Evaluation of the diagnostic value of the first-trimester maternal serum high-sensitivity C-reactive protein level for prediction of pre-eclampsia

Authors


Abstract

Aim

The purpose of the present study was to evaluate the diagnostic value of maternal serum high-sensitivity C-reactive protein (hs-CRP) measurement during the first trimester of pregnancy for predicting pre-eclampsia.

Material and Methods

A prospective cohort study was performed on 394 pregnant women who were at the gestational age of 8–13 weeks. In all women, serum hs-CRP was measured by latex agglutination test. The women were then monitored to delivery. We compared the hs-CRP of the two groups, those with and without pre-eclampsia. We used the receiver–operator curve for finding the optimum cut-off points.

Results

Out of 394 women, 42 cases (10.7%) were complicated by pre-eclampsia, of whom 23 women (56.1%) had severe pre-eclampsia. Mean serum hs-CRP of the pre-eclamptic group was higher than that of the normotensive group (7.06 ± 2.6 mg/L vs 3.6 ± 2.3 mg/L, P = 0.001). The receiver–operator curve showed a significant difference between the under-curve zone for the hs-CRP level with the reference line. Serum hs-CRP of 4 mg/L showed sensitivity, specificity, positive predictive value, negative predictive value and diagnostic accuracy as 78.1%, 72.1%, 25%, 96.5% and 72.8%, respectively. Hs-CRP of more than 7 mg/L was found in 26 (61.9%) cases of pre-eclampsia and 22 (6.25%) normotensive pregnancies, which showed a significant difference (P = 0.001, relative risk = 12.1, 95% confidence interval: 6.91–21.15).

Hs-CRP of more than 7 mg/L was found in 17 (73.91%) cases of severe pre-eclampsia and 22 (6.25%) normotensive pregnancies, which showed a significant difference (P = 0.001, relative risk = 9.35, 95% confidence interval: 4.48–19.52).

Conclusion

Hs-CRP measurements during the first trimester of pregnancy are helpful in predicting pre-eclampsia.

Introduction

Pre-eclampsia is one of the most serious complications during pregnancy and is a major cause of maternal mortality and morbidity. It is also one of the complications during pregnancy that we know little about. Despite many studies on its pathogenesis, there are still many unanswered questions. It is seen in about 5–10 percent of all pregnancies.[1]

Abnormal trophoblastic invasion, immunologic maladaptation between fetal, maternal and paternal tissue, and also genetic factors all have been reported as causative factors.[1, 2] An inflammatory change happens following defective placentation,[2, 3] therefore it seems that antiangiogenic and metabolic factors and the other inflammatory mediators cause endothelial cell damage. It has been suggested that endothelial cell dysfunction is due to overactivity of leucocytes in the maternal circulation.[4, 5] High-sensitivity C-reactive protein (hs-CRP) is a sensitive marker that shows an inflammatory reaction and tissue damage,[6] which increases during pregnancy.[6]

Human placenta produces hs-CRP and releases it predominantly into the maternal blood.[7] Hs-CRP is a systemic inflammatory marker and it has been shown that it can be found in amniotic fluid and fetal urine; its increased amount directly relates to poor pregnancy outcome.[7] However, the real source of amniotic fluid hs-CRP, its regulation and its function during pregnancy, is not clear.

Previous studies[8-11] have reported that maternal hs-CRP level increases during pre-eclampsia. At the same time it has been shown that not only is its level higher in pre-eclamptic women, but higher levels have also been related to the clinical parameters of pre-eclampsia.[10] A previous study has suggested that its plasma level may be used as a marker for evaluation of pre-eclampsia severity.[10, 11]

Regarding the fact that this test is fast and inexpensive, and screening methods of prediction of pre-eclampsia in the first trimester have shown little success,[2] it seems necessary to perform more studies on serum hs-CRP in order to determine its role in predicting pre-eclampsia, and to also find a suitable cut-off point for serum hs-CRP if applicable.

The purpose of the present study was to evaluate the diagnostic value of maternal serum hs-CRP measurement during the first trimester of pregnancy in order to predict pre-eclampsia and determine the optimal cut-off point.

Methods

A prospective cohort study was performed on pregnant women who had been referred to the prenatal clinic of Akbarabadi Teaching Hospital in Tehran, Iran, between March 2009 and March 2010. This hospital is a very busy referral hospital where the most complicated pregnancies are referred. It is located in downtown Tehran, with a very poor population, and the rate of pre-eclampsia is high.

Written informed consent was obtained from all participants who were fully informed about the study. Institutional review board approval and institutional ethics committee approval were also given to the study.

Inclusion criteria were gestational age of 8–13 weeks of pregnancy (according to a reliable last menstrual period and ultrasound confirmation) and singleton pregnancy.

Exclusion criteria were any maternal systemic disorders or drug use (except usual supplementation, including folic acid), chronic hypertension, diabetes mellitus, collagen vascular diseases, renal disorders, any recent or present fever or infectious disease, malignancies or autoimmune diseases, multiple pregnancies, smoking and any vaginal bleeding.

Regarding the prevalence of 5–10% for pre-eclampsia, a sample size of 390 was considered enough, with a confidence of 95%, α = 0.05 and P = 0.03.

A blood sample was obtained from all participants for hs-CRP measurement (latex agglutination test in which latex has been covered with a layer of anti-human CRP),[12] and blood pressure was measured (Korotkoff phase V was used to define diastolic pressure).[1] All women were then monitored to delivery. The first-trimester serum hs-CRP level of pre-eclamptic women was compared with normotensive women.

Mild pre-eclampsia was defined as blood pressure of at least 140/90 mmHg on two occasions with an interval of 6 h, plus 24-h urine protein of more than 300 mg or random urine protein of 3+ on two occasions, 6 h apart.1 Severe pre-eclampsia was defined as diastolic blood pressure ≥110 mm hg, systolic blood pressure ≥160 mm hg, proteinuria 2.0 g/24 h or ≥3+ dipstick, persistent headache, visual disturbances, persistent epigastric or upper abdominal pain, oliguria (urine output <30 mL/h), convulsion (eclampsia), elevated serum creatinine (more than 1.2 mg/dL, thrombocytopenia [platelet count <100 000/μL]), serum transaminase elevation (serum AST > 70 U/L), fetal-growth restriction, and pulmonary edema.[1]

Statistical analysis was performed using spss 16. A P-value of less than 0.05 was considered as statistically significant. Receiver–operator curve (ROC) was used to show the optimal cut-off points, sensitivity, specificity, positive predictive value and negative predictive value. Area under the ROC of 1 has the maximum diagnostic accuracy with a sensitivity of 100% and specificity of 100%. The least diagnostic accuracy is when the area under the ROC is 0.5.

Results

A total of 394 women completed the study. In 42/394 (10.7%), pre-eclampsia was diagnosed and 23 of these 42 women (56.1%) suffered from severe pre-eclampsia. Mean age of all women was 25.7 ± 5.4 years (range: 15–44).

The women of the two groups (both with and without pre-eclampsia) did not show any significant difference according to gravidity, leukocyte count and hemoglobin concentration (Table 1); however age (P = 0.018), body mass index (BMI) (P = 0.001), and systolic and diastolic blood pressure (P = 0.001) were higher in the pre-eclamptic group. Gestational age (P = 0.001), neonatal weight (P = 0.001) and platelet count were significantly lower in the pre-eclamptic group (Table 1). Serum CRP of the first trimester was significantly higher in the pre-eclamptic women (Table 1).

Table 1. Demographic and clinical characteristics of the pregnant women in the two study groups
CharacteristicsPre-eclamptic group n = 42Normotensive group n = 352P-value
  1. †Mean ± standard deviation. ‡Number(%).
Age(years)27.5 ± 5.525.4 ± 5.30.018
Body mass index(kg/m2)31.1 ± 4.328.2 ± 4.50.001
Gravidity121(50%)178(51%)0.532
210(23.8%)106(29.5%)
≥311(26.2%)68(19.5%)
Gestational age(weeks)34.8 ± 3.838.1 ± 3.10.001
Neonatal weight(gram)2242.1 ± 879.823084.1 ± 627.30.001
Systolic blood pressure(mmHg)155.7 ± 19.2117.8 ± 54.40.001
Diastolic blood pressure(mmHg)100.1 ± 11.574.5 ± 44.10.001
White blood cells(103 cell/ml)13.3 ± 6.212.1 ± 3.20.249
Hemoglobin(mg/dl)12.9 ± 1.712.4 ± 1.40.171
Platelet(103 cell/ml)185.2 ± 52.3203.3 ± 55.50.049
High-sensitivity C-reactive protein(mg/L)7.06 ± 2.63.6 ± 2.30.001

Serum hs-CRP of more than 7 mg/L was found in 26 (61.9%) cases of pre-eclampsia versus 22 (6.25%) normotensive pregnancies, which showed a significant difference (P = 0.001, relative risk [RR] = 12.1, 95% confidence interval [CI]: 6.91–21.15). After adjusting for BMI, RR was calculated to be 10.7 (95% CI: 7.96–14.22).

Serum hs-CRP of more than 7 mg/L was found in 17 (73.91%) cases of severe pre-eclampsia versus 22 (6.25%) normotensive pregnancies, which showed a significant difference (P = 0.001, RR = 9.35, 95% CI: 4.48–19.52). After adjusting for BMI, RR was calculated to be 10.66 (95% CI: 7.32–15.22).

According to Figure 1, the area under the ROC is 0.855, which shows an appropriate diagnostic accuracy for the prediction of pre-eclampsia.

Figure 1.

Receiver–operator curve for determining the diagnostic value of first-trimester serum high-sensitivity C-reactive protein for prediction of pre-eclampsia.

Tables 2 and 3 show the sensitivity, specificity, positive and negative predictive value, and diagnostic accuracy of different concentrations of serum hs-CRP for prediction of pre-eclampsia and severe pre-eclampsia.

Table 2. Sensitivity, specificity, PPV and NPV and diagnostic accuracy of different cut-off points of serum high-sensitivity C-reactive protein during first trimester of pregnancy for prediction of pre-eclampsia
Cut-off pointSensitivity (%)Specificity (%)PPV (%)NPV (%)Diagnostic accuracy (%)
  1. NPV, negative predictive value; PPV, positive predictive value.
1.2597.50.610.566.610.9
390.242.615.897.347.7
3.585.357.219.297.160.05
478.172.12596.572.8
578.183.936.796.683.9
675.688.944.396.887.4
763.493.554.195.590.34
Table 3. Sensitivity, specificity, PPV and NPV and diagnostic accuracy of different cut-off points of serum high-sensitivity C-reactive protein during first trimester of pregnancy for prediction of severe pre-eclampsia
Cut-off pointSensitivity (%)Specificity (%)PPV (%)NPV (%)Diagnostic accuracy (%)
  1. NPV, negative predictive value; PPV, positive predictive value.
1.251000.910.710011.4
392.343.516.397.948.7
3.588.457.319.897.660.6
476.971.122.496.571.5
576.977.528.996.577.4
669.282.532.195.781.1
765.388.941.495.786.4

Figure 1 shows the ROC of diagnostic accuracy of hs-CRP in the first trimester of pregnancy for prediction of pre-eclampsia. The area under the ROC was 0.855, which shows an appropriate diagnostic accuracy of hs-CRP.

Figures 2 and 3 show the ROC of diagnostic accuracy of hs-CRP for the prediction of mild and severe pre-eclampsia, respectively. In both, the area under the ROC is 0.870, which shows an appropriate diagnostic accuracy of hs-CRP of the first trimester of pregnancy for prediction of both mild and severe pre-eclampsia.

Figure 2.

Receiver–operator curve for determining of diagnostic value of first-trimester serum high-sensitivity C-reactive protein for prediction of mild pre-eclampsia. Diagonal segments are produced by ties.

Figure 3.

Receiver–operator curve for determining of diagnostic value of first-trimester serum high-sensitivity C-reactive protein for prediction of severe pre-eclampsia.

Discussion

In the present study, first-trimester maternal serum hs-CRP level was significantly higher in pre-eclamptic women than normotensives, which shows an increased inflammatory response in pre-eclampsia compared to normal pregnancy and is in agreement with other studies.[8, 10, 11, 13] However, even at the highest CRP cut-off, only 41.4% of women with high CRP developed pre-eclampsia in this cohort; thus, the majority of women with high CRP did not develop pre-eclampsia. Therefore, the predictive ability of CRP is quite limited, and if it is to be clinically useful, it will probably need to be combined with additional markers.

Due to the importance of pre-eclampsia during pregnancy, many studies have been performed on the different methods of predicting its occurrence.[2] In a systematic review,[14] a diagnostic value of 27 different tests (including alpha-fetoprotein, uterine artery Doppler, body mass index of more than 34, etc.) have been evaluated. Some of these methods have high sensitivity but low specificity or they are very expensive. This study has concluded that these tests do not have enough accuracy. Therefore, it seems that studies should try to identify the tests that are accurate and at the same time inexpensive and applicable.[15] Then the optimal cut-off points have to be determined in order to perform them in practice.

Inflammatory factors, including hs-CRP, increase during pregnancy,[6] which suggests an increased inflammatory response during pregnancy. In different studies, these inflammatory factors were higher in pre-eclampsia rather than normal pregnancy.[8-11, 13] If the studies can show an optimal cut-off point for these inflammatory factors, like hs-CRP, they can probably be used as screening tests for the prediction of pre-eclampsia, and may benefit women a great deal.

In a study by Adali et al. on pre-eclamptic women,[16] the relation between maternal serum hs-CRP level and uterine artery Doppler velocimetry was evaluated. This study showed that a serum level of hs-CRP is higher in pre-eclamptic than normotensive women. Hs-CRP also related positively with the mean arterial pressure and its level was higher in pre-eclamptic women who had abnormal uterine artery Doppler velocimetry than pre-eclamptic women with normal Doppler. The study concluded that the hs-CRP level of maternal serum has a correlation with severity of pre-eclampsia, which shows an endothelial dysfunction, and may be considered as a probable marker of pathologic uteroplacental perfusion.

Serum hs-CRP level has shown a correlation with the severity of pre-eclampsia,[10, 16-19] intrauterine growth restriction (IUGR),[16] birthweight[11, 20] and pregnancy duration[13] in different studies, but did not show a significant correlation with recurrence of pre-eclampsia.[21] In the present study, serum hs-CRP level showed a correlation with the severity of pre-eclampsia too. Unfortunately, there are few prospective cohort studies[22, 23] evaluating the diagnostic value of serum hs-CRP for the prediction of pre-eclampsia.

In a study by Tjoa et al.,[22] the CRP level of women was measured at 23–25 weeks of pregnancy and compared in two groups of women with and without pre-eclampsia who had normal or impaired Doppler velocimetry and babies with and without IUGR. They reported no difference between the pre-eclamptic, IUGR group and normal-outcome pregnancies. Savvidou et al.[23] reported elevated CRP level at 10–14 weeks of gestation in women who developed pre-eclampsia or delivered IUGR babies. Most of these studies are case–control studies.[10-19, 22, 23]

In a study by Ertas et al.,[18] an optimal cut-off point of 9.66 mg/L has been suggested for the prediction of severe pre-eclampsia and the study by Qiu et al.[24] concluded that hs-CRP level of ≥4.9 mg/L may increase the risk of pre-eclampsia in thin women up to 2.5-fold. This study was a nested case–control study. The weak point of the present study is the fact that the patients were not matched for BMI.

A study by Cebesoy et al.[19] reported an optimal cut-off point of 5 mg/L for serum hs-CRP. Another study[25] has compared the serum hs-CRP level and pregnancy-associated plasma protein A (PAPP-A) in normotensive and pre-eclamptic women in the third trimester of pregnancy and showed that both items were higher in pre-eclampsia, but could not show a correlation between these two markers in pre-eclamptic and non-pre-eclamptic women. However, in this study, serum hs-CRP level showed a positive correlation with the mean arterial pressure.

In conclusion, regarding the above-mentioned studies, it seems reasonable to perform further cohort studies for determining the predictive value of serum hs-CRP for pre-eclampsia and to try to find an optimum cut-off point.

Acknowledgment

This work has been supported by Tehran University of Medical Sciences, Deputy of Research and Technology, Project no. 14144.

Disclosure

None of the authors has anything to disclose.

Ancillary