Pre‐eclampsia and the cardiovascular–placental axis

Pre-eclampsia (PE) is a major cause of maternal mortality and morbidity worldwide, with an estimated incidence of between 3% and 10% of all pregnancies. Although its precise etiology remains elusive, the most commonly accepted theory thus far is that it develops as a consequence of poor or inadequate placentation in early pregnancy. The resulting placental hypoperfusion is thought to be responsible for fetal growth restriction as well as the systemic endothelial 'stress' response associated with the maternal phenotype. However, whilst the placenta certainly plays an important role in the pathophysiology of PE, there is now compelling evidence that the maternal cardiovascular system may make an equally important contribution. Established risk factors for PE include advanced maternal age, obesity, ethnicity and prepregnancy comorbidities such as diabetes and chronic hypertension. These predisposing factors have always been considered to be related to poor placentation, despite all being well-recognized risk factors for cardiovascular disease. Many consider the overlap of risk factors for PE and cardiovascular disease to be spurious, but it lends support to the involvement of the maternal cardiovascular system in the pathophysiology of PE. In fact, a large Scandinavian epidemiological registry study concluded that the association of PE with post-pregnancy cardiovascular morbidity may be due largely to shared prepregnancy risk factors rather than reflecting a direct influence of PE on the cardiovascular system1. It is in this context that we should interpret the findings of the systematic review of Giannakou et al.2, in this issue of the Journal. They undertook an umbrella review of 58 systematic reviews, including 1466 primary studies with data on 130 risk factors associated with PE. They demonstrate that the only non-genetic risk factor with convincing evidence of an association with PE was having ovum donation and the genetic factor with strong evidence of a contribution to the pathogenesis of PE was plasminogen activator inhibitor-1 (PAI-1). Although the presumed mechanism of ovum donation predisposing to PE has always relied on disrupted decidual receptivity and immunological tolerance, more recent data demonstrate that women who receive donated ova tend to be older, to be postmenopausal or to have mosaic Turner syndrome – all factors that predispose to increased cardiovascular risk3, 4. PAI-1 is produced mainly by the vascular endothelium and it has a role that is principally cardiovascular as an inhibitor of fibrinolysis. Furthermore, Mohseni et al.5 recently reviewed the overlap between microRNA (miRNA) expression in concentric (aberrant) cardiac remodeling in response to the volume load of pregnancy and in PE. They highlight nine miRNAs relevant to both aberrant maternal cardiac remodeling and PE – exposing a potential mechanistic link between PE and the development of remote postpartum cardiovascular disease. It has long been presumed by researchers interested in PE that defects of trophoblast development, motility, invasion and function are intrinsic defects which influence the persistence of a high-resistance uterine artery circulation by way of inadequate spiral (uterine) artery invasion. Rodriguez and colleagues6 show that, in women with late PE, uterine artery Doppler indices are useful in the classification of clinical disease and correlate with placental histological findings, i.e. the apoptotic index. They propose that the latter associations provide evidence of a subgroup of late PE cases whose disease is of placental origin. Closer examination of this proposition suggests a unique human disease mechanism, in which end-organ placental tissue damage somehow dictates upstream perfusion. It is far more biologically plausible – and typical of human cardiovascular disorders – that disruption in vascular perfusion leads to end-organ damage. The latter hypothesis is supported by the systematic review of Kalafat et al.7, which demonstrates that Doppler assessment of two seemingly unrelated maternal vessels (ophthalmic and uterine arteries) can be used for the prediction of disease considered a 'placental disorder'. The finding of a relationship between ophthalmic Doppler indices and PE cannot be a consequence of impaired trophoblast invasion and is far more likely to be related to general hemodynamic adaptation to the pregnancy. Recent basic preliminary research indeed supports the more plausible scenario that primary impaired uterine perfusion leads to inadequate trophoblast invasion and function8. The development of a sensitive screening model for PE has been at the forefront of research priorities for several decades. Recently, researchers from The Fetal Medicine Foundation (FMF) conducted a double-blind, placebo-controlled trial9 in which women screened as high risk for preterm PE were randomized to receive either aspirin (150 mg per day) or placebo from 11 to 36 weeks of gestation. The study demonstrated a 60% reduction in preterm PE, without any demonstrable difference in the incidence of neonatal or maternal adverse outcome in either arm of the study. In this issue of the Journal, these authors report that, in women who are screen positive for preterm PE according to the American College of Obstetricians and Gynecologists (ACOG) or National Institute for Health and Care Excellence (NICE) criteria, but screen negative according to the FMF algorithm, the risk of preterm PE is reduced to within or below background levels10. They go on to report a validation study for the FMF algorithm in a National Institute for Health Research (NIHR)-funded UK study11, demonstrating poor performance of and compliance with NICE guidelines, but substantial improvement in screening performance using the FMF PE algorithm. The FMF algorithm combines known (cardiovascular) risk factors with mean arterial pressure, uterine artery pulsatility index and serum placental growth factor (PlGF) level measured at 11–13 weeks' gestation. The use of this algorithm more than doubles the sensitivity for preterm PE, to 82%, thereby making a compelling case for the routine implementation of this screening protocol. It should be noted that the screening test utilizes cardiovascular markers, and PE prophylaxis is achieved using a drug with established vascular effects. These findings are all further support for involvement of the maternal cardiovascular system in both the development and the amelioration of PE. It is remarkable that, for a disease which has been recognized clinically for over a century and that has a significant human toll, no well-established prognostic tests exist. The ability to predict the risk of maternal complications, especially for women with early-onset PE, would be highly beneficial to guide management in care facilities. The PIERS (Pre-eclampsia Integrated Estimate of RiSk) and PREP (Prediction models for Risks of complications in Early-onset Pre-eclampsia) models are better able to predict adverse maternal outcome in women admitted with a diagnosis of early-onset PE compared with the use of blood pressure or other clinical markers in isolation12, 13. However, although PIERS and PREP are of value in established cases of PE, considerable clinical resources are expended monitoring women considered to be at high risk of PE. This issue contains several publications demonstrating the added clinical value of repeat soluble fms-like tyrosine kinase-1 (sFlt-1):PlGF ratio testing and use of additional (N-terminal pro-B natriuretic peptide (NT-proBNP) and C-terminal pro-endothelin-1 (CT-pro-ET1)) biomarkers in the high-risk population14-16. Whilst these are encouraging clinical findings, it is important to bear in mind that, as for early-pregnancy PE screening tests, combining biochemical and biophysical markers may allow for even better prediction of adverse pregnancy outcome17. The overlap between PE and maternal cardiovascular function is further highlighted when one considers that all of the aforementioned biomarkers have definitive roles as cardiovascular biomarkers outside pregnancy18. The long-held belief that PE is a placental disorder is based on the association between PE and placental histological abnormalities. A recent systematic review of placental histology demonstrated that histological lesions assumed previously to be characteristic of PE are neither sensitive nor specific markers for the disorder19. Furthermore, impaired fetal growth is an expected consequence of impaired placentation in PE, but over 80% of PE occurs at term – when the majority of newborns are appropriate- or even large-for-gestational age20. As a consequence, late-onset PE is frequently attributed to 'maternal' factors, which is neither explanatory nor acceptable when considering the significant morbidity and mortality of PE. Placental dysfunction is fundamental to the pathophysiology of pregnancy complications such as PE, but, to date, the placenta has been considered only in isolation, without regard to the fact that its function is dependent on adequate perfusion by the maternal circulation. An alternative explanation for PE is that placental dysfunction is secondary to maternal cardiovascular maladaptation in pregnancy (Figure 1). The concept that placental dysfunction is secondary to a maternal disorder is not new when one considers the clinical similarities between PE and gestational diabetes (Table 1), both pregnancy-specific conditions that are 'cured' by birth. It is accepted that gestational diabetes develops when the maternal pancreas is unable to manage the increasing glucose load of pregnancy. Similarly, that cardiovascular dysfunction precedes PE, cardiovascular signs predominate in the clinical syndrome and risk of cardiovascular disease persists for several decades postpartum are compelling evidence for its cardiovascular origins21, 22. GTT assesses pancreatic reserve; improved sensitivity when performed later in pregnancy Reduce load (dietary control of blood glucose levels); treat biological defect (insulin) The evidence that placental dysfunction in PE may be secondary to failure of the maternal cardiovascular system to adapt to pregnancy has significant implications for helping to understand PE and to manage the condition. Future research should concentrate on understanding the mechanisms involved in the cardiovascular–placental axis and dysfunction related to the development of so-called placental syndromes.