Preeclampsia at term: evidence of disease heterogeneity based on the profile of circulating cytokines and angiogenic factors

Abstract

Background:

Intravascular inflammation and an anti-angiogenic state have been implicated in the pathophysiology of preeclampsia. Based on the profiles of their angiogenic/anti-angiogenic factors, women with preeclampsia at term may be classified into 2 subgroups with different characteristics and prevalence of adverse outcomes. This study was undertaken to examine whether these 2 subgroups of preeclampsia at term also show differences in their profiles of intravascular inflammation.

Objective:

To determine the plasma profiles of cytokines and chemokines in women with preeclampsia at term who had a normal or an abnormal angiogenic profile.

Study Design:

A nested case-control study was conducted to include women classified into 3 groups: women with an uncomplicated pregnancy (n=213), and women with preeclampsia at term with a normal (n=55) or an abnormal (n=41) angiogenic profile. An abnormal angiogenic profile was defined as a plasma ratio of placental growth factor and soluble fms-like tyrosine kinase-1 multiple of the median below the 10^th percentile for gestational age. Concentrations of cytokines were measured by multiplex immunoassays.

Results:

Women with preeclampsia at term and an abnormal angiogenic profile showed evidence of the greatest intravascular inflammation among the study groups. These women had higher plasma concentrations of 5 cytokines [interleukin (IL)-6, IL-8, IL-12/IL-23p40, IL-15, and IL-16] and 7 chemokines [eotaxin, eotaxin-3, interferon-γ inducible protein (IP)-10, monocyte chemotactic protein-4 (MCP-4), macrophage inflammatory protein-1β, macrophage-derived chemokine, and thymus and activation-regulated chemokine] compared to uncomplicated pregnancies. By contrast, women with preeclampsia at term and a normal angiogenic profile, compared to uncomplicated pregnancies, only had a higher plasma concentration of MCP-4. A significant correlation between severity of the anti-angiogenic state, blood pressure, and plasma concentrations of a subset of cytokines was observed.

Conclusions:

Term preeclampsia can be classified into 2 clusters. One is characterized by an anti-angiogenic state coupled with an excessive inflammatory process while the other has neither of these features. These findings further support the heterogeneity of preeclampsia at term and may explain the distinct clinical outcomes.

Introduction

Preeclampsia, a pregnancy-specific disorder that complicates 3% to 7% of pregnancies,^1–9 is one of the leading causes of maternal and perinatal morbidity and mortality worldwide.^{9, 10} In the United States (US), the incidence of new-onset hypertensive disorders of pregnancy has increased dramatically in the last decade,¹¹ and this has been partially attributed to the increasing rate of obesity,¹² a major risk factor for preeclampsia, especially at term.^{13, 14} The estimated cost of preeclampsia to the US healthcare system in 2012 was $2.18 billion.¹⁵

Several lines of evidence suggest that preeclampsia is a heterogeneous disease,^16–22 and sub-classifications into “early- or late-onset”¹⁶ as well as into “placental or maternal preeclampsia” ²¹ (abnormal placentation or maternal vascular predisposition) have been proposed to better reflect the syndromic nature of this condition. Despite the progress in the prediction and prevention of early preeclampsia,^23–26 preeclampsia at term, which accounts for the majority of all cases,¹⁵ remains a clinical challenge.^{27, 28}

An anti-angiogenic state is central to the pathogenesis of preeclampsia^29–40; however, the predictive and diagnostic performance of angiogenic biomarkers for women with preeclampsia at term is limited.^41–54 A possible explanation is that an abnormal angiogenic profile is present in only about half of women with preeclampsia at term.⁵⁵ Of interest, demographics, clinical characteristics, and prognosis differ between the two subgroups of women with preeclampsia at term with and without angiogenic abnormalities.⁵⁵

Intravascular inflammation leading to endothelial cell dysfunction and multi-organ damage has also been implicated in the pathophysiology of preeclampsia.^56–64 However, the extent to which an exaggerated intravascular inflammation and an anti-angiogenic state coexist in women with preeclampsia at term has not yet been elucidated. To address this question and to gain further insight into the pathophysiology of preeclampsia at term, we determined the plasma profiles of cytokines and chemokines in women with preeclampsia at term who had a normal or an abnormal angiogenic profile.

Materials and methods

Study design and participants

A nested case-control study was conducted to include women classified into 3 groups: women with preeclampsia at term with a normal (n=55) and with an abnormal (n=41) angiogenic profile and women with an uncomplicated pregnancy (n=213). All patients were selected from a previously described longitudinal cohort of 4,006 women, enrolled between 2006 and 2010⁶⁵ and did not meet any of the following exclusion criteria: multiple gestation, major obstetrical complication at enrollment, serious medical illness, chronic hypertension requiring medication, asthma requiring systemic steroids, the requirement of anti-platelet or non-steroidal anti-inflammatory drugs, active hepatitis, and identified fetal anomalies. Women with preeclampsia at term who had an available maternal blood sample collected within 7 days before delivery (considered to be the peak of an anti-angiogenic state) and had adequate sample volume to perform multiplex cytokine assays were included in the study. Women with an uncomplicated pregnancy (n=213) were selected from the control group of 540 women matched (2:1) for gestational age at venipuncture with patients diagnosed with preeclampsia. An abnormal angiogenic profile was defined as a plasma ratio of placental growth factor and soluble fms-like tyrosine kinase-1 (PlGF/sFlt-1) multiple of the median (MoM) below the 10^th percentile for gestational age, using previously described cut-off values.⁵⁵ All patients provided written informed consent prior to the collection of samples, and the use of clinical data and biological specimens for research purposes was approved by the Institutional Review Boards of Wayne State University and the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, United States Department of Health and Human Services (NICHD/NIH/DHHS). Many of these samples had been used in past studies that examined angiogenic and anti-angiogenic factors and other candidate biomarkers for preeclampsia.

Clinical Definitions

Preeclampsia was defined as new-onset hypertension developing after 20 weeks of gestation and proteinuria.⁶⁶ The control group comprised women who delivered at term without major obstetrical complications defined as any of the following diagnoses at delivery: preeclampsia, eclampsia, fetal death, small-for-gestational-age (SGA) neonate, preterm labor, preterm premature rupture of the membranes, acute pyelonephritis, placental abruption, gestational hypertension, and gestational or pre-gestational diabetes mellitus.⁶⁵ Hypertension was defined as systolic ≥140 and/or diastolic ≥90 mmHg blood pressure, measured on 2 occasions, 4 hours to 1 week apart. Chronic hypertension was defined as hypertension present before pregnancy or diagnosed before the 20^th week of gestation. Proteinuria was defined as a urine protein level ≥300 mg in a 24-hour urine collection or a random urine specimen, showing ≥1+ proteins by dipstick.⁶⁶ Severe preeclampsia was defined as previously described.⁴⁷ Superimposed preeclampsia was diagnosed if there was an increase in blood pressure from the baseline associated with new-onset proteinuria or a sudden increase in proteinuria if already present in early gestation.⁶⁶ A composite adverse maternal outcome was defined as any of the following complications: eclampsia, blindness, stroke, myocardial ischemia, pulmonary edema, elevated liver enzymes (alanine aminotransferase or aspartate aminotransferase ≥70 IU/L), hepatic hematoma, thrombocytopenia (platelet count <100 × 109/L), acute kidney injury (serum creatinine >1.1 mg/dL), placental abruption, disseminated intravascular coagulopathy, or maternal death.⁶⁷ A composite neonatal outcome was defined as the presence of any of the following complications: requirement of oxygen supplementation, continuous positive airway pressure or mechanical ventilation, respiratory distress syndrome, confirmed sepsis, necrotizing enterocolitis, intraventricular hemorrhage, periventricular leukomalacia, meconium aspiration syndrome, chronic lung disease, or neonatal death. An SGA newborn was defined as one whose birthweight was below the 10^th percentile for gestational age at delivery, according to a US reference population.⁶⁸

Sample collection and multiplex immunoassays

Blood was obtained by venipuncture and collected into tubes containing ethylenediaminetetraacetic acid. Samples were centrifuged and stored at −70°C. Maternal plasma concentrations of sFlt-1 and PlGF were determined by immunoassays (R&D Systems, Minneapolis, MN, USA), as previously described.⁶⁹ The results of angiogenic and anti-angiogenic factors from these patients have been reported in a previous publication.⁵⁵

The V-PLEX Pro-inflammatory Panel 1 (human), Cytokine Panel 1 (human), and Chemokine Panel 1 (human) immunoassays (Meso Scale Discovery, Gaithersburg, MD, USA) were used to measure the concentrations of cytokines and chemokines described in Table 1. Briefly, 50 μL of each sample were dispensed into separate wells of the plates and incubated for 2 hours with vigorous shaking at room temperature. The samples and calibrators were discarded, and the plates were washed 3 times with PBS and 0.05% Tween-20, followed by adding 25 μL of the 1X Detection Antibody Solution into each well. Plates were then incubated for 2 hours with vigorous shaking. The detection antibody was removed, and plates were washed 3 times. One hundred and fifty microliters of 2X Read Buffer T were added to each well, and signals were read by the SECTOR^® Imager 2400 (Meso Scale Discovery). Standard curves were generated, and the assay values of the samples were interpolated based on these curves. The limit of detection and the assay performance are displayed in Table 1. The laboratory personnel were masked to the clinical information.

Table 1.

The assay characteristics of each analyte

Analytes	Alternative names	Sensitivity (pg/mL)	Intra-assay coefficient of variation (%)	Inter-assay coefficient of variation (%)
MSD V-PLEX Pro-inflammatory Panel 1
IFN-γ		0.262	3.330	4.369
TNF-α		0.069	6.318	6.804
IL-1β		0.164	2.188	14.322
IL-2		0.249	3.020	2.183
IL-4		0.018	2.459	6.710
IL-6		0.099	2.764	5.852
IL-8	CXCL8	0.095	5.625	8.281
IL-10		0.045	3.101	4.577
IL-12p70		0.104	3.371	9.870
IL-13		0.115	3.284	7.135
MSD V-PLEX Cytokines Panel 1
TNF-β		0.072	3.034	8.368
IL-1α		0.093	7.532	9.243
IL-5		0.108	1.971	9.450
IL-7		0.110	2.389	4.139
IL-12/IL-23p40		0.310	2.462	1.243
IL-15		0.092	5.053	10.216
IL-16		0.479	6.087	11.599
IL-17A		0.309	3.559	2.058
VEGF		0.126	2.921	7.625
GM-CSF		0.160	2.909	5.848
MSD V-PLEX Chemokines Panel 1
Eotaxin	CCL11	0.540	7.252	6.046
Eotaxin-3	CCL26	0.932	4.256	7.358
IP-10	CXCL10	0.099	3.164	4.461
MCP-1	CCL2	0.113	4.469	8.753
MCP-4	CCL13	0.203	3.972	8.273
MDC	CCL22	1.021	2.895	5.389
MIP-1α	CCL3	0.277	7.221	14.226
MIP-1β	CCL4	0.193	4.898	4.746
TARC	CCL17	0.409	8.153	7.872

GM-CSF: granulocyte-macrophage colony-stimulating factor; IFN: interferon; IL: interleukin; IP: interferon gamma-induced protein; MCP: monocyte chemotactic protein; MDC: macrophage-derived chemokine; MIP: macrophage inflammatory protein; MSD: Mesoscale discovery; TARC: thymus and activation-regulated chemokine; TNF: tumor necrosis factor; VEGF: vascular endothelial growth factor.

Histological Examination

Placentas were examined histologically according to standardized protocols by perinatal pathologists masked to clinical diagnoses and obstetrical outcomes, as previously described.⁶⁵ Diagnoses of placental lesions consistent with maternal vascular malperfusion were made according to criteria established by the Perinatology Section of the Society for Pediatric Pathology ⁷⁰ and the Amsterdam Placental Workshop Group Consensus.⁷¹

Statistical Analysis

Demographic categorical data were summarized as proportions, and continuous variables were summarized in medians and interquartile ranges. Differences were examined using the chi-square or Fisher’s exact and Mann-Whitney U tests. The expected median values of PlGF, sFlt-1, and the PlGF/sFlt-1 ratio by gestational age were previously defined based on a longitudinal study.⁵⁵ MoM values of PlGF, sFlt-1, and PlGF/sFlt-1 were calculated by dividing the observed value for each study participant by the median value.⁵⁵ Comparisons of analyte data (cytokines and chemokines) were performed among controls and women with preeclampsia with and without an abnormal angiogenic profile by using Wilcoxon tests. The correlations between pairs of continuous variables, such as systolic and diastolic blood pressure, PlGF MoM, sFlt-1 MoM, PlGF/sFlt-1 MoM, and analytes were assessed with Spearman’s correlation tests. The p-values of between-group comparisons and correlations were adjusted for multiple testing by using the False Discovery Rate method. Adjusted p-values (q-values) <0.1 were considered significant, hence controlling the rate of false positives among all positive findings at the 10% level. An alternative analysis for between-group comparisons was based on linear models after log (base 2) transformation of the data and included maternal body mass index (BMI) and nulliparity as covariates in the models. Statistical analyses were performed by using the R statistical language version 4.1.2 and IBM SPSS version 19.0 (IBM Corporation, Armonk, NY, USA).

Results

Demographics and clinical characteristics

The demographics and clinical characteristics of the study groups are displayed in Table 2. Women with preeclampsia were more likely to be nulliparous and to have a higher BMI than patients with an uncomplicated pregnancy (p<0.05). Moreover, patients with preeclampsia at term with a normal angiogenic profile were more likely to be obese than women with an uncomplicated pregnancy (Table 2).

Table 2.

Demographics and clinical characteristics of the study population

Characteristic	Uncomplicated Pregnancies (n=213)		Preeclampsia
			With a normal angiogenic profile (n=55)		With an abnormal angiogenic profile (n=41)
Maternal age, (y)	23	(20–26)	23	(19–27)	21	(19–30)
African American	196	92%	52	94.5%	39	95.1%
Nulliparity	71	33%	29	52.7%*	21	51.2%*
Tobacco use	32	15%	14	25.5%	9	22%
Pre-pregnancy body mass index (kg/m2)	26.3	(22.6–31.3)	31.3	(24.7–39.4)*	28.7	(24.5–37.9)*
≥30 kg/m2	67	31.5%	29/54	53.7%*	14/39	35.9%
≥35 kg/m2	36	16.9%	19/54	35.2%*	11/39	28.2%
Gestational age at venipuncture (weeks)	39.1	(38.4–39.7)	38.9	(37.9–39.7)	38.1	(37.6–39.0)
Gestational age at delivery (weeks)	40.1	(39.1–41.1)	39.0	(38.0–39.9)*	38.4	(37.8–39.1)*
Neonatal birthweight (grams)	3375	(3165–3655)	3235	(2977–3580)	3070	(2520–3275)*
Birthweight percentile	44.8	(31.8–65.2)	43.6	(22.4–61.2)	29.4	(5.1–51.7)*

Values expressed as median (Interquartile range) or number (%)

^*:p < 0.05 compared to uncomplicated pregnancy

Kg: kilogram; m: meter

Clinical characteristics, severity features, and maternal outcomes of women diagnosed with preeclampsia at term are displayed in Table 3. Women with an abnormal angiogenic profile had significantly higher median systolic and diastolic blood pressure values (p=0.01 and p=0.003, respectively), a higher frequency of severe hypertension (56% vs. 27%; p=0.004), and a higher frequency of SGA newborns (39% vs. 11%; p=0.001) as well as placental lesions of maternal vascular malperfusion (48% vs. 21%; p=0.008) than those with a normal angiogenic profile (Table 3).

Table 3.

Clinical characteristics, severity features, and adverse maternal outcomes of women with preeclampsia

	Preeclampsia with a normal angiogenic profile (n=55)	Preeclampsia with an abnormal angiogenic profile (n=41)	p
Pre-gestational diabetes	0	2 (4.9%)	0.2
Gestational diabetes	7 (12.7%)	4 (9.8%)	0.8
Chronic hypertension	17 (30.9%)	9 (22%)	0.3
Preeclampsia with severe features	29 (52.7%)	30 (73.2%)	0.04*
Highest systolic blood pressure (mmHg)	156 (150–160)	160 (154–171)	0.01*
≥ 160 mmHg	15 (27.3%)	23 (56.1%)	0.004*
Highest diastolic blood pressure (mmHg)	92 (86–98)	97 (91–104)	0.003*
≥ 110 mmHg	--	--	--
Composite adverse maternal outcome	2 (3.6%)	4 (9.8%)	0.4
Highest serum AST (IU/mL)	24 (19–30); n=53	27 (20–32); n=41	0.1
≥ 40 IU/mL	5 (9.4%)	6 (14.6%)	0.5
≥ 70 IU/mL	--	1 (2.4%)	0.4
Highest serum ALT (IU/mL)	17 (11–25); n=44	17 (12–26); n=28	0.9
≥ 40 IU/mL	1 (2.3%)	3 (10.7%)	0.3
≥ 70 IU/mL	--	--	--
Lowest platelets count (×109/L)	204 (170–255)	214 (161–253)	0.8
< 100 ×109/L	1 (1.8%)	3 (7.3%)	0.3
Serum creatinine > 1.1 mg/mL	1 (1.9%); n=53	1 (2.8%); n=36	1
Birthweight <10th percentile	6 (10.9%)	16 (39%)	0.001*
Birthweight <5th percentile	3 (5.5%)	10 (24.4%)	<0.007*
Birthweight >90th percentile	3 (5.5%)	1 (2.4%)	0.6
Composite neonatal complications	--	2 (4.9%)	0.2
Maternal vascular malperfusion lesions	11 (21.2%); n=52	19 (47.5%); n=40	0.008*
Angiogenic index-1 (PlGF/sFlt-1 MoM )	0.6 (0.3–1.4)	0.07 (0.04–0.12)	<0.001*

ALT: alanine transaminase; AST: aspartate transaminase; GA: gestational age; HELLP: hemolysis, elevated liver enzymes, low platelet count syndrome; MoM: multiple of the median; PlGF: placental growth factor; sFlt-1; soluble fms-like tyrosine kinase-1

Values are expressed as number (percentage), or median (interquartile range)

^*:p < 0.05.

Cytokine profiles of term preeclamptic women with angiogenic abnormalities

Patients with preeclampsia at term and an abnormal angiogenic profile had significantly higher median plasma concentrations of 5 cytokines [interleukin (IL)-6, IL-8, IL-12/IL-23p40, IL-15, and IL-16] and 7 chemokines [eotaxin, eotaxin-3, interferon-inducible protein (IP)-10, monocyte chemotactice protein (MCP)-4, macrophage inflammatory protein (MIP)-1β, macrophage-derived chemokine (MDC), and thymus and activation-regulated chemokine (TARC)] than women with an uncomplicated pregnancy (Table 4, Figures 1–3). Moreover, these patients had significantly higher median plasma concentrations of 2 cytokines (IL-12/IL-23p40 and IL-15), 3 chemokines (eotaxin, eotaxin-3, and IP-10) [q <0.1; Table 5, Figures 1–3], and a significantly lower concentration of the growth factor granulocyte-macrophage colony-stimulating factor (GM-CSF) [q=0.018] than those with term preeclampsia and a normal angiogenic profile. Only significant results after adjustment for BMI and nulliparity and accounting for multiple testing were presented above.

Table 4.

Plasma concentrations (pg/mL) of cytokines and chemokines in women with uncomplicated pregnancies and in those with preeclampsia with an abnormal angiogenic profile

Analytes	Uncomplicated pregnancies	Preeclampsia with an abnormal angiogenic profile	Fold changes	Direction	p	q
MSD V-PLEX Pro-inflammatory Panel 1
IFN-γ	3 (2.2–4.7)	3.2 (2–5.1)	1.1	↑	0.702	0.783
TNF-α	1.4 (1.2–1.7)	1.6 (1.3–2)	1.1	↑	0.12	0.171
IL-1β	0.1 (0.1–0.1)	0.1 (0.1–0.1)	1.03	↓	0.098	0.158
IL-2	0.2 (0.1–0.2)	0.2 (0.1–0.2)	1.12	↓	0.491	0.593
IL-4	0 (0–0.1)	0 (0–0.1)	1.27	↓	0.292	0.385
IL-6 *	0.7 (0.5–0.9)	1.1 (0.7–1.7)	1.56	↑	< 0.001	< 0.001
IL-8 *	2.4 (1.7–3.4)	2.9 (2.2–4.5)	1.32	↑	0.009	0.024
IL-10	0.2 (0.2–0.3)	0.3 (0.2–0.4)	1.17	↑	0.024	0.051
IL-12p70	0.1 (0.1–0.2)	0.1 (0.1–0.2)	1.0	↑	0.972	0.981
IL-13	0.8 (0.8–0.8)	0.8 (0.8–0.8)	1.05	↓	0.533	0.619
MSD V-PLEX Cytokines Panel 1
TNF-β	0.1 (0.1–0.1)	0.1 (0.1–0.1)	1.14	↓	0.124	0.171
IL-1α	1.5 (0.9–2.6)	1 (0.6–2.2)	1.13	↓	0.11	0.167
IL-5	0.2 (0.1–0.2)	0.2 (0.1–0.2)	1.14	↓	0.956	0.981
IL-7	5.6 (3.9–8.5)	6.7 (4.1–9.8)	1.2	↑	0.092	0.157
IL-12/IL-23p40 *	59.4 (42.5–73.6)	83.3 (61.1–108.7)	1.58	↑	< 0.001	< 0.001
IL-15 *	1.7 (1.4–2)	2 (1.6–2.7)	1.23	↑	0.001	0.002
IL-16 *	79.9 (64.9–98.5)	107.4 (85.5–128.5)	1.27	↑	< 0.001	< 0.001
IL-17A	0.7 (0.5–1)	0.6 (0.5–0.9)	1.09	↓	0.444	0.559
VEGF	7.3 (4.8–11.4)	4.8 (3–8.6)	1.41	↓	0.016	0.037
GM-CSF	0.2 (0.1–0.3)	0.2 (0.1–0.3)	1.12	↓	0.981	0.981
MSD V-PLEX Chemokines Panel 1
Eotaxin *	28.7 (22.6–35.3)	33 (28.1–43.9)	1.24	↑	0.001	0.004
Eotaxin-3 *	4.4 (1.9–6.7)	6.3 (4.6–10.7)	1.66	↑	0.001	0.002
IP-10 *	151.5 (107.5–225.6)	190.8 (155.1–293.3)	1.33	↑	< 0.001	0.002
MCP-1	53.8 (43.9–67.1)	59.5 (50.8–81.4)	1.12	↑	0.046	0.084
MCP-4 *	35.8 (25.1–52.5)	54.9 (41–82.7)	1.58	↑	< 0.001	< 0.001
MDC *	671.2 (536.6–794.9)	760.2 (590.3–1016.1)	1.17	↑	0.016	0.037
MIP-1α	7.5 (5.3–9.5)	8.7 (6.3–12.3)	1.12	↑	0.031	0.06
MIP-1β *	53.3 (43.5–69.6)	70.9 (55.9–91.7)	1.28	↑	< 0.001	0.001
TARC *	57.7 (36.1–80.4)	79.2 (61.5–120.2)	1.58	↑	< 0.001	0.001

Value presented as median (interquartile range); p: nominal p-value; q: adjustment for false discovery rate

Dark gray: p < 0.05 and q < 0.1; Light gray: p < 0.05 and q > 0.1

^*:Significance remained after adjustment for body mass index and nulliparity (p<0.05 and q<0.1)

GM-CSF: granulocyte macrophage colony-stimulating factor; IFN: interferon gamma; IL: interleukin; IP: interferon gamma-induced protein; MCP: monocyte chemotactic protein; MDC: macrophage-derived chemokine; MIP: macrophage inflammatory protein; MSD: Mesoscale discovery; TARC: thymus and activation-regulated chemokine; TNF: tumor necrosis factor; VEGF: vascular endothelial growth factor

Figure 1.

Violin plot for plasma concentrations (pg/mL) of cytokines from V–PLEX pro-inflammatory panel 1 in the controls, term preeclampsia with and without an abnormal angiogenic profile. Cytokines highlighted are IL-1β, IL-6, IL-8, and IL-10. Dash line displays median. IL: interleukin

Figure 3.

Violin plot for plasma concentrations (pg/mL) of V-PLEX chemokines panel 1 in the controls, term preeclampsia with and without an abnormal angiogenic profile. Chemokines include eotaxin, eotaxin-3, IP-10, MCP-1, MCP-4, MDC, MIP-1α, MIP-1β and TARC. Dash line displays median. IP: interferon gamma-induced protein; MCP: monocyte chemotactic protein; MDC: macrophage-derived chemokine; MIP: macrophage inflammatory protein; TARC: thymus and activation-regulated chemokine

Table 5.

Plasma concentrations (pg/mL) of cytokines and chemokines in preeclampsia at term with and without an abnormal angiogenic profile

Analytes	Preeclampsia without an abnormal angiogenic profile	Preeclampsia with an abnormal angiogenic profile	Fold changes	Direction	p	q
MSD V-PLEX Pro-inflammatory Panel 1
IFN-γ	3.1 (2.3–4)	3.2 (2–5.1)	1.24	↑	0.579	0.799
TNF-α	1.5 (1.2–1.9)	1.6 (1.3–2)	1.12	↑	0.463	0.747
IL-1β	0.1 (0.1–0.1)	0.1 (0.1–0.1)	1.23	↓	0.767	0.903
IL-2	0.2 (0.1–0.2)	0.2 (0.1–0.2)	1.02	↑	0.964	0.964
IL-4	0 (0–0.1)	0 (0–0.1)	1.3	↓	0.335	0.648
IL-6	0.9 (0.5–1.3)	1.1 (0.7–1.7)	1.32	↑	0.042	0.132
IL-8	2.8 (1.7–3.9)	2.9 (2.2–4.5)	1.29	↑	0.1	0.242
IL-10	0.2 (0.1–0.3)	0.3 (0.2–0.4)	1.24	↑	0.021	0.088
IL-12p70	0.1 (0.1–0.2)	0.1 (0.1–0.2)	1.04	↑	0.929	0.962
IL-13	0.8 (0.8–0.8)	0.8 (0.8–0.8)	1.11	↓	0.297	0.614
MSD V-PLEX Cytokines Panel 1
TNF-β	0.1 (0.1–0.1)	0.1 (0.1–0.1)	1.01	↓	0.767	0.903
IL-1α	0.9 (0.6–1.6)	1 (0.6–2.2)	1.17	↑	0.778	0.903
IL-5	0.2 (0.1–0.2)	0.2 (0.1–0.2)	1.22	↓	0.561	0.799
IL-7	7.1 (4.7–10.2)	6.7 (4.1–9.8)	1.04	↑	0.894	0.96
IL-12/IL-23p40 *	58.1 (41.8–82.9)	83.3 (61.1–108.7)	1.59	↑	0.001	0.018
IL-15 *	1.6 (1.3–2)	2 (1.6–2.7)	1.31	↑	< 0.001	0.014
IL-16	92 (69.7–117.1)	107.4 (85.5–128.5)	1.18	↑	0.063	0.166
IL-17A	0.5 (0.5–0.9)	0.6 (0.5–0.9)	1.03	↑	0.439	0.747
VEGF	6.2 (3.7–7.5)	4.8 (3–8.6)	1.16	↓	0.566	0.799
GM-CSF *	0.3 (0.2–0.3)	0.2 (0.1–0.3)	1.48	↓	0.004	0.018
MSD V-PLEX Chemokines Panel 1
Eotaxin *	25.3 (19.1–36.2)	33 (28.1–43.9)	1.4	↑	0.002	0.018
Eotaxin-3 *	3.7 (1.3–6.4)	6.3 (4.6–10.7)	1.81	↑	0.003	0.018
IP-10 *	154.6 (111.8–234.7)	190.8 (155.1–293.3)	1.61	↑	0.004	0.018
MCP-1	50.9 (42.1–71.4)	59.5 (50.8–81.4)	1.26	↑	0.046	0.132
MCP-4	61.1 (40.6–73.1)	54.9 (41–82.7)	1.09	↑	0.894	0.96
MDC	721.8 (592.7–901.1)	760.2 (590.3–1016.1)	1.19	↑	0.378	0.685
MIP-1α	8.9 (5–12.2)	8.7 (6.3–12.3)	1.16	↓	0.633	0.834
MIP-1β	63.4 (48.6–85.6)	70.9 (55.9–91.7)	1.2	↑	0.122	0.271
TARC	57.9 (40.9–119.5)	79.2 (61.5–120.2)	1.42	↑	0.039	0.132

Value presented as median (interquartile range); p: nominal p-value; q: adjustment for false discovery rate

Dark gray: p < 0.05 and q < 0.1; Light gray: p < 0.05 and q > 0.1

^*:Significance remained after adjustment for body mass index and nulliparity (p<0.05 and q<0.1)

GM-CSF: granulocyte macrophage colony-stimulating factor; IFN: interferon gamma; IL: interleukin; IP: interferon gamma-induced protein; MCP: monocyte chemotactic protein; MDC: macrophage-derived chemokine; MIP: macrophage inflammatory protein; MSD: Mesoscale discovery; TARC: thymus and activation-regulated chemokine; TNF: tumor necrosis factor; VEGF: vascular endothelial growth factor

Cytokine profiles of term preeclamptic women without angiogenic abnormalities

When compared to uncomplicated pregnancies, after adjusting for multiple testing, women with preeclampsia at term and a normal angiogenic profile had significantly higher median plasma concentrations of MCP-4 and GM-CSF and a lower plasma concentration of IL-1α than women with an uncomplicated pregnancy (all q<0.1, Table 6, Figures 1–3). However, after adjustment for BMI and nulliparity, only a higher plasma concentration of MCP-4 remained significant (q<0.1).

Table 6.

Plasma concentrations (pg/mL) of cytokines and chemokines in women with uncomplicated pregnancies and in those with preeclampsia with a normal angiogenic profile

Analytes	Uncomplicated pregnancies	Preeclampsia with a normal angiogenic profile	Fold changes	Direction	p	Q
MSD V-PLEX Pro-inflammatory Panel 1
IFN-γ	3 (2.2–4.7)	3.1 (2.3–4)	1.12	↓	0.804	0.897
TNF-α	1.4 (1.2–1.7)	1.5 (1.2–1.9)	1.02	↓	0.471	0.625
IL-1β	0.1 (0.1–0.1)	0.1 (0.1–0.1)	1.19	↑	0.028	0.124
IL-2	0.2 (0.1–0.2)	0.2 (0.1–0.2)	1.15	↓	0.515	0.625
IL-4	0 (0–0.1)	0 (0–0.1)	1.02	↑	0.904	0.937
IL-6	0.7 (0.5–0.9)	0.9 (0.5–1.3)	1.18	↑	0.03	0.124
IL-8	2.4 (1.7–3.4)	2.8 (1.7–3.9)	1.02	↑	0.495	0.625
IL-10	0.2 (0.2–0.3)	0.2 (0.1–0.3)	1.06	↓	0.248	0.423
IL-12p70	0.1 (0.1–0.2)	0.1 (0.1–0.2)	1.04	↓	0.736	0.854
IL-13	0.8 (0.8–0.8)	0.8 (0.8–0.8)	1.05	↑	0.224	0.423
MSD V-PLEX Cytokines Panel 1
TNF-β	0.1 (0.1–0.1)	0.1 (0.1–0.1)	1.14	↓	0.236	0.423
IL-1α	1.5 (0.9–2.6)	0.9 (0.6–1.6)	1.33	↓	0.008	0.073
IL-5	0.2 (0.1–0.2)	0.2 (0.1–0.2)	1.08	↑	0.35	0.535
IL-7	5.6 (3.9–8.5)	7.1 (4.7–10.2)	1.16	↑	0.074	0.216
IL-12/IL-23p40	59.4 (42.5–73.6)	58.1 (41.8–82.9)	1.0	↓	0.904	0.937
IL-15	1.7 (1.4–2)	1.6 (1.3–2)	1.07	↓	0.307	0.494
IL-16	79.9 (64.9–98.5)	92 (69.7–117.1)	1.08	↑	0.029	0.124
IL-17A	0.7 (0.5–1)	0.5 (0.5–0.9)	1.12	↓	0.149	0.359
VEGF	7.3 (4.8–11.4)	6.2 (3.7–7.5)	1.22	↓	0.018	0.124
GM-CSF	0.2 (0.1–0.3)	0.3 (0.2–0.3)	1.32	↑	0.001	0.017
MSD V-PLEX Chemokines Panel 1
Eotaxin	28.7 (22.6–35.3)	25.3 (19.1–36.2)	1.13	↓	0.179	0.372
Eotaxin-3	4.4 (1.9–6.7)	3.7 (1.3–6.4)	1.09	↓	0.475	0.625
IP-10	151.5 (107.5–225.6)	154.6 (111.8–234.7)	1.21	↓	0.969	0.969
MCP-1	53.8 (43.9–67.1)	50.9 (42.1–71.4)	1.12	↓	0.517	0.625
MCP-4 *	35.8 (25.1–52.5)	61.1 (40.6–73.1)	1.45	↑	< 0.001	< 0.001
MDC	671.2 (536.6–794.9)	721.8 (592.7–901.1)	1.02	↓	0.053	0.17
MIP-1α	7.5 (5.3–9.5)	8.9 (5–12.2)	1.29	↑	0.125	0.328
MIP-1β	53.3 (43.5–69.6)	63.4 (48.6–85.6)	1.07	↑	0.041	0.149
TARC	57.7 (36.1–80.4)	57.9 (40.9–119.5)	1.11	↑	0.171	0.372

Value presented as median (interquartile range); p: nominal p-value; q: adjustment for false discovery rate

Dark gray: p < 0.05 and q < 0.1; Light gray: p < 0.05 and q > 0.1

^*:Significance remained after adjustment for body mass index and nulliparity (p<0.05 and q<0.1)

GM-CSF: granulocyte macrophage colony-stimulating factor; IFN: interferon gamma; IL: interleukin; IP: interferon gamma-induced protein; MCP: monocyte chemotactic protein; MDC: macrophage-derived chemokine; MIP: macrophage inflammatory protein; MSD: Mesoscale discovery; TARC: thymus and activation-regulated chemokine; TNF: tumor necrosis factor; VEGF: vascular endothelial growth factor

Correlation between degree of anti-angiogenic state and severity of intravascular inflammation

In the whole study population (n=309), there was a positive correlation between maternal plasma sFlt-1 MoM values and most of the cytokine and chemokine concentrations [IL-6, IL-8, tumor necrosis factor (TNF)-α, IL-10, IL-7, IL-12/IL-23p40, IL-15, IL-16, eotaxin, eotaxin-3, IP-10, MCP-1, MCP-4, MIP-1α, MIP-1β, MDC, TARC; all q<0.1; Table 7]. By contrast, there was a negative correlation between plasma PlGF MoM values and maternal plasma concentrations of the same subset of cytokines and chemokines (all q<0.1; Table 7), except for IL-10 and eotaxin.

Table 7.

Correlations between maternal plasma concentrations of cytokines and chemokines and the severity of angiogenic/anti-angiogenic profiles or blood pressure

Population	Correlation	Analyte	Spearman r	Spearman r2	p	q
Cases and controls	sFlt-1 MoM	IP_10	0.46	0.21	<0.001	<0.001
Cases and controls	sFlt-1 MoM	Eotaxin	0.31	0.10	<0.001	<0.001
Cases and controls	sFlt-1 MoM	IL_6	0.31	0.10	<0.001	<0.001
Cases and controls	sFlt-1 MoM	Eotaxin_3	0.28	0.08	<0.001	<0.001
Cases and controls	sFlt-1 MoM	TNF_alpha	0.27	0.07	<0.001	<0.001
Cases and controls	sFlt-1 MoM	IL_8	0.27	0.07	<0.001	<0.001
Cases and controls	sFlt-1 MoM	IL_15	0.26	0.07	<0.001	<0.001
Cases and controls	sFlt-1 MoM	IL_16	0.25	0.06	<0.001	<0.001
Cases and controls	sFlt-1 MoM	TARC	0.24	0.06	<0.001	<0.001
Cases and controls	sFlt-1 MoM	IL12/IL23p40	0.23	0.05	<0.001	<0.001
Cases and controls	sFlt-1 MoM	MCP_4	0.22	0.05	<0.001	<0.001
Cases and controls	sFlt-1 MoM	MCP_1	0.21	0.05	<0.001	<0.001
Cases and controls	sFlt-1 MoM	MIP_1b	0.20	0.04	<0.001	0.001
Cases and controls	sFlt-1 MoM	IL_10	0.18	0.03	0.002	0.003
Cases and controls	sFlt-1 MoM	MDC	0.17	0.03	0.003	0.005
Cases and controls	sFlt-1 MoM	MIP_1a	0.15	0.02	0.010	0.018
Cases and controls	sFlt-1 MoM	IL_7	0.14	0.02	0.012	0.021
Cases and controls	PlGF MoM	IL_6	−0.35	0.12	<0.001	<0.001
Cases and controls	PlGF MoM	MCP_4	−0.35	0.12	<0.001	<0.001
Cases and controls	PlGF MoM	IP_10	−0.27	0.07	<0.001	<0.001
Cases and controls	PlGF MoM	Eotaxin_3	−0.26	0.07	<0.001	<0.001
Cases and controls	PlGF MoM	TARC	−0.25	0.06	<0.001	<0.001
Cases and controls	PlGF MoM	IL_16	−0.24	0.06	<0.001	<0.001
Cases and controls	PlGF MoM	MDC	−0.22	0.05	<0.001	<0.001
Cases and controls	PlGF MoM	IL_7	−0.20	0.04	<0.001	0.001
Cases and controls	PlGF MoM	MIP_1b	−0.20	0.04	<0.001	0.001
Cases and controls	PlGF MoM	IL_8	−0.20	0.04	0.001	0.002
Cases and controls	PlGF MoM	Eotaxin	−0.18	0.03	0.002	0.005
Cases and controls	PlGF MoM	MCP_1	−0.17	0.03	0.002	0.006
Cases and controls	PlGF MoM	IL_15	−0.16	0.03	0.006	0.013
Cases and controls	PlGF MoM	IL12/IL23p40	−0.15	0.02	0.008	0.016
Cases and controls	PlGF MoM	MIP_1a	−0.14	0.02	0.015	0.030
Cases and controls	PlGF MoM	TNF_alpha	−0.12	0.02	0.032	0.058
Cases	systolic BP	IL_6	0.39	0.15	<0.001	0.003
Cases	diastolic BP	Eotaxin	0.35	0.12	<0.001	0.013
Cases	diastolic BP	IL_17A	0.30	0.09	0.003	0.024
Cases	diastolic BP	IL12/IL23p40	0.30	0.09	0.003	0.024
Cases	diastolic BP	IL_8	0.30	0.09	0.003	0.024
Cases	diastolic BP	IP_10	0.29	0.08	0.005	0.027
Cases	diastolic BP	IL_6	0.25	0.06	0.015	0.074
Cases	diastolic BP	TNF_alpha	0.23	0.05	0.024	0.099

The subset of patients involved in each analysis is displayed in the first column (Population). The correlations considered are the severity of the anti-angiogenic profiles (multiple of the median values of soluble fms-like tyrosine kinase-1 or sFlt-1 MoM, the severity of angiogenic profiles (multiples of the median value of placental growth factor, PlGF MoM), and systolic or diastolic blood pressure (BP). For each correlation and each cytokine/chemokine, Spearman’s correlation coefficient (r) and r² values are shown together with the p-values and adjusted p-values (q).

Correlation between blood pressure and magnitude of intravascular inflammation

Among patients with preeclampsia (n=96), systolic blood pressure values were correlated with plasma IL-6 concentrations (q<0.1, Spearman’s r = 0.4; Table 7). By contrast, diastolic blood pressure values were correlated with plasma TNF-α, IL-6, IL-8, IL-10, IL-12/IL-23p40, IL-17A, and eotaxin concentrations (q<0.1, Table 7). Correlation analysis in each subgroup of preeclampsia yielded results significant at a nominal level (p<0.05) but did not pass adjustment for multiple testing.

Discussion

Principal findings of this study

1. Women with preeclampsia at term and an abnormal angiogenic profile show evidence of the greatest intravascular inflammation among the study groups; 2) only a mild degree of intravascular inflammation is present in women with preeclampsia at term with a normal angiogenic profile; and 3) there is a significant correlation between the severity of the anti-angiogenic state as well as blood pressure level and the severity of intravascular inflammation.

Results in the context of what is known

We recently reported that a possible explanation for the low performance of angiogenic biomarkers in women with preeclampsia at term is that only about half of women with this condition have an abnormal angiogenic profile, compared to 90% to 98% of those with early preeclampsia.⁵⁵ Based on these observations, we defined 2 clusters of preeclampsia at term, according to the presence or absence of abnormalities in the angiogenic and anti-angiogenic factors, with different clinical characteristics and outcomes.⁵⁵

An exaggerated intravascular inflammation is one of the features of women with early preeclampsia, whereas the role of inflammation in the pathophysiology of late preeclampsia is still under investigation.^72–75 According to a recent systematic review on this topic, the majority of the available studies reported that women with preeclampsia have higher blood concentrations of cytokines (IL-6, IL-8, TNF-α, and C-reactive protein) than normal pregnant women.⁷⁶ The behavior of other cytokines, such as IL-10, was inconsistently reported among studies.^76–78 Herein, the study was undertaken to examine whether these 2 subgroups of preeclampsia at term also have differences in their profiles of intravascular inflammation.

Research implications

Term preeclampsia with angiogenic abnormalities: signature of intense intravascular inflammation

Women with preeclampsia at term and an anti-angiogenic state have a greater degree of intravascular inflammation than those with an uncomplicated pregnancy and those with preeclampsia and a normal angiogenic profile. Differences in cytokine and chemokine profiles between the 2 subgroups of preeclampsia at term suggest that they may derive from 2 different pathophysiologic processes. Preeclampsia at term with an abnormal angiogenic profile is characterized by an anti-angiogenic state coupled with excessive intravascular inflammation, whereas preeclampsia at term without angiogenic abnormalities is not.

The anti-angiogenic state of women with preeclampsia has been proposed to derive from the placenta: in early-onset preeclampsia, it is a consequence of an incomplete remodeling of the spiral arteries and a defective deep placentation,^{74, 79–84} whereas, in a subset of women with preeclampsia at term, it occurs following “syncytiotrophoblast stress” related to physical constraints on placental growth and cytotrophoblast senescence.⁷ Indeed, in our previous study of 258 women with preeclampsia at term, patients with angiogenic abnormalities showed evidence of placental involvement such as a higher frequency of lesions consistent with maternal vascular malperfusion (42%) and of delivery of SGA neonates (38%) than those with a normal angiogenic profile.⁵⁵

According to previous observations, an anti-angiogenic state can induce an inflammatory response.^{85, 86} However, the precise sources of the circulating inflammatory cytokines and chemokines in women with preeclampsia at term and an anti-angiogenic profile remain to be elucidated. Soluble Flt-1 can stimulate placental explants to release TNF-α and IL-10.⁸⁶ Moreover, both sFlt-1 and soluble endoglin administration to pregnant animals can induce an increase of circulating CD4+ T cells, TNF-α, and IL-17.⁸⁵ Alternatively, the placenta can release pro-inflammatory cytokines under conditions of reduced uterine perfusion⁸⁷ and hypoxia,⁸⁸ although these mechanisms are more frequently described in early than in late preeclampsia.^{79, 89} Another potential source of inflammatory cytokines in the context of preeclampsia with angiogenic/anti-angiogenic abnormalities is represented by peripheral blood mononuclear cells. These cells, upon stimulation by syncytiotrophoblast-derived extracellular vesicles, isolated from placental explants of women with early preeclampsia, can release IL-6, TNF-α, IL-8, IL-10, IL-17, granulocyte colony-stimulating factor, and MIP-1α as well as MIP-1β.^{90 91} However, this response has not been replicated when exposing macrophages to placental explants of women with late-onset preeclampsia.⁹² Whether the patients diagnosed with late-onset disease in this study had an abnormal angiogenic profile is unknown. The existence of the 2 subgroups of preeclampsia at term should be taken into account when designing future studies on this condition.

Several lines of evidence suggest that intravascular inflammation in preeclampsia involves interaction between circulating monocytes and placenta-derived microparticles, the numbers of which are greatly increased in women with preeclampsia,⁹³ or with alarmins that released upon cellular damage, such as hyaluronan and heat shock protein −70.⁹⁴ Such activation potentially drives the production of inflammatory mediators, such as IL-8, IL-6, IL-1β, TNF, and IL-12,^{75, 93–98} by circulating monocytes, together with downregulation of IL-10 production.⁹⁶ Thus, the upregulated cytokine/chemokine profile identified in the current study supports the systemic activation of monocytes with an M1-like phenotype as characteristic of term preeclampsia with an anti-angiogenic profile, which is potentially driven by activating signals originating from the placenta.

Clinical implications

Clinical severity, magnitude of inflammation, and degree of an anti-angiogenic state

In the current study of women with preeclampsia, we observed a significant correlation between blood pressure levels and maternal plasma concentrations of a subset of cytokines: IL-6, TNF-α, and IL-8. Pro-inflammatory mediators and anti-angiogenic factors may be responsible for endothelial cell dysfunction, thus contributing to the loss of endothelia-dependent vasodilatation and proteinuria.^{29, 99–101} Consistent with this hypothesis, experimental evidence has shown that the administration of IL-6,¹⁰² TNF-α,⁸⁷ or IL-17^{103, 104} can elicit some features of preeclampsia in animal models, including hypertension as well as a reduction in the glomerular filtration rate.

IP-10 is one of the chemokines found to be significantly higher in women with preeclampsia at term and an abnormal angiogenic profile; it has pro-inflammatory and anti-angiogenic properties and has been proposed to be a link between inflammation and an anti-angiogenic state in preeclampsia.¹⁰⁵ Moreover, IP-10 has been implicated in allograft rejection ^106–108 and atherosclerosis,^{109, 110} 2 mechanisms of diseases related to preeclampsia. Of note, the production of this chemokine is increased by IL-12/IL-23p40, which was also higher in the maternal plasma of this subgroup of women.^111–113

We recently reported that women with preeclampsia at term and an anti-angiogenic profile, similar to those with preterm preeclampsia,¹¹⁴ are at greater risk of adverse maternal-neonatal outcomes and more likely to deliver an SGA neonate than those with a normal angiogenic profile.⁵⁵ Given the association between an anti-angiogenic state and exaggerated intravascular inflammation, it is tempting to suggest that a blockade of pro-inflammatory cytokines^115–118 or the administration of anti-inflammatory cytokines,^{119, 120} coupled with a tight control of vascular endothelial growth factors, could optimize endothelial cell functions¹²¹ and benefit a subset of women with preeclampsia and an anti-angiogenic profile.

Term preeclampsia without angiogenic abnormalities: an alternative mechanism with a milder inflammatory state?

Our findings suggest that women with preeclampsia at term and a normal angiogenic profile have a mild degree of intravascular inflammation. The mechanism of disease in this cluster of women with preeclampsia at term requires further investigation. The prevalence of SGA neonates and placental lesions of maternal vascular malperfusion in this subgroup is similar to what is observed in uncomplicated pregnancies,¹²² suggesting the possibility of little or no involvement of the placenta. Alternatively, the placenta in women with metabolic syndrome or high insulin resistance can release substances (eg, inositol phosphoglycans), which can contribute to endothelial cell damage.¹²³ The findings that more than half of women in this group were obese and that about one-third had underlying chronic hypertension are consistent with our previous observations.⁵⁵ Of interest, the adipose tissue is one of the potential sources of MCP-4, previously proposed to be a link between obesity, chronic low-grade inflammation, and atherosclerosis in non-pregnant individuals.^124–129 However, the plasma MCP-4 concentration was significantly higher in this subgroup of term preeclampsia than an uncomplicated pregnancy even after adjusting for BMI. Given the high prevalence of obesity, metabolic changes such as insulin resistance or perturbations in adipokines (i.e., adiponectin, leptin, visfatin, resistin, retinol-binding proteins),^130–134 released from either the adipose tissues or the placenta, may be involved in the disease process of this subgroup of term preeclampsia. Indeed, although the patients in this group have a lower risk of maternal and neonatal complications than those with an anti-angiogenic profile, they can still encounter major adverse outcomes such as pulmonary edema and fetal death.⁵⁵ These complications could result from the metabolic syndrome also referred to as ”maternal preeclampsia.”²¹ In a recent case-series of women with eclampsia, an abnormal angiogenic profile was present in 95% of women with preterm eclampsia, while one-third of women diagnosed with eclampsia at term had a normal angiogenic profile in maternal blood.¹³⁵

Strengths and limitations of the study

Our findings demonstrate differences in the profiles of cytokines and chemokines between women with preeclampsia at term with and without abnormalities in angiogenic/anti-angiogenic factors. The inclusion of consecutive cases of preeclampsia at term, enrolled in a longitudinal case-cohort study, reduced the probability of selection bias. Moreover, the definition of an abnormal angiogenic profile was derived on the distribution of MoM values in our population, allowing replication in other populations.

The low prevalence of cases with adverse maternal and neonatal outcomes in the groups of preeclampsia at term as well as the exclusion of patients with severe medical illness at enrollment of the original longitudinal cohort is a limitation. Collectively, this may explain why in the current study we observed a higher prevalence of adverse maternal (9.8% vs. 3.6%) and neonatal (4.9% vs. 0) outcomes in women with an abnormal angiogenic profile than in those with a normal profile, although the difference did not reach statistical significance. Future studies that include other biomarkers (i.e. adipokines, cardiovascular markers, etc.) and more patients with adverse outcomes from both subgroups may improve our understanding of the intravascular inflammation profiles of term preeclampsia with a normal angiogenic profile. Since our study specifically examined patients with term preeclampsia, these findings may not be directly applicable to those with early-onset disease.

Conclusions

We provided evidence that women with preeclampsia at term have a greater degree of intravascular inflammation when compared to normal pregnant women and that this feature is more prominent in the presence of an anti-angiogenic state. The intensity of these processes may contribute to the clinical severity of the syndrome. Our observations suggest that accurate sub-classification of patients with preeclampsia is important in experimental⁹² and in clinical research.^136–138 Moreover, awareness that preeclampsia is a heterogeneous condition with multiple pathophysiologic mechanisms of disease is required to advance biomarker discovery and to develop and implement personalized preventive and therapeutic interventions.

Figure 2.

Violin plot for plasma concentrations (pg/mL) of V-PLEX cytokines panel 1 in the controls, term preeclampsia with and without an abnormal angiogenic profile. Cytokines highlighted are IL-1α, IL-12/IL-23p40, IL-15, IL-16, VEGF and GM-CSF. Dash line displays median. Dash line displays median. IL: interleukin, VEGF: vascular endothelial growth factor; GM-CSF: granulocyte-macrophage colony-stimulating factor

Condensation:

Term preeclampsia can be classified into 2 clusters. One is characterized by an anti-angiogenic state with an excessive inflammatory response while the other has neither of these features.

AJOG at a Glance.

A. Why was this study conducted?

Recent evidence indicates that about one-half of women diagnosed with preeclampsia at term have an abnormal angiogenic profile and that this condition correlates with a higher risk for adverse maternal and neonatal outcomes. This study was undertaken to examine the profiles of intravascular inflammation in women with preeclampsia at term with and without an abnormal angiogenic profile.

B. What are the key findings?

Women with preeclampsia at term and an anti-angiogenic profile had higher plasma concentrations of 5 cytokines and 7 chemokines than those with an uncomplicated pregnancy. By contrast, women with preeclampsia at term and a normal angiogenic profile, compared to uncomplicated pregnancies, only had a higher plasma concentration of monocyte chemotactic protein-4. A significant correlation between severity of the anti-angiogenic state as well as blood pressure and plasma concentrations of a subset of cytokines was observed.

C. What does this study add to what is already known?

Term preeclampsia can be classified into 2 clusters. One is characterized by an anti-angiogenic state coupled with an excessive inflammatory process while the other has neither of these features. These findings further support the heterogeneity of preeclampsia at term, explain the distinct clinical outcomes, and provide the background for personalized diagnosis, prevention, and treatment.