Objective
Decreased levels of serum placenta growth factor (PlGF) are associated with preeclampsia. We sought to determine whether serum and placental levels of PlGF (sPlGF and pPlGF) are associated with preeclampsia and whether there is a correlation between serum and placental PlGF levels.
Study Design
These analyses were part of a larger, prospective, case-control study. Cases were women with preeclampsia. Controls were women without preeclampsia who delivered at term. Analyses included nonparametric tests to compare medians, logistic regression to estimate odds, and calculation of correlation coefficients.
Results
Twenty-four cases (10 preterm, 14 term) were compared with 14 controls. Median levels of PlGF were significantly lower in cases than controls (pPlGF: 232.6 vs 363.4 pg/mL, P = .02; sPlGF: 85.5 vs 274.4 pg/mL, P < .001). Serum and placental PlGF were correlated (overall: 39%, P = .006; cases with preterm preeclampsia and growth restriction: 87%, P = .02).
Conclusion
Serum and placental PlGF are independently associated with preeclampsia and correlated with each other.
Preeclampsia is a hypertensive disease that complicates 3-8% of all pregnancies. Preeclampsia accounts for approximately 20% of maternal deaths in the United States and is a leading cause of medically indicated preterm birth. Despite decades of research, the etiology of preeclampsia remains elusive. However, alterations in angiogenesis leading to abnormal placentation and subsequent systemic endothelial dysfunction are thought to contribute to the disease process.
Placenta growth factor (PlGF) is a proangiogenic protein and member of the vascular endothelial growth factor (VEGF) family produced by villous syncytiotrophoblasts in the placenta. Previous studies have demonstrated that levels of circulating PlGF in the serum of patients with preeclampsia are significantly decreased compared with normotensive controls. This association has been demonstrated throughout gestation and even as early as the first trimester, a time when angiogenesis is critical for placental invasion. Furthermore, PlGF has been identified as a marker of disease severity, with lower levels being associated with increased blood pressure, early-onset vs late-onset preeclampsia, severe vs mild preeclampsia, small-for-gestational-age infants, and laboratory evidence of the HELLP (hemolysis, elevated liver enzymes, and low platelet count) syndrome. Based on the collective findings of these and other studies, serum PlGF has been posited as a promising biomarker for preeclampsia.
Biologically, if preeclampsia occurs because of alterations of key angiogenic factors at the level of the placenta, such as PlGF, then changes in the levels of these angiogenic factors should plausibly be apparent at the level of the placenta. Furthermore, because preeclampsia is a systemic disease, it is plausible that changes in the level of this angiogenic factor would also be present in the serum. This notion has been previously explored in human placentas and maternal serum with the VEGF receptor known as soluble fms-like tyrosine kinase-1. No studies to date have explored the relationship between serum and placental PlGF levels and preeclampsia. Correlation of serum and placental PlGF levels would provide additional biologic rationale for using PlGF as a biomarker of preeclampsia.
The objectives of our study were to determine whether both serum and placental levels of PlGF are independently associated with preeclampsia and to assess whether there is a correlation between PlGF levels in the placenta and serum.
Materials and Methods
A prospective, case-control study, Preeclampsia: Mechanisms and Consequences II, was conducted at a single urban tertiary care center between January and October 2009. The institutional review board at the University of Pennsylvania (Philadelphia, PA) approved this study.
Women were classified as cases and controls based on investigator developed a priori definitions of hypertensive disorders of pregnancy and not based on individual physician diagnosis. Specifically, cases were identified based on prespecified maternal criteria for preeclampsia consistent with American Congress of Obstetricians and Gynecologists guidelines and included elevated blood pressure (≥140/90 mm Hg on 2 measurements ≥6 hours apart) with 1+ or greater proteinuria, the majority of which were by sterile catheter. Patients with either gestational hypertension (GHTN), defined as elevated blood pressure (≥140/90 mm Hg on 2 measurements ≥6 hours apart) without proteinuria or laboratory abnormalities or HELLP syndrome, defined as preeclampsia in the setting of otherwise unexplained laboratory abnormalities (including elevated liver enzymes aspartate aminotransferase >45 U/L, aminotransferase >60 U/L, and/or decreased platelets <100,000/μL) were also included as cases. Controls were prospectively enrolled from women presenting for delivery at term (≥37 weeks) as either a scheduled induction of labor, a scheduled cesarean section, spontaneous rupture of membranes, or term labor.
Exclusion criteria included fetal anomalies and aneuploidy. Additionally, patients who did not have blood drawn prior to delivery and patients who did not have both serum and placental PlGF specimens were also excluded from this study. Preterm preeclampsia was defined as delivery occurring at less than 37 weeks’ gestation. A diagnosis of small for gestational age (SGA) was made based on birthweight less than 10% for gestational age according to the Alexander growth curve.
Consecutive eligible patients were enrolled in the study by trained clinical research coordinators who obtained informed consent at the time of enrollment. Once a patient was enrolled in the study, all management decisions were made by the treating physician according to the standard of care at our institution. Management typically includes delivery for any woman diagnosed with a hypertensive disorder of pregnancy at term as well as for severe preeclampsia and HELLP at 34 weeks or longer. Patients less than 34 weeks with severe preeclampsia and HELLP were managed according to recommended guidelines. At the time of enrollment, patients were interviewed to obtain self-reported demographic information (including age, height, and race) and pertinent clinical information (including history of prior preeclampsia, hypertension, and/or family history of hypertension). All other medical and obstetric history, such as renal disease and diabetes mellitus, as well as delivery information, was obtained through chart abstraction.
Within 24 hours of enrollment into the study, prior to delivery, peripheral blood samples were collected from all pregnant patients. After sitting at room temperature for 30-60 minutes, samples were centrifuged using a SERO-FUGE 2001 (Becton Dickinson, Franklin Lakes, NJ) at 3100-3500 rpm for 20 minutes. Serum was extracted from whole blood and placed into cryovial tubes that were immediately placed in liquid nitrogen and stored at –80°C until biomarker analyses were performed.
Immediately after delivery (mean time: 20.35 ± 8.69 minutes), 4 random placental biopsies were taken: 1 from each quadrant of the placenta. Once the 4 placental biopsies were collected, the chorionic plate, including the amnion and chorion, was removed. Additionally, any remaining decidual tissue found on the maternal side of the placenta was removed. The remaining placental tissue, consisting mostly of trophoblasts and fetal vessels, was washed 4 times in phosphate-buffered saline to remove all remaining blood. These samples were then snap frozen in liquid nitrogen and stored at −80°C.
To extract protein from the placental tissues, 50 mg of tissue were homogenized on ice in 2 mL T-Per tissue extraction buffer (Thermo Scientific, Rockford, IL) supplemented with protease inhibitors (complete protease inhibitor cocktail tables; Roche Diagnostics, Indianapolis, IN). Homogenates were centrifuged at 10,000 × g for 10 minutes and supernatants were stored at −80°C until biomarker analyses were performed. Levels of both serum and placental PlGF were determined using ligand-specific commercially available enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN). The PlGF ELISA assay sensitivity is 7.0 pg/mL. The intraassay coefficient of variation (CV) for PlGF is 3.6% and the interassay CV is 11.0%. The PlGF assay recognizes both natural and recombinant human PlGF.
Statistical analyses were performed using STATA version 10.1 (Stata Corp, College Station, TX). χ 2 tests were used to determine associations between categorical demographic variables and case status. Rank-sum tests were performed to compare sample distributions of nonparametric data. Logistic regression was used to estimate odds of case status as a function of PlGF levels. Receiver-operator analyses were performed to determine the area under the curve (AUC) for each logistic model. Finally, Kendall rank correlation coefficients were calculated to measure the correlation between serum PlGF (sPlGF) and placental PlGF (pPlGF).
Results
Sixty-two cases and 32 controls were enrolled. We analyzed data from 24 cases (10 preterm, 14 term) and 14 controls who met inclusion criteria and had both serum obtained prior to delivery and placental tissue available for analysis. Demographic characteristics comparing cases and controls are included in Table 1 . All patients included in these analyses received prenatal care, had insurance, and were pregnant with singletons. There was no statistically significant difference between time from serum collection to delivery between cases and controls (10.85 vs 7.54 hours, P = .28). Cases were more likely to have a prior pregnancy affected with preeclampsia ( P = .09) and a family history of hypertension ( P = .04). Gestational age at delivery was significantly lower among cases compared with controls ( P = .002).
| Demographic variable | Cases (n = 24) a | Controls (n = 14) a | P value b |
|---|---|---|---|
| Gestational age at delivery, wks | 37.50 [34.50–39.00] | 39.50 [39.00–40.00] | .002 |
| Maternal age, y | 24.46 [19.77–34.65] | 29.09 [20.09–33.63] | .86 |
| Primiparous | 14 (58.33) | 6 (42.86) | .36 |
| African American race | 21 (87.50) | 11 (71.57) | .47 |
| Initial BMI, kg/m 2 | 28.35 [23.92–33.08] | 30.05 [21.13–36.61] | .80 |
| History of PEC c | 4 (40.00) | 0 (0.00) | .09 |
| Diabetes (any class) | 3 (12.50) | 1 (7.14) | .60 |
| CHTN | 4 (16.67) | 0 (0.00) | .11 |
| Renal disease | 1 (4.17) | 0 (0.00) | .44 |
| Family history of HTN | 22 (91.67) | 9 (64.29) | .04 |
| Tobacco use | 4 (16.67) | 3 (21.43) | .72 |
| Birthweight, g | 2807.50 [2172.50–3370.00] | 3087.50 [2950.00–3765.00] | .06 |
| SGA <10% | 10 (41.67) | 2 (14.29) | .08 |
a Categorical data expressed as n (%). Continuous data expressed as median [interquartile range];
b P value determined by χ 2 for categorical variables and nonparametric tests of medians for continuous variables as appropriate for overall case to control comparison;
Overall, median levels of sPlGF were significantly lower in cases compared with controls (85.45 vs 274.41 pg/mL, P < .001) ( Figure 1 ). For every 50 U increase in sPlGF, there was a 34.6% reduction in odds of preeclampsia (95% confidence interval [CI], 0.47–0.91). Serum PlGF demonstrated good discriminative ability between cases and controls at the time of disease manifestation (AUC = 0.83). Similarly, median levels of pPlGF were significantly lower in cases compared with controls (232.6 vs 363.4 pg/mL, P = .02) ( Figure 2 ). For every 50 U increase in pPlGF, there was a 23.2% reduction in odds of preeclampsia (95% CI, 0.61–0.97). Placental PlGF also demonstrated a good ability to discriminate case status at the time of disease manifestation (AUC = 0.72). The association between both sPlGF (odds ratio [OR], 0.68; 95% CI, 0.48–0.94) and pPlGF (OR, 0.78; 95% CI, 0.61–0.99) and preeclampsia remained consistent after controlling for SGA.
To account for the impact of the gestational age difference between cases and controls, we next restricted our analyses to term patients only (14 cases and 14 controls). The difference in sPlGF levels between cases and controls remained significant (97.91 vs 274.41 pg/mL, P < .001) such that for every 50 U increase in sPlGF, there was a 62.7% reduction in the odds of preeclampsia (95% CI, 0.17–0.83). Similarly, pPlGF levels remained significantly lower in cases compared with controls (245.96 vs 363.35 pg/mL, P = .04) such that for every 50 U increase in pPlGF, there was a 27.2% reduction in the odds of preeclampsia (95% CI, 0.53–0.99).
We then evaluated the association between sPlGF and pPlGF and disease severity among cases only. There was a nonsignificant decrease in median sPlGF levels associated with both progressively worsening disease (GHTN: 129.63 pg/mL, preeclampsia: 97.91 pg/mL, HELLP: 57.54 pg/mL; P = .62) and term vs preterm gestational age at diagnosis (97.91 vs 80.38 pg/mL, P = .77). Similarly, there was a nonsignificant decrease in median pPlGF levels associated with disease progression from GHTN to preeclampsia (291.67 vs 211.15 pg/mL, P = .60) and term vs preterm gestational age at diagnosis (245.96 vs 232.57 pg/mL, P = .86). In contrast, median pPlGF levels were similar between patients with preeclampsia and HELLP (preeclampsia: 211.15 pg/mL, HELLP: 211.60 pg/mL; P = .63).
Finally, the correlation coefficients between sPlGF and pPlGF were calculated. Among all patients, there was a 39% correlation ( P = .006) between cases and controls ( Figure 3 ). The strength of this correlation increased with progressing severity of the preeclampsia phenotype ( Table 2 ). The subgroup with preterm preeclampsia and growth restriction demonstrated the highest correlation between sPlGF and pPlGF (correlation coefficient 0.87, P = .02).
