Preeclampsia/eclampsia: the conceptual evolution of a syndrome

Abstract

Preeclampsia, one of the most enigmatic complications of pregnancy, is considered a pregnancy-specific disorder caused by the placenta and cured only by delivery. This article traces the condition from its origin—once thought to be a disease of the central nervous system, recognized by the occurrence of seizures (ie eclampsia)—to the present time when preeclampsia is conceptualized primarily as a vascular disorder. We review the epidemiologic data that led to the recommendation to use diastolic hypertension and proteinuria as diagnostic criteria, given that their combined presence was associated with an increased risk of fetal death and the birth of small-for-gestational-age neonates. However, preeclampsia is a multi-systemic disorder with protean manifestations, and the condition can be present even in the absence of hypertension and proteinuria. Toxin(s) that gain access to the maternal circulation have been proposed to mediate the clinical manifestations—hence, the term “toxemia of pregnancy,” which was used for several decades. The search for putative toxins has challenged investigators for more than a century, and a growing body of evidence suggests that products of an ischemic or a stressed placenta are responsible for the vascular changes that characterize the syndrome. The discovery that the placenta can produce anti-angiogenic factors, which regulate endothelial cell function and induce intravascular inflammation, has been a major step forward in the understanding of preeclampsia. We view the release of anti-angiogenic factors by the placenta as an adaptive response to improve uterine perfusion by modulating endothelial function and maternal blood pressure. However, this homeostatic response can become maladaptive and lead to the damage of target organs during pregnancy or the postpartum period. Early-onset preeclampsia has many features in common with atherosclerosis, while late-onset disease appears to result from a mismatch of fetal demands and maternal supply: in other words, a metabolic crisis. Preeclampsia, as it is understood today, is essentially vascular dysfunction unmasked, or caused, by pregnancy that results in a multi-systemic disorder. A subset of patients diagnosed with preeclampsia are at greater risk for the subsequent development of hypertension, ischemic heart diseases, heart failure, vascular dementia, and end-stage renal disease. However, these adverse events may be the result of a preexisting vascular pathologic process; it is not known if the occurrence of preeclampsia by itself increases the baseline risk. The understanding of preeclampsia is a healthcare priority.

Condensation:

Conceptual evolution of preeclampsia/eclampsia as a syndrome

Introduction

Preeclampsia and eclampsia are a syndrome considered unique to pregnant women; its causes, pathophysiology, prediction, management, and prevention present formidable challenges. This supplement of the American Journal of Obstetrics & Gynecology summarizes the state of the science on this disorder, and this article reviews the evolution of the concept of disease.

An enigmatic pregnancy-specific disorder

Convulsions in pregnant women: epilepsy versus eclampsia

For centuries, ancient texts from Egypt, China, India, and Europe documented the greater risk of seizures among pregnant women (Figure 1).¹ Convulsions were reported to occur more frequently during or after first pregnancies, and such episodes had poor prognostic significance. A fundamental issue was whether they represented epilepsy or a unique, specific complication of pregnancy. The term eclampsia (derived from the Greek eklampsis, meaning “a shining forth” ^{2, 3}) is attributed to François Boissier de Sauvages de Lacroix, a French physician and botanist interested in the taxonomy of disease.^{4, 5} He proposed that eclampsia differed from epilepsy because the latter was chronic and recurred throughout the lifespan, while Eclampsia parturientium did not.¹

Figure 1.

(A) The Kahun Gynaecological Papyrus. A medical text from the late Middle Kingdom (1850–1700 BC) addressing women’s health. The Papyrus was found near the modern-day Egyptian town of Lehun in 1889 by Flinders Petrie. The Kahun Gynaecological Papyrus (UC 32057) is housed at University College London, London, UK. 21367707). In this Figure, pages 1, 2, and 3 of Plate VI are shown. Modified from https://en.wikipedia.org/wiki/Kahun_Gynaecological_Papyrus (B) The Kahun Gynaecological Papyrus was translated in 1893 by F. Griffiths and published as The Petrie Papyri: Hieratic Papyri from Kahun and Gurob (Principally of the Middle Kingdom). This figure shows the translation of Prescription No XXXIII from page 3 of the Plate VI (Medical Papyrus). It describes a cure to prevent a woman from biting her tongue the day of birth. Modified from Griffith FL. Hieratic Papyri from Kahun and Gurob:(principally of the Middle Kingdom). B. Quaritch; 1898.

Convulsions and albuminuria

The association of convulsions, edema, and albuminuria was recognized in 1843 by John C. W. Lever at Guy’s Hospital in London, England,⁶ and James Young Simpson at the University of Edinburgh, Scotland.⁷ Lever was interested in the resemblance between patients with eclampsia and those with glomerulonephritis, who were cared for by Richard Bright at Guy’s Hospital (glomerulonephritis was known as “Bright’s disease” at the time).⁶ Lever tested the urine of 10 women diagnosed with puerperal convulsions and found albumin in most cases (9/10) (Figure 2). ⁶ The exception was a patient who died from meningitis. Lever proposed that eclampsia differed from glomerulonephritis because albuminuria disappeared after delivery.

Figure 2.

The title page of Cases of Puerperal Convulsions, with Remarks, in which Dr. John C. W. Lever, a British obstetric physician, reported the association between puerperal convulsions and albuminuria. Modified from Lever JC. Cases of puerperal convulsions, with remarks. Guys Hosp Rep 1843;2:495–517.

The triad: hypertension, albuminuria, and edema

Hypertension was recognized as a feature of eclampsia in 1885 by John William Ballantyne at the University of Edinburgh, who reported hypertension by using sphygmographic tracings (Figure 3) in three cases of pregnant women with “Bright’s disease.”⁸ The tracings were obtained during pregnancy, labor, and the postpartum period. Glomerulonephritis had been suspected in these cases given the combination of edema and albuminuria. Others have credited Louis Henri Vaquez and Pierre Nobécourt in 1897 with the discovery of eclamptic hypertension.⁹

Figure 3.

Sphygmographic Tracings in Puerperal Eclampsia. The sphygmographic tracings record blood pressure from women with preeclampsia (A) during pregnancy and (B) in the postpartum period. Modified from Ballantyne JW. Sphygmographic Tracings in Puerperal Eclampsia. Trans Edinb Obstet Soc. 1885;10:56–70.

The origin of the term “preeclampsia”

The association of hypertension, proteinuria, and convulsions was subsequently confirmed by other investigators, and importantly, some noted that hypertension and proteinuria were present before seizures, hence the name “pre-eclampsia.”⁵ Leon Chesley credited the introduction of this term to John Clarence Webster¹⁰ in 1903 in the United States and to Bar¹¹ in France (eclampsisme, meaning eclampsia without convulsions).¹¹ The concept took hold and has since been a driving force in the organization of prenatal care, which is largely structured to detect preeclampsia and to prevent eclampsia by measuring blood pressure, detecting proteinuria, and increasing the frequency of prenatal visits as term approaches.

“Toxemia of pregnancy”

Eclampsia and preeclampsia were originally attributed to “toxins” or “poisons” believed to enter the maternal circulation—hence, the term toxemia of pregnancy,^{5, 11–13} although there were dissenting opinions.¹⁴ The source of the toxins was considered exogenous or endogenous. One exogenous source was bacteria, and Gerdes proposed that eclampsia was caused by products of a bacillus, which he named Bacillus eclampsiae.^{15, 16} The endogenous sources (auto-toxicity) were thought to be caused by metabolic products of the fetus, mother, or placenta.¹⁷ The term “toxemias of pregnancy” included eclampsia, hyperemesis gravidarum, acute yellow atrophy of the liver, pruritus gravidarum, and ptyalism.¹⁸

The search for the toxin(s) responsible for preeclampsia and eclampsia has lasted for more than a century. In 1914, James Young reported an association between placental infarctions and eclampsia and showed that the administration of placental extracts from patients with eclampsia could induce seizures and other pathologic abnormities when injected into guinea pigs.¹⁹ Subsequently, Hunter and Howard^{20, 21} demonstrated the presence of a pressor substance in placental and decidual extracts and in the plasma of patients with preeclampsia: they named this substance “hysterotonin.” Importantly, Tatum and Mulé obtained whole blood from patients with severe preeclampsia and re-transfused it into the same patients postpartum, resulting in a transient, but significant, increase in systolic and diastolic blood pressure.²² Similar observations were made by Pirani and MacGillivray,²³ who collected plasma from normal pregnant women as well as from those with hypertension and proteinuria and then re-transfused aliquots on postpartum day 6 and at 6 weeks postpartum (Figure 4). Plasma from patients with preeclampsia—but not normal pregnancy—elicited a hypertensive response on Day 6 postpartum. This finding led to the conclusion that a soluble factor present in the plasma of patients with preeclampsia could induce hypertension—supporting the concept of a circulating “toxin.” However, when a similar aliquot of plasma was re-transfused at 6 weeks postpartum, hypertension did not occur: this was interpreted as an indication that, with time, a change took place in the vascular reactivity of the patient to the circulating pressor substances (toxins), which disappeared over time. This conclusion was strengthened by the observation that women with preeclampsia and those destined to develop preeclampsia were more sensitive to the pressor effects of angiotensin-II.²⁴ The search for the “toxin(s)” has continued to the present time, but the use of the term “toxemia” has progressively been abandoned. The anti-angiogenic factor [soluble fms-like tyrosine kinase 1 (sFlt-1)] has emerged as a major candidate for one of the “toxin(s)” responsible for preeclampsia, and this topic is discussed later in this article.

Figure 4.

Changes in diastolic blood pressure in patients with preeclampsia and controls (A) on postpartum day 6 and (B) at 6 weeks postpartum after plasma autotransfusion. Modified from Pirani BB, MacGillivray I. The effect of plasma retransfusion on the blood pressure in the puerperium. Am J Obstet Gynecol. 1975 Jan 15;121(2):221–6.

A shift in focus: from maternal signs to fetal and neonatal adverse outcomes

A major change in the conceptualization of preeclampsia and eclampsia occurred when investigators refocused the emphasis from maternal health outcomes (eg seizures, death) to fetal and neonatal outcomes [eg fetal death, fetal growth restriction, and small for gestational age (SGA)]. Two major studies conducted in the United States shaped the classification of hypertensive disorders and the interpretation of risk as a function of blood pressure and proteinuria in the antepartum period.^{25, 26}

Page and Christianson²⁵ reported the results of a prospective study of nearly 13,000 pregnancies between 1959 and 1967, which were part of the Child and Health Development studies performed in the United States. Patients were classified according to the mean arterial blood pressure measured in the second and third trimesters and to the presence or absence of proteinuria in the third trimester. Preeclampsia was defined as an elevated blood pressure in the third trimester with proteinuria. If the patient presented an elevated blood pressure in the second trimester, the disorder was considered to represent borderline hypertension with or without proteinuria.²⁵ The major finding of the study indicated that the rate of fetal death was higher in patients diagnosed with preeclampsia or chronic hypertension, but not in those with gestational hypertension (Figure 5).²⁵ Of interest, proteinuria (defined as 2+ or greater) was also a risk factor for fetal death, regardless of the presence of hypertension.

Figure 5.

Risk of stillbirth in patients with normotensive, gestational hypertension, chronic hypertension, and proteinuria by ethnic group. Modified from Page EW, Christianson R. Influence of blood pressure changes with and without proteinuria upon outcome of pregnancy. Am J Obstet Gynecol. 1976 Dec 1;126(7):821–33.

A major epidemiologic effort to examine the relationship between blood pressure, proteinuria, and adverse pregnancy outcome was undertaken as part of the Collaborative Perinatal Project, sponsored by the National Institutes of Health, which began in 1958 and prospectively enrolled 58,806 pregnancies at 12 university centers in the United States.²⁶ The results of the systematic evaluation of blood pressure, proteinuria, and pregnancy outcome were published in the book Pregnancy Hypertension written by Friedman and Neff (Figure 6).²⁶ One of the conclusions of this comprehensive study indicated the minimal influence of edema on the outcome, whether edema was recorded by history or by physical examination. This observation strengthened the case to remove edema from the triad of hypertension, proteinuria, and edema, which had formerly been used to diagnose preeclampsia. The study was also key in generating important data about the relationships among hypertension, proteinuria, and adverse pregnancy outcome.

Figure 6.

The book cover of Pregnancy Hypertension: A Systematic Evaluation of Clinical Diagnostic Criteria authored by Emanuel A. Friedman and Raymond K. Neff and published in 1977. In this book, the authors reported on the Collaborative Perinatal Project, in which the relationship between blood pressure, proteinuria, and adverse pregnancy outcomes was analyzed. Modified from Friedman EA, Neff RK. Pregnancy hypertension: a systematic evaluation of clinical diagnostic criteria. PSG Publishing Company, Inc.: Littleton, MA, 1977.

Table 1 shows rate ratios of fetal mortality by diastolic pressure according to the gestational age.²⁶ Before 36 weeks of gestation, fetal mortality was associated with values of 85 mmHg or more (diastolic blood pressure).²⁶ After 36 completed weeks, fetal mortality was associated with diastolic pressure of 95 mmHg or higher.²⁶ Moreover, increased fetal mortality was associated with proteinuria exceeding 1+, and the more intense the proteinuria, the greater the risk of fetal mortality (Table 2). Furthermore, the synergistic effect was noted between diastolic blood pressure and proteinuria (Table 3 and Figure 7). For example, fetal mortality was increased 9-fold (red circle in Table 3) if diastolic blood pressure (95 mmHg to 104 mmHg) and proteinuria 2+ were present. The data from the Collaborative Perinatal Project indicated that an elevation in systolic blood pressure was associated with virtually the same fetal outcome as an elevation in diastolic blood pressure.²⁶ However, diastolic blood pressure was considered a better risk indicator than systolic pressure, given that low systolic pressure was not associated with a poor outcome but low diastolic pressure was.

Table 1.

Rate ratios of fetal mortality by diastolic pressure according to the gestational age

Gestational ages, weeks	Diastolic blood pressure (mmHg)
	No.	<65	65–74 (Reference)	75–84	85–94	95–104	>105
28–32	24,640	0.7	1.0	1.0	2.3	5.9	10.2
33–34	19,340	0.9	1.0	1.1	2.3	7.1	8.3
35–36	20,593	0.8	1.0	1.3	2.0	8.8	7.8
37–38	20,243	1.0	1.0	1.2	1.8	2.8	-
39–41	15,797	0.8	1.0	1.0	1.2	3.2	2.6

Modified from Friedman EA, Neff RK. Pregnancy hypertension: a systematic evaluation of clinical diagnostic criteria. PSG Publishing Company; 1977, Page 132.

Table 2.

Fetal mortality rate and rate ratios by proteinuria maxima

Proteinuria	Fetal mortality rate, %	Fetal mortality rate ratio
None (reference)	0.9	1.0
Trace	0.9	1.0
1+	1.2	1.3
2+	2.3	2.6
3+	4.4	4.9
4+	5.7	6.3

Modified from Friedman EA, Neff RK. Pregnancy hypertension: a systematic evaluation of clinical diagnostic criteria. PSG Publishing Company; 1977, Page 131.

Table 3.

Rate ratios for fetal mortality by diastolic pressure and proteinuria combination

Diastolic pressure, mmHg	Proteinuria
	None	Trace	1+	2+	3+	4+
<65	2.5	2.3	1.0	-	-	-
65–74	1.5	1.3	0.8	5.5	7.0	-
75–84	1.0*	1.3	1.0	3.2	-	-
85–94	1.5	1.5	4.0	-	3.7	-
95–104	3.2	2.8	4.5	9.3	19.2	23.8
≥105	3.3	4.7	10.5	11.5	20.8	18.5

^*Reference rate of fetal mortality: 0.6% or 6 per 1000 births

Modified from Friedman EA, Neff RK. Pregnancy hypertension: a systematic evaluation of clinical diagnostic criteria. PSG Publishing Company; 1977, Page 136.

Figure 7.

Rate ratios for fetal mortality by diastolic pressure and proteinuria combinations. The synergistic effect of diastolic blood pressure and proteinuria is evident, and it determines a considerable increase in the risk of fetal mortality. Modified from Friedman EA, Neff RK. Pregnancy hypertension: a systematic evaluation of clinical diagnostic criteria. PSG Publishing Company, Inc.: Littleton, MA, 1977, page 170.

The clinical implementation of these observations followed the recommendation of the World Health Organization and the American College of Obstetricians and Gynecologists (ACOG). The selection of 90 mmHg as a cut-off value for diastolic blood pressure was made because it is a midpoint between 85 mmHg and 95 mmHg; 85 mmHg is associated with fetal mortality when combined with proteinuria and 95 mmHg regardless of proteinuria.²⁶ The recent recommendations of the American College of Cardiology (ACC) and the American Heart Association (AHA) to lower the threshold for the diagnosis of hypertension to 130/80 mmHg²⁷ has now stimulated a dialogue about whether the threshold should also be applied to the diagnosis of preeclampsia.^{28, 29} Early evidence suggests that perinatal outcomes in pregnant women with stage I hypertension before 20 weeks of gestation are worse than those with normal blood pressure.³⁰ However, more work is necessary to determine whether the diagnostic criteria should be modified.

Preeclampsia: more than pregnancy-induced hypertension

The hallmark of the clinical diagnosis of preeclampsia has been hypertension. However, the involvement of other organs has been known for decades based on autopsy findings and clinical reports.^{31, 32} The most frequent organs involved are the kidney (proteinuria),^{32, 33} liver (elevation of transaminases, liver hematoma, and rupture),^{32, 34–37} hematopoietic system (hemolysis, leukocytosis, thrombocytopenia),^38–44 brain (seizures, cortical blindness, intracranial hemorrhage, infarction),^45–47 and uteroplacental circulation (fetal growth restriction, abruption, fetal death).^{19, 48–56} Other organs/systems that may be involved include the lung (ventilation-perfusion mismatch, adult respiratory distress syndrome, pulmonary edema),⁵⁷ heart (systolic and diastolic dysfunction),^58–63 pancreas (pancreatitis),⁶⁴ eyes (retinal problems leading to detachment), ^{65, 66} small and large intestines (ischemia),⁶⁷ endocrine organs (adrenal glands, thyroid, parathyroid),^68–72 and immune system (exaggerated intravascular coagulation, changes in B and T cells as well as T regulatory cells).^73–76

An atypical form of preeclampsia: HELLP syndrome

Clinicians and investigators realized that an elusive complication eventually recognized as preeclampsia can exist without hypertension. Indeed, about 15% of patients with HELLP (Hemolysis, Elevated Liver enzymes, and Low Platelets) syndrome have normal diastolic blood pressure at admission,^{38, 77} and 10%−15% of patients with eclampsia^78–80 do not develop hypertension. Robert Goodlin⁸¹ emphasized this point based on a case series of pregnant patients with atypical presentation and referred to preeclampsia as another “great imitator” (other great imitators include syphilis, tuberculosis, and Lyme disease) (Figure 8). In 1982, Louis Weinstein coined the term HELLP syndrome to describe the combination of hemolysis, elevated liver enzymes, and low platelet count and proposed it to be a severe consequence of hypertension in pregnancy (Figure 9).⁸² However, this cluster of laboratory findings often associated with abdominal pain can be present in the absence of hypertension and proteinuria:^{83, 84} sometimes thrombocytopenia and liver dysfunction can resolve,⁸³ and only later do patients develop hypertension and proteinuria occur.⁸³ Indeed, thrombocytopenia and an elevated SGOT (serum glutamic-oxaloacetic transaminase) level are independent risk factors for adverse pregnancy outcomes after adjusting for hypertension and proteinuria. Patients may present with unusual clinical manifestations, such as visual disturbances,^85–87 renal failure,^{88, 89} congestive heart failure,⁶³ abdominal pain,^{90, 91}, and headaches.^{92, 93} Criteria for the clinical definition of preeclampsia’s “severe features” are indicative of the extent of the maternal multisystem disorder. The abnormal laboratory tests associated with preeclampsia reflect damage or dysfunction of the tissues and organs involved (Figure 10). Thus, hypertension is only one clinical sign of a multi-systemic disorder.

Figure 8.

“Severe pre-eclampsia: another great imitator.” Modified from Goodlin RC. Severe pre-eclampsia: another great imitator. Am J Obstet Gynecol. 1976 Jul 15;125(6):747–53.

Figure 9.

The original paper in which the term HELLP was coined. “Syndrome of hemolysis, elevated liver enzymes, and low platelet count: a severe consequence of hypertension in pregnancy” Am J Obstet Gynecol. 1982 Jan 15;142(2):159–67.

Figure 10.

Preeclampsia as a multi-systemic disease that involves virtually every organ system in humans.

Preeclampsia without proteinuria?

Proteinuria has been a requirement for the diagnosis of preeclampsia for decades, as it indicates renal involvement; however, in the 1990s, several groups of investigators began to question whether proteinuria should be a necessary criterion.^94–97 Multiple studies reported that patients with severe gestational hypertension had a high rate of adverse pregnancy outcomes, despite the absence of proteinuria.^94–96 Indeed, the frequency of preterm birth, SGA, placental abruption, respiratory distress syndrome, and perinatal death had been reported to be higher among women with severe gestational hypertension compared to those with mild preeclampsia, defined as the combination of hypertension and proteinuria.⁹⁵ Moreover, about 50% of women with a prospective diagnosis of gestational hypertension will subsequently develop proteinuria or end-organ damage. ^{96, 98} Two additional arguments favored abandoning the requirement of proteinuria for the diagnosis of preeclampsia: the standard definition (≥ 300 mg in a 24-hour urine collection or a ≥ 0.3 protein/creatinine ratio) is based on limited data,^{99, 100} and dipstick testing for proteinuria is unreliable.^{101, 102}

Collectively, this set of observations, coupled with the lack of difference in the clinical management of gestational hypertension and conventionally defined preeclampsia, led the ACOG Task Force on Hypertension in Pregnancy to modify the definition for preeclampsia in 2013 as follows:¹⁰³ hypertension that develops after 20 weeks of gestation with proteinuria or evidence of end-organ damage, such as an abnormal renal function test, an elevation of liver enzymes, or thrombocytopenia without proteinuria. The modification reflected a consensus who agreed that the crucial factor was blood pressure coupled with other manifestations of end-organ damage, one of which could be renal involvement. While this recommendation aimed to optimize the clinical management of patients, it has also led to an increase in the frequency of diagnoses of preeclampsia; however, it is unclear whether this will translate into improved maternal and perinatal outcomes because most of the newly diagnosed cases present a mild form of the disease.¹⁰⁴ A more objective definition of the vascular dysfunction we call preeclampsia today is required to supersede the current approach of relying largely on blood-pressure measurement. The identification of biomarkers for the early detection of the different forms of the syndrome is crucial for improved diagnosis, taxonomy, prediction, and prevention.

The discovery of anti-angiogenic factors in preeclampsia

A major breakthrough occurred with the discovery that placentas from patients with preeclampsia overexpressed the mRNA and protein for sFlt-1 (Figure 11).^{105, 106} This story is a wonderful example of the power of discovery tools to gain insights into biological processes in medicine. The experiment was simple: the transcriptomes of placentas from patients with preeclampsia was compared to those of a control group. Karumanchi’s group found that two anti-angiogenic factors—sFlt-1 and soluble endoglin (sEng)—were overexpressed in the placentas of women with preeclampsia.^106–109 Around the same time, it was known that the blockade of angiogenesis with VEGF (vascular endothelial growth factor) antagonists (monoclonal antibodies against VEGF) in non-pregnant patients with cancer would lead to hypertension and proteinuria.¹¹⁰ Encouraged by this finding, the investigators pursued a systematic series of studies that demonstrated that overproduction of sFlt-1 in pregnant animals recapitulated the features of preeclampsia and the renal lesions associated with this condition (glomerular endotheliosis).^{106, 111} This information, coupled with previous observations in patients with a low maternal serum concentration of the angiogenic factor PlGF (placental growth factor), strengthened the case for a role of an anti-angiogenic imbalance in preeclampsia.¹¹²

Figure 11.

(A) Remodeling of the spiral arteries increases blood supply to the fetus. (B) In preeclampsia, sFlt-1 is overexpressed in the placenta, leading to hypertension and proteinuria. Modified from Luttun A, Carmeliet P. Soluble VEGF receptor Flt1: the elusive preeclampsia factor discovered? J Clin Invest. 2003 Mar;111(5):600–2.

The discovery that anti-angiogenic factors are linked to preeclampsia has improved the understanding of the pathophysiology, allowed classification according to the presence or absence of an abnormal anti-angiogenic profile, and identified biomarkers for the prediction of term and preterm preeclampsia. We believe that anti-angiogenic factors meet the criteria to be one, if not the major, “toxin” responsible for preeclampsia and eclampsia. A full account of the scientific basis for this claim is provided in the companion articles of this supplement.^{113, 114}

Preeclampsia is one of the “great obstetrical syndromes”

Obstetric disorders, by contrast with diseases in the non-pregnant state, develop in the context of a unique biological situation—two individuals with different genomes coexisting, one inside the other. The common interest of the mother and her fetus is successful reproduction; however, conflict can occur when the interests of the mother and the fetus diverge, perhaps as the result of an insult (such as an infection or a compromised blood supply). The term “great obstetrical syndromes” describes the unique nature of obstetric diseases. These features include 1) multiple etiologies, 2) a long subclinical phase, 3) fetal involvement, 4) the adaptive nature of clinical manifestations, and 5) complex gene-environment interactions.^115–117 This section will summarize the evidence that supports the recognition of these features in the case of preeclampsia.

Multiple etiologies

Preeclampsia is not one disorder but rather different entities recognized by a common phenotype of hypertension and proteinuria, which represents the manifestation of one or more insults to the uteroplacental unit. Maternal, fetal, and placental causes of preeclampsia have now been identified. The causative risk factors, or etiology, of preeclampsia are reviewed in detail by Jung et al.¹¹⁸

Long subclinical phase

Multiple lines of evidence indicate that a subset of patients who will develop preeclampsia can be identified weeks or months before the diagnosis. Indeed, some women may present an abnormal pressor response to angiotensin II,¹¹⁹ an abnormal uterine artery Doppler velocimetry measurement, and abnormal angiogenic and anti-angiogenic profiles before the clinical diagnosis.^{108, 120–124}

Fetal involvement

Although preeclampsia is diagnosed by clinical signs in the mother (hypertension and proteinuria), fetal involvement can manifest as fetal growth restriction, and a subset of neonates may have subtle abnormalities¹²⁵ such as thrombocytopenia or neutropenia. Neonates, delivered by their respective preeclamptic mother, have recently been reported to present dilatation of the right coronary artery and the increased expression of vascular cell adhesion molecule 1 (VCAM-1) in the umbilical arteries.¹²⁶ Although the changes in the right coronary artery are transient, the offspring of mothers with preeclampsia are at risk for long-term cardiovascular disease and for the development of attention-deficit/hyperactivity disorder later in life.¹²⁵ Collectively, the findings indicate fetal involvement that goes beyond growth restriction.

Clinical manifestations are adaptive

Hypertension can be considered an adaptive response of an injured placenta, which signals to the mother the need to maintain perfusion; this is accomplished by increasing maternal cardiac output, an elevation of maternal blood pressure, or, in some cases, a combination of both. The concept of the adaptive nature is supported by a set of clinical observations including the resolution of maternal hypertension following the death of a growth-restricted fetus in a twin gestation,^127–130 after SARS-CoV-2 infection,¹³¹ or after transfusion to correct fetal anemia due to fetal parvovirus infection.¹³² Indeed, pharmacologic treatment of maternal hypertension does not improve fetal outcomes. The adaptive responses can become maladaptive; for example, a hypertensive crisis can result in cerebrovascular accidents and maternal death. Therefore, maternal death can be viewed as the price paid for a maladaptive response.

Preeclampsia, nonetheless, could be considered to have survival value. A compromised uterine blood supply can lead to fetal growth restriction, an adaptive mechanism whereby the fetus slows its growth to avoid outstripping the delivery of oxygen and nutrients. This adaptation alone might be sufficient for the fetus to reach the end of pregnancy. However, when fetal growth restriction alone is insufficient to cope with an insult, signals originating from an ischemic placenta increase the maternal blood pressure to sustain the fetus and prevent death.

Complex gene-environment interaction

Obstetrical syndromes take place in a unique situation in which there are two genomes and two environments (maternal and fetal). The occurrence of preeclampsia is probably determined by a combination of genetic and environmental factors. Preeclampsia clusters in families, and genetic association studies have been conducted to identify DNA variants that predispose to preeclampsia. Although most studies have focused on mothers,^{133, 134} it is important to note that an association between a fetal DNA variant and the syndrome has been recently replicated.¹³⁵ Indeed, the result of a genome-wide association study of 2,658 offspring from pregnancies complicated with preeclampsia and of 308,292 controls indicated that a fetal genome DNA variant near FLT1 (the gene encoding fms-like tyrosine kinase 1, sFlt-1) is associated with the risk of preeclampsia.¹³⁶ This result highlights the complex relationship between the maternal and fetal genomes and the occurrence of preeclampsia. In addition to the contribution of the fetal and maternal DNA variants by themselves to a particular obstetrical syndrome, the interaction and incompatibility of genotypes may also operate to confer risk. The role of genetic incompatibility in preeclampsia has been illustrated by the increased risk conferred by specific combinations of HLA-C (major histocompatibility complex, Class I, C) genotypes in the fetus and KIRs by the mother.^137–139 This was also illustrated by the increased risk conferred by other combinations in the genes encoding for the von Willebrand factor (VWF), alpha 2 chain of type IV collagen (COL4A2), and lymphotoxin alpha (LTA).¹⁴⁰

Implications of preeclampsia as a great obstetrical syndrome

Considering preeclampsia as one of the great obstetrical syndromes has several important consequences. The etiological explanations, at once multiple and very diverse, include the preexisting cardiovascular disease of mothers, fetal and maternal infection, placental ischemia, aging of the placenta, and endocrine disorders, among others. It follows that the prediction, optimal management, and prevention of preeclampsia will vary according to the etiology. Therefore, it is unlikely that the challenge of preeclampsia will be solved with a single test, treatment, or preventive strategy. We anticipate that different biomarkers will be required to identify, predict, and monitor the various types of preeclampsia.

The long subclinical phase creates a window of opportunity for prediction. Current evidence indicates this possibility for patients with preterm preeclampsia diagnosed at the end of the first trimester and those with term preeclampsia occurring later in gestation (ie 36 weeks), and it relies on different sets of biophysical/biochemical parameters and analytes.^{141, 142} Serial testing may be required in a subset of patients. Strategies for treatment and prevention are also likely to differ. For example, while aspirin, low molecular weight heparin, or statin may have a role in some forms of the disease, late-onset preeclampsia may be best handled by the induction of labor.

Although the diagnosis and classification of preeclampsia are largely based on maternal symptoms and signs, the assessment of fetal involvement and the degree of compromise should be a part of the evaluation of the patient. The adaptive nature of the hypertensive response aids the understanding of why antihypertensive treatment may protect the mother but does not alter the inexorable course of the disease.

Classification: Early versus late, mild versus severe disease

Preeclampsia has been classified according to gestational age at the time of diagnosis or delivery as early onset (<34 weeks) or late onset (≥34 weeks). Although several other gestational age cut-off values have been suggested (such as 32 weeks and 36 weeks), 34 weeks remains the most commonly used, presumably because the rate of neonatal morbidities declines considerably after this gestational age.^{143, 144} For instance, in women with severe preeclampsia, expectant management is not considered, and induction of labor is recommended after 34 weeks of gestation. Preeclampsia can also be classified according to its severity (Table 4); nonetheless, to avoid conveying a false sense of security to clinicians, some professional organizations have recommended against the use of the terms “mild” or “severe” in favor of “preeclampsia with or without severe features.”^{103, 145}

Table 4.

Severe features of pre-eclampsia (one or more of these findings)

• Systolic blood pressure of 160 mm Hg or higher, or diastolic blood pressure of 110 mm Hg or higher on two occasions at least 4 hours apart while the patient in on bed rest (unless antihypertensive therapy is initiated before this time)

• Thrombocytopenia (platelet count less than 100,000/microliter)

• Impaired liver function as indicated by abnormally elevated blood concentrations of liver enzymes (to twice normal concentration), severe persistent right upper quadrant or epigastric pain unresponsive to medication and not accounted for by alternative diagnoses, or both

• Progressive renal insufficiency (serum creatinine concentration greater than 1.1 mg/dL or a doubling of the serum creatinine concentration in the absence of other renal disease)

• Pulmonary edema

• New-onset cerebral or visual disturbances

Adapted from Hypertension in pregnancy. Report of the American College of Obstetricians and Gynecologists’ Task Force on Hypertension in Pregnancy. Obstet Gynecol. 2013 Nov;122(5):1122–1131.

Early-onset and late-onset preeclampsia: two different conditions

The early literature did not distinguish between early-onset and late-onset disease. However, multiple lines of evidence have now coalesced, indicating that early and late preeclampsia are different conditions. Table 5 shows the main differences between the two conditions. Early-onset preeclampsia is characterized by a higher frequency of the HELLP syndrome,^{142, 146} an abnormal uterine artery Doppler waveform,¹⁴² atherosis,^{147, 148} placental lesions consistent with maternal vascular malperfusion,^{142, 149} SGA,^{141, 142} and fetal growth restriction.¹⁴² Moreover, 80%−90% of patients with early-onset preeclampsia have abnormal maternal plasma placental growth factor (PlGF):sFlt-1 or PlGF:sEng ratios, while this finding occurs only in 40%−50% of those with late-onset preeclampsia.^{141, 150, 151} We consider that early-onset preeclampsia is a clinical manifestation of atherosclerosis in pregnancy while late-onset disease represents a metabolic crisis in which there is a mismatch between fetal demands and maternal supply.

Table 5.

Early-onset vs. late-onset preeclampsia

	Early	Late
Gestational age at onset	<34 weeks*	≥34 weeks*
Prevalence169	0.38% (12% of all preeclampsia)	2.72% (88% of all preeclampsia)
HELLP syndrome142	40.0%	11.1%
Fetal growth restriction142	60.0%	25.0%
Small for gestational age142	66.7%	31.9%
Neonatal death/severe neonatal morbidity169	27.7%	2.2%
Combined adverse maternal outcomes146	20.9%	9.2%
Altered uterine artery Doppler PI142	88.6%	48.6%
Abnormal maternal plasma PlGF:sFlt-1141	~ 80–90%	~ 40–50%

^*Gestational age at diagnosis or delivery

HELLP: Hemolysis, Elevated Liver enzymes and Low Platelets, PI: Pulsatility Index, PlGF: placental growth factor, sFlt-1: soluble fms-like tyrosine kinase

Postpartum preeclampsia

Some patients are diagnosed only after delivery, and postpartum preeclampsia and eclampsia could be potentially dangerous because it may occur in the late postpartum period after the patient has been discharged. The condition has remained an enigma since delivery of the placenta is considered to be curative. Retained fragments of the placenta have been implicated in postpartum preeclampsia and eclampsia for decades, although evidence to support this view is difficult to substantiate.¹⁵² Interestingly, in 1960, while investigating the etiology of hypertension in “toxemia of pregnancy,” Hunter and Howard²⁰ described that the decidua of patients with toxemia and a molar pregnancy produced a pressor substance that they called “hysterotonin.” Although the molecule responsible for this pressor response was never characterized, Hunter et al. ¹⁵³ proposed that postpartum curettage of the decidua could help improve the condition. Indeed, this procedure was reported to ameliorate hypertension in 69 patients. In one patient with postpartum eclampsia and three episodes of convulsions (Figure 12), curettage was performed, and maternal hypertension improved and convulsions did not recur.¹⁵³ Approximately 30 years later, Magann et al.¹⁵⁴ reported a randomized clinical trial of immediate postpartum curettage in 32 patients with severe preeclampsia and observed that patients who had undergone curettage had not only significantly lower blood pressure (Figure 13A) but also a significantly higher urinary output and platelet count (Figure 13B) than those who did not undergo a curettage. Collectively, these observations suggest that material present in the uterus after the delivery of the placenta can still have biological properties. While Hunter et al.¹⁵³ attributed this to the decidua, trophoblasts may also be present in the uterus after delivery of the placenta and are consistently observed in histologic examinations of hysterectomy specimens. Therefore, it is possible that trophoblasts continue to be a source of bioactive material. However, why some patients develop hypertension and proteinuria only after delivery remains a riddle.

Figure 12.

Effect of uterine curettage on blood pressure in a woman with postpartum eclampsia. In one patient with postpartum eclampsia and three episodes of convulsions, curettage was performed, maternal hypertension improved, and convulsions did not recur. Modified from Hunter CA Jr, Howard WF, McCormick CO Jr. Amelioration of the hypertension of toxemia by postpartum curettage. Am J Obstet Gynecol. 1961 May;81:884–9.

Figure 13.

(A) Postpartum uterine curettage not only has an effect on blood pressure (B) but also on platelet count in patients with preeclampsia. Modified from Magann EF, Martin JN Jr, Isaacs JD, Perry KG Jr, Martin RW, Meydrech EF. Immediate postpartum curettage: accelerated recovery from severe preeclampsia. Obstet Gynecol. 1993 Apr;81(4):502–6.

One conceptualization of preeclampsia is an adaptive homeostatic response to maintain uterine perfusion of an ischemic placenta during pregnancy; yet, in a small subset of patients, this response becomes maladaptive only after delivery. We propose that postpartum preeclampsia and eclampsia belong to a group of conditions of unknown etiology that occur after delivery in a fraction of patients, which include peripartum cardiomyopathy, renal failure, uremic hemolytic syndrome, and acute fatty liver. Insights into the pathophysiology of these conditions can be gleaned by recent progress made in peripartum cardiomyopathy, which is now recognized as the deleterious effect of the anti-angiogenic factor sFlt-1.¹⁵⁵ This condition is thought to result from a two-hit process.¹⁵⁵ The first is related to a high concentration of sFlt-1 in the maternal circulation, which can impair cardiac function, and the second is represented by a lack of local pro-angiogenic defenses in the maternal heart. Peripartum cardiomyopathy can be induced in pregnant mice by the combination of increased sFlt-1 maternal plasma concentrations (induced through an adenovirus vector) and gene deletion of a powerful regulator of angiogenesis called peroxisome proliferator-activated receptor-gamma coactivator (PGC-1α).¹⁵⁵ Similarly, pregnant women with peripartum cardiomyopathy present an increased concentration of sFlt-1 in the postpartum period, which has been detected 4 to 6 weeks after delivery (elevated sFlt-1 typically returns to normal range within three days of delivery).¹⁵⁵

A similar pathophysiologic process can be responsible for other postpartum syndromes, including postpartum renal failure. VEGF is essential for the maintenance of glomerular health, and its blockade damages the fenestrated endothelium causing proteinuria.^{156, 157} It is now recognized that a subset of patients with preeclampsia has impaired subclinical renal function postpartum and proteinuria, which may persist for months. An antepartum sFlt-1 concentration correlates with a lower glomerular filtration rate, and a high concentration is a risk factor for renal impairment at 6 and 12 months postpartum.¹⁵⁸ Recent observations suggest that the pathophysiology of acute fatty liver of pregnancy may also be related to an excess concentration of sFlt-1 (Figure 14).^{159, 160} The reasons why this excess may target the heart, kidney, liver, or, perhaps, the brain (in some cases of postpartum eclampsia) is unknown but probably results from gene-environment interactions. We envision that several enigmatic postpartum syndromes represent a part of the spectrum of disease of vascular dysfunction mediated, at least in part, by an excess of anti-angiogenic factors that alter VEGF function. The protean manifestations of this vascular dysfunction in the postpartum period reflect the target organ involved.

Figure 14.

Plasma concentration of sFlt-1 in women with normal pregnancy, preeclampsia, HELLP syndrome, and acute fatty liver of pregnancy. Women with acute fatty liver of pregnancy have a higher plasma concentration of sFlt-1 compared to the other conditions. Data are presented as individual values (dot) and median (bar). Modified from Neuman RI, Saleh L, Verdonk K, et al. Accurate Prediction of Total PlGF (Placental Growth Factor) From Free PlGF and sFlt-1 (Soluble Fms-Like Tyrosine Kinase-1): Evidence for Markedly Elevated PlGF Levels in Women With Acute Fatty Liver of Pregnancy. Hypertension. 2021 Aug;78(2):489–498. HELLP: Hemolysis, Elevated Liver enzymes and Low Platelets; sFlt-1: Soluble Fms-Like Tyrosine Kinese-1.

Vascular dysfunction of pregnancy: unmasked, induced, protean

The most important adaptation for a successful pregnancy is the establishment and development of an adequate blood supply to the placenta and conceptus. The clinical consequences of suboptimal perfusion range from fetal growth restriction, SGA, preeclampsia, abruptio placenta, and fetal death (Figure 15). A fundamental question remains: why does maternal malperfusion lead to one particular syndrome over another? We propose that the timing and magnitude of the insult and the genetic makeup of the mother, fetus, and placenta determine the clinical presentation. Extreme compromise of the blood supply may result in fetal death, while lesser degrees of maternal malperfusion could be compensated for by a reduction in fetal growth, maternal hypertension, or a combination of both; in some cases, the adaptive response is spontaneous preterm birth. Why some cases of fetal growth restriction due to placental malperfusion are associated with preeclampsia and others are not is likely to reflect the nature of the molecular and cellular dialogue between the mother, fetus, and placenta. It is now clear that a subset of women who develop preeclampsia have preexisting vascular dysfunction, which manifests clinically during pregnancy and remains operative in the postpartum period. Such recognition offers unique opportunities to improve the health care of women by implementing strategies to prevent cardiovascular disease. Figure 16 illustrates the long-term adverse events associated with preeclampsia. These include not only maternal hypertension and coronary artery disease but also vascular dementia and end-stage renal disease.¹⁶¹ While this article has focused on the subject of preeclampsia as a distinct clinical syndrome in obstetrics, we are cognizant that patients with other complications of pregnancy, such as fetal death, are also at risk for subsequent cardiovascular diseases. Therefore, the recognition of vascular dysfunction during pregnancy diagnosed through preeclampsia or other syndromes has important implications for women’s health.¹⁶²

Figure 15.

Vascular dysfunction during pregnancy may result in one of several obstetrical syndromes. The timing and magnitude of the insult may determine the clinical syndromes.

Figure 16.

Relative risk of cardiometabolic disease in women with a history of preeclampsia. Modified from Ramlakhan KP, Johnson MR, Roos-Hesselink JW. Pregnancy and cardiovascular disease. Nat Rev Cardiol. 2020 Nov;17(11):718–731.

Is preeclampsia a pregnancy specific disorder, caused by the placenta, and cured only by delivery?

Textbooks,¹⁶³ websites,¹⁶⁴ and medical articles^{165, 166} frequently make these three statements about preeclampsia. If the condition is considered as a vascular dysfunction associated with pregnancy or the postpartum period, they are a good approximation to the syndrome as it is understood today. However, it is now clear that hypertension and proteinuria can occur with a blockade of the VEGF pathway in non-pregnant oncologic patients treated with VEGF antagonists.¹¹⁰ Therefore, if the definition of preeclampsia rests on the occurrence of hypertension and proteinuria, it is non-specific as it can be present in patients with chronic renal disease and other conditions outside of pregnancy. We accept that a placenta is necessary because the diagnosis is made only in pregnant women. Nonetheless, it can occur in the postpartum period after the placenta has been delivered.¹⁶⁷ Perhaps what the requirement of a placenta means is that the source of the anti-angiogenic or other mediators responsible for hypertension and proteinuria originates in the placenta during pregnancy. We, along with other investigators, have observed hypertension and proteinuria in SARS-CoV-2 infection, and both disappear when the viral infection improves,¹³¹ as well as in other conditions such as a twin gestation in which one of the twins dies^{128, 129} or after the recovery of a fetal viral infection.¹³² What is clear is that reliance on a sphygmomanometer to identify this complex, consequential vascular syndrome seems less than ideal. Discovery of biological markers that detect the pathophysiologic derangements in an early phase is an important priority for advances in the classification, prediction, and prevention of preeclampsia; further, we believe that obstetrics should be unlocked from the ball-and-chain created by an overreliance on blood pressure measurements. We will review the etiology¹¹⁸ and the pathophysiology of preeclampsia¹⁶⁸ in two companion articles with the hope that this information may contribute to this urgent need for further research and data.

Conclusion

What began as a disorder of the central nervous system, or eclampsia, subsequently attracted the attention of nephrologists, and it is now recognized as a cardiovascular disorder. The concept of “preeclampsia and eclampsia” has helped shape prenatal and intrapartum care. Recognizing that the fundamental problem is vascular dysfunction unmasked, or induced, by pregnancy (through preeclampsia or other obstetrical syndromes) has profound implications. Framed in this way, the understanding of the nature of vascular dysfunction during pregnancy would allow the health care system to focus on the prevention of cardiovascular disease in women throughout their lifespan.