Predictive value of soluble fms‐like tyrosine kinase‐1 against placental growth factor for preeclampsia in a Chinese pregnant women population

Objective The purpose of the present study was to explore the predictive effects of soluble fms‐like tyrosine kinase‐1 (sFlt‐1) and placental growth factor (PlGF) for preeclampsia. Methods A total of 1580 singleton pregnant women aged 18‐45 years were included in this study. Serum samples were collected and stored frozen during their regular obstetric examinations. A total of 48 women who were eventually diagnosed with preeclampsia among them were defined as the preeclampsia group, other 134 women who were matched with age and sample collecting gestational weeks and finally diagnosed without preeclampsia were selected as control. The concentration of sFlt‐1 and PlGF in prestored serum samples was examined. The optimal cut‐off of sFlt‐1, PlGF, and sFlt‐1/PlGF ratio in predicting preeclampsia was determined by establishing the receiver operating characteristic curve (ROC). Results Serum PlGF levels in patients with preeclampsia were significantly lower than those in normal pregnancy (P < 0.05), On the contrary, sflt‐1 levels and sflt‐1/PlGF ratios were significantly higher than those in the normal pregnant women (P < 0.05). The ROC curve study showed that using the sFlt‐1/PlGF ratio to predict preeclampsia was better than using PlGF alone but no difference with sFlt‐1. When the cut‐off of the sFlt‐1/PlGF ratio was 26.6, the area under the ROC curve was 0.918, and high sensitivity (85.42%) and specificity (96.27%) for predicting preeclampsia were obtained. Conclusion The cut‐off of sflt‐1/PlGF ratio determined by ROC curve has a good predictive value for the occurrence of preeclampsia.

cardiovascular disease, and even death in pregnant women, but also can cause premature birth, slow growth, and low birthweight for fetus. Approximately, 60 000 pregnant women die from preeclampsia in the world every year, with an incidence of about 2% to 10% of all pregnancies, while the incidence in China is 9.4%. [3][4][5] The clinical manifestations of preeclampsia are diverse, and effective diagnostic testing methods for preeclampsia are absent, which result in no reliable early warning indicators and make preeclampsia becoming a serious threat to the health of pregnant and infant. The main purpose of this study is discussing the early predictive role of sFlt-1/ PlGF ratio on preeclampsia in Western China and establishing a predictive cut-off for the population in this area by using the ROC curve.

| Participants
The complete data and peripheral blood samples of 1580 singleton pregnant women aged 18-45 years who had routine prenatal examination and did not occur preeclampsia yet between 12

| Samples collection and examination
All peripheral blood samples were collected during regular obstetric examinations of pregnant women before diagnosis of preeclampsia.
Serum was isolated by centrifuge (2700 g for 5 minutes) from the peripheral blood samples and stored at a temperature of −80°C.
After determining the experimental group, the sflt-1 and PlGF (from Roche, Germany) in the retained serum of the selected participants were detected by electrochemiluminescence.

| Statistical analysis
All data were analyzed with SPSS21.0. Measurement data complying with the normal distribution were presented by means plus or minus standard deviation (X ± s). The differences between the two groups were analyzed by t test or Mann-Whitney U test, and the chi-square test was used to compare the rates between groups. The optimal cut-off value, specificity and sensitivity of sFlt-1, PlGF, and sFlt-1/PlGF ratios for predicting preeclampsia were determined by

| Clinical characteristics
There were 48 cases in preeclampsia group, with an average age of 32.5 ± 4.8 years and an average gestational age of 29.1 ± 3.1 weeks.
The control group consisted of 134 patients, with an average age of 31.9 ± 4.3 years and an average gestational age of 28.5 ± 5.3 weeks.
There was no difference in age, gestational age, BMI, number of pregnancy, or multipara percentage between the preeclampsia patient group and the control group (shown in Table 1).

| Comparison of observation indicators
Because the measurement data of the observed indicators are nonnormal distribution, nonparametric Mann-Whitney U test was used for analysis. As shown in Table 2, comparing with the control group, the factors of sFlt-1, P1GF, and sFlt-1/PlGF were significantly different with PE group (shown in Table 2).

| D ISCUSS I ON
PlGF levels with preeclampsia. Hertig et al found that the increasing serum sFlt-1 at an average of 6.5 weeks earlier than the clinical symptoms by continuous detection. 6 The sensitivity and specificity of predicting preeclampsia were 80% and 100%, respectively, when the threshold was 0.975 g/L. Lee et al found that the serum levels of sFlt-1 was 934.5 ± 235.5 ng/L in preeclampsia patients and was 298.0 ± 168.2 ng/L in normal pregnancy control group. 7 We assessed the serum levels of sFlt-1 in pregnant women with mediumterm and late period of preeclampsia. The results showed that the serum levels of sFlt-1 in preeclampsia group were significantly higher than that in control group (P < 0.05), and it indicates that the soluble fms-like tyrosine kinase-1 was a risk factor for preeclampsia, which is consistent with the results reported in the literature above.

Mijal et al found that the serum level of PlGF in preeclampsia
pregnant women was significantly decreased by comparing with normal pregnant women. 8 Cetin I et al found that PlGF level could be used to predict the risk associated with adverse pregnancy such as very low birthweight infants and emergency childbirth. 9 However, our results showed that the level of serum PlGF in preeclampsia group was significantly lower than that in normal pregnancy control group (P < 0.05) and indicated that placental growth factor was a protective factor for preeclampsia. Many researches advocated that low level of PlGF in pregnant women could auxiliary diagnosis preeclampsia by specific binding sFlt-1.
Stepan et al and Verlohren et al found that predicted preeclampsia more accurately by combining sFlt-1 and PlGF. 10,11 And the predictive specificity of early-onset preeclampsia is 8.3%, and sensitivity is 95%. Maynard

TA B L E 2 Comparison of observation indicators of the two groups
could predict the occurrence of preeclampsia; sensitivity was 94% and specificity was 77%. 12 Our study combined sFlt-1 and PIGF and found that the ratio of sFlt-1/PLGF in preeclampsia group was significantly difference from that in normal group (P < 0.05), which increased several dozen times. The data showed that the sFlt-1/PLGF ratio was more sensitive than the single index. It was recommended that patients should seek medical help as soon as possible when sFlt-1 increased, and/or PlGF decreased, especially the ratio of sFlt-1/PlGF was increased gradually after repeated examinations.
Another key objective of this study is using the receiver operating characteristics curve (ROC) to determine the best cut-off value of sFlt-1/PlGF ratio for predicting preeclampsia in a regional population. ROC curve was a method to evaluate the

CO N FLI C T S O F I NTE R E S T
The authors declare that they have no competing interests.

AUTH O R S' CO NTR I B UTI O N S
Fan Yu was responsible for the statistical analysis and prepared the manuscript. Yongmei Jiang was responsible for the study design and coordination, guided the statistical analysis, and revised the manuscript. Shihong Zhang was responsible for the study design and coordination and reviewed the manuscript critically. Qianjin Bai collected the data and reviewed the manuscript. All authors read and approved the final manuscript. The excel data used to support the findings of this study are available from the corresponding author upon request.