An imbalance between angiogenic and anti-angiogenic factors precedes fetal death in a subset of patients: results of a longitudinal study

Figures & data

Table I.  Demographic and clinical characteristics of the study groups.

	Normal pregnancy (n = 124)	Fetal death (n = 19)	p
Maternal age (years)	24 (16–47)	24 (16–44)	0.4
Body mass index (BMI) (kg/m^2)	23.5 (16–37)	23.7 (19–44)	0.4
Smoking	14 (11.3%)	1 (5.3%)	0.7
Nulliparity	58 (46.8%)	11 (57.9%)	0.4
GA at enrollment (weeks)	12.2 (7.9–14.9)	11.1 (6.3–19.4)	0.3
GA at delivery (weeks)	40.0 (37.3–41.9)	36.0 (20.6–40.3)	<0.001
Birthweight (g)*	3390 (2580–4090)	2280 (420–3650)	<0.001
Birthweight <10th percentile	0	6 (33.3%)	<0.001
Delivery < 37 weeks	0	11 (57.9%)	<0.001

Data expressed as median (range) and number (percentage). GA, gestational age. *: n of fetal death group = 18.

Table II.  Median (interquartile range) plasma concentrations of sVEGFR-1, sEng, and PlGF in each gestational age interval.

		Controls	Fetal death
1st interval (6–10 weeks)	GA	9.3 (8.9–9.4)	8.8 (8.5–9.5)
	sVEGFR-1 (pg/ml)	1363 (1006–1746)	1258 (886–2999)
	sEng (ng/ml)	7.1 (6.3–7.6)	7.3 (6.6–7.9)
	PlGF (pg/ml)	9.8 (6.3–24.4)	18.2 (13.3–21.3)
	n	16	8
2nd interval (10.1–14 weeks)	GA	12.3 (11.3–13.0)	12.3 (11.2–13.6)
	sVEGFR-1 (pg/ml)	1732 (1287–2238)	1810 (1549–2333)
	sEng (ng/ml)	7.3 (6.3–8.6)	7.5 (6.6–9.1)
	PlGF (pg/ml)	34.3 (24.5–48.9)	38.0 (26.1–45.1)
	n	98	10
3rd interval (14.1–18 weeks)	GA	16.7 (16.0–17.5)	15.4 (15.0–17.0)
	sVEGFR-1 (pg/ml)	1628 (991–2378)	1931 (1514–2772)
	sEng (ng/ml)	7.0 (6.0–7.7)	6.4 (6.0–7.8)
	PlGF (pg/ml)	99.1 (79.1–121.7)	88.9 (45.8–108.9)
	n	99	17
4th interval (18.1–22 weeks)	GA	20.7 (19.3–21.4)	20.4 (19.3–21.1)
	sVEGFR-1 (pg/ml)	1606 (1192–2357)	1658 (1362–2986)
	sEng (ng/ml)	6.0 (5.2–7.0)	7.1 (6.0–8.4)
	PlGF (pg/ml)	229.2 (154.5–323.4)	142.6 (99.0–235.8)
	n	95	17
5th interval (22.1–26 weeks)	GA	24.0 (22.7–25.4)	24.3 (23.9–25.1)
	sVEGFR-1 (pg/ml)	1697 (1124–2695)	1802 (1184–2460)
	sEng (ng/ml)	6.1 (5.1–7.4)	6.8 (5.8–8.9)
	PlGF (pg/ml)	399.9 (275.3–565.0)	318.1 (175.6–442.6)
	n	108	12
6th interval (26.1–30 weeks)	GA	27.6 (26.8–29.6)	28.4 (27.2–29.0)
	sVEGFR-1 (pg/ml)	1672 (1236–2390)	3320 (2592–4873)
	sEng (ng/ml)	7.2 (6.1–8.8)	10.0 (7.1–11.4)
	PlGF (pg/ml)	565.2 (362.4–853.1)	238.2 (113.6–525.6)
	n	108	10
7th interval (30.1–34 weeks)	GA	31.1 (30.6–31.7)	32.7 (32.0–33.0)
	sVEGFR-1 (pg/ml)	2069 (1473–2712)	6435 (4830–7998)
	sEng (ng/ml)	8.6 (6.9–11.9)	16.2 (11.0–20.3)
	PlGF (pg/ml)	535.1 (368.3–891.4)	203.7 (60.7–394.8)
	n	78	11
8th interval (34.1–38 weeks)	GA	35.6 (34.9–36.1)	35.3 (34.4–36.4)
	sVEGFR-1 (pg/ml)	3686 (2621–4977)	5580 (3803–7535)
	sEng (ng/ml)	11.9 (9.1–17.1)	14.8 (12.8–24.6)
	PlGF (pg/ml)	312.9 (185.8–654.5)	443.1 (109.1–1001.5)
	n	108	7
9th interval (38.1–42 weeks)	GA	39.9 (39.4–40.7)	39.4 (39.3–39.4)
	sVEGFR-1 (pg/ml)	7571 (5855–10965)	9166 (2958–12936)
	sEng (ng/ml)	13.1 (9.8–19.3)	18.1 (15.0–23.4)
	PlGF (pg/ml)	128.2 (86.8–194.7)	184.1 (15.6–438.8)
	n	112	5

Figure 1.  Maternal plasma concentrations (actual data on a log_e scale) of soluble vascular endothelial growth factor receptor-1 (sVEGFR-1) in women with a normal pregnancy (ˆ) and patients destined to have a fetal death (•). Each curve represents a quadratic model fit of the concentrations as a function of gestational age in normal pregnant women (dashed line) and those with a fetal death (solid line) without adjusting for covariates. The short vertical lines on the solid curve denote statistical significance between the two groups at the corresponding gestational age according to a linear mixed-effects model adjusting for covariates.

Figure 2.  Maternal plasma concentrations (actual data on a log_e scale) of soluble endoglin (sEng) in women with a normal pregnancy (ˆ) and patients destined to have a fetal death (•). Each curve represents a quadratic model fit of the concentrations as a function of gestational age in normal pregnant women (dashed line) and those with a fetal death (solid line) without adjusting for covariates. The short vertical lines on the solid curve denote statistical significance between the two groups at the corresponding gestational age according to a linear mixed-effects model adjusting for covariates.

Figure 3.  Maternal plasma concentrations (actual data on a log_e scale) of placental growth factor (PlGF) in women with a normal pregnancy (ˆ) and patients destined to have a fetal death (•). Each curve represents a quadratic model fit of the concentrations as a function of gestational age in normal pregnant women (dashed line) and those with a fetal death (solid line) without adjusting for covariates. The short vertical lines on the solid curve denote statistical significance between the two groups at the corresponding gestational age according to a linear mixed-effects model adjusting for covariates.

Figure 4.  Individual profiles of maternal plasma concentrations (actual data on a log_e scale) of sVEGFR-1, sEng and PlGF as a function of gestational age in women destined to have a normal pregnancy (controls) and those destined to have a fetal death.

Table III.  Longitudinal analysis of the association between anti-angiogenic (sVEGFR-1 and sEng) or angiogenic factor (PIGF) and fetal death after adjusting for confounding factors.

	sVEGFR-1					s-Eng					PlGF
	Value	Std. error	DF	t	p	Value	Std. error	DF	t	p	Value	Std. error	DF	t	p
(Intercept)	8.1391	0.2522	825	32.28	0.0000	2.3068	0.1687	825	13.67	0.0000	2.2029	0.3295	825	6.69	0.0000
Fetal death	−0.6585	0.1722	137	−3.82	0.0002	−0.2350	0.1169	137	−2.01	0.0464	1.2877	0.2278	137	5.65	0.0000
GA	−0.1015	0.0064	825	−15.78	0.0000	−0.0507	0.0044	825	−11.49	0.0000	0.4430	0.0086	825	51.63	0.0000
GA^2	0.0039	0.0002	825	22.19	0.0000	0.0020	0.0001	825	17.06	0.0000	−0.0100	0.0002	825	−43.02	0.0000
Maternal age	−0.0022	0.0058	137	−0.39	0.7007	0.0058	0.0039	137	1.49	0.1380	−0.0097	0.0076	137	−1.28	0.2026
BMI	−0.0172	0.0079	137	−2.19	0.0300	−0.0150	0.0053	137	−2.85	0.0050	−0.0112	0.0103	137	−1.09	0.2774
smoking	−0.0730	0.1082	137	−0.68	0.5007	0.0132	0.0722	137	0.18	0.8550	0.2394	0.1411	137	1.70	0.0920
Nulliparity	0.2112	0.0749	137	2.82	0.0055	0.0768	0.0500	137	1.54	0.1271	−0.1544	0.0977	137	−1.58	0.1164
Duration of storage	0.0002	0.0001	825	3.45	0.0006	0.0000	0.0000	825	0.97	0.3342	−0.0003	0.0001	825	−4.15	0.0000
Fetal death × GA	0.0975	0.0188	825	5.19	0.0000	0.0368	0.0129	825	2.86	0.0043	−0.1514	0.0251	825	−6.04	0.0000
Fetal death × GA^2	−0.0022	0.0005	825	−4.05	0.0001	−0.0005	0.0004	825	−1.23	0.2203	0.0030	0.0007	825	4.14	0.0000

GA, gestational age; BMI, body mass index.

Fetal death × GA, interaction between diagnosis and GA; Fetal death × GA², interaction between diagnosis and GA².

Table IV.  The statistical differences (p) and percentage changes [(fetal death − control)/control × 100] in the maternal plasma concentrations of angiogenic and anti-angiogenic factors between patients with a fetal death and normal pregnant women according to gestational age.

sVEGFR-1			sEng			PlGF
GA (weeks)	p	% change	GA (weeks)	p	% change	GA (weeks)	p	% change
7–11	<0.05	− (26–48)	7	0.04	−21	7–14	<0.05	+ (46–262)
12–19	>0.05	–	8–19	>0.05	–	15–21	>0.05	–
20–37	<0.05	+ (26–51)	20–40	<0.05	+ (18–62)	22–39	<0.05	− (26–46)
38–40	>0.05	–				40	0.06	–

The full version of this table is displayed in Supplementary Table I.

Table V.  The statistical differences (p) and percentage changes [(fetal death − control)/control × 100] in the ratio of maternal plasma concentrations of angiogenic and anti-angiogenic factors between patients with a fetal death and normal pregnant women according to gestational age.

PlGF/(sVEGFR-1 × sEng)			PlGF/sEng			PlGF/sVEGFR-1
GA (weeks)	p	% change	GA (weeks)	p	% change	GA (weeks)	p	% change
7–13	<0.05	+ (94–781)	7–14	<0.05	+ (47–362)	7–14	<0.05	+ (56–593)
14–19	>0.05	–	15–20	>0.05	–	15–19	>0.05	–
20–40	<0.05	− (44–75)	21–40	<0.05	− (35–64)	20–39	<0.05	− (34–64)
						40	0.06	–

The full version of this table is displayed in Supplementary Table II.

Figure 5.  The ratio (log_e scale) of maternal plasma concentrations of PlGF/(sEng × sVEGFR-1) in women with a normal pregnancy (ˆ) and patients destined to have a fetal death (•). Each curve represents a quadratic model fit of the concentrations as a function of gestational age in normal pregnant women (dashed line) and those with a fetal death (solid line) without adjusting for covariates. The short vertical lines on the solid curve denote statistical significance between the two groups at the corresponding gestational age according to a linear mixed-effects model adjusting for covariates.

Table VI.  The statistical differences (p) and percentage changes [(fetal death − control)/control ×100] in the ratio of maternal plasma concentrations of angiogenic and anti-angiogenic factors [PlGF/ (sVEGFR-1 × sEng)] between patients with a fetal death and normal pregnant women stratified by gestational age at which fetal death was diagnosed.

Fetal death at < 37 weeks			Fetal death at > 37 weeks
GA (weeks)	p	% change	GA (weeks)	p	% change
7–11	<0.05	+ (111–328)	7–14	<0.05	+ (119–1146)
12–19	>0.05	–	15–24	>0.05	–
20–37	<0.05	− (49–94)	25–34	<0.05	− (48–54)
			35–40	>0.05

The full version of this table is displayed in Supplementary Table III.

Figure 6.  The ratio (log_e scale) of maternal plasma concentrations of PlGF/(sEng × sVEGFR-1) in women with a normal pregnancy (ˆ) and patients destined to have a fetal death before 37 weeks of gestation (•). Each curve represents a quadratic model fit of the concentrations as a function of gestational age in normal pregnant women (dashed line) and those with a fetal death (solid line) without adjusting for covariates. The short vertical lines on the solid curve denote statistical significance between the two groups at the corresponding gestational age according to a linear mixed-effects model adjusting for covariates.

Figure 7.  The ratio (log_e scale) of maternal plasma concentrations of PlGF/(sEng × sVEGFR-1) in women with a normal pregnancy (ˆ) and patients destined to have a fetal death after 37 weeks of gestation (•). Each curve represents a quadratic model fit of the concentrations as a function of gestational age in normal pregnant women (dashed line) and those with a fetal death (solid line) without adjusting for covariates. The short vertical lines on the solid curve denote statistical significance between the two groups at the corresponding gestational age according to a linear mixed-effects model adjusting for covariates.