Adjusting for the level of the other analyte and for maternal characteristics had a minimal effect on the nature and strength of associations (Table 2). There were no statistically significant interactions between decile of either analyte and maternal characteristics (all P>.05, adjusted for number of comparisons). There was a statistically significant interaction between decile of sFlt-1 and decile of PlGF in relation to all adverse outcome (odds ratio for the interaction 1.015, 95% CI 1.004–1.026, P=.007). The interaction was weaker for preeclampsia but stronger for all other outcomes. The interaction between sFlt-1 and PlGF was then assessed for a composite adverse outcome, namely all cases where there was no diagnosis of preeclampsia. This analysis demonstrated a highly statistically significant interaction between the two analytes (odds ratio for the interaction 1.018, 95% CI 1.006–1.031, P=.003). When the analysis of the relationship between sFlt-1 and this composite outcome was stratified by the level of PlGF, it was apparent that the association was stronger when PlGF was lower (Fig. 3).
The association between PlGF and preeclampsia was more marked among women with complicated preeclampsia (defined as being associated with preterm delivery or with an SGA infant; odds ratio 0.88, 95% CI 0.81 to 0.96) than among women with preeclampsia who delivered an infant with birth weight appropriate for gestational age at term (odds ratio 0.99, 95% CI 0.95 to 1.04). There was no significant association between sFlt-1 and either complicated (odds ratio 0.96, 95% CI 0.89–1.04) or uncomplicated preeclampsia (odds ratio 1.01, 95% CI 0.96–1.06). When the analyses were confined to women whose blood was obtained before 13 weeks of gestation, the pattern of associations was very similar to the findings for the whole population (Fig. 4).
We found no association between circulating levels of sFlt-1 at 10–14 weeks of gestation and the risk of preeclampsia, but a decreased risk of the disease among women with higher levels of PlGF. These findings are in complete agreement with a previous large-scale nested case–control study that found no association between maternal serum levels of sFlt-1 and the risk of preeclampsia in early pregnancy but a reduced risk of the disease among women with high levels of PlGF from 13 weeks onward.3 However, we demonstrate that women with higher circulating levels of sFlt-1 at 10–14 weeks of gestation were at lower risk of other complications of pregnancy, specifically delivery of an SGA infant, spontaneous preterm birth, and stillbirth. Higher levels of PlGF at 10–14 weeks of gestation were associated with a decreased risk of delivery of an SGA infant but were not associated with the other outcomes.
The finding of improved outcome in association with higher levels of sFlt-1 is of particular interest. It could be that high circulating levels of sFlt-1 have some physiologic role in early pregnancy. Alternatively, it could be that high levels of sFlt-1 reflect some other aspect of placental function that is in turn associated with better outcome. It is unlikely that sFlt-1 is acting simply as a marker of placental mass, because we have previously related maternal serum levels of the free β-hCG at the same gestational age to the risk of later adverse outcome and failed to demonstrate any association.6
Previous studies have related maternal serum levels of sFlt-1 in pregnancy to the risk of adverse outcome for the offspring. Two demonstrated no association between maternal levels of sFlt-1 and the risk of delivering an SGA infant.16,17 A further study demonstrated elevated levels of sFlt-1 at 23–25 weeks gestational age among women with high-resistance patterns of uterine artery flow velocimetry and established early onset growth restriction.18 None of these studies corrected sFlt-1 levels for the gestational age at the time of sampling or for maternal weight, and none included more than 30 cases. Previous studies had demonstrated inconsistent relationships between levels of PlGF and the risk of delivering an SGA infant, with both higher19 and lower20,21 levels of PlGF reported.
We observed a statistically significant interaction between sFlt-1 and PlGF. There was no prior hypothesis that these factors would interact. We found that among women with low PlGF, elevated levels of sFlt-1 were significantly protective against adverse perinatal outcome, whereas there was no association between levels of sFlt-1 among women with PlGF levels in the upper two quintiles (Fig. 3). The assay employed in this study measured free PlGF. Soluble Flt-1 binds PlGF, and high levels of sFlt-1 would be anticipated to lead to low levels of free PlGF. The interaction could indicate that low circulating PlGF due to high levels of sFlt-1 is associated with better outcome than low circulating PlGF due to other causes, such as reduced placental production. Alternatively, it could indicate that a protective effect of sFlt-1 is antagonized by high circulating levels of PlGF.
Higher levels of sFlt-1 in late pregnancy are associated with an increased risk of preeclampsia. In contrast, we show that higher levels of sFlt-1 in very early pregnancy are associated with a decreased risk of other pregnancy complications. This finding is potentially clinically relevant. It has been hypothesized that there is a causal association between elevated maternal levels of sFlt-1 and preeclampsia.3 It is proposed that sFlt-1 is released by the placenta into the maternal circulation and binds maternal PlGF and VEGF-A, leading to maternal endothelial dysfunction,3 and animal studies are consistent with this model.1,4 These data suggest that administration of VEGF-A121, or alternative approaches to reducing maternal circulating levels of sFlt-1, may be therapeutically useful among women with established preeclampsia or among women who are at increased risk of the disease. The current data indicate that this has the potential to be harmful in early pregnancy. It may be prudent to conduct further observational studies of the associations in late pregnancy before evaluation of such treatments.
The biologic basis for the association between sFlt-1 and preterm birth warrants further study. There are some data that indicate an association is biologically plausible. Studies of pregnant human uterus have demonstrated expression of VEGF-A in the myometrium.23 Studies in the pregnant rat demonstrated expression of VEGF-A and its main functional receptor in the cervix. Moreover, VEGF-A expression peaks in association with ripening of the cervix.24 If VEGF-A also had a role in human cervical ripening, low levels of sFlt-1 could result in increased free VEGF-A and hence promote preterm birth. There are no data on the possible effects of PlGF on the myometrium and cervix, and this should also be addressed.
In conclusion, higher levels of sFlt-1 in early pregnancy were associated with a decreased risk of intrauterine growth restriction, spontaneous preterm labor, and stillbirth. Strategies aimed at reducing the risk of preeclampsia by inactivating sFlt-1 in the maternal circulation may adversely affect perinatal outcome.
1. Maynard SE, Min JY, Merchan J, Lim KH, Li J, Mondal S, et al. Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 2003;111:649–58.
2. Tjoa ML, Oudejans CB, van Vugt JM, Blankenstein MA, van Wijk IJ. Markers for presymptomatic prediction of preeclampsia and intrauterine growth restriction. Hypertens Pregnancy 2004;23:171–89.
3. Levine RJ, Maynard SE, Qian C, Lim KH, England LJ, Yu KF, et al. Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med 2004;350:672–83.
4. Zhang Y, Ma J, Almirez R, de Forest N, Schellenberger U, Pollitt S, et al. Overexpression of soluble fms-like tyrosine kinase-1 (sFlt-1) induced experimental pre-eclampsia in rats and attenuation of pre-eclamptic phenotype by recombinant vascular endothelial growth factor 121 (VEGF121). J Soc Gynecol Investig 2006;13:287A.
5. Crossley JA, Aitken DA, Cameron AD, McBride E, Connor JM. Combined ultrasound and biochemical screening for Down's syndrome in the first trimester: a Scottish multicentre study. BJOG 2002;109:667–76.
6. Smith GC, Stenhouse EJ, Crossley JA, Aitken DA, Cameron AD, Connor JM. Early pregnancy levels of pregnancy-associated plasma protein a and the risk of intrauterine growth restriction, premature birth, preeclampsia, and stillbirth. J Clin Endocrinol Metab 2002;87:1762–7.
7. Cole SK. Scottish maternity and neonatal records. In: Chalmers I, McIlwaine GM, editors. Perinatal audit and surveillance. London (UK): Royal College of Obstetricians and Gynaecologists; 1980. p 39–51.
8. Information and Statistics Division NHS Scotland. Scottish perinatal and infant mortality report 2000. Edinburgh (UK): ISD Scotland Publications; 2001.
9. McLoone P, Boddy FA. Deprivation and mortality in Scotland, 1981 and 1991. BMJ 1994;309:1465–70.
10. Smith GC, Crossley JA, Aitken DA, Pell JP, Cameron AD, Connor JM, et al. First-trimester placentation and the risk of antepartum stillbirth. JAMA 2004;292:2249–54.
11. Crossley JA, Aitken DA, Berry E, Connor JM. Impact of a regional screening programme using maternal serum alpha fetoprotein (AFP) and human chorionic gonadotrophin (hCG) on the birth incidence of Down's syndrome in the west of Scotland. J Med Screen 1994;1:180–3.
12. Neveux LM, Palomaki GE, Larrivee DA, Knight GJ, Haddow JE. Refinements in managing maternal weight adjustment for interpreting prenatal screening results. Prenat Diagn 1996;16:1115–9.
13. van Belle G, Fisher LD, Heagerty PJ, Lumley T. Biostatistics: a methodology for the health sciences. Hoboken (NJ): Wiley-Interscience; 2004.
14. Hosmer DW, Lemeshow S. Applied logistic regression. New York (NY): Wiley; 2000.
15. Mehta CR, Patel NR. Exact logistic regression: theory and examples. Stat Med 1995;14:2143–60.
16. Wathen KA, Tuutti E, Stenman UH, Alfthan H, Halmesmaki E, Finne P, et al. Maternal serum-soluble vascular endothelial growth factor receptor-1 in early pregnancy ending in preeclampsia or intrauterine growth retardation. J Clin Endocrinol Metab 2006;91:180–4.
17. Shibata E, Rajakumar A, Powers RW, Larkin RW, Gilmour C, Bodnar LM, et al. Soluble fms-like tyrosine kinase 1 is increased in preeclampsia but not in normotensive pregnancies with small-for-gestational-age neonates: relationship to circulating placental growth factor. J Clin Endocrinol Metab 2005;90:4895–903.
18. Savvidou MD, Yu CK, Harland LC, Hingorani AD, Nicolaides KH. Maternal serum concentration of soluble fms-like tyrosine kinase 1 and vascular endothelial growth factor in women with abnormal uterine artery Doppler and in those with fetal growth restriction. Am J Obstet Gynecol 2006;195:1668–73.
19. Ong CY, Liao AW, Cacho AM, Spencer K, Nicolaides KH. First-trimester maternal serum levels of placenta growth factor as predictor of preeclampsia and fetal growth restriction. Obstet Gynecol 2001;98:608–11.
20. Tjoa ML, van Vugt JM, Mulders MA, Schutgens RB, Oudejans CB, van Wijk IJ. Plasma placenta growth factor levels in midtrimester pregnancies. Obstet Gynecol 2001;98:600–7.
21. Bersinger NA, Odegard RA. Second- and third-trimester serum levels of placental proteins in preeclampsia and small-for-gestational age pregnancies. Acta Obstet Gynecol Scand 2004;83:37–45.
22. Haig D. Genetic conflicts in human pregnancy. Q Rev Biol 1993;68:495–532.
23. Oh MJ, Lee JK, Lee NW, Shin JH, Yeo MK, Kim A, et al. Vascular endothelial growth factor expression is unaltered in placentae and myometrial resistance arteries from pre-eclamptic patients. Acta Obstet Gynecol Scand 2006;85:545–50.
24. Mowa CN, Jesmin S, Sakuma I, Usip S, Togashi H, Yoshioka M, et al. Characterization of vascular endothelial growth factor (VEGF) in the uterine cervix over pregnancy: effects of denervation and implications for cervical ripening. J Histochem Cytochem 2004;52:1665–74.
Regressed median of soluble fms-like tyrosine kinase-1 (sFlt-1) in relation to gestational age:
sFlt-1 (in pg/mL)=e(6.389+33.366/x), where x is the gestational age in days.
Regressed multiple of the median of sFlt-1 in relation to maternal weight:
Multiple of the median of sFlt-1=0.6944+18.2822/x, where x is the maternal weight in kilograms.
Regressed median of placental growth factor (PlGF) in relation to gestational age:
PlGF (in pg/mL)=e(2.071+0.02288/x), where x is the gestational age in days.
Regressed multiple of the median of PlGF in relation to maternal weight:
Multiple of the median of PlGF=0.8609+9.8629/x, where x is the maternal weight in kilograms.
Correction of multiple of the median of PlGF for maternal smoking:
Smokers: Multiples of the median (MoM) value divided by 1.29. Nonsmokers: MoM value divided by 0.95
Cut points for deciles of sFlt-1 (expressed as MoM):
0.572, 0.702, 0.814, 0.909, 1.016, 1.117, 1.219, 1.366, and 1.611.
Cut points for deciles of PlGF (expressed as MoM):
0.610, 0.720, 0.824, 0.907, 1.003, 1.092, 1.197, 1.332, and 1.575
Cut point for quintiles of PlGF (expressed as MoM):
0.720, 0.907, 1.092, and 1.332.
Eleven hospitals recruiting 98.6% of patients:
The Queen Mother's Hospital, Glasgow; Simpson Maternity, Edinburgh; Southern General Hospital, Glasgow; Royal Maternity Hospital, Glasgow; Royal Alexandria Hospital, Paisley; Ninewells, Dundee; Ayrshire Central Hospital, Irvine; Forth Park, Kikcaldy; Stirling Royal Infirmary, Stirling; Falkirk and District Royal Infirmary, Falkirk, and Inverclyde Royal Hospital, Greenock.