Kidney International, Vol. 67 (2005), pp. 2101–2113

PERSPECTIVES IN RENAL MEDICINE

Preeclampsia: A renal perspective

S. ANANTH KARUMANCHI, SHARON E. MAYNARD, ISAAC E. STILLMAN, FRANKLIN H. EPSTEIN,and VIKAS P. SUKHATME

Renal Division and Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston,Massachusetts; Department of Obstetrics & Gynecology, Beth Israel Deaconess Medical Center and Harvard Medical School,Boston, Massachusetts; and Department of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston,Massachusetts

Preeclampsia: A renal perspective. Preeclampsia is a syndromethat affects 5% of all pregnancies, producing substantial mater-nal and perinatal morbidity and mortality. The aim of this reviewis to summarize our current understanding of the pathogenesisof preeclampsia with special emphasis on the recent discoverythat circulating anti-angiogenic proteins of placental origin mayplay an important role in the pathogenesis of proteinuria andhypertension of preeclampsia.

Preeclampsia, the syndrome of hypertension and pro-teinuria that heralds the seizures of eclampsia, remainsone of the great mysteries in the field of obstetrics.Although our understanding of the pathophysiologyof preeclampsia has increased over the past 50 years,it is quite incomplete, and management remains sup-portive: close observation, treatment with antihyperten-sive agents and magnesium sulfate, and if progressivesigns and symptoms occur, urgent delivery of the fetus.Preeclampsia is still one of the leading causes of maternaland neonatal mortality in the world.

Preeclampsia is characterized by a constellation ofsigns and symptoms, including the new onset of hy-pertension and proteinuria during the last trimester ofpregnancy, usually associated with edema and hyper-uricemia [1, 2]. It occurs only in the presence of theplacenta, even when there is no fetus (as in hydatidi-form mole) and remits dramatically postpartum [3]. Theplacenta in preeclampsia is usually abnormal, with ev-idence of hypoperfusion and ischemia. Although theseplacental changes are neither universal nor specific forpreeclampsia, there appears to be a correlation between

Key words: pregnancy, HELLP syndrome, angiogenesis, pseudovascu-logenesis, VEGF, PlGF, soluble Flt-1, soluble VEGFR-1, proteinuria,edema.

Received for publication February 9, 2004and in revised for April 23, 2004Updated on November 23, 2004Accepted on January 12, 2005

C© 2005 by the International Society of Nephrology

the severity of the disease and the extent of placental ab-normalities. The clinical findings of severe preeclampsiaare unified by the presence of systemic endothelial dys-function and microangiopathy, in which the target organmay be the brain (seizures or eclampsia), the liver [thehemolysis, elevated liver function tests, and low plateletcount (HELLP) syndrome], or the kidney (glomerularendotheliosis and proteinuria). Severe preeclampsia isalso associated with small for gestational age (SGA) fe-tuses. Because of the strong clinical and experimental ev-idence of early placental involvement and dysfunction ofthe maternal endothelium, it is currently believed thatpreeclampsia has its origin in disordered vascular devel-opment of the placenta, which in turn leads to widespreadmaternal vascular endothelial effects [4, 5] (Fig. 1). Thisreview will discuss mechanisms underlying the clinicalmanifestations of preeclampsia. In addition, we will de-scribe evidence suggesting that placental secretion of anantiangiogenic protein may contribute to the endothelialdysfunction of preeclampsia.

RISK FACTORS FOR THE DEVELOPMENT OFPREECLAMPSIA

The risk factors for the development of preeclampsiaare listed in Table 1 [1, 5–7]. Preeclampsia is morecommon not only in first pregnancies, but also in multi-gravidas who have a new partner, suggesting that priorexposure to paternal antigens may be protective [8, 9].However, recent evidence from a large Norwegian birthregistry suggests that prolonged interpregnancy interval,rather than primipaternity, accounts for this increase inrisk [10], though why this occurs is unclear. Althoughpreeclampsia is traditionally not considered to be agenetic disease, it is clear that genetic factors contributeto the susceptibility to preeclampsia [6]. A family history(mother, sister, or both) of preeclampsia is associatedwith a fourfold increased risk for preeclampsia [6]. Otherepidemiologic studies suggest that paternal geneticcontributions to the zygotic genotype in addition to

2101

2102 Karumanchi et al: Preeclampsia and angiogenic factors

Table 1. Risk factors for the development of preeclampsia

Family history of preeclampsiaNulliparityMultiple gestationMolar pregnanciesOlder maternal ageObesityPreexisting hypertensionChronic renal diseaseDiabetes mellitusThrombotic vascular disease

maternal genes may contribute susceptibility topreeclampsia [11]. Polymorphisms of genes involved inthe regulation of blood pressure or coagulation, such asrenin, angiotensinogen (T235), endothelial nitric oxidesynthase (eNOS), prothrombin, factor V Leiden, andmethyltetrahydrofolate (MTHFR), though promising inearly studies [12–15], have not been confirmed in largerstudies [16–20]. Genome-wide scanning of Icelandicfamilies revealed a significant locus on chromosome 2p13[logarithm of the odds (LOD) score 4.70] [21]. Morerecently, the 2p locus was confirmed in a study of patientsfrom New Zealand and Australia [22]. Finally, a Dutchstudy reported linkage of HELLP syndrome, but notpreeclampsia, with a locus on 12q, suggesting that geneticfactors important in HELLP (hemolysis, elevated liverfunction tests, low platelets) syndrome may be distinctfrom those in preeclampsia [23]. Women with trisomy 13fetuses have a higher incidence of preeclampsia, regard-less of parity, suggesting that a gene on chromosome 13may be important in preeclampsia [24]. An activatingmineralocorticoid receptor mutation was described in arare group of patients who have only pregnancy-inducedhypertension without proteinuria [25]. The incidence ofpreeclampsia is also higher in women who live at highaltitudes and in the third world, suggesting that hypoxiaand/or hitherto unknown environmental factors mayalso contribute to the development of preeclampsia [26,27].

CLINICAL MANIFESTATIONS AND THEIRPATHOPHYSIOLOGIC UNDERPINNINGS

Hypertension

While in normal pregnancy peripheral vascular resis-tance and blood pressure are decreased, in preeclamp-sia these changes are reversed. Increased peripheralvascular resistance, rather than increased cardiac output,is the chief cause of hypertension [28]. Sympathetic ac-tivation is noted in preeclampsia as it is in other formsof hypertension, a conclusion supported by electricalrecordings of sympathetic nerve impulses [29] and byreports of increased concentrations of circulating cate-cholamines [30]. Sympathetic activation may be respon-

sible for the increase in cardiac output noted by some inthe early stages of preeclampsia [31]. Preeclampsia is alsonotable for an exaggerated response to angiotensin II,catecholamines, and other hypertensive stimuli whencompared to normal pregnant controls [32, 33]. Onegroup has reported that this response may precede theonset of overt hypertension by weeks to months [34](a finding disputed by others [35]), reminiscent of theexaggerated response to vasoconstrictors described innormotensive relatives of patients with essential hyper-tension [36].

Although total plasma volume has been reported tobe low in preeclampsia [37], there may be increased“effective circulating volume” as evidenced by sup-pressed renin and aldosterone [38, 39] and elevated brainnatriuretic hormone [40] relative to normal pregnancy.These findings are reminiscent of the fall in plasma vol-ume and rise in blood pressure produced by infusionof vascoconstrictors suggesting that peripheral vasocon-striction, especially of small venules, shifts blood to thearteries and central veins while elevating capillary pres-sure [41]. Blood pressure rises, and renin and aldosteronelevels fall as a secondary phenomenon.

Generalized vascular constriction is universallypresent in preeclampsia, at least compared to the physi-ologic vasodilation of normal pregnancy [28]. There issubstantial evidence that this may be due to endothelialdysfunction. A myriad of markers for endothelial acti-vation and dysfunction, including endothelin, cellularfibronectin, plasminogen activator inhibitor-1 (PAI-1),and von Willebrand’s factor, are altered in preeclampsia.Women with preeclampsia have enhanced responsive-ness to vasopressors as compared with normal pregnantwomen. Women with a history of preeclampsia exihibitevidence of impaired endothelial-dependent vasore-laxation as measured by brachial artery flow–mediateddilatation up to 3 years after delivery, implying thesechanges in the maternal endothelium may be more thantransient [42, 43]. Alterations of endothelial functionhave been noted in preeclamptic vessels examined invitro, supporting the hypothesis that endothelial dys-function may underlie the hypertension of preeclampsia[4, 5]. The increased incidence of preeclampsia in womenwith chronic diseases such as diabetes and hypertensionalso suggests some factor in the maternal milieu may alsolend susceptibility to preeclampsia. In addition to in-creased vascular reactivity, the vasoconstriction appearsto be mediated at least in part by alterations in localconcentrations of several vasoactive molecules, includingthe vasoconstrictors norepinephrine, endothelin, andperhaps thromboxane, and the vasodilators prostacyclinand perhaps nitric oxide.

Prostaglandin I2 (PGI2) (prostacyclin), a circu-lating vasodilator produced primarily by the en-dothelial and smooth muscle layers of blood vessels, is

Karumanchi et al: Preeclampsia and angiogenic factors 2103

Geneticfactors

Environmentalfactors

Other

Placentaldysfunction

Circulatingfactor(s)

Vascular endothelial dysfunction

Hypertension Glomerularendotheliosisproteinuria edemarenal insufficiency

HeadacheCerebral edemaSeizures

↑LFTsHELLP syndrome

Fig. 1. Placental dysfunction and endothelialdysfunction in the pathogenesis of preeclamp-sia. Placental dysfunction, triggered by poorlyunderstood mechanisms, which may includegenetic, immunologic, and environmental fac-tors, plays an early and primary role in thedevelopment of preeclampsia. The diseasedplacenta in turn secretes a factor(s) intothe maternal circulation, causing systemicendothelial cell dysfunction. Most of themanifestations of preeclampsia, includinghypertension, proteinuria (glomerular en-dotheliosis), seizures (cerebral edema and/orvasospasm), and the hemolysis, elevatedliver function tests, and low platelet count(HELLP) syndrome can be attributed to vas-cular and endothelial effects.

Fig. 2. Glomerular endotheliosis. (A) Nor-mal human glomerulus (hematoxylin andeosin stain). (B) Human preeclampticglomerulus (hematoxylin and eosin stain). A33-year-old woman with twin gestation andsevere preeclampsia at 26 weeks’ gestationwith urine protein/creatinine ratio of 26 atthe time of biopsy. (C) Electron microscopyof glomerulus of the above patient describedin (B). Note occlusion of capillary lumen cy-toplasm and expansion of the subendothelialspace with some electron-dense material.Podocyte cytoplasms show protein resorptiondroplets and relatively intact foot processes(original magnification 1500×). (D) Controlrat glomerulus (hematoxylin and eosinstain). Note normal cellularity and opencapillary loops. (E) Soluble Flt-1–treated rat(hematoxylin and eosin stain). Note occlusionof capillary loops by swollen cytoplasm withminimal increase in cellularity. (F) Electronmicroscopy of sFlt-1–treated rat. Note occlu-sion of capillary loops by swollen cytoplasmwith relative preservation of podocyte footprocesses (original magnification 2500×). Alllight micrographs taken at identical originalmagnification of 40×).

2104 Karumanchi et al: Preeclampsia and angiogenic factors

increased in normal pregnancy [44]. In preeclampsia,prostacyclin production is lower than in normal preg-nancy well before the onset of hypertension and protein-uria [45, 46]. Endothelial cells incubated with serum fromwomen with preeclampsia (vs. serum from normotensivepregnant women) produce less prostacyclin in vitro, sug-gesting the presence of a circulating factor suppressingprostacyclin synthesis [47]. Prostacyclin concentrationsare normal in pregnant women with chronic hyperten-sion but without preeclampsia [48], providing furthercircumstantial evidence that decreased prostacyclin is acontributing cause of preeclampsia rather than a conse-quence of hypertension.

Thromboxane A2 (TXA2) is a potent vasoconstric-tor synthesized by endothelial cells, activated platelets,macrophages, and other organs. Urinary excretion ofTXA2 metabolites has been reported to be increased inpreeclampsia by some [49] but not all investigators [46,50]. TXA2 production appears to parallel the severity ofpreeclampsia, being higher in patients with coagulopathyand pronounced platelet activation [50]. Some attemptsto prevent preeclampsia in high risk patients by inhibitingplatelet thromboxane synthesis using aspirin were suc-cessful but others found no benefit, either in low risk orhigh risk populations [51, 52].

There is some evidence that nitric oxide, a vascularsmooth muscle relaxant synthesized by vascular endothe-lium, may also mediate the generalized vasodilation ofnormal pregnancy [53, 54]. In rodent experiments, ni-tric oxide inhibition during pregnancy induces hyper-tension [55–57] and proteinuria [58] and reverses thenorepinephrine and angiotensin II resistance character-istic of pregnancy [59, 60]. Thus, damage to vascular en-dothelium with decreased nitric oxide production mightcontribute to vascular constriction in preeclampsia. Al-though some human studies have reported decreasednitric oxide production in preeclampsia [61–63], othersreport nitric oxide production to be unchanged [64, 65]or even increased [66, 67]. Unfortunately, studies of ni-tric oxide production are difficult due to its extremelyshort circulating half-life and other challenges to itsmeasurement.

Endothelin-1, a potent vasoconstrictor that can be re-leased by vascular endothelial cells in response to in-jury [68], has also been implicated in the hypertension ofpreeclampsia. Reports of plasma endothelin concentra-tions in preeclampsia have been mixed, with most [69–71]but not all [50] investigators reporting a modest increasein circulating concentrations. The release of endothelinby cultured endothelial cells is enhanced by exposureto plasma from preeclamptic patients, as compared withplasma from women with normal pregnancies [72]. Lim-ited animal data suggest that hypertension induced byendothelin-1 administration during pregnancy producesproteinuria [73].

Renal dysfunction, proteinuria, and renal pathologyIn normal pregnancy, glomerular filtration rate (GFR)

as measured by inulin clearance and renal plasma flow(para-aminohippurate clearance) increases by 40% to60% during the first trimester [74, 75], resulting in a fall inserum markers of renal clearance, including blood ureanitrogen (BUN), creatinine, and uric acid. In preeclamp-sia, both GFR and renal plasma flow decrease by 30%to 40% compared with normal pregnancy of the sameduration [76, 77]. Rarely, prolonged renal hypoperfusionwith resulting “acute tubular necrosis” can occur in severepreeclampsia. It is important to note that in preeclamp-sia, BUN and creatinine often remain in the normal rangefor nonpregnant women despite a significant decrease inGFR from the high level of normal pregnancy.

Proteinuria may rarely precede hypertension but usu-ally accompanies or follows it. After pregnancy is ter-minated, proteinuria commonly disappears within 3 to 8weeks, but occasionally persists for months. The quantityof protein excreted in the urine varies widely from lessthan a gram to 8 to 10 g per day. The urinary sediment isusually bland; red blood cells and cellular casts are rare.Preeclampsia is the leading cause of nephrotic syndromeduring pregnancy. Recent data suggest that a loss of bothsize and charge selectivity of the glomerular barrier con-tribute to the development of albuminuria [77].

Preeclampsia is associated with a characteristicglomerular lesion, “glomerular capillary endothelio-sis” (Fig. 2A to C) [78–80]. By light microscopy, theglomeruli are enlarged and the glomerular capillary lu-men is “bloodless” due to endothelial and mesangial cellswelling and hypertrophy. While these changes may befocally present in other conditions (such as abruptio pla-centae), only in preeclampsia is endotheliosis so promi-nent and widespread. Glomerular cellularity is at mostslightly increased. Mesangial interposition may occur insevere cases or in the healing stages. Glomerular vis-ceral epithelial cells (podocytes) usually appear swollenwith periodic acid-Schiff (PAS)-positive hyaline droplets.Immunofluorescence may reveal deposition of fibrinor fibrinogen derivatives particularly in biopsies donewithin 2 weeks postpartum [81]. Electron microscopyis important to confirm endotheliosis and the loss ofendothelial fenenstrae. Remarkably, despite heavy pro-teinuria the podocyte foot processes are relatively pre-served [82]. Glomerular subendothelial and occasionalmesangial electron-dense deposits can be seen. Theselikely relate to fibrin or related breakdown products. Mildglomerular endotheliosis has been noted in up to 50%of patients with pregnancy-induced hypertension with-out proteinuria [83], suggesting that pregnancy-inducedhypertension may in some cases reflect an earlier ormilder form of the same syndrome. Glomerular enlarge-ment and endothelial swelling usually disappear within8 weeks of delivery, coinciding with resolution of the

Karumanchi et al: Preeclampsia and angiogenic factors 2105

hypertension and proteinuria. Focal segmental glomeru-losclerosis (FSGS) can accompany the generalizedglomerular endotheliosis of preeclampsia in 50% or moreof cases [84, 85].

Edema

Although the presence of edema is not necessary forthe diagnosis, sudden weight gain, with edema of thefeet, hands, and face, is a common presenting symp-tom in preeclampsia [32]. Salt loads are excreted moreslowly than in normal pregnancy [86, 87]. The edema ofpreeclampsia is unlike that in congestive heart failure,hepatic cirrhosis, and nephrotic syndrome, where loweffective circulating volume leads to high plasma reninand aldosterone concentrations and secondary renalsodium retention (“underfill” edema) [88]. Instead, reninand aldosterone concentrations are suppressed relativeto normal pregnancy [39]. The edema of preeclampsiathus resembles the “overfill” edema of acute glomeru-lonephritis or of acute ischemic renal failure in whichGFR is decreased out of proportion to the drop in renalplasma flow and in which glomerular-tubular imbalancehas been invoked as a cause of salt retention.

Decreased GFR, increased capillary permeability, andhypoalbuminemia may contribute to the edema for-mation. Albumin-bound Evan’s blue dye disappearsmore quickly from the intravascular space in preeclamp-tic women compared with normal pregnant women,suggesting endothelial permeability is increased [89].However, this phenomenon is also seen in non-pregnantpatients with heart failure and the nephrotic syndrome[90, 91]. Hypoalbuminemia is common and may con-tribute to edema. However, correction of hypoalbumine-mia with albumin infusions in preeclampsia patientsusually does not produce a diuresis, suggesting that thehypoalbuminemia alone is not responsible for the edema[92].

Hyperuricemia

The association between preeclampsia and elevatedserum uric acid was first noted more than 80 yearsago [93]. Serum uric acid levels are usually elevated inpreeclampsia, and the degree of uric acid elevation hasbeen correlated with the severity of proteinuria [94], renalpathologic changes [83], maternal morbidity [95], and fe-tal demise [96]. Often, the elevation of uric acid precedesthe onset of proteinuria and fall in GFR [97]. Althoughit has been proposed that uric acid generation may be in-creased in preeclampsia as a result of tissue ischemia [98],most evidence suggests that decreased renal clearance isthe more important mechanism [99]. Similar decreases inuric acid clearance have been noted by infusions of vaso-constrictors in humans [100]. It has recently been sug-

gested that hyperuricemia may also directly contributeto vascular damage and hypertension [101, 102].

Coagulopathy and HELLP syndrome

Preeclampsia is sometimes complicated by consump-tive coagulopathy and thrombotic microangiopathy cul-minating in the HELLP syndrome. HELLP syndromedevelops in approximately 10% to 20% of women withsevere preeclampsia. The tendency of blood to coagulateis increased in normal pregnancy at least in part becauseof changes in the circulating concentrations of coagulantfactors. In the indolent microangiopathy of preeclamp-sia, it is mainly the factors that are synthesized in thevascular endothelium that are abnormal. These includeprostacyclins (PGI2), von Willebrand factor, thrombo-modulin, cellular fibronectin, and PAI-1 [103–106]. Evenin the absence of overt coagulopathy, serum and urine fib-rin degradation products are increased, implicating sub-clinical activation of the coagulation cascade.

These changes are not simply epiphenomena in re-sponse to hypertension. Plasma concentrations of cellu-lar fibronectin may be increased weeks before the onsetof hypertension [107]. Exposure of cultured endothelialcells to serum from women with preeclampsia triggersincreased cellular filbronectin [108, 109] and thrombo-modulin [110] release compared with serum from nor-motensive pregnant women, again suggesting that afactor in preeclamptic serum is directly “activating”endothelial cells. It is interesting to note that otherendothelial disorders, including thrombotic thrombocy-topenic purpura/hemolytic uremic syndrome, vasculitis,and disseminated intravascular coagulation can producesimilar alterations in markers of endothelial activation[111–113]. There is also evidence that platelet activationis present in preeclampsia as a consequence of endothe-lial damage. Markers of platelet activation such as betathromboglobulin have been found to be elevated prior tothe onset of clinical disease [114]. Furthermore, adhesionmolecules such as soluble P-selectin, soluble E-selectin,and soluble vascular cell adhesion molecule-1 (VCAM-1)that reflect endothelial, platelet, and leukocyte activationhave also been reported to be elevated in preeclampsia[115]. The severe thrombocytopenia that is occasionallyseen is likely due to consumption during intravascularcoagulation.

Neurologic abnormalities

Preeclampsia is occasionally complicated by the devel-opment of seizures (eclampsia). Other neurologic symp-toms include headache, blurred vision, and temporaryloss of vision. The neurologic changes in eclampsia andpreeclampsia have been attributed to cerebral edema andvasoconstriction. Brain edema by magnetic resonance

2106 Karumanchi et al: Preeclampsia and angiogenic factors

(MR) imaging and abnormalities in circulating endothe-lial markers correlate closely with the occurrence ofeclampsia, suggesting that endothelial damage and dis-ruption of the blood-brain barrier may be an underlyingmechanism [116]. The cerebral edema of eclampsia in-volves predominantly the posterior portions of the whitematter, and has been referred to as reversible poste-rior leukoencephalopathy syndrome (RPLS) [117]. In-terestingly, patients with thrombotic thrombocytopenicpurpura (TTP), another disorder with microangiopathyalso have features of RPLS on MR imaging examina-tions [118], as do some patients with hypertensive en-cephalopathy.

ABNORMAL PLACENTATION INPREECLAMPSIA

It was 1939, when E.W. Page first proposed thatplacenta plays a central role in preeclampsia [3]. Pla-centas from severe preeclamptic pregnancies classicallyhave numerous infarcts, sclerotic narrowing of arter-ies and arterioles, and fibrin deposition and thrombosis.Physiologic studies have confirmed that uteroplacentalblood flow is diminished and uterine vascular resistance isincreased in preeclamptic women. In addition, placentalischemia induced by mechanical constriction of the uter-ine arteries or aorta produces hypertension, proteinuria,and, variably, glomerular endotheliosis, in a variety ofspecies [119–121]. Placental ischemia may be contributedto by a decrease in placental production of nitric oxide[122]. However, placental ischemia alone, as in intrauter-ine growth restriction, does not appear to be sufficientto produce preeclampsia. Thus, although uteroplacentalischemia is an important trigger of preeclampsia, the re-sponse to this ischemia—either at the placental or thematernal systemic level—must be variable.

Evidence suggests that placental ischemia inpreeclampsia occurs as a result of defective placen-tal vascular remodeling. Early in normal placentaldevelopment, cytotrophoblast cells, fetal in origin, attachto the uterine deciduas via anchoring villi (see Fig. 3).Floating villi, containing fetal vessels which participate innutrient exchange, are bathed in maternal blood whichfills the intravillous space via uterine spiral arteries.During placental vasculogenesis, a small percentage ofcytotrophoblasts in the anchoring villi break throughthe syncytium and migrate into the endometrium.These extravillous trophoblasts invade the uterinespiral arteries of the myometrium, a process that peaksat about 12 weeks’ gestation. By 18 to 20 weeks, theendometrial and superficial myometrial segments ofthe spiral arteries are lined by cells of cytotrophoblastorigin, with transformation of these vessels from smallresistance vessels to flaccid, high-caliber capacitancevessels [123, 124]. This dramatic transformation allows

the increase in placental blood flow needed to sustainthe fetus through the pregnancy.

The characteristic pathologic placental lesion in severepreeclampsia is diminished endovascular invasion by cy-totrophoblasts and failure of uterine spiral arteriolar re-modeling [123, 125]. Cytotrophoblast invasion is limitedto the proximal decidua, and the myometrial segmentsof the spiral arteries remain narrow and undilated [126],resulting in uterine hypoperfusion. Zhou et al [127, 128]have shown that invasive cytotrophoblasts down-regulatethe expression of adhesion molecules characteristic oftheir epithelial cell origin and adopt a cell-surface ad-hesion phenotype typical of endothelial cells, a processreferred to as pseudovasculogenesis. They hypothesizedand confirmed that in preeclampsia, cytotrophoblast cellsfail to undergo this switching of cell-surface integrinsand adhesion molecules [129]. This work suggests thatcytotrophoblast differentiation is abnormal in severepreeclampsia, and may be an early defect that eventuallyleads to placental ischemia. Others have demonstratedthat hypoxia-inducible factor-1 (HIF-1) is up-regulated inpreeclampsia and have suggested that HIF-1 target genessuch as transforming growth factor-beta 3 (TGF-b3) mayblock the cytotrophoblast invasion [130–132]. More re-cently, heparin-binding epidermal growth factor (EGF)-like growth factor (HB-EGF) has been demonstrated tobe decreased in preeclamptic cytotrophoblasts; however,its role in cytotrophoblast differentiation and invasion isunclear [133]. Interestingly, uteroplacental ischemia pro-duced in monkeys by aortic constriction late in gestationdoes not produce the defective placental cytotrophoblastinvasion seen in preeclampsia, though it is associated withproteinuria and hypertension [134].

THE PLACENTAL FACTOR IN PREECLAMPSIA

All of the clinical manifestations of preeclampsia canbe attributed to endothelial cell dysfunction leading toend-organ damage and hypoperfusion. Based on thisobservation, it has been suggested that there may bea circulating factor, likely placental in origin, whichaffects systemic endothelial endothelial cell functionand leads to the clinical syndrome of preeclampsia [4,135]. Intense investigation has focused on the searchfor a placental factor which might induce the maternalsyndrome. Although many candidate factors, includingtumor necrosis factor-a (TNF-a), interleukin (IL)-6,IL-1a, IL-1b , Fas ligand, neurokinin B, and asymmetricdimethy L-arginine (ADMA) have been suggested, noneso far has been proven to be etiologic [1, 2, 5, 136, 137].Recent data suggest that increased syncytiotrophoblastshedding as a consequence of placental apoptosis maycontribute to endothelial dysfunction [138–140]; how-ever, there is no in vivo evidence so far that circulating

Karumanchi et al: Preeclampsia and angiogenic factors 2107

Placenta Anchoring villuscytotrophoblastcolumn

Decidua Myometrium

Cytotrophoblaststem cells

Cytotrophoblast

Cytotrophoblast Endovascularcytotrophoblast

Maternalendothelialcells

Syncytiotrophoblast

Maternal blood

Floating villus

Blood flow Spiralartery

Tunica mediasmooth muscle

Placenta Anchoring villuscytotrophobastcolumn

Decidua Myometrium

Cytotrophoblaststem cells

Cytotrophoblast

Cytotrophoblast

Maternalendothelial cells

SyncytiotrophoblastMaternal blood

Floating villus

Fetal bloodvessel

Spiral artery

Tunica mediavascular smoothmuscle layer

Lessbloodflow

Fig. 3. Abnormal placentation in preeclamp-sia. Exchange of oxygen, nutrients, and wasteproducts between the fetus and the motherdepends on adequate placental perfusion bymaternal vessels. In normal placental devel-opment, invasive cytotrophoblasts of fetalorigin invade the maternal spiral arteries,transforming them from small-caliber resis-tance vessels to high-caliber capacitance ves-sels capable of providing placental perfusionadequate to sustain the growing fetus. Duringthe process of vascular invasion, the cytotro-phoblasts differentiate from an epithelial phe-notype to an endothelial phenotype, a processreferred to as “pseudovasculogenesis” (upperpanel). In preeclampsia, cytotrophoblasts failto adopt an invasive endothelial phenotype.Instead, invasion of the spiral arteries is shal-low and they remain small caliber, resistancevessels (lower panel). This may result in theplacental ischemia.

VEGF

PIGF

FLT-1

sFLT-1

Healthyplacenta

Preeclampsiaplacenta

Blood vessel

Blood vessel

Anticoagulant andvasodilatory factors

Procoagulant andvasoconstricting factors

Healthy endothelial cell

Dysfunctional endothelial cell

Fig. 4. Soluble Flt-1 (sFlt-1) causes endothelial dysfunction by antagonizing vascular endothelial growth factor (VEGF) and placental growthfactor (PlGF). There is mounting evidence that VEGF, and possibly PlGF, are required to maintain endothelial health in several tissues includingthe kidney and perhaps the placenta. In normal pregnancy, the placenta produces modest concentrations of VEGF, PlGF, and soluble Flt-1. Inpreeclampsia, excess placental soluble Flt-1 binds circulating VEGF and PlGF and prevents their interaction with endothelial cell-surface receptors.This results in endothelial cell dysfunction, including decreased prostacyclin, nitric oxide production and release of procoagulant proteins such asvon Willebrand factor, endothelin, cellular fibronectin, and thrombomodulin.

2108 Karumanchi et al: Preeclampsia and angiogenic factors

placental debris induces preeclampsia-like features. Re-cently, preeclampsia-like features were noted in p57kip2 (acell-cycle inhibitor) knockout mice, however, the mech-anisms responsible for this phenotype are unclear [141].

Some [142], but not all [143], investigators have foundmarkers of oxidative stress to be elevated in womenwith preeclampsia, suggesting that generation of reactiveoxygen-free radicals may contribute to endothelial dys-function [2]. Decreased intake of the antioxidant vitaminC and low circulating ascorbic acid concentrations areassociated with an increased risk of preeclampsia [144].A small randomized controlled trial found that antiox-idant therapy decreased the incidence of preeclampsia,suggesting that oxidative stress may be a cause, ratherthan a consequence, of the syndrome [145]. Additionalstudies are needed to test this hypothesis.

As noted earlier, increased sensitivity to the vasopres-sor effects of angiotensin is a well-established featureof preeclampsia. Two recent advances have illuminatedpossible mechanisms for this phenomenon. AbdAllaet al [146] reported increased angiotensin-1/bradykininB2 receptor heterodimers in women with preeclampsia,and showed that such heterodimers can result in en-hanced angiotensin II responsiveness [146]. In a similarvein, Wallukat et al [147] noted increased concentrationsof agonistic antibodies to the angiotensin II type 1 (AT-1) receptor in women with preeclampsia. These autoan-tibodies may induce the production of reactive oxygenspecies and block cytotrophoblast invasion in vitro [148,149], thus accounting for several of the clinical features ofpreeclampsia. Further work is needed to establish if thesealterations are etiologic or epiphenomenona that mightbe encountered in other examples of microangiopathy.

Recently, gene expression profiling has been used tosearch for candidate factors produced by the placenta inpreeclampsia. Using this approach, we found that placen-tal sFlt-1 mRNA (soluble fms-like tyrosine kinase-1) isup-regulated in preeclampsia [150]. sFlt-1 is a splice vari-ant of the vascular endothelial growth factor (VEGF))receptor Flt-1, lacking the transmembrane and cyto-plasmic domains. sFlt is made in large amounts by theplacenta and is released into the maternal circulation[151–153]. sFlt-1 acts as a potent VEGF and placen-tal growth factor (PlGF) antagonist by binding thesemolecules in the circulation [154]. Circulating sFlt-1 con-centrations are increased in women with establishedpreeclampsia [150, 155–157]. Consistent with the antag-onistic effect of sFlt-1, free (or unbound) VEGF andfree (or unbound) PlGF concentrations are decreased inpreeclamptic women at disease presentation and evenbefore the onset of clinical symptoms [150, 158, 159] (seeFig. 4). When administered to pregnant and nonpregnantrats, sFlt-1 produces a syndrome of hypertension, pro-teinuria, and glomerular endotheliosis that mimics thehuman syndrome of preeclampsia (Fig. 2D to F) [150].

Recently, our laboratory as well as others have observedthat there is a marked rise in circulating sFlt-1 concen-tration beginning about 5 to 6 weeks before the onset ofclinical preeclampsia, accompanied by decreases in thecirculating free PlGF and VEGF [160, 161]. Moreover,alterations in these circulating angiogenic proteins corre-lated with disease severity, earlier onset of preeclampsia,and the birth of an infant small for gestational age. Finally,we have also recently reported that decreased urinaryconcentrations of free PlGF during midgestation pre-dict the subsequent development of clinical preeclampsia[162]. These data lend further support to the hypothe-sis that circulating sFlt-1 may have a pathogenic role inpreeclampsia.

The possibility that antagonism of VEGF and PlGFmight play a role in preeclampsia has sound physiologicunderpinnings (Fig. 4). In addition to being a potentpromoter of angiogenesis, VEGF is known to inducenitric oxide and vasodilatory prostacyclins in endothe-lial cells, decreasing vascular tone and blood pressure[163]. There is evidence from animal models that VEGFis important in maintaining glomerular endothelial cellhealth and healing [164, 165], and its absence induces pro-teinuria and glomerular endotheliosis [166, 167]. In an-tiangiogenic oncology trials, antagonism of VEGF usingneutralizing antibodies and VEGF receptor inhibitorscan produce headaches, hypertension, proteinuria, andcoagulopathy in human subjects [168–170]. Furthermore,exogenous VEGF and PlGF can reverse the antiangio-genic properties of preeclamptic serum, as assessed by invitro angiogenesis assays [150]. Thus, the antiangiogeniceffects of sFlt-1 may account for many of the manifes-tations of preeclampsia, including the unique glomeru-lar effects. Some tissues targeted in preeclampsia (i.e.,renal glomeruli, hepatic sinusoids) have fenestrated en-dothelia, which are necessary for physiologic diffusion ofwater and solutes. It has been shown that VEGF, whichis expressed constitutively in these organs, is necessaryto induce and maintain the health of the fenestrated en-dothelium [167, 171]. Thus, it is intuitive that tissues withfenestrated endothelium, especially the renal glomeru-lus, might be particularly susceptible to damage by sFlt-1–mediated VEGF blockade.

How placental dysfunction is related to placental sFlt-1 production, and why placental perfusion is derangedin preeclampsia remains unknown. Recent data usingin vitro primary cytotrophoblast cultures suggest thatplacental hypoxia may play an important role in up-regulating sFlt-1 production [172]. It remains to be shownif sFlt-1 can induce other features of preeclampsia suchas the HELLP syndrome or if additional synergistic fac-tors are needed. Additionally, the Flt-1 locus at 13q12 hasnot been localized within the genomic region identifiedby genetic linkage studies for preeclampsia. However, ifthe Flt-1 locus contributed to preeclampsia, one could

Karumanchi et al: Preeclampsia and angiogenic factors 2109

hypothesize that patients who carry trisomy 13 fetusesshould have a 50% increased risk of preeclampsia. Twocase-control studies done several years ago did in factdemonstrate higher incidence of preeclampsia in moth-ers who carry trisomy 13 fetuses as compared to other tri-somies or control pregnant patients [24, 173]. AlthoughsFlt-1 levels were not measured in the patients with tri-somy 13, these data are consistent with the hypothesis thatelevated production of sFlt-1 may lead to preeclampsia.

IMPACT OF PREECLAMPSIA ON LONG-TERMMATERNAL HEALTH

Women with hypertensive disorders of pregnancy havean increased long-term incidence of dyslipidemias, insulinresistance, and cardiovascular diseases [174–176]. Sibai,el-Nazer, and Gonzalez-Ruiz [177] reported an 18-fold in-crease in the incidence of chronic salt-sensitive hyperten-sion in women with a history of preeclampsia. It has beenspeculated that this late-onset hypertension may play arole in the eightfold increase in cardiovascular mortalitythat is also observed in these patients [176]. However, it isdifficult to determine causality, because many of the riskfactors for preeclampsia (such as diabetes mellitus, hyper-tension, and renal insufficiency) predispose to hyperten-sion and cardiovascular disease as well. It has also beensuggested that subtle renal injury caused by preeclampsiacan eventually lead to the development of chronic hyper-tension [178, 179]. It is tempting to speculate that the long-term cardiovascular complications noted in some patientswho have had preeclampsia may be due to a chronic an-tiangiogenic state resulting from polymorphisms in genessuch as sFlt-1. Additionally, patients with preeclampsiaare said to have a decreased long-term incidence of malig-nancy [180], a provocative observation disputed by some[181], that may suggest that the antiangiogenic state ofpreeclampsia may reflect a permanent state of the mater-nal milieu.

CONCLUSION

Preeclampsia is a systemic disorder characterized bywidespread endothelial dysfunction. Substantial animaland human experimental evidence fulfilling Koch’s pos-tulates [182] supports the hypothesis that placental sFlt1,an antiangiogenic protein, may be etiologic. Several im-portant questions remain to be explored, and many ofthese can now be approached experimentally. Why docertain conditions (see Table 1) predispose to preeclamp-sia? Are there additional synergistic factors which mightinduce/predispose to HELLP syndrome? What is the re-lationship of sFlt-1 with other postulated mediators ofthe maternal syndrome such as hyperuricemia, ADMA,or antibodies to angiotensin II AT-1 receptor? Whatgoverns regulation of sFlt1 transcription and splicing,

and why is it abnormal in preeclampsia? Is placental is-chemia a cause or a consequence of excessive placen-tal sFlt1? Large genetic trials such as GOPEC (Geneticsof Preeclampsia) [6], ongoing proteomic/genomic stud-ies of tissue and blood specimens from patients withpreeclampsia, and further characterization of the sFlt-1induced animal model of preeclampsia may help to an-swer these questions.

The implications of these advances on our clinicalmanagement of preeclampsia may be profound. Large,prospective studies are needed to describe in detailthe changes in sFlt1, VEGF, and PlGF that precedepreeclampsia onset; measurement of these factors inpregnant women may prove to be useful diagnosti-cally. Pharmacologic intervention aimed at restoringsFlt1/PlGF/VEGF balance may prevent or modify thecourse of preeclampsia. Safely prolonging pregnancyby only a few weeks could substantially reduce ma-ternal and neonatal morbidity and mortality resultingfrom preeclampsia, the world’s most common glomerulardisease.

ACKNOWLEDGMENTS

We would like to thank Ben Sachs and Kee-Hak Lim of the De-partment of Obstetrics and Gynecology at the Beth Israel DeaconessMedical Center (BIDMC) and Marshall Lindheimer of the Univer-sity of Chicago for helpful discussions and strong support. This workwas supported by NIH grants (DK02825, DK64255, and DK65997),the American Society of Nephrology Carl W. Gottschalk award, andBIDMC seed funds to S.A.K. S.A.K, S.E.M, and V.P.S are listed as co-inventors in a patent filed by BIDMC for the use of angiogenesis-relatedproteins for the diagnosis and treatment of preeclampsia. Therapeuticclaims of the patent have been licensed by BIDMC to Scios, Inc, CA.

Reprint requests to S. Ananth Karumanchi, M.D., Beth Israel Dea-coness Medical Center, Renal Division, 330 Brookline Avenue, Dana517, Boston, MA 02215.E-mail: sananth@bidmc.harvard.edu

REFERENCES

1. WALKER JJ: Pre-eclampsia. Lancet 356:1260–1265, 20002. ROBERTS JM, COOPER DW: Pathogenesis and genetics of pre-

eclampsia. Lancet 357:53–56, 20013. PAGE EW: The relation between hydatid moles, relative ischemia

of the gravid uterus and the placental origin of eclampsia. Am JObstet Gynecol 37:291–293, 1939

4. ROBERTS JM, TAYLOR RN, MUSCI TJ, et al: Preeclampsia: An en-dothelial cell disorder. Am J Obstet Gynecol 161:1200–1204, 1989

5. ROBERTS JM: Preeclampsia: What we know and what we do notknow. Semin Perinatol 24:24–28, 2000

6. LACHMEIJER AM, DEKKER GA, PALS G, et al: Searching forpreeclampsia genes: The current position. Eur J Obstet GynecolReprod Biol 105:94–113, 2002

7. THADHANI R, STAMPFER MJ, HUNTER DJ, et al: High body massindex and hypercholesterolemia: risk of hypertensive disorders ofpregnancy. Obstet Gynecol 94:543–550, 1999

8. ROBILLARD PY, HULSEY TC, ALEXANDER GR, et al: Paternity pat-terns and risk of preeclampsia in the last pregnancy in multiparae.J Reprod Immunol 24:1–12, 1993

9. TRUPIN LS, SIMON LP, ESKENAZI B: Change in paternity: A riskfactor for preeclampsia in multiparas. Epidemiology 7:240–244,1996

2110 Karumanchi et al: Preeclampsia and angiogenic factors

10. SKJARVEN R, WILCOX AJ, LIE RT: The interval between pregnanciesand the risk of preeclampsia. N Engl J Med 346:33–38, 2002

11. ESPLIN MS, FAUSETT MB, FRASER A, et al: Paternal and maternalcomponents of the predisposition to preeclampsia. N Engl J Med344:867–872, 2001

12. WARD K, HATA A, JEUNEMAITRE X, et al: A molecular variant ofangiotensinogen associated with preeclampsia. Nat Genet 4:59–61,1993

13. ARNGRIMSSON R, HAYWARD C, NADAUD S, et al: Evidence for afamilial pregnancy-induced hypertension locus in the eNOS-generegion. Am J Hum Genet 61:354–362, 1997

14. DIZON-TOWNSON DS, NELSON LM, EASTON K, WARD K: The factor VLeiden mutation may predispose women to severe preeclampsia.Am J Obstet Gynecol 175:902–905, 1996

15. SOHDA S, ARINAMI T, HAMADA H, et al: Methylenetetrahydrofolatereductase polymorphism and pre-eclampsia. J Med Genet 34:525–526, 1997

16. ARNGRIMSSON R, GEIRSSON RT, COOKE A, et al: Renin gene re-striction fragment length polymorphisms do not show linkage withpreeclampsia and eclampsia. Acta Obstet Gynecol Scand 73:10–13,1994

17. LADE JA, MOSES EK, GUO G, et al: The eNOS gene: A candidatefor the preeclampsia susceptibility locus? Hypertens Preg 18:81–93,1999

18. LACHMEIJER AM, ARNGRIMSSON R, BASTIAANS EJ, et al: Muta-tions in the gene for methylenetetrahydrofolate reductase, ho-mocysteine levels, and vitamin status in women with a history ofpreeclampsia. Am J Obstet Gynecol 184:394–402, 2001

19. O’SHAUGHNESSY KM, FU B, FERRARO F, et al: Factor V Leiden andthermolabile methylenetetrahydrofolate reductase gene variantsin an East Anglian preeclampsia cohort. Hypertension 33:1338–1341, 1999

20. GUO G, WILTON AN, FU Y, et al: Angiotensinogen gene variationin a population case-control study of preeclampsia/eclampsia inAustralians and Chinese. Electrophoresis 18:1646–1649, 1997

21. ARNGRIMSSON R, SIGURARDTTIR S, FRIGGE ML, et al: A genome-wide scan reveals a maternal susceptibility locus for pre-eclampsiaon chromosome 2p13. Hum Mol Genet 8:1799–1805, 1999

22. MOSES EK, LADE JA, GUO G, et al: A genome scan in families fromAustralia and New Zealand confirms the presence of a maternalsusceptibility locus for pre-eclampsia, on chromosome 2. Am JHum Genet 67:1581–1585, 2000

23. LACHMEIJER AM, ARNGRIMSSON R, BASTIAANS EJ, et al: A genome-wide scan for preeclampsia in the Netherlands. Eur J Hum Genet9:758–764, 2001

24. TUOHY JF, JAMES DK: Pre-eclampsia and trisomy 13. Br J ObstetGynaecol 99:891–894, 1992

25. GELLER DS, FARHI A, PINKERTON N, et al: Activating mineralocorti-coid receptor mutation in hypertension exacerbated by pregnancy.Science 289:119–123, 2000

26. KEYES LE, ARMAZA JF, NIERMEYER S, et al: Intrauterine growth re-striction, preeclampsia, and intrauterine mortality at high altitudein Bolivia. Pediatr Res 16:16, 2003

27. CHESLEY LC: History and epidemiology of preeclampsia-eclampsia. Clin Obstet Gynecol 27:801–820, 1984

28. WALLENBURG HCS: Hemodynamics of hypertensive pregnancy, inHandbook of Hypertension, edited byRubin PC, New York, Else-vier Science, 1988, pp 66–101

29. SCHOBEL HP, FISCHER T, HEUSZER K, et al: Preeclampsia—Astate of sympathetic overactivity. N Engl J Med 335:1480–1485,1996

30. MANYONDA IT, SLATER DM, FENSKE C, et al: A role for nora-drenaline in pre-eclampsia: Towards a unifying hypothesis for thepathophysiology. Br J Obstet Gynaecol 105:641–648, 1998

31. EASTERLING TR, BENEDETTI TJ, SCHMUCKER BC, MILLARD SP: Ma-ternal hemodynamics in normal and preeclamptic pregnancies: Alongitudinal study. Obstet Gynecol 76:1061–1069, 1990

32. CHESLEY L: Hypertensive Disorders of Pregnancy, New York, NY,Appleton Century Crafts, 1978

33. VICTOR R, MARK A: The sympathetic nervous system in humanhypertension, in Hypertension: Pathophysiology, Diagnosis, andManagement, edited byLaragh J, Brenner B, 2nd ed., New York,NY, Raven Press, 1995, pp 865–887

34. GANT NF, DALEY GL, CHAND S, et al: A study of angiotensin IIpressor response throughout primigravid pregnancy. J Clin Invest52:2682–2689, 1973

35. KYLE PM, BUCKLEY D, KISSANE J, et al: The angiotensin sensitivitytest and low-dose aspirin are ineffective methods to predict andprevent hypertensive disorders in nulliparous pregnancy. Am JObstet Gynecol 173:865–872, 1995

36. LUFT FC, MILLER JZ, COHEN SJ, et al: Heritable aspects of saltsensitivity. Am J Cardiol 61:1H–6H, 1988

37. REDMAN CW: Maternal plasma volume and disorders of pregnancy.Br Med J (Clin Res Ed) 288:955–956, 1984

38. TAPIA HR, JOHNSON CE, STRONG CG: Renin-angiotensin system innormal and in hypertensive disease of pregnancy. Lancet 2:847–850, 1972

39. AUGUST P, LENZ T, ALES KL, et al: Longitudinal study of therenin-angiotensin-aldosterone system in hypertensive pregnantwomen: deviations related to the development of superimposedpreeclampsia. Am J Obstet Gynecol 163:1612–1621, 1990

40. OKUNO S, HAMADA H, YASUOKA M, et al: Brain natriuretic peptide(BNP) and cyclic guanosine monophosphate (cGMP) levels in nor-mal pregnancy and preeclampsia. J Obstet Gynaecol Res 25:407–410, 1999

41. FINNERLY FAJ, BUCHOLZ JH, GUILLAUDEU RL: The blood volumeand plasma proteins during levarterenol-induced hypertension. JClin Invest 37:425–429, 1958

42. AGATISA PK, NESS RB, ROBERTS JM, et al: Impairment of endothe-lial function in women with a history of preeclampsia: an indicatorof cardiovascular risk. Am J Physiol Heart Circ Physiol 286:H1389–H1393, 2004

43. CHAMBERS JC, FUSI L, MALIK IS, et al: Association of maternalendothelial dysfunction with preeclampsia. JAMA 285:1607–1612,2001

44. GOODMAN RP, KILLAM AP, BRASH AR, BRANCH RA: Prostacyclinproduction during pregnancy: Comparison of production duringnormal pregnancy and pregnancy complicated by hypertension.Am J Obstet Gynecol 142:817–822, 1982

45. FITZGERALD DJ, ENTMAN SS, MULLOY K, FITZGERALD GA: De-creased prostacyclin biosynthesis preceding the clinical manifesta-tion of pregnancy-induced hypertension. Circulation 75:956–963,1987

46. MILLS JL, DERSIMONIAN R, RAYMOND E, et al: Prostacyclin andthromboxane changes predating clinical onset of preeclampsia: Amulticenter prospective study. JAMA 282:356–362, 1999

47. BAKER PN, DAVIDGE ST, BARANKIEWICZ J, ROBERTS JM: Plasmaof preeclamptic women stimulates and then inhibits endothelialprostacyclin. Hypertension 27:56–61, 1996

48. MOUTQUIN JM, LINDSAY C, ARIAL N, et al: Do prostacyclin andthromboxane contribute to the “protective effect” of pregnancieswith chronic hypertension? A preliminary prospective longitudinalstudy. Am J Obstet Gynecol 177:1483–1490, 1997

49. FITZGERALD DJ, ROCKI W, MURRAY R, et al: Thromboxane A2synthesis in pregnancy-induced hypertension. Lancet 335:751–754,1990

50. PAARLBERG KM, DE JONG CL, VAN GEIJN HP, et al: Vasoactive medi-ators in pregnancy-induced hypertensive disorders: a longitudinalstudy. Am J Obstet Gynecol 179:1559–1564, 1998

51. CARITIS S, SIBAI B, HAUTH J, et al: Low-dose aspirin to preventpreeclampsia in women at high risk. N Engl J Med 338:701–705,1998

52. CLASP: A randomised trial of low-dose aspirin for the preven-tion and treatment of pre-eclampsia among 9364 pregnant women.CLASP (Collaborative Low-dose Aspirin Study in Pregnancy)Collaborative Group. Lancet 343:619–629, 1994

53. WILLIAMS DJ, VALLANCE PJ, NEILD GH, et al: Nitric oxide-mediatedvasodilation in human pregnancy. Am J Physiol 272:H748–H752,1997

54. CONRAD KP, KERCHNER LJ, MOSHER MD: Plasma and 24-h NO(x)and cGMP during normal pregnancy and preeclampsia in womenon a reduced NO(x) diet. Am J Physiol 277:F48–F57, 1999

55. MOLNAR M, HERTELENDY F: N omega-nitro-L-arginine, an inhibitorof nitric oxide synthesis, increases blood pressure in rats and re-verses the pregnancy-induced refractoriness to vasopressor agents.Am J Obstet Gynecol 166:1560–1567, 1992

Karumanchi et al: Preeclampsia and angiogenic factors 2111

56. BUHIMSCHI I, YALLAMPALLI C, CHWALISZ K, GARFIELD RE: Pre-eclampsia-like conditions produced by nitric oxide inhibition: ef-fects of l-arginine, d-arginine and steroid hormones. Hum Reprod10:2723–2730, 1995

57. YALLAMPALLI C, GARFIELD RE: Inhibition of nitric oxide synthe-sis in rats during pregnancy produces signs similar to those ofpreeclampsia. Am J Obstet Gynecol 169:1316–1320, 1993

58. BAYLIS C, MITRUKA B, DENG A: Chronic blockade of nitric oxidesynthesis in the rat produces systemic hypertension and glomerulardamage. J Clin Invest 90:278–281, 1992

59. MARTINEZ-ORGADO J, GONZALEZ R, ALONSO MJ, SALAICES M: Im-pairment of fetal endothelium-dependent relaxation in a rat modelof preeclampsia by chronic nitric oxide synthase inhibition. J SocGynecol Investig 11:82–88, 2004

60. MAEDA T, YOSHIMURA T, OHSHIGE A, et al: Nitric oxide af-fects angiotensin II pressor response: Possible mechanism ofattenuated pressor response during pregnancy and etiology ofpregnancy-induced hypertension. Gynecol Obstet Invest 49:84–87,2000

61. RANTA V, VIINIKKA L, HALMESMAKI E, YLIKORKALA O: Nitric oxideproduction with preeclampsia. Obstet Gynecol 93:442–445, 1999

62. SHAAMASH AH, ELSNOSY ED, MAKHLOUF AM, et al: Maternal andfetal serum nitric oxide (NO) concentrations in normal pregnancy,pre-eclampsia and eclampsia. Int J Gynaecol Obstet 68:207–214,2000

63. SMARASON AK, ALLMAN KG, YOUNG D, REDMAN CW: Elevated lev-els of serum nitrate, a stable end product of nitric oxide, in womenwith pre-eclampsia. Br J Obstet Gynaecol 104:538–543, 1997

64. DAVIDGE ST, STRANKO CP, ROBERTS JM: Urine but not plasma ni-tric oxide metabolites are decreased in women with preeclampsia.Am J Obstet Gynecol 174:1008–1013, 1996

65. SILVER RK, KUPFERMINC MJ, RUSSELL TL, et al: Evaluation of nitricoxide as a mediator of severe preeclampsia. Am J Obstet Gynecol175:1013–1017, 1996

66. SELIGMAN SP, BUYON JP, CLANCY RM, et al: The role of nitric oxidein the pathogenesis of preeclampsia. Am J Obstet Gynecol 171:944–948, 1994

67. GARMENDIA JV, GUTIERREZ Y, BLANCA I, et al: Nitric oxide in dif-ferent types of hypertension during pregnancy. Clin Sci (Lond)93:413–421, 1997

68. BATTISTINI B, DUSSAULT P: The many aspects of endothelins inischemia-reperfusion injury: emergence of a key mediator. J InvestSurg 11:297–313, 1998

69. SLOWINSKI T, NEUMAYER HH, STOLZE T, et al: Endothelin system innormal and hypertensive pregnancy. Clin Sci (Lond) 103 (Suppl48):446S–449S, 2002

70. SCHIFF E, BEN-BARUCH G, PELEG E, et al: Immunoreactive circu-lating endothelin-1 in normal and hypertensive pregnancies. Am JObstet Gynecol 166:624–628, 1992

71. CLARK BA, HALVORSON L, SACHS B, EPSTEIN FH: Plasma endothe-lin levels in preeclampsia: Elevation and correlation with uric acidlevels and renal impairment. Am J Obstet Gynecol 166:962–968,1992

72. SCALERA F, SCHLEMBACH D, BEINDER E: Production of vasoactivesubstances by human umbilical vein endothelial cells after incuba-tion with serum from preeclamptic patients. Eur J Obstet GynecolReprod Biol 99:172–178, 2001

73. GREENBERG SG, BAKER RS, YANG D, CLARK KE: Effects of con-tinuous infusion of endothelin-1 in pregnant sheep. Hypertension30:1585–1590, 1997

74. DAVISON JM, DUNLOP W: Renal hemodynamics and tubular func-tion normal human pregnancy. Kidney Int 18:152–161, 1980

75. SIMS E, KRANTZ K: Serial studies of renal function during pregnancyand the puerperium in normal women. J Clin Invest 37:1764–1774,1958

76. LAFAYETTE RA, DRUZIN M, SIBLEY R, et al: Nature of glomerulardysfunction in pre-eclampsia. Kidney Int 54:1240–1249, 1998

77. MORAN P, BAYLIS PH, LINDHEIMER MD, DAVISON JM: Glomerularultrafiltration in normal and preeclamptic pregnancy. J Am SocNephrol 14:648–652, 2003

78. SPARGO B: The renal lesions in preeclampsia, in Hypertension inPregnancy, edited byLindheimer M, New York, Wiley Medical,1976, pp 129–137

79. FISHER KA, LUGER A, SPARGO BH, LINDHEIMER MD: Hyperten-sion in pregnancy: clinical-pathological correlations and remoteprognosis. Medicine (Baltimore) 60:267–276, 1981

80. POLLAK V, NETTLES J: The kidney in toxemia of pregnancy; a clinicaland pathological study based on renal biopsies. Medicine 39:469–474, 1960

81. MORRIS R, VASSALDI P, BILLER F, MCCLUSKEY R: Immunofluores-cent studies of renal biopsies in the diagnosis of toxemia of preg-nancy. Obstet Gynecol 24:32–46, 1964

82. MAUTNER W, CHURG J, GRISHMANN E, DACHS S: Preeclampticnephropathy. An electron microscopic study. Lab Invest 11:518–530, 1962

83. NOCHY BP, HINGLAIS N, FREUND M, et al: Renal lesions in the hy-pertensive syndromes of pregnancy: Immunomorphological andultrastructural studies in 114 cases. Clinl Nephrol 13:155–162, 1980

84. GABER LW, SPARGO BH: Pregnancy-induced nephropathy: The sig-nificance of focal segmental glomerulosclerosis. Am J Kidney Dis9:317–323, 1987

85. HEATON JM, TURNER DR: Persistent renal damage following pre-eclampsia: A renal biopsy study of 13 patients. J Pathol 147:121–126, 1985

86. STRAUSS M: Observations on the etiology of the toxemia ofpregnancy-II. Production of acute exacerbations of toxemia bysodium salts in pregnant women with hypoproteinemia. Am J MedSci 6:772–783, 1937

87. GALLERY ED, BROWN MA: Control of sodium excretion in humanpregnancy. Am J Kidney Dis 9:290–295, 1987

88. SCHRIER RW: Pathogenesis of sodium and water retention in high-output and low-output cardiac failure, nephrotic syndrome, cirrho-sis, and pregnancy (2). N Engl J Med 319:1127–1134, 1988

89. BROWN MA, ZAMMIT VC, LOWE SA: Capillary permeability and ex-tracellular fluid volumes in pregnancy-induced hypertension. ClinSci (Lond) 77:599–604, 1989

90. GALATIUS S, BENT-HANSEN L, WROBLEWSKI H, et al: Plasma disap-pearance of albumin and impact of capillary thickness in idiopathicdilated cardiomyopathy and after heart transplantation. Circula-tion 102:319–325, 2000

91. ROSTOKER G, BEHAR A, LAGRUE G: Vascular hyperpermeability innephrotic edema. Nephron 85:194–200, 2000

92. ORLOFF J, WELT L, STOWE L: The effects of concentrated salt-pooralbumin on the metabolism of water and electrolytes in nephrosisand toxemia of pregnancy. J Clin Invest 29:770–780, 1950

93. SLEMMONS J, BOGERT L: The uric acid content of maternal and fetalblood. J Biol Chem 32:63–69, 1917

94. REDMAN CW, BONNAR J: Plasma urate changes in pre-eclampsia.Br Med J 1:1484–1485, 1978

95. MARTIN JN, Jr., MAY WL, MAGANN EF, et al: Early risk assessmentof severe preeclampsia: Admission battery of symptoms and labo-ratory tests to predict likelihood of subsequent significant maternalmorbidity. Am J Obstet Gynecol 180:1407–1414, 1999

96. SAGEN N, HARAM K, NILSEN ST: Serum urate as a predictor offetal outcome in severe pre-eclampsia. Acta Obstet Gynecol Scand63:71–75, 1984

97. GALLERY ED, GYORY AZ: Glomerular and proximal renal tubu-lar function in pregnancy-associated hypertension: A prospectivestudy. Eur J Obstet Gynecol Reprod Biol 9:3–12, 1979

98. MANY A, HUBEL CA, ROBERTS JM: Hyperuricemia and xanthineoxidase in preeclampsia, revisited. Am J Obstet Gynecol 174:288–291, 1996

99. SCHAFFER N, DILL L, CADDEN J: Uric acid clearance in normalpregnancy and preeclampsia. J Clin Invest 22:201–206, 1943

100. FERRIS TF, GORDEN P: Effects of angiotensin and norepinephrineupon urate clearance in man. Am J Med 44:359–365, 1968

101. WATANABE S, KANG DH, FENG L, et al: Uric acid, hominoid evolu-tion, and the pathogenesis of salt-sensitivity. Hypertension 40:355–360, 2002

102. KANG DH, FINCH J, NAKAGAWA T, et al: Uric acid, endothelial dys-function and pre-eclampsia: Searching for a pathogenetic link. JHypertens 22:229–235, 2004

103. TAYLOR RN, CROMBLEHOLME WR, FRIEDMAN SA, et al: High plasmacellular fibronectin levels correlate with biochemical and clinicalfeatures of preeclampsia but cannot be attributed to hypertensionalone. Am J Obstet Gynecol 165:895–901, 1991

2112 Karumanchi et al: Preeclampsia and angiogenic factors

104. FRIEDMAN SA, SCHIFF E, EMEIS JJ, et al: Biochemical corroborationof endothelial involvement in severe preeclampsia. Am J ObstetGynecol 172:202–203, 1995

105. HSU CD, IRIYE B, JOHNSON TR, et al: Elevated circulating throm-bomodulin in severe preeclampsia. Am J Obstet Gynecol 169:148–149, 1993

106. ESTELLES A, GILABERT J, AZNAR J, et al: Changes in the plasmalevels of type 1 and type 2 plasminogen activator inhibitors in nor-mal pregnancy and in patients with severe preeclampsia. Blood74:1332–1338, 1989

107. CHAVARRIA ME, LARA-GONZALEZ L, GONZALEZ-GLEASON A, et al:Maternal plasma cellular fibronectin concentrations in normal andpreeclamptic pregnancies: A longitudinal study for early predic-tion of preeclampsia. Am J Obstet Gynecol 187:595–601, 2002

108. ROBERTS JM, EDEP ME, GOLDFIEN A, TAYLOR RN: Sera frompreeclamptic women specifically activate human umbilical vein en-dothelial cells in vitro: Morphological and biochemical evidence.Am J Reprod Immunol 27:101–108, 1992

109. TAYLOR RN, CASAL DC, JONES LA, et al: Selective effects ofpreeclamptic sera on human endothelial cell procoagulant proteinexpression. Am J Obstet Gynecol 165:1705–1710, 1991

110. KOBAYASHI H, SADAKATA H, SUZUKI K, et al: Thrombomodulinrelease from umbilical endothelial cells initiated by preeclampsiaplasma-induced neutrophil activation. Obstet Gynecol 92:425–430,1998

111. BOEHME MW, SCHMITT WH, YOUINOU P, et al: Clinical relevance ofelevated serum thrombomodulin and soluble E-selectin in patientswith Wegener’s granulomatosis and other systemic vasculitides.Am J Med 101:387–394, 1996

112. GANDO S, NAKANISHI Y, KAMEUE T, NANZAKI S: Soluble thrombo-modulin increases in patients with disseminated intravascular co-agulation and in those with multiple organ dysfunction syndromeafter trauma: role of neutrophil elastase. J Trauma 39:660–664,1995

113. MORI Y, WADA H, OKUGAWA Y, et al: Increased plasma throm-bomodulin as a vascular endothelial cell marker in patients withthrombotic thrombocytopenic purpura and hemolytic uremic syn-drome. Clin Appl Thromb Hemost 7:5–9, 2001

114. BALLEGEER VC, SPITZ B, DE BAENE LA, et al: Platelet activationand vascular damage in gestational hypertension. Am J ObstetGynecol 166:629–633, 1992

115. CHAIWORAPONGSA T, ROMERO R, YOSHIMATSU J, et al: Soluble ad-hesion molecule profile in normal pregnancy and pre-eclampsia. JMatern Fetal Neonatal Med 12:19–27, 2002

116. SCHWARTZ RB, FESKE SK, POLAK JF, et al: Preeclampsia-eclampsia:Clinical and neuroradiographic correlates and insights into thepathogenesis of hypertensive encephalopathy. Radiology 217:371–376, 2000

117. HINCHEY J, CHAVES C, APPIGNANI B, et al: A reversible posteriorleukoencephalopathy syndrome. N Engl J Med 334:494–500, 1996

118. BAKSHI R, SHAIKH ZA, BATES VE, KINKEL PR: Thrombotic throm-bocytopenic purpura: brain CT and MRI findings in 12 patients.Neurology 52:1285–1288, 1999

119. KUMAR D: Chronic placetal ischemia in relation to toxemias ofpregnancy. A preliminary report. Am J Obstet Gynecol 84:1323–1329, 1962

120. COMBS CA, KATZ MA, KITZMILLER JL, BRESCIA RJ: Experimentalpreeclampsia produced by chronic constriction of the lower aorta:Validation with longitudinal blood pressure measurements in con-scious rhesus monkeys. Am J Obstet Gynecol 169:215–223, 1993

121. PODJARNY E, BAYLIS C, LOSONCZY G: Animal models of preeclamp-sia. Semin Perinatol 23:2–13, 1999

122. NORIS M, TODESCHINI M, CASSIS P, PASTA F, et al: l-arginine deple-tion in preeclampsia orients nitric oxide synthase toward oxidantspecies. Hypertension 43:614–622, 2004

123. BROSENS IA, ROBERTSON WB, DIXON HG: The role of the spiralarteries in the pathogenesis of preeclampsia. Obstet Gynecol Annu1:177–191, 1972

124. DE WOLF F, DE WOLF-PEETERS C, BROSENS I, ROBERTSON WB: Thehuman placental bed: Electron microscopic study of trophoblas-tic invasion of spiral arteries. Am J Obstet Gynecol 137:58–70,1980

125. ROBERTSON WB, BROSENS I, DIXON HG: The pathological response

of the vessels of the placental bed to hypertensive pregnancy. JPathol Bacteriol 93:581–592, 1967

126. MEEKINS JW, PIJNENBORG R, HANSSENS M, et al: A study of pla-cental bed spiral arteries and trophoblast invasion in normal andsevere pre-eclamptic pregnancies. Br J Obstet Gynaecol 101:669–674, 1994

127. ZHOU Y, DAMSKY CH, CHIU K, et al: Preeclampsia is associatedwith abnormal expression of adhesion molecules by invasive cy-totrophoblasts. J Clin Invest 91:950–960, 1993

128. ZHOU Y, FISHER SJ, JANATPOUR M, et al: Human cytotrophoblastsadopt a vascular phenotype as they differentiate. A strategy forsuccessful endovascular invasion? J Clin Invest 99:2139–2151, 1997

129. ZHOU Y, DAMSKY CH, FISHER SJ: Preeclampsia is associated withfailure of human cytotrophoblasts to mimic a vascular adhesionphenotype. One cause of defective endovascular invasion in thissyndrome? J Clin Invest 99:2152–2164, 1997

130. CANIGGIA I, GRISARU-GRAVNOSKY S, KULISZEWSKY M, et al: Inhibi-tion of TGF-beta 3 restores the invasive capability of extravilloustrophoblasts in preeclamptic pregnancies. J Clin Invest 103:1641–1650, 1999

131. CANIGGIA I, MOSTACHFI H, WINTER J, et al: Hypoxia-induciblefactor-1 mediates the biological effects of oxygen on humantrophoblast differentiation through TGFbeta(3). J Clin Invest105:577–587, 2000

132. RAJAKUMAR A, WHITELOCK KA, WEISSFELD LA, et al: Selectiveoverexpression of the hypoxia-inducible transcription factor, HIF-2alpha, in placentas from women with preeclampsia. Biol Reprod64:499–506, 2001

133. LEACH RE, ROMERO R, KIM YM, et al: Pre-eclampsia and expres-sion of heparin-binding EGF-like growth factor. Lancet 360:1215–1219, 2002

134. ZHOU Y, CHIU K, BRESCIA RJ, et al: Increased depth of trophoblastinvasion after chronic constriction of the lower aorta in rhesusmonkeys. Am J Obstet Gynecol 169:224–229, 1993

135. FERRIS TF: Pregnancy, preeclampsia, and the endothelial cell. NEngl J Med 325:1439–1440, 1991

136. PAGE NM, WOODS RJ, GARDINER SM, et al: Excessive placentalsecretion of neurokinin B during the third trimester causes pre-eclampsia. Nature 405:797–800, 2000

137. SAVVIDOU MD, HINGORANI AD, TSIKAS D, et al: Endothelial dys-function and raised plasma concentrations of asymmetric dimethy-larginine in pregnant women who subsequently develop pre-eclampsia. Lancet 361:1511–1517, 2003

138. SMARASON AK, SARGENT IL, STARKEY PM, REDMAN CW: The effectof placental syncytiotrophoblast microvillous membranes fromnormal and pre-eclamptic women on the growth of endothelialcells in vitro. Br J Obstet Gynaecol 100:943–949, 1993

139. REDMAN CW, SARGENT IL: The pathogenesis of pre-eclampsia.Gynecol Obstet Fertil 29:518–522, 2001

140. COCKELL AP, LEARMONT JG, SMARASON AK, et al: Human placen-tal syncytiotrophoblast microvillous membranes impair maternalvascular endothelial function. Br J Obstet Gynaecol 104:235–240,1997

141. KANAYAMA N, TAKAHASHI K, MATSUURA T, et al: Deficiency inp57Kip2 expression induces preeclampsia-like symptoms in mice.Mol Hum Reprod 8:1129–1135, 2002

142. HUBEL CA: Oxidative stress in the pathogenesis of preeclampsia.Proc Soc Exp Biol Med 222:222–235, 1999

143. REGAN CL, LEVINE RJ, BAIRD DD, et al: No evidence for lipid per-oxidation in severe preeclampsia. Am J Obstet Gynecol 185:572–578, 2001

144. ZHANG C, WILLIAMS MA, KING IB, et al: Vitamin C and the risk ofpreeclampsia–results from dietary questionnaire and plasma assay.Epidemiology 13:409–416, 2002

145. CHAPPELL LC, SEED PT, BRILEY AL, et al: Effect of antioxidantson the occurrence of pre-eclampsia in women at increased risk: Arandomised trial. Lancet 354:810–816, 1999

146. ABDALLA S, LOTHER H, EL MASSIERY A, QUITTERER U: IncreasedAT(1) receptor heterodimers in preeclampsia mediate enhancedangiotensin II responsiveness. Nat Med 7:1003–1009, 2001

147. WALLUKAT G, HOMUTH V, FISCHER T, et al: Patients with preeclamp-sia develop agonistic autoantibodies against the angiotensin AT1receptor. J Clin Invest 103:945–952, 1999

Karumanchi et al: Preeclampsia and angiogenic factors 2113

148. XIA Y, WEN H, BOBST S, et al: Maternal autoantibodies frompreeclamptic patients activate angiotensin receptors on human tro-phoblast cells. J Soc Gynecol Investig 10:82–93, 2003

149. DECHEND R, VIEDT C, MULLER DN, et al: AT1 receptor agonisticantibodies from preeclamptic patients stimulate NADPH oxidase.Circulation 107:1632–1639, 2003

150. MAYNARD SE, MIN JY, MERCHAN J, et al: Excess placental solublefms-like tyrosine kinase 1 (sFlt1) may contribute to endothelialdysfunction, hypertension, and proteinuria in preeclampsia. J ClinInvest 111:649–658, 2003

151. KENDALL RL, THOMAS KA: Inhibition of vascular endothelial cellgrowth factor activity by an endogenously encoded soluble recep-tor. Proc Natl Acad Sci USA 90:10705–10709, 1993

152. CLARK DE, SMITH SK, HE Y, et al: A vascular endothelial growthfactor antagonist is produced by the human placenta and releasedinto the maternal circulation. Biol Reprod 59:1540–1548, 1998

153. ZHOU Y, MCMASTER M, WOO K, et al: Vascular endothelial growthfactor ligands and receptors that regulate human cytotrophoblastsurvival are dysregulated in severe preeclampsia and hemolysis,elevated liver enzymes, and low platelets syndrome. Am J Pathol160:1405–1423, 2002

154. BANKS RE, FORBES MA, SEARLES J, et al: Evidence for the existenceof a novel pregnancy-associated soluble variant of the vascularendothelial growth factor receptor, Flt-1. Mol Hum Reprod 4:377–386, 1998

155. KOGA K, OSUGA Y, YOSHINO O, et al: Elevated serum soluble vas-cular endothelial growth factor receptor 1 (sVEGFR-1) levels inwomen with preeclampsia. J Clin Endocrinol Metab 88:2348–2351,2003

156. TSATSARIS V, GOFFIN F, MUNAUT C, et al: Overexpression of thesoluble vascular endothelial growth factor receptor in preeclamp-tic patients: Pathophysiological consequences. J Clin EndocrinolMetab 88:5555–5563, 2003

157. CHAIWORAPONGSA T, ROMERO R, ESPINOZA J, et al: Evidence sup-porting a role for blockade of the vascular endothelial growth fac-tor system in the pathophysiology of preeclampsia. Young Inves-tigator Award. Am J Obstet Gynecol 190:1541–1547, 2004

158. TAYLOR RN, GRIMWOOD J, TAYLOR RS, et al: Longitudinal serumconcentrations of placental growth factor: evidence for abnormalplacental angiogenesis in pathologic pregnancies. Am J Obstet Gy-necol 188:177–182, 2003

159. POLLIOTTI BM, FRY AG, SALLER DN, et al: Second-trimester mater-nal serum placental growth factor and vascular endothelial growthfactor for predicting severe, early-onset preeclampsia. Obstet Gy-necol 101:1266–1274, 2003

160. LEVINE RJ, MAYNARD SE, QIAN C, et al: Circulating angiogenicfactors and the risk of preeclampsia. N Engl J Med 350:672–683,2004

161. HERTIG A, BERKANE N, LEFEVRE G, et al: Maternal serum sFlt1 con-centration is an early and reliable predictive marker of preeclamp-sia. Clin Chem 50:1702–1703, 2004

162. LEVINE RJ, THADHANI R, QIAN C, et al: Urinary placental growthfactor and risk of preeclampsia. JAMA 293:77–85, 2005

163. HE H, VENEMA VJ, GU X, et al: Vascular endothelial growth factorsignals endothelial cell production of nitric oxide and prostacyclinthrough flk-1/KDR activation of c-Src. J Biol Chem 274:25130–25135, 1999

164. MASUDA Y, SHIMIZU A, MORI T, et al: Vascular endothelialgrowth factor enhances glomerular capillary repair and accelerates

resolution of experimentally induced glomerulonephritis. Am JPathol 159:599–608, 2001

165. OSTENDORF T, KUNTER U, EITNER F, et al: VEGF(165) mediatesglomerular endothelial repair. J Clin Invest 104:913–923, 1999

166. SUGIMOTO H, HAMANO Y, CHARYTAN D, et al: Neutralization ofcirculating vascular endothelial growth factor (VEGF) by anti-VEGF antibodies and soluble VEGF receptor 1 (sFlt-1) inducesproteinuria. J Biol Chem 278:12605–12608, 2003

167. EREMINA V, SOOD M, HAIGH J, et al: Glomerular-specific alterationsof VEGF-A expression lead to distinct congenital and acquiredrenal diseases. J Clin Invest 111:707–716, 2003

168. KABBINAVAR F, HURWITZ HI, FEHRENBACHER L, et al: PhaseII, randomized trial comparing bevacizumab plus fluorouracil(FU)/leucovorin (LV) with FU/LV alone in patients with metastaticcolorectal cancer. J Clin Oncol 21:60–65, 2003

169. YANG JC, HAWORTH L, SHERRY RM, et al: A randomized trial ofbevacizumab, an anti-vascular endothelial growth factor antibody,for metastatic renal cancer. N Engl J Med 349:427–434, 2003

170. KUENEN BC, LEVI M, MEIJERS JC, et al: Analysis of coagulation cas-cade and endothelial cell activation during inhibition of vascularendothelial growth factor/vascular endothelial growth factor re-ceptor pathway in cancer patients. Arterioscler Thromb Vasc Biol22:1500–1505, 2002

171. ESSER S, WOLBURG K, WOLBURG H, et al: Vascular endothelialgrowth factor induces endothelial fenestrations in vitro. J Cell Biol140:947–959, 1998

172. NAGAMATSU T, FUJII T, KUSUMI M, et al: Cytotrophoblastsup-regulate soluble fms-like tyrosine kinase-1 expression un-der reduced oxygen: an implication for the placental vasculardevelopment and the pathophysiology of preeclampsia. En-docrinology 145:4838–4845, 2004

173. BOYD PA, LINDENBAUM RH, REDMAN C: Pre-eclampsia and trisomy13: A possible association. Lancet 2:425–427, 1987

174. LAIVUORI H, TIKKANEN MJ, YLIKORKALA O: Hyperinsulinemia 17years after preeclamptic first pregnancy. J Clin Endocrinol Metab81:2908–2911, 1996

175. HUBEL CA, SNAEDAL S, NESS RB, et al: Dyslipoproteinaemia inpostmenopausal women with a history of eclampsia. Br J ObstetGynaecol 107:776–784, 2000

176. IRGENS HU, REISAETER L, IRGENS LM, LIE RT: Long term mortal-ity of mothers and fathers after pre-eclampsia: population basedcohort study. Br Med J 323:1213–1217, 2001

177. SIBAI BM, EL-NAZER A, GONZALEZ-RUIZ A: Severe preeclampsia-eclampsia in young primigravid women: Subsequent pregnancyoutcome and remote prognosis. Am J Obstet Gynecol 155:1011–1016, 1986

178. EPSTEIN FH: Late vascular effects of toxemia of pregnancy. NewEngl J Med 271:391–395, 1964

179. JOHNSON RJ, HERRERA-ACOSTA J, SCHREINER GF, RODRIGUEZ-ITURBE B: Subtle acquired renal injury as a mechanism of salt-sensitive hypertension. N Engl J Med 346:913–923, 2002

180. VATTEN LJ, ROMUNDSTAD PR, TRICHOPOULOS D, SKJAERVEN R: Pre-eclampsia in pregnancy and subsequent risk for breast cancer. BrJ Cancer 87:971–973, 2002

181. MOGREN I, STENLUND H, HOGBERG U: Long-term impact of repro-ductive factors on the risk of cervical, endometrial, ovarian andbreast cancer. Acta Oncol 40:849–854, 2001

182. CHIEN KR: Meeting Koch’s postulates for calcium signaling in car-diac hypertrophy. J Clin Invest 105:1339–1342, 2000