Objective
The purpose of this study was to compare longitudinally sampled maternal angiogenic proteins between singleton and twin pregnancies.
Study Design
Placental growth factor (PlGF), soluble feline McDonough sarcoma (fms)-like tyrosine kinase (sFlt)-1, and soluble endoglin from healthy pregnant women were quantified at 10, 18, 26, and 35 weeks’ gestation (n = 91), and during the third trimester (31-39 weeks) and at delivery (33-41 weeks; n = 41). Geometric means and 95% confidence intervals were calculated for gestational age-adjusted angiogenic protein concentrations and compared between matched twin and singleton pregnancies.
Results
Maternal sFlt-1 concentrations and the sFlt-1/PlGF ratio were higher in twins than singletons across pregnancy and at delivery, with the greatest differences at week 35 (sFlt-1: 36,916 vs 10,151 pg/mL; P < .0001; sFlt-1/PlGF: 168.4 vs 29.0; P < .0001). Maternal concentrations of soluble endoglin also were higher in the third trimester and delivery. Maternal PlGF concentrations were lower in twin than singleton pregnancies at week 35 only (219.2 vs 350.2 pg/mL; P < .0001). Placental weight appeared to be inversely correlated with maternal sFlt-1/PlGF ratio at the end of pregnancy in both twins and singletons.
Conclusion
Higher maternal antiangiogenic proteins in twin than singleton pregnancies does not appear to be due to greater placental mass in the former, and may be one explanation for the increased risk of preeclampsia in women carrying multiple gestations. Determining whether women with a history of multiple gestations have an altered cardiovascular disease and breast cancer risk, like those with a history of preeclampsia, is warranted.
Women carrying twins or other higher-order multiples are at 2-3 times the risk of developing preeclampsia, a common cause of maternal and fetal morbidity, than women with singleton pregnancies. Preeclampsia is marked by shallow trophoblast invasion into the maternal endometrium resulting in a less extensive vascular network supporting the pregnancy. Alterations in angiogenic proteins, as well as inflammatory cytokines and other immune-modulating molecules, have been demonstrated in preeclamptic pregnancies, with elevations in soluble feline McDonough sarcoma (fms)-like tyrosine kinase (sFlt)-1 and soluble endoglin (s-endoglin), 2 antiangiogenic proteins, typically preceding the clinical manifestation of maternal disease.
Maternal angiogenic factors also appear altered in pregnancies involving multiples compared with singletons, with elevated concentrations of sFlt-1 in the former. The timing of these changes in angiogenic balance may provide insight into the mechanism whereby preeclampsia risk is elevated in women carrying multiple gestations. Therefore, we followed women longitudinally through pregnancy and delivery to examine circulating maternal concentrations of placental growth factor (PlGF), sFlt-1, and s-endoglin in twin and singleton pregnancies.
Materials and Methods
Study subjects
The data for the analysis derive from 2 sources: the BIRTH cohort and a study of twins at the Geisel School of Medicine (Dartmouth College). The study protocols were approved by each institutional review board and at the US National Cancer Institute, and written informed consent was obtained from all participating women.
BIRTH cohort
Participants were enrolled at 3 US tertiary care academic centers from October 2007 through June 2009. Eligible women initiated routine prenatal care at <15 weeks’ gestation, were >18 years of age, and planned to deliver at the enrolling institution. Women who developed preeclampsia in the index pregnancy or a prior one (gestational hypertension defined as a blood pressure elevation of >140/90 mmhg on 2 occasions with concomitant proteinuria defined as positive urine protein test result >300 mg/24 hours or protein/creatinine >0.20 mg/mg) were excluded from the present analysis. A total of 2230 singleton and 93 twin gestations were enrolled, and 2193 and 91 singletons and twins, respectively, met the inclusion criteria for analysis.
Geisel School of Medicine twin study
Eligible were pregnant women ≥18 years of age who intended to deliver at the facility. Women carrying twin gestations and presenting for prenatal care or hospitalized for antenatal surveillance from 2003 through 2007 were approached in the third trimester of pregnancy. The next singleton pregnancy that met the eligibility criteria and could be matched to the twin pregnancy on gestational age (within 1 week), parity (nulliparous/parous), and maternal age (±5 years), in that order, was recruited for the study (prenatal controls; n = 40). Another group of women with singleton pregnancies were recruited at admission for labor and delivery (labor controls) and matched to twin mothers according to the criteria above. Five twin pregnancies and 3 singleton pregnancies were excluded because they developed preeclampsia after enrollment, leaving 41 twins and a total of 62 singleton controls (40 with blood samples in the third trimester and 52 with blood samples at labor and delivery). Placentas were routinely examined by the pathology department.
Biospecimen collection and processing
BIRTH cohort
Maternal blood samples were obtained at the following median (interquartile range) weeks of gestation: 9.7 (8.4–11.6), 17.8 (16.8–18.7), 25.9 (24.8–28.1), and 35.1 (34.6–35.9). Approximately 10 mL of blood was drawn in EDTA plasma tubes; the samples were kept at 4°C until processing for storage within 4 hours of venipuncture. The specimens were centrifuged for 20 minutes and stored at −80°C. Samples were shipped in batches on dry ice to Abbott Diagnostics (Abbott Park, IL) where they were stored at −80°C.
Geisel School of Medicine twin study
Blood samples were collected in the third trimester (31-39 weeks) and at the earliest possible time after admission for labor and before any administration of medication (33-41 weeks). A 10-mL red-top tube of whole blood was allowed to clot at room temperature, was centrifuged, and the sera were stored at −70°C. Samples were shipped on dry ice to a biorepository in Rockville, MD, where they were stored at −80°C.
Laboratory assays
BIRTH cohort
PlGF and sFlt-1 were measured with prototype Abbot Diagnostics immunoassays (Abbott Diagnostics, Abbott Park, IL). The PlGF immunoassay measures the free form of PlGF-1, with a lower limit of detection of 1 pg/mL, and a range up to 1500 pg/mL. The sFlt-1 immunoassay measures both free and bound sFlt-1, with a lower limit of detection of 0.10 ng/mL and a range up to 150 ng/mL. The combined intraassay and interassay coefficients of variation reported by the laboratory were <7% for PlGF and sFlt-1.
Geisel School of Medicine twin study
Serum levels of sFlt-1 and PlGF were determined in a blinded fashion using commercially available enzyme-linked immunosorbent assay kits (R&D Systems, Minneapolis, MN) as described elsewhere. Interassay CVs for the sFlt-1 kit ranged from 7.0–8.1% and for PlGF ranged from 10.9–11.8%. S-endoglin was also measured using commercially available enzyme-linked immunosorbent assay kits (R&D Systems) as described elsewhere. Interassay coefficients of variation for the s-endoglin kit ranged from 6.3–6.7%.
In a subset of BIRTH cohort mothers with singleton pregnancies, PlGF and sFlt-1 concentrations were measured using both the Abbot Diagnostics immunoassay (used for study samples in the BIRTH cohort) and the R&D Systems assay (used for study samples in the Geisel School of Medicine twin study) at each time point. Pearson correlation was used to describe the concordance between logarithm-transformed values, as well as Cronbach alpha, a measure of interrater reliability. PlGF and sFlt-1 values measured by the different assays showed high concordance ( Table 1 ). Because absolute values for the angiogenic factors differ between assays, levels between studies could not be directly compared.
| Time | Assay | n | Pearson correlation | Cronbach Alpha |
|---|---|---|---|---|
| Time 1 | PlGF | 639 | 0.79 | 0.88 |
| sFlt-1 | 632 | 0.86 | 0.93 | |
| Time 2 | PlGF | 608 | 0.91 | 0.96 |
| sFlt-1 | 607 | 0.87 | 0.93 | |
| Time 3 | PlGF | 602 | 0.92 | 0.96 |
| sFlt-1 | 602 | 0.86 | 0.93 | |
| Time 4 | PlGF | 568 | 0.95 | 0.98 |
| sFlt-1 | 568 | 0.90 | 0.95 |
Clinical data
BIRTH cohort
Maternal age, parity, conception by assisted reproductive technologies (ART), and baby’s birth anthropometrics were abstracted from medical records. The participants completed a brief questionnaire that ascertained information on race/ethnicity, medical history, and history of preeclampsia in a previous pregnancy. Gestational age was confirmed by ultrasound scanning at <15 weeks’ gestation.
Geisel School of Medicine twin study
Data on mother’s age and race/ethnicity; parity; ART conception; and baby’s sex and birth anthropometrics were abstracted from medical records and a form completed at delivery. The pathology report provided data on chorionicity of the twins and placental weight. Gestational age was confirmed by ultrasound.
Statistical methods
Clinical characteristics were compared between twin and singleton pregnancies using Student t test for continuous variables, and χ 2 analysis for categorical variables. Z-scores were developed for birthweight values to account for fetal sex and gestational age using an external standard. Angiogenic protein values and their ratio were adjusted for gestational week in models with logarithm-transformed angiogenic factor as the dependent variable; geometric means and 95% confidence intervals were calculated by taking the exponent of the logarithm-transformed mean. Geometric mean values for the angiogenic factors and the ratio of sFlt-1/PlGF and 95% confidence intervals from an analysis of variance model (PROC MIXED; SAS Institute, Cary, NC) that adjusted for weeks of gestation were plotted by gestational week at blood collection. Statistical analyses were performed using SAS, version 9.2 (SAS Institute).
Results
Mothers of twins were slightly older on average than mothers of singletons in both studies ( Table 2 ). The majority of women in the BIRTH cohort were white as were nearly all mothers in the Dartmouth twin study. In the BIRTH cohort, white mothers contributed a greater proportion of twin than singleton pregnancies (68.8% vs 59%), and black mothers contributed a higher proportion of singleton than twin pregnancies (21.0% vs 9.7%). As expected, gestational week at delivery was earlier among the twin pregnancies in both studies. Approximately a third of the women in both studies were nulliparous, and ART was more likely to have been used among the mothers of twins (BIRTH cohort 68.8%; Geisel School of Medicine study 39.0%) than among the mothers of singletons (6.2% and 8.1%, respectively). Birthweight was lower in the twin pregnancies, even when accounting for gestational age and fetal sex.
| Characteristic | Birth cohort | Geisel School of Medicine twin study | ||||
|---|---|---|---|---|---|---|
| Singletons | Twins | P value a | Singletons | Twins | P value a | |
| n = 2193 | n = 91 | n = 62 | n = 41 | |||
| Maternal and gestational | ||||||
| Maternal age, y (SD) | 31.4 (5.7) | 35.1 (5.8) | < .0001 | 31.0 (5.7) | 33.4 (5.9) | .05 |
| Gestational weeks (SD) at delivery | 39.0 (1.8) | 36.3 (2.3) | < .0001 | 38.5 (1.8) | 37.1 (2.1) | .002 |
| Nulliparous, n (%) | 648 (29.1) | 27 (29.0) | .9957 | 21 (33.9) | 12 (29.3) | .62 |
| Artificial reproductive technologies used (%) | 138 (6.2) | 64 (68.8) | < .0001 | 5 (8.1) | 16 (39.0) | < .0001 |
| Race, n (%) | ||||||
| White | 1318 (59.1) | 64 (68.8) | 58 (95.1) | 40 (97.6) | ||
| Black | 469 (21.0) | 9 (9.7) | ||||
| Asian | 145 (6.5) | 9 (9.7) | ||||
| Hispanic | 213 (9.6) | 9 (9.7) | ||||
| Other/mixed/unknown | 85 (3.8) | 2 (2.2) | .06 | 3 (4.9) | 1 (2.4) | .65 a |
| Infant b | ||||||
| Birthweight, g | 3312 (538) | 2418 (214) | < .0001 | 3257 (597) | 2635 (412) | < .0001 |
| Z-score for birthweight | 0.09 (0.97) | −0.84 (0.70) | < .0001 | 0.20 (1.1) | −0.72 (0.57) | < .0001 |
a Means are presented for continuous variables, percentages for categorical variables– P values are from t tests and χ 2 tests, respectively, except for race among Geisel School of Medicine study women, which is based on Fisher exact test
Table 3 shows the values for angiogenic factors at the 4 time points in pregnancy among women in the BIRTH cohort. In singleton and twin mothers, PlGF concentrations increased from early to midpregnancy and then declined by the third trimester (as noted in the fourth time point), while s-Flt1 concentrations increased throughout the pregnancy. Accordingly, the s-Flt1/ PlGF ratio decreased through midpregnancy but increased in the third trimester. In twins, maternal PlGF concentrations were higher than in singleton pregnancies until the third trimester, when values became lower than those in singleton pregnancies ( Table 3 ). Maternal sFlt-1 concentrations and the sFlt-1/PlGF ratio were higher in twins compared with singletons throughout the pregnancy, with the greatest differences demonstrated in the third trimester ( Table 3 ). Results for mean concentrations from repeated measures models that accounted for the correlations among angiogenic factors over the pregnancy were similar but the statistical significance of the differences, particularly for PlGF, were attenuated ( Figures 1-3 ). Results were similar with additional adjustment for maternal age, race/ethnicity, and parity (data not shown).
| Variable | Gestational weeks, mean (range) | |||||
|---|---|---|---|---|---|---|
| 9.7 (8.4–11.6) | 17.8 (16.8–18.7) | |||||
| Singleton | Twin | P value | Singleton | Twin | P value | |
| n = 2193 | n = 91 | n = 2058 | n = 89 | |||
| PlGF, pg/mL a | 22.8 (22.4–23.2) | 24.9 (23.0–26.9) | .0323 | 138.4 (135.6–141.3) | 213.5 (193.3–235.8) | < .0001 |
| sFlt-1, pg/mL a | 4485 (4347–4628) | 7037 (6031–8212) | < .0001 | 6131 (5974–6291) | 12,543 (11,074–14,207) | < .0001 |
| sFlt-1:PlGF ratio | 202.4 (196.0–209.0) | 294.6 (251.5–335.5) | < .0001 | 44.3 (42.9–45.7) | 58.7 (50.6–68.2) | .0003 |
| 25.9 (24.8–28.1) | 35.1 (34.6–35.9) | |||||
|---|---|---|---|---|---|---|
| Singleton | Twin | P value | Singleton | Twin | P value | |
| n = 2089 | n = 91 | n = 2064 | n = 68 | |||
| PlGF, pg/mL a | 445.9 (434.1–458.1) | 668.0 (587.2–759.9) | < .0001 | 350.2 (335.6–365.5) | 219.2 (173.2–277.3) | < .0001 |
| sFlt-1, pg/mL a | 5898 (5736–6065) | 12,968 (11,347–14,821) | < .0001 | 10,151 (9875–10,435) | 36,916 (31,714–42,971) | < .0001 |
| sFlt-1:PlGF ratio | 13.2 (12.7–13.7) | 19.4 (16.2–23.2) | < .0001 | 29.0 (27.3–30.7) | 168.4 (121.9–232.8) | < .0001 |
a Means (95% confidence intervals) are adjusted for gestational age from models with logarithm-transformed angiogenic factor as outcome–presented are geometric means derived from taking exponent of adjusted mean; data are from the BIRTH cohort.
