Angiogenic Factors in the Pathogenesis of Preeclampsia

Comp. by:dshanthi
Date:1/8/05
Time:15:46:44
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
9 ____________________________________________________________________________
Angiogenic Factors in the Pathogenesis
of Preeclampsia
Hai‐Tao Yuan,* David Haig,{ and S. Ananth Karumanchi*
*Renal, Molecular, and Vascular Medicine Division, Departments of OF
Medicine
Obstetrics and Gynecology, Beth Israel Deaconess Medical Center and Harvard
Medical School, Boston, MA
{Department of Organismic and Evolutionary Biology, Harvard University
Cambridge, MA
I. Introduction
II. Abnormal Placentation and Placental Ischemia (Stage 1)
III. Systemic Endothelial Dysfunction (Stage 2)
A. Excess Circulating sFlt1, Impaired VEGF Signaling, andPRO
Antiangiogenic State
B. sFlt1 and Maternal–Fetal Conflict
C. Speculations about the Mechanisms of Preeclampsia
D. Unanswered Questions
IV. Conclusions
Acknowledgments
References
Preeclampsia aVects 5–10% of pregnancies and is responsible for
substantial maternal and neonatal morbidity and mortality. It is believed
to be a two‐stage disease with an initial placental trigger with no maternal
symptoms followed by a maternal syndrome characterized by hyperten-
sion, proteinuria, and endothelial dysfunction. The first stage is thought to
be due to shallow cytotrophoblast invasion of maternal spiral arterioles
leading to placental insuYciency. The diseased placenta in turn releases
soluble angiogenic factors that induce systemic endothelial dysfunction
and clinical preeclampsia during the second stage. This review will discuss
the role of circulating angiogenic factors of placental origin as potential
mediators of the systemic endothelial dysfunction and the clinical syndrome
of preeclampsia and provide an evolutionary explanation for this
phenomenon. ß 2005, Elsevier Inc.
I.
UNCORRECTED
Introduction
Preeclampsia is characterized by the new onset of hypertension and protein-
uria after 20 weeks of gestation (Roberts, 2000; Roberts and Cooper, 2001;
Walker, 2000). Preeclampsia is also frequently associated with edema and
Current Topics in Developmental Biology, Vol. 71
0070-2153/05 $35.00
Copyright 2005, Elsevier Inc. All rights reserved.
297
DOI: 10.1016/S0070-2153(05)71009-7

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:44
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
298
Yuan et al.
hyperuricemia and it usually remits when the placenta is delivered. The
placenta in preeclampsia is often abnormal, with evidence of hypoperfusion
and ischemia. Vascular endothelial dysfunction and microangiopathy are
present in the mother, but not in the fetus. Severe complications of pre-
eclampsia can include acute renal failure, cerebral edema, cerebral hemor-
rhage, seizures (eclampsia), pulmonary edema, thrombocytopenia, hemolytic
anemia, coagulopathy, and liver injury—including HELLP, the syndrome
of Hemolysis, Elevated Liver enzymes, and Low Platelets (Sibai OF
et al., 2005).
When preeclampsia threatens to lead to severe maternal complications,
urgent delivery of the fetus and placenta are often undertaken to preserve
maternal health.
Although preeclampsia has been traditionally thought of as a pregnancy‐
specific syndrome, recent data suggests that women with a history of pre-
eclampsia have an eightfold increased risk of
PRO
cardiovascular death when
followed over their lifetime (Irgens et al., 2001). Risk factors for preeclamp-
sia include nulliparity, preexisting hypertension, obesity, diabetes mellitus,
thrombophilias, and a family history of preeclampsia (Dekker, 1999).
Observations to date suggest that the earliest pathologic change in pre-
eclampsia occurs in the uteroplacental circulation resulting in placental
insuYciency or ischemia, which may be considered stage 1 of the disease
(Roberts, 2000). In stage 2, the diseased placental tissue (ischemic placenta)
in turn secretes circulating factors that cause generalized endothelial cell
injury in the mother resulting in the clinical syndrome of preeclampsia
(Roberts, 2000; Roberts et al., 1989). Current understanding of the patho-
genesis of preeclampsia will be reviewed with an emphasis on evidence that
an imbalance in circulating angiogenic factors (Bdolah et al., 2004) and their
interaction with the maternal vasculature may be responsible for the clinical
phenotype of preeclampsia.
II. Abnormal Placentation and Placental Ischemia (Stage 1)
It is widely accepted that the placenta plays a central role in the pathogenesis
of preeclampsia as it occurs only in the presence of the placenta and the
clinical symptoms remit dramatically postpartum after the delivery of the
placenta (Page, 1939). In a case of preeclampsia with extrauterine pregnancy,
removal of the fetus alone was not suYcient; symptoms persisted until the
placenta
UNCORRECTED
was delivered (Shembrey and Noble, 1995). A recent report suggests
that in cases of preeclampsia with discordant twins, selective feticide reverses
preeclampsia, with the attenuation of symptoms occurring in a timeframe
consistent with placental involution (Heyborne and Porreco, 2004).
Several lines of evidence suggest that placental insuYciency is central to
the pathogenesis of preeclampsia (Karumanchi et al., 2004): (1) Pathologic

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:44
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
9. Angiogenic Factors in the Pathogenesis of Preeclampsia
299
examination of placentas from preeclamptic pregnancies reveals numerous
placental infarcts and sclerotic narrowing of arterioles (De Wolf et al., 1975;
Khong et al., 1986); (2) placental bed biopsies in preeclamptics have
been noted to have inadequate trophoblastic invasion of maternal decidual
arterioles leading to tight and constricted vessels (Gerretsen et al., 1981;
Robertson et al., 1967); (3) maternal risk factors for preeclampsia include
medical conditions that predispose a patient to underlying vascular insuY-
ciency such as chronic hypertension, diabetes, systemic lupus
OF
erythematosus,
as well as acquired and inherited thrombophilias (Dekker, 1999); (4) obstet-
rical conditions such as multiple gestations or hydatiform moles that increase
placental mass but with a relative decrease of placental blood flow increase
the risk of preeclampsia (Dekker, 1999); and (5) animal models of preeclamp-
sia involve creating placental insuYciency by disrupting uterine blood flow
(Casper and Seufert, 1995; Kumar, 1962).
During normal placental development, cytot PRO
rophoblasts invade the
maternal spiral arterioles and completely remodel the maternal spiral arter-
ioles into large capacitance vessels with low resistance (Gerretsen et al., 1981;
Robertson et al., 1967). This endovascular cytotrophoblast invasion in-
volves replacement of not only the endothelium but also the highly muscular
tunica media. In preeclampsia, there is shallow placental cytotrophoblast
invasion of uterine spiral arterioles leading to reduced placental perfusion
and consequently placental insuYciency (Brosens et al., 1972). Paradoxical-
ly, even in the normal placenta, there is little or no invasion of the uterine
venules. The primary event that contributes to the failed trophoblast inva-
sion/diVerentiation in preeclampsia is unknown but genetic, immunological,
and environmental factors (such as hypoxia and nutritional deficiencies) are
thought to play a role. Extensive studies (by Fisher et al., 1981) suggest that
diVerences in O2 tension may be the governing factor that regulates the
invasive behavior of the cytotrophoblasts (Genbacev et al., 1996, 1997).
The remodeling of the spiral arterioles is thought to begin in late first
trimester and is complete by 18 to 20 weeks. Although the exact gestational
age at which the trophoblast invasion of these arterioles ceases is unclear,
histological studies show that fewer invasive trophoblasts are seen in the
decidua with increasing gestational age.
DiVerentiation of trophoblasts along the invasive pathway involves alter-
ation in expression of a number of diVerent classes of molecules, including
cytokines, adhesion molecules and extracellular matrix molecules, metallo-
protei
UNCORRECTED
nases and class Ib major histocompatibility complex molecule, HLA‐G
(Damsky et al., 1992, 1994; Fisher and Damsky, 1993). During normal
diVerentiation, invasive trophoblasts alter their adhesion molecule expression
from those that are characteristic of epithelial cells (integrin 6/4, v/5, and
E‐cadherin) to those of endothelial cells (integrin 1/1, v/3, PECAM,
and VE‐cadherin), a process referred to as pseudo‐vasculogenesis (Zhou

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:44
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
300
Yuan et al.
et al., 1993, 1997b). Both in vitro and in vivo studies show that trophoblasts
obtained from patients with preeclampsia fail to undergo these alterations of
adhesion molecules and pseudo‐vasculogenesis (Lim et al., 1997; Zhou et al.,
1997a), a finding that is disputed by some groups (Kaufmann et al., 2003). The
molecular pathways that regulate pseudo‐vasculogenesis may involve a vast
array of transcription factors, growth factors, and cytokines (Zhou et al.,
2003). Considerable attention has recently been focused on angiogenesis‐
related gene products such as VEGF, Angiopoietin/Tie, and
AU1
OF
Ephrine family
proteins and their role in regulating pseudo‐vasculogenesis and invasiveness.
Interestingly, invasive trophoblasts have been found to express VEGF,
PlGF, VEGF‐C, and their receptors. Furthermore, blocking their signaling
pathways decreases the expression of integrin 1 (a marker of pseudo‐
vasculogenesis) in vitro (Zhou et al., 2002). However, in vivo evidence directly
linking abnormalities of VEGF signaling to impai
PRO
red pseudo‐vasculogenesis
is lacking at the present time. More recently, the invasive trophoblasts were
also found to express L‐selectin, an adhesion molecule that mediates leuko-
cyte migration from blood to tissues (Genbacev et al., 2003). It has been
hypothesized that abnormalities of the selectin system at the fetal‐
maternal interface may account for implantation failures and preeclampsia.
Finally, trophoblast expression of HLA‐G, a nonclassical class I molecule
that has shown to be decreased in preeclampsia, has been hypothesized
to protect trophoblasts from NK cell attack at the implantation site
(MoVett‐King, 2002).
Long‐standing and severe preeclampsia is associated with placental
changes such as atherosis, fibrinoid necrosis, thrombosis, and placental in-
farction (De Wolf et al., 1975). Although not all these lesions are uniformly
found in patients with preeclampsia, there appears to be a correlation
between the severity of the disease and the extent of the lesions. Further-
more, in about one third of preeclamptic women (especially in those with
term preeclampsia) these placental changes are not present. Abnormal re-
modeling of the spiral arterioles results in placental ischemia, which in turn
is thought to lead to the secretion of soluble factors into the maternal
bloodstream. However evidence establishing a causal relationship between
abnormal placentation and the maternal syndrome is lacking.
III. Systemic Endothelial Dysfunction (Stage 2)
Generalize UNCORRECTED
d endothelial dysfunction can account for most of the clinical
aspects of preeclampsia (Roberts, 1998): hypertension through disturbed
endothelial control of vascular tone, proteinuria from increased glomerular
vascular permeability, coagulopathy as a result of abnormal endothelial
expression of procoagulants, and liver dysfunction from hepatic ischemia.

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:44
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
9. Angiogenic Factors in the Pathogenesis of Preeclampsia
301
Data from several studies support the theory that the maternal response in
preeclampsia is secondary to generalized endothelial dysfunction. Studies
have reported increased circulating fibronectin, Factor VIII antigen, and
thrombomodulin, all markers of endothelial cell injury in patients with
preeclampsia (Friedman et al., 1995; Hsu et al., 1993; Taylor et al., 1991).
Flow‐mediated vasodilation has also been found to be impaired in pre-
eclamptic vessels further suggesting abnormal endothelial function (Cockell
and Poston, 1997; McCarthy et al., 1993). Decreased produ
OF
ction of endo-
thelial‐derived vasodilators such as prostacyclins, increased production of
endothelins as well as enhanced vascular reactivity to angiotensin II also
suggest abnormal endothelial function (Clark et al., 1992; Gant et al., 1973;
Mills et al., 1999). Renal biopsies from patients with preeclampsia reveal
diVuse glomerular endothelial swelling referred to as glomerular endothelio-
sis (Fisher et al., 1981). Finally, serum from
PRO
preeclamptic women causes
endothelial activation in human umbilical vein endothelial cells in vitro
(Roberts et al., 1992).
The identification of circulating factors mediating endothelial dysfunction
has been the source of great research interest for decades. Several groups
have reported alterations of cytokines/growth factors/chemicals such as
TNF‐, IL‐6, IL‐1, IL‐1, Fas ligand, neurokinin‐B, oxidized lipid pro-
ducts, and ADMA (asymmetric dimethyl arginine) that are released by the
placenta and/or other maternal sources in preeclampsia (Benyo et al., 2001;
Conrad et al., 1998; Page et al., 2000; Roberts and Cooper, 2001; Savvidou
et al., 2003). However, there is no evidence that any of these molecules are
etiological. Increased sensitivity to angiotensin II, a consistent feature of
preeclampsia, has been found to be secondary to increased bradykinin (B2)
receptor upregulation in preeclamptic patients (AbdAlla et al., 2001). This,
in turn, was found to lead to heterodimerization of B2 receptors with
angiotensin II type I receptors (AT1), and this AT1/B2 heterodimer was
shown to increase responsiveness to angiotensin II in vitro. It is unclear,
however, whether these alterations observed are pathophysiological or epi-
phenomena. Along similar lines, increased circulating concentrations of
agonistic antibodies to the angiotensin‐1 (AT‐1) receptor have been reported
in women with preeclampsia (Wallukat et al., 1999). Stimulation of the AT‐1
receptor by these autoantibodies might contribute to the vascular damage
and the enhanced angiotensin II sensitivity noted in preeclampsia (Dechend
et al., 2003; Xia et al., 2003). These antibodies have also been encountered
in other
UNCORRECTED
examples of vascular injury such as vascular rejection (Dragun
et al., 2005), suggesting that they may be secondary to the generalized
microangiopathy of preeclampsia.
Studies from several laboratories have demonstrated an increased placen-
tal expression and secretion of sFlt1 (soluble fms–like tyrosine kinase 1, see
following), a naturally occurring circulating vascular endothelial growth

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:44
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
302
Yuan et al.
factor (VEGF) antagonist in patients with preeclampsia (Koga et al., 2003;
Maynard et al., 2003; Zhou et al., 2002). Importantly, when administered
exogenously to rats, sFlt1 alone has been shown to be suYcient to induce a
preeclampsia‐like phenotype (Maynard et al., 2003).
A. Excess Circulating sFlt1, Impaired VEGF Signaling, and
Antiangiogenic State
VEGF is an endothelial‐specific mitogen that plays a key role OF
in promoting
angiogenesis. VEGF’s activities are mediated primarily by its interaction
with two high‐aYnity receptor tyrosine kinases: kinase‐insert domain region
(KDR) and fms‐like tyrosine kinase‐1 (Flt‐1) that are selectively expressed
on vascular endothelial cell surface (Dvorak, 2002). PRO
Alternative splicing of
Flt‐1 results in the production of an endogenously secreted protein referred
to as sFlt1, which lacks the cytoplasmic and transmembrane domain, but
retains the ligand‐binding domain (He et al., 1999b; Kendall and Thomas,
1993). Thus, sFlt1 can antagonize circulating VEGF by binding to it and
preventing VEGF’s interaction with its endogenous receptors located in the
vasculature (Fig. 1). sFlt1 also binds and antagonizes placental growth
factor (PlGF), another member of the VEGF family that is made in the
placenta predominantly (Fig. 1).
In vitro studies confirm that excess placental sFlt1 production induces an
antiangiogenic state in the serum of preeclamptic women that can be rescued
by exogenous VEGF and PlGF (Maynard et al., 2003). sFlt1 alone, when
administered to pregnant rats, induced albuminuria, hypertension, and renal
pathological changes of glomerular endotheliosis by antagonizing circulat-
ing VEGF and PlGF and inducing endothelial dysfunction. In addition,
circulating levels of free VEGF and free PlGF were found to be decreased in
conjunction with elevated sFlt1 in the bloodstream at the time of disease
presentation (Chaiworapongsa et al., 2004; Koga et al., 2003; Maynard
et al., 2003; Tsatsaris et al., 2003). Although sFlt1 is made in small amounts
by other tissues (endothelial cells and monocytes), the placenta seems to be the
major source of circulating sFlt1 during pregnancy as evidenced by a dramat-
ic fall in circulating concentrations of sFlt1 after delivery of the placenta
(Maynard et al., 2003). The increase in sFlt1 precedes the onset of clinical
disease by at least 5 weeks (Hertig et al., 2004; Levine et al., 2004) and appears
to be more UNCORRECTED
pronounced in severe and early‐onset preeclampsia (Levine et al.,
2004). More recently, when free PlGF and free VEGF was measured through-
out pregnancy, it was found to be decreased in preeclamptics well before the
onset of clinical disease (Levine et al., 2004; Polliotti et al., 2003; Taylor et al.,
2003). Finally, decreased urinary PlGF has also been reported to precede
clinical preeclampsia (Levine et al., 2005).

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:44
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
9. Angiogenic Factors in the Pathogenesis of Preeclampsia
303
OF
PRO
Figure 1 Mechanism of action of sFlt‐1. sFlt‐1 protein, derived from alternative splicing of
Flt‐1 lacks the transmembrane and cytoplasmic domains, but still has the intact VEGF and
PlGF binding extracellular domain. During normal pregnancy, VEGF and PlGF signal through
the VEGF receptors (Flt‐1) and maintain endothelial health. In preeclampsia, excess sFlt‐1
binds to circulating VEGF and PlGF thus impairing normal signaling of both VEGF and PlGF
through their cell‐surface receptors. Thus excess sFlt‐1 leads to maternal endothelial
dysfunction. Reproduced with permission from Bdolah et al., 2004. (See Color Insert.)
VEGF is known to stimulate angiogenesis, as well as promote vasodila-
tion by stimulating NO and prostacyclin formation (signaling molecules that
are decreased in preeclampsia) (He et al., 1999a). Furthermore, a significant
percentage of cancer patients receiving VEGF‐signaling antagonists develop
hypertension and proteinuria (Kabbinavar et al., 2003; Yang et al., 2003).
Even a 50% reduction of renal VEGF production in genetically modified
mice resulted in glomerular endotheliosis and proteinuria (Eremina et al.,
2003). These data suggest that excess sFlt1, by neutralizing VEGF and
PlGF, may play a causal role in the pathogenesis of the maternal syndrome
in preeclampsia. Data suggests that VEGF may be particularly important in
maintaining the health of fenestrated endothelium (Risau, 1998), which is
found in
UNCORRECTED
the renal glomerulus, choroid plexus, and the hepatic sinusoids—
organs disproportionately aVected in preeclampsia. It has been shown that
VEGF induces endothelial fenestrae in vitro (Esser et al., 1998) and even a
50% decrease in VEGF production in the glomerulus in mice leads to
not only glomerular endotheliosis but also loss of glomerular endothelial
fenestrae (Eremina et al., 2003).

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:45
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
304
Yuan et al.
B. sFlt1 and Maternal–Fetal Conflict
Substantial evidence suggests that sFlt1 is, at least partially, responsible for
the endothelial dysfunction of preeclampsia. Why then does the placenta
release a factor that damages the maternal endothelium? A conventional
interpretation would be that sFlt1 performs some other function and
that endothelial damage is an occasional maladaptive side eVect of its release
into the maternal circulation, perhaps in women who are parti OF
cularly ‘‘sus-
ceptible.’’ Another possibility should also be considered. The placenta may
release sFlt1 into the maternal circulation to cause endothelial dysfunc-
tion because the associated vasoconstriction benefits the fetus by directing
a greater share of maternal cardiac output to the intervillous space of the
placenta. In this view, stage II of preeclampsia is an adaptive response of the
conceptus to the placental insuYciency arising from PRO
stage I of preeclampsia
(Haig, 1993, 1996).
During pregnancy, the maternal systemic circulation can be conceptua-
lized as consisting of two subcirculations, placental and nonplacental,
arranged in parallel: the placental subcirculation consists of all the vessels
that supply maternal blood to the intervillous space of the placenta; the
nonplacental subcirculation consists of all vessels that supply other tissues
of the maternal body (Fig. 2). Increases in the nonplacental resistance (Rn),
as occurs in preeclampsia, would result in increased blood flow through
the placental subcirculation, other things being equal (i.e., for unchanged
cardiac output and placental resistance, Rp). More generally, any increase
in the ratio of nonplacental to placental resistance (Rn/Rp) will result in a
larger fractional share of maternal cardiac output flowing through the
intervillous space (Haig, 1999).
Figure 2 A simple model of the maternal circulation during pregnancy. Maternal systemic
cardiac
UNCORRECTED
output is shared between placental and nonplacental subcirculations. In stage I of
preeclampsia, inadequate modification of maternal spiral arterioles results in increased
placental resistance (Rp). In stage II of preeclampsia, placental release of sFlt1 causes a
disproportionate increase in the nonplacental resistance (Rn) relative to the placental resistance
(Rp). As a result, an increased share of maternal cardiac output is directed to the placental
subcirculation (modified from Haig, 1999).

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:46
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
9. Angiogenic Factors in the Pathogenesis of Preeclampsia
305
The hypothesis that preeclampsia is an adaptation of malnourished fe-
tuses to increase their supply of nutrients posits that the endothelial dysfunc-
tion of preeclampsia disproportionately increases nonplacental resistance
relative to placental resistance. This seems plausible—given the remodeling
of spiral arterioles that occurs during the first half of pregnancy—but is yet
to be experimentally demonstrated. The hypothesis is based on the evolu-
tionary theory of parent–oVspring conflict (Trivers, 1974) that what is
‘‘best’’ for a parent is not always ‘‘best’’ for an oVspring,
OF
and vice versa.
This conflict is illustrated in the clinical dilemmas of treating preterm pre-
eclampsia: the longer that induction of delivery is delayed, the greater the
risk to a mother’s health but the greater the benefits to the fetus.
If the induction of preeclampsia is an adaptation to enhance fetal nutri-
tion, then the adaptation need not be simple and could involve the release of
multiple placental factors that target diVerent
PRO
physiological systems of the
mother. That is, sFlt1 may be just one component, albeit an important
component, of a cocktail of substances that are released into maternal blood
in nutritionally compromised pregnancies.
C. Speculations about the Mechanisms of Preeclampsia
If sFlt‐1 is an important cause of preeclampsia, there might be at least
two kinds of predisposing factors. One might involve the overproduction
of sFlt‐1. Conditions falling in this category might include multiple gesta-
tion, hydatiform mole, trisomy 13, and, possibly, first pregnancy. Another
set of predisposing factors would include disorders that sensitize the mater-
nal vascular endothelium to the antiangiogenic eVects of sFlt‐1. Such factors
might include obesity, preexisting hypertension or renal disease, diabetes,
and preexisting vasculitis. It is interesting that the alteration in the angio-
genic factors in the serum of obese patients with preeclampsia was somewhat
lower than that in lower‐weight preeclamptic patients (Levine et al., 2004;
Thadhani et al., 2004a,b). It is not yet known whether diabetes, hyperten-
sion, and preexisting renal disease predispose to preeclampsia by increasing
the production of sFlt‐1 or by sensitizing the vascular endothelium to its
presence.
D. Unans UNCORRECTED
wered Questions
There are limitations and several unanswered questions to the sFlt1 story.
The precise mechanisms of excess sFlt1 production by the placenta are not
known and importantly, the role of sFlt1 in normal placental development
and in placental pseudo‐vasculogenesis is not clear. No coagulation or liver

---

Comp. by:dshanthi
Date:1/8/05
Time:15:46:46
Stage:First Proof
File Path://
spsind002s/serials/PRODENV/000000~1/001BFD~1/S00000~1/00D23E~1/000000~3/
000009218.3D
Proof by:
QC by:
ProjectAcronym:bs:CTDB
Volume:71009
306
Yuan et al.
function abnormalities or brain abnormalities (eclampsia) were reported in
sFlt1‐treated animals. Moreover, genetic studies provide little support for a
role for sFlt1. For example, both an Australian/New Zealand cohort (Moses
et al., 2000) and an Icelandic cohort (Arngrimsson et al., 1999) have sug-
gested a maternal susceptibility locus on chromosome 2, bearing no known
relationship to sFlt1. One possible interpretation is that these studies detect a
locus responsible for susceptibility to stage I of preeclampsia, whereas
placental production of sFlt1 is responsible for stage II of OF
the disease.
Although, it is also possible that such loci are associated with transcription
factors or splicing factors aVecting sFlt1 production, it seems more likely
that there are other yet unidentified genetic factors that contribute to this
multifactorial disease.
On the other hand, the hypothesis that excessive production of sFlt1 may
play a causal role in preeclampsia is supported by
PRO
studies of the occurrence
of this syndrome in mothers of infants with trisomy 13. The genes for sFlt1
and Flt‐1 are carried on chromosome 13. Fetuses with an extra copy of
this chromosome should theoretically produce more of these gene products
than their normal counterparts. The incidence of preeclampsia in mothers
who carry fetuses with trisomy 13 is in fact greatly increased, as compared
with all other trisomies or with control pregnant patients (Tuohy and James,
1992). Recently it has been noted the women carrying trisomy 13 fetuses
have a greater concentration of circulating sFlt1 as compared to normal
karyotype controls thus providing a molecular explanation for the increased
risk of preeclampsia observed in these patients (Bdolah and Karumanchi,
unpublished observations).
Serum concentrations of sFlt‐1 have been found to be modestly elevated in
patients with IUGR without preeclampsia (Tsatsaris et al., 2003), a finding
that has not been confirmed by others (Shibita et al., 2004). Finally, although
sFlt‐1 was elevated in most patients with preeclampsia, it was not elevated
in some patients with mild preeclampsia (Levine et al., 2004). Hence it is likely
that additional synergistic factors that are elaborated by the placenta may
yet be identified that play a role in the pathogenesis of the generalized
endothelial dysfunction and vascular damage noted in preeclampsia.
IV. Conclusions
In summ
UNCORRECTED
ary, preeclampsia is believed to be a two‐stage disease with an initial
placental syndrome that is followed by the maternal syndrome (Fig. 3). The
maternal syndrome in preeclampsia is a state of generalized endothelial
dysfunction secondary to excessive amounts of circulating antiangiogenic
factors (such as sFlt1) that are released by the diseased placenta (Fig. 3). The
excess sFlt1 theory in the pathogenesis of preeclampsia fits very well with

---