Preeclampsia/eclampsia 1

Description

Preeclampsia, which along with chronic hypertension and gestational hypertension comprise the hypertensive disorders of pregnancy, is characterized by new hypertension (blood pressure 140/90 or greater) presenting after 20 weeks' gestation with clinically relevant proteinuria. Preeclampsia is 1 of the top 4 causes of maternal mortality and morbidity worldwide (summary by Payne et al., 2011).

Preeclampsia is otherwise known as gestational proteinuric hypertension (Davey and MacGillivray, 1988). A high proportion of patients with preeclampsia have glomerular endotheliosis, the unique histopathologic feature of the condition (Fisher et al., 1981). A distinct form of severe preeclampsia is characterized by hemolysis, elevated liver enzymes, and low platelets (HELLP syndrome) (Brown et al., 2000).

Genetic Heterogeneity of Preeclampsia/Eclampsia

Susceptibility loci for preeclampsia/eclampsia include PEE1 on chromosome 2p13, PEE2 (609402) on chromosome 2p25, and PEE3 (609403) on chromosome 9p13. PEE4 (609404) is caused by mutation in the STOX1 gene (609397) on chromosome 10q22. PEE5 (614595) is caused by mutation in the CORIN gene (605236) on chromosome 4p12. An association with PEE has been found with the EPHX1 gene (132810) on chromosome 1q.

Inheritance

Humphries (1960) presented a systematic study of hypertensive toxemia of pregnancy in mother-daughter pairs delivered at The Johns Hopkins Hospital. Toxemia occurred in 28% of daughters of women who had toxemia in the pregnancy in which they were delivered as compared with 13% in a comparison group. Chesley et al. (1968) did a similar study with similar results. In cases in which 2 or more daughters of an eclamptic woman have been tested in pregnancy, toxemia developed in the first pregnancy of at least 1 daughter in 53% of the families.

No definite conclusions on genetic heterogeneity, role of maternal versus fetal genotype, and possible genotype-genotype interaction were reached by Cooper and Liston (1979). Editorials published in The Lancet (Anonymous (1980, 1988)) and in the British Medical Journal (Anonymous, 1980) gave excellent reviews of genetic studies on eclampsia.

Cooper et al. (1988) reported several examples of 3- and 4-generation involvement. Despite this, pedigree analysis by Chesley and Cooper (1986) suggested autosomal recessive inheritance with a frequency of the 'abnormal' allele of approximately 0.25.

Arngrimsson et al. (1990) did a study through 3 or 4 generations in 94 families in Iceland patterned after the Humphries (1960) study. The families were descended from index women who were delivered in the years 1931-47 and who had either eclampsia or severe preeclampsia. Inheritance was followed through both sons and daughters. They concluded that either a recessive or a dominant model could fit the data.

Esplin et al. (2001) found that both men and women who were the product of a pregnancy complicated by preeclampsia were significantly more likely than control men and women to have a child who was the product of a pregnancy complicated by preeclampsia. Their findings were consistent with the suggestion of Liston and Kilpatrick (1991) that the single-gene model of inheritance of preeclampsia that best explains the frequency of preeclampsia in a low-risk population (3 to 6%) is the presence of homozygosity for the same recessive gene in both the mother and the fetus. In accordance with this model, the fetus must have 1 recessive paternally derived allele for preeclampsia to develop.

Cnattingius et al. (2004) analyzed pregnancy outcomes from Swedish families joined by full sibs, including information from 244,564 sib pairs who had 701,488 pregnancies. The authors found that 35% of the variance in risk of preeclampsia was attributable to maternal genetic effects, 20% to fetal genetic effects (with equal contribution of maternal and paternal genetic effects), 13% to the couple effect, less than 1% to shared sib environment, and 32% to unmeasured factors. Cnattingius et al. (2004) concluded that genetic factors account for more than half of the risk of preeclampsia, and that maternal genes contribute more than fetal genes. They suggested that the couple effect is due to a genetic interaction between mother and father.

Thornton and Macdonald (1999) performed a cohort study of female twins with information on hypertensive diseases of pregnancy obtained by questionnaire screening, and verified the diagnosis from hospital or general practitioner records. Self-reported preeclampsia was found to have a heritability of 0.221 and nonproteinuric hypertension of 0.198. However, none of the pairs who were self-reported as concordant for preeclampsia was confirmed from hospital records. Using hospital records, the heritability of preeclampsia was 0.0 and that for nonproteinuric hypertension was 0.375. Using a model treating preeclampsia as a separate disease from nonproteinuric hypertension, and assuming that the next pair identified was both monozygotic and concordant for preeclampsia, the estimated heritability of preeclampsia remained at 0.0. Using a threshold model in which nonproteinuric hypertension was treated as a mild form of preeclampsia, heritability was estimated at 0.247. They concluded that neither nonproteinuric hypertension nor preeclampsia is inherited in a simple mendelian fashion, and that the genetic contribution to the multifactorial inheritance of these 2 traits is smaller than hitherto believed.

Berends et al. (2008) analyzed familial aggregation, consanguinity, and parent-of-origin effects in 106 women from a genetically isolated population in the Netherlands, 50 who had previous preeclampsia and 56 with previous pregnancies complicated by intrauterine growth retardation (IUGR). Eight-six of the women, 39 preeclampsia and 47 IUGR cases, could be linked to 1 common ancestor within 14 generations. The proportion of related women with previous preeclampsia or pregnancies complicated by IUGR was significantly greater than that expected by chance, and the proportion of women born from consanguineous marriages was increased in women with previous preeclampsia and those with IUGR compared to controls (p less than 0.001 for both). Berends et al. (2008) stated that the observed cosegregation of preeclampsia and IUGR supported a common genetic etiology, and that the high proportion of parental consanguineous marriages suggested the possibility of an underlying recessive mutation. No evidence was found for a parent-of-origin effect in either disorder.

Population Genetics

Preeclampsia complicates 3 to 8% of pregnancies in Western countries and causes 10 to 15% of maternal deaths. Incidence ranges from 3 to 7% for nulliparas and 1 to 3% for multiparas (summary by Uzan et al., 2011).

Pathogenesis

Napolitano et al. (2000) investigated the interactions between ET1 (131240) and the NO system in the fetoplacental unit. They examined the mRNA expression of ET1, inducible NO synthase (iNOS; 163730), and eNOS in human cultured placental trophoblastic cells obtained from preeclamptic (PE) and normotensive pregnancies. ET1 expression was increased in PE cells, whereas iNOS, which represents the main source of NO synthesis, was decreased; conversely, eNOS expression was increased. ET1 was able to influence its own expression as well as NOS isoform expression in normal and PE trophoblastic cultured cells. The findings suggested the existence of a functional relationship between ET(s) and NOS isoforms that could constitute the biologic mechanism leading to the reduced placental blood flow and increased resistance to flow in the fetomaternal circulation that are characteristic of the pathophysiology of preeclampsia.

Cross (2003) reviewed various interpretations of the genetics of preeclampsia and 3 different mouse models suggesting that it can be initiated by 3 independent mechanisms: preexisting borderline maternal hypertension that is exacerbated by pregnancy, elevated levels of the vasoconstrictor angiotensin II (see 106150) in the maternal circulation by placental overproduction of renin (179820), and placental pathology. He stated that the potential contributions of both maternal and fetal genes to the onset of the disorder complicate its genetic analysis in humans.

Maynard et al. (2003) found that soluble FMS-related tyrosine kinase-1 (FLT1; 165070) was upregulated in preeclampsia, leading to increased systemic levels of sFLT1 that fell after delivery. Increased sFLT1 in preeclamptic women was associated with decreased circulating levels of free vascular endothelial growth factor (VEGF; 192240) and placental growth factor (PGF; 601121), resulting in endothelial dysfunction in vitro that was rescued by exogenous VEGF and PGF. Administration of sFLT1 to pregnant rats induced hypertension, proteinuria, and glomerular endotheliosis, the classic lesion of preeclampsia. Maynard et al. (2003) suggested that excess circulating sFLT1 contributes to the pathogenesis of preeclampsia.

In 120 preeclamptic women and 120 matched, normotensive controls, Levine et al. (2004) measured serum levels of the angiogenic factors sFLT1, PGF, and VEGF throughout pregnancy. Beginning at 13 to 16 weeks of gestation, PGF levels were significantly lower in women who later had preeclampsia than in controls (p = 0.01), with the greatest difference occurring during the weeks before the onset of preeclampsia, coincident with an increase in the sFLT1 level which was also more pronounced in the preeclamptic women. Levine et al. (2004) concluded that increased levels of sFLT1 and reduced levels of PGF predict the subsequent development of preeclampsia.

Page et al. (2000) sought vasoactive placental neuropeptides using mRNA fingerprinting and human databases and identified neurokinin B (162330). In female rats, concentrations of NKB several-fold above that of an animal 20 days into pregnancy caused substantial pressor activity. In human pregnancy, the expression of NKB was confined to the outer syncytiotrophoblast of the placenta, significant concentrations of NKB could be detected in plasma as early as week 9, and plasma concentrations of NKB were grossly elevated in pregnancy-induced hypertension and preeclampsia. Page et al. (2000) suggested that elevated levels of NKB in early pregnancy may be an indicator of hypertension and preeclampsia and that treatment with certain neurokinin receptor antagonists may be useful in alleviating symptoms.

In a review of the pathogenesis and genetics of preeclampsia, Roberts and Cooper (2001) stated that aberration of the interaction between placental and maternal tissue is probably the primary cause, but the exact nature of the differences from normal pregnancy remained elusive. There are genetic components to susceptibility, but the relative contributions of maternal and fetal genotypes were unclear.

Pipkin (2001) reviewed the evidence on risk factors for preeclampsia.

Wallukat et al. (1999) reported that patients with preeclampsia develop autoantibodies against the angiotensin II receptor type 1 (AGTR1; 106165), and suggested that these antibodies may participate in the angiotensin II-induced vascular lesions seen in patients with preeclampsia. Zhou et al. (2008) injected pregnant mice with either total IgG or affinity-purified angiotensin AT1 receptor antibodies from women with preeclampsia and observed the development of key features of preeclampsia in the mice, including hypertension, proteinuria, glomerular endotheliosis, placental abnormalities, and small fetus size. These features were prevented by coinjection with the AT1 receptor antagonist losartan or by an antibody-neutralizing 7-amino-acid epitope peptide. Zhou et al. (2008) concluded that preeclampsia may be a pregnancy-induced autoimmune disease in which key features of the disease result from autoantibody-induced angiotensin receptor activation.

Mapping

PEE1

Unlike most other human disorders, preeclampsia impacts 2 individuals, the mother and her child, both of whom can be severely affected. Although the pathophysiology of the disorder is incompletely understood, familial clustering is apparent. Arngrimsson et al. (1999) reported the results of a genomewide screen of Icelandic families representing 343 affected women. Including those patients with nonproteinuric preeclampsia (gestational hypertension), proteinuric preeclampsia, and eclampsia, they detected a locus on 2p13 with a lod score of 4.70.

Moses et al. (2000) reported the results of a medium-density genome scan in 34 families, representing 121 women with preeclampsia, from Australia and New Zealand. Multipoint nonparametric linkage analysis showed suggestive evidence of linkage to chromosome 2 (lod = 2.58), at 144.7 cM, between D2S112 and D2S151. Somewhat weaker linkage to chromosome 11q23-q24 was found. Given the limited precision of estimates of the map location of disease-predisposing loci for complex traits, Moses et al. (2000) concluded that the findings on chromosome 2 were consistent with the findings from the Icelandic study of Arngrimsson et al. (1999), and that their results may represent evidence of the same locus segregating in the population from Australia and New Zealand. They proposed that the chromosome 2 locus should be symbolized PREG1 for preeclampsia, eclampsia gene-1.

Oudejans et al. (2015) identified a SNP, rs34174194, in the INO80B gene (616456) as a susceptibility allele for preeclampsia in Icelandic families with linkage to chromosome 2p13. The T-G SNP altered a highly conserved 7-nucleotide sequence in the 3-prime UTR of INO80B. The risk allele (G) of rs34174194 reduced binding of MIR1324 to the 3-prime UTR of the INO80B transcript and was predicted to increase INO80B translation.

Other Linkage Associations

See NOS3 (163729) for a discussion of the possible role of endothelial nitric oxide synthetase, also called eNOS, in the pathogenesis of pregnancy-induced hypertension and a study by Arngrimsson et al. (1997) providing evidence for a preeclampsia susceptibility locus in the region of 7q36 encoding the NOS3 gene.

Harrison et al. (1997) reported results of a genomewide linkage search for preeclampsia/eclampsia (PEE) susceptibility genes, using 15 informative pedigrees. The 2.8-cM region between D4S450 and D4S610 on 4q was identified as a strong candidate region for a PEE susceptibility locus. The maximum multipoint lod score within this interval was 2.9. Analysis of markers in the region affected-member method also supported the possibility of a susceptibility locus in this region.

Lachmeijer et al. (2001) performed a genome scan including 293 polymorphic markers in 67 Dutch sib-pair families affected by preeclampsia, eclampsia, or HELLP syndrome. A total of 12 regions showed nominal lod score peaks (lod scores between 0.6 and 2.2), with the highest lod score of 1.99 on chromosome 12q at 109.5 cM. Analysis in 38 preeclampsia families showed suggestive evidence for linkage on chromosome 22q at 32.4 cM (lod score of 2.41) and on chromosome 10q at 93.9 cM (lod score of 2.38). In 34 HELLP families, these peaks were absent, but the peak on 12q increased to a lod score of 2.1. The authors suggested that this may indicate that HELLP syndrome has a different genetic background than preeclampsia, which they noted was in contrast to the consensus statement of the Australasian Society on the Study of Hypertension in Pregnancy (Brown et al., 2000), in which HEELP syndrome was classified as a severe form of preeclampsia. A comparison between the Dutch genome scan of Lachmeijer et al. (2001) and the Icelandic scan by Arngrimsson et al. (1999) showed overlapping regions on chromosomes 3p and 15q. Another overlapping area on chromosome 11 was revealed when comparing the Dutch preeclampsia families with the study from Australia/New Zealand (Moses et al., 2000). Lachmeijer et al. (2001) concluded that these overlapping areas may harbor maternal susceptibility genes that increase a woman's risk of preeclampsia.

Using additional microsatellite markers, van Dijk et al. (2012) reanalyzed the cohort of 34 families with HELLP syndrome originally studied by Lachmeijer et al. (2001) and found that the lod score for the region on chromosome 12q23 increased from 2.1 to 2.37. Van Dijk et al. (2012) then tested 57 individuals, including 7 families with affected sib pairs, 4 families with affected cousin pairs, and 2 discordant monozygous twin sisters with their partners, of which 36 females were affected, with 26 microsatellite markers in the 23.6-Mb region on 12q23; pedigree analysis narrowed the region to 2 minimal critical intervals: D12S1607 to PAH (612349) and D12S338 to D12S317. No mutations were found in the coding sequences of 38 known or predicted genes within or near those 2 intervals; rather, the HELLP locus was found to reside in a 154-kb intergenic region between C12ORF48 (613687) and IGF1 (147440) (chr12:101,114,674-101,268,434, NCBI36), and this region was confirmed by haplotype association analysis and deep sequencing. Van Dijk et al. (2012) screened the intergenic region and identified a long intergenic noncoding RNA (lincRNA) transcript with expression in the placental extravillous trophoblast (HELLPAR; 614985).

Associations Pending Confirmation

McGinnis et al. (2017) reported the first genomewide association study (GWAS) of offspring from preeclamptic pregnancies and discovery of the first genomewide significant susceptibility locus (rs4769613; p = 5.4 x 10(-11)) in 4,380 cases and 310,238 controls. This locus is near the FLT1 gene (165070), encoding Fms-like tyrosine kinase-1, providing biologic support, as a placental isoform of this protein (sFlt1) is implicated in the pathology of preeclampsia (Maynard et al., 2003). The association was strongest in offspring from pregnancies in which preeclampsia developed during late gestation and offspring birth weights exceeded the 10th centile. An additional nearby variant, rs12050029, associated with preeclampsia independently of rs4769613.

Molecular Genetics

In 627 families with preeclampsia (including 398 maternal triads and 536 fetal triads), the GOPEC Consortium (2005) analyzed 7 candidate genes previously reported as conferring susceptibility to preeclampsia: angiotensinogen (AGT; 106150), the angiotensin receptors AGTR1 (106165) and AGTR2 (300034), factor V Leiden variant (612309.0001), methylenetetrahydrofolate reductase (MTHFR; 607093), nitric oxide synthase (NOS3; 163729), and tumor necrosis factor-alpha (TNF; 191160). Using the transmission disequilibrium test, no genotype risk ratio achieved the prespecified criteria for statistical significance (posterior probability less than 0.05). The GOPEC Consortium (2005) concluded that none of the genetic variants tested confers a high risk of preeclampsia.

Uz et al. (2007) found extremely skewed X-inactivation (greater than or equal to 90:10) in peripheral blood cells of 10 (22%) of 46 Caucasian women with preeclampsia and in 2 (2.33%) of 86 controls, suggesting a role for the X chromosome in the pathogenesis of the disorder in some patients.

Association with HLA

Kilpatrick et al. (1989) studied a group of 56 women who had had proteinuric preeclampsia and who had parous sisters. In the first pregnancy, proteinuric preeclampsia was more common in the sisters than in the maternity hospital population; the relative risk was 6.0. Frequency of HLA-DR4 (see 142860) was higher in sisters with pregnancy-induced hypertension than in sisters with normotensive pregnancies and more of them shared HLA-DR4 with their spouses. Kilpatrick et al. (1989) referred to a study of unrelated women in which they confirmed the association between DR4 and proteinuric preeclampsia. They proposed the hypothesis that preeclampsia occurs when both mother and fetus are homozygous for an HLA-linked recessive gene. Wilton et al. (1990), however, excluded close linkage of an eclampsia susceptibility gene with HLA. Liston and Kilpatrick (1991) examined 6 simple mendelian models of inheritance and rejected all except the one in which both mother and fetus must express the same recessive gene to confer susceptibility. They considered this model to be consistent with the putative association with HLA-DR4.

Based on the hypothesis that preeclampsia occurs in women who are homozygous for a relatively common susceptibility gene, Hayward et al. (1992) constructed an exclusion map by using both candidate genes and random DNA markers on a panel of 2-generation families in which preeclampsia was rigorously defined. No evidence was found for linkage to the HLA region or to several genes implicated in the pathogenesis of hypertension, e.g., pronatriodilatin (108780), sodium-hydrogen ion antiporter (107310), mineralocorticoid receptor (600983), or glucocorticoid receptor (138040). Van Meter and Weaver (1993) commented on the study of Hayward et al. (1992). See 106150.0001 for information concerning the association of preeclampsia with a met235-to-thr mutation of the angiotensinogen gene, which maps to 1q.

Preeclampsia is a pregnancy complication in which the fetus receives an inadequate blood supply due to failure of trophoblast invasion. Hiby et al. (2004) noted that the only polymorphic histocompatibility antigens on the trophoblast surface are HLA-C molecules (142840), including the paternal allele, which are recognized by members of the highly polymorphic KIR (see KIR2DL1; 604936) family of natural killer cell receptors. There are 2 distinct KIR haplotypes, termed A and B. Haplotype A has 1 activating and 6 inhibitory KIRs, whereas haplotype B has 5 activating and 2 inhibitory KIRs. Hiby et al. (2004) found that mothers with an AA KIR genotype and a fetus with HLA-C2 were at greatly increased risk of preeclampsia. KIR2DL5 (605305), a haplotype B gene encoding an inhibitory receptor, was significantly less frequent in preeclampsia mothers. The KIR-HLA-C2 interaction appeared to be physiologic rather than immunologic, in that maternal HLA-C type was of no consequence. Hiby et al. (2004) found that different human populations have reciprocal relationships between KIR AA frequency and HLA-C2 frequency, suggesting that this combination may be selected against and that reproductive success may have influenced the evolution and maintenance of KIR and HLA-C polymorphisms.

Association with MTHFR

Sohda et al. (1997) studied the 677C-T polymorphism of the methylenetetrahydrofolate reductase gene (MTHFR; 607093.0003) in preeclampsia. They found an increased frequency of the 677T allele and the 677T homozygous genotype in patients as compared with controls. The 677T variant of MTHFR had been identified as a risk factor in vascular disease in other studies.

Rajkovic et al. (2000) found no statistically significant association between the maternal MTHFR genotype at the 677C-T polymorphism (607093.0003) and risk of preeclampsia. Conversely, Rajkovic et al. (2000) found a strong graded association between maternal plasma folate concentration and risk of preeclampsia. Women with plasma folate concentrations of less than 5.7 nmol/L experienced a 10.4-fold increase in risk of preeclampsia. There was no clear pattern of preeclampsia risk and vitamin B12 concentrations.

Association with EPHX1

Zusterzeel et al. (2001) studied genetic variability in the EPHX1 gene (132810) in women with a history of preeclampsia. They found that the high activity genotype tyr113/tyr113 (132810.0001) was significantly more common in women with a history of preeclampsia (OR, 2.0, 95% CI, 1.2-3.7) as compared to controls. No difference in the frequency of the polymorphism was found between groups who did or did not develop the syndrome of hemolysis, elevated liver enzymes, and low platelets (HELLP syndrome).

Laasanen et al. (2002) studied 2 single-nucleotide polymorphisms (SNPs) in the EPHX1 gene in 133 Finnish preeclamptic and 115 healthy control women with at least 2 normal pregnancies. The T allele of the exon 3 T-C polymorphism (tyr113 to his; 132810.0001) was overrepresented among the preeclampsia group (0.74) when compared with the control group (0.66), displaying a borderline association (P = 0.05). Haplotype analysis using this polymorphism and the exon 4 A-G polymorphism (his139 to arg; 132810.0002) showed that the high activity haplotype T/A (tyr113/his139) was significantly overrepresented in the preeclampsia group (P = 0.01; odds ratio 1.61, 95% C.I. 1.12-2.32). The authors supported the feasibility of haplotype estimation analysis for detecting association more efficiently than single-point association analysis in terms of detection power.

Association with GSTP1

Zusterzeel et al. (1999) found that glutathione S-transferase P1 (GSTP1; 134660) is the main GST isoform in normal placental and decidual tissue. In preeclamptic women, they found lower median placental and decidual GSTP1 levels compared to those in controls. Zusterzeel et al. (1999) suggested that reduced levels of GSTP1 in preeclampsia may indicate a decreased capacity of the detoxification system, resulting in a higher susceptibility to preeclampsia. Among 113 preeclampsia trios (mother, father, and baby), Zusterzeel et al. (2002) found an increased frequency of the GSTP1 val105 polymorphism (see 134660.0002) in mothers, fathers, and offspring of preeclamptic pregnancies compared to controls. There was no significant difference of the GSTP1 allele frequencies in preeclamptic mothers, fathers, and offspring. The authors emphasized the paternal contribution to the risk for preeclampsia.

Association with Coagulation Factor V

Brenner et al. (1996) identified the factor V Leiden mutation (R506Q; 612309.0001) in 2 patients with the HELLP syndrome, and Kupferminc et al. (1999) found an association between that mutation and a variety of obstetrical complications, including preeclampsia. Lindqvist et al. (1998), however, found no significant difference in the prevalence of the Leiden mutation between women with preeclampsia and/or intrauterine growth retardation and a control group. In a study of Finnish women, Faisel et al. (2004) found that susceptibility to preeclampsia was associated with a factor V R485K polymorphism but not with the Leiden mutation.

Association with NOS3

In a study of 150 'coloured' South African patients, 50 with normal pregnancies, 50 with severe preeclampsia, and 50 with abruptio placentae, Hillermann et al. (2005) found that the combined frequency of the GT and TT NOS3 variant genotypes was significantly higher in the abruptio placentae group than in the control group (p = 0.006). Among preeclamptic patients who subsequently developed abruptio placentae, the T allele emerged as a major risk factor for the development of abruptio placentae (p less than 0.0001); the T variant did not seem to affect the risk of preeclampsia itself, however.

Animal Model

Kanasaki et al. (2008) showed that pregnant mice deficient in catechol-O-methyltransferase (COMT; 116790) showed a preeclampsia-like phenotype resulting from an absence of 2-methoxyestradiol (2-ME), a natural metabolite of estradiol that is elevated during the third trimester of normal human pregnancy. Administration of 2-ME ameliorated all preeclampsia-like features without toxicity in Comt -/- pregnant mice and suppressed placental hypoxia, Hif1a (603348) expression, and soluble Flt1 (165070) elevation. The levels of COMT and 2-ME were significantly lower in women with severe preeclampsia. Kanasaki et al. (2008) suggested that Comt-null mice may provide a model for preeclampsia and that 2-ME may serve as a diagnostic marker as well as a therapeutic agent for preeclampsia.