol 1992;79:351–7.
36. Romero R, Espinoza J, Goncalves LF, Kusanovic JP, Friel L, Hassan S. The role of inflammation and infection in preterm birth. Semin Reprod Med 2007;25:21–39.
37. Wilcox AJ, Weinberg CR, O'Connor JF, Baird DD, Schlatterer JP, Canfield RE, et al. Incidence of early loss of pregnancy. N Engl J Med 1988;319:189–94.
38. Andrews WW, Sibai BM, Thom EA, Dudley D, Ernest JM, McNellis D, et al. Randomized clinical trial of metronidazole plus erythromycin to prevent spontaneous preterm delivery in fetal fibronectin-positive women. Obstet Gynecol 2003;101:847–55.
39. Wegmann TG, Lin H, Guilbert L, Mosmann TR. Bidirectional cytokine interactions in the maternal–fetal relationship: is successful pregnancy a TH2 phenomenon? Immunol Today 1993;14:353–6.
40. Chaouat G. The Th1/Th2 paradigm: still important in pregnancy? Semin Immunopathol 2007;29:95–113.
41. Blank V, Hirsch E, Challis JR, Romero R, Lye SJ. Cytokine signaling, inflammation, innate immunity and preterm labour—a workshop report. Placenta 2008;29(suppl A):S102–4.
42. Lockwood CJ, Kuczynski E. Markers of risk for preterm delivery. J Perinat Med 1999;27:5–20.
43. Whitcomb BW, Schisterman EF, Klebanoff MA, Baumgarten M, Luo X, Chegini N. Circulating levels of cytokines during pregnancy: thrombopoietin is elevated in miscarriage. Fertil Steril 2008;89:1795–802.
44. Redman CW, Sargent IL. Pre-eclampsia, the placenta and the maternal systemic inflammatory response—a review. Placenta 2003;24(suppl A):S21–7.
45. Eunice Kennedy Shriver
National Institute of Child Health and Human Development. EAGeR trial. Available at: www.eagertrial.org/
. Retrieved August 20, 2010.
46. Walport MJ. Complement. Second of two parts. N Engl J Med 2001;344:1140–4.