Cardiac dysfunction and remodeling regulated by anti-angiogenic environment in patients with preeclampsia: the ANGIOCOR prospective cohort study protocol

Cardiovascular diseases (CVD) are cause of increased morbidity and mortality in spite of advances for diagnosis and treatment. Manifestations of CVD, risk factors and treatment are different in women than in men; women are affected by traditional CVD risk factors, but also by gender-specific ones, such as polycystic ovarian syndrome or pregnancy-associated disorders such as preeclampsia (PE) [1]. In 2011, the American Heart Association (AHA) guideline for preventing CVD in women included PE, gestational hypertension, diabetes mellitus, and systemic autoimmune collagen-vascular disease (i.e., lupus or rheumatoid arthritis) as factors associated with increased risk of CVD [2]. Later, in 2018 and 2019, the AHA/ACC Multisociety Guideline on the Management of Blood Cholesterol [3]; and the American College of Cardiology (ACC)/AHA guidelines [4] introduced the concept of women-specific factors to consider for risk, diagnosis and treatment of CVD; these factors now included PE.

One of the most prevalent complications of pregnancy is PE. Epidemiological studies have shown that PE is associated with development of CV risk factors, hypertension, myocardial infarction, stroke, renal complications, vascular and metabolic dysfunction during pregnancy and in later years. Pregnant women that develop preeclampsia (PE) have higher risk (up to 4 times) of clinical CVD in the short- and long-term [5, 6]. When PE is accompanied by preterm birth and/or small for gestational age, the adjusted risk for CVD is increased by 45% for the next 5 years post-partum [7]; risk comparable to that of smoking. Therefore, as history of PE has recently been incorporated by the aforementioned societies as an independent risk factor for CVD [8], follow-up starting within the fourth decade of life is recommended by some authors [9]. All these facts highlight the need to prevent and diagnose PE during pregnancy, and follow-up these women after delivery.

Changes during pregnancy affect importantly the maternal CV system, in order to maintain adequate fetal blood perfusion; maternal cardiac output and heart rate increase, while blood pressure and vascular resistances decrease [10]. In PE, CV adaptation in pregnancy is abnormal, exceeding physiological levels, which occurs more frequently if there is a hypertensive disorder before pregnancy. In fact, CV findings in PE resemble those due to cardiac diastolic dysfunction; i.e., improper left ventricle (LV) relaxation, increased LV chamber stiffness, and increased cardiac filling pressures [11], all of which can be detected by echocardiography.

PE is also characterized by endothelial dysfunction, which results from an imbalance between maternal circulating angiogenic factors, like pro-angiogenic placental growth factor (PlGF), and anti-angiogenic factors such as soluble fms-like tyrosine kinase 1 (sFlt-1) [12]. Placental expression and maternal circulating concentrations of sFlt-1 are increased in PE [13], whereas PlGF is decreased. This occurs due to a combination of decreased expression because of poor placentation, and reduced free PlGF due to sFlt-1 binding [14]. This leads to an increase in the sFlt-1/PlGF ratio, which has been described as better predictor for PE severity than either marker alone [15]. Recently, our group reported an association between angiogenic factors during pregnancy and CV disease 12 years post-partum, suggesting angiogenic factors are important players in PE-associated CV risk later in life [16].

In addition to maternal CV effects, fetal cardiac programming during pregnancy and its relationship with angiogenic environment has not been previously assessed. Fetal programming refers to the impact on fetal health due to intrauterine environment [17]. It occurs when the fetal environment is modified by an adverse condition in a critical moment of development. Pregnancy complications such as PE and intrauterine growth restriction are associated with fetal programming [18, 19]. However, little is known on the role of angiogenic biomarkers for early detection of fetal cardiac abnormalities and programming in PE pregnancies.

Multiple biomarkers have been assessed to evaluate CVD risk or myocardial damage in PE. Arginine vasopressin (AVP) has been shown to reproduce main maternal and fetal PE features in animal models [20] and some studies propose its more stable metabolite, copeptin, as an early predictor of PE [21]. Thus, AVP concentration is estimated measuring copeptin. Cardiac troponins (cTn) are isoforms solely produced by myocardial cells. When measured by high-sensitivity (hs) methods, cTn can detect small amounts released in myocardial ischemia. Few studies have analyzed hs-cTn in PE [22]; thus, it remains to be determined whether hs-cTn could detect minor myocardial damage in PE women. Natriuretic peptides are biomarkers of cardiac volume overload, vascular hypertension and myocardial elongation. All these conditions are observed in PE, and particularly those of the B-family (B-type natriuretic peptide, BNP), have been analyzed in normal and hypertensive pregnancies and found to be biomarkers for CV alterations in PE [23]. BNP is synthetized as a prehormone (pro-BNP), which is cleaved to an N-terminal fragment (NT-proBNP), and both are associated with hemodynamic stress and heart failure.

We hypothesize that an association exists between the anti-angiogenic environment and cardiac dysfunction/remodeling in women at risk for PE during pregnancy and their offspring, and that this relationship could determine an increased CV risk later in life.

In the current study, we hypothesize that in PE, anti-angiogenic imbalance may be associated to maternal and fetal cardiac dysfunction and remodeling, and that both conditions could induce myocardial damage detected by cardiac biomarkers and imaging techniques. Finally, we postulate that such cardiac abnormalities could be predictive of future cardiovascular health in patients with history of PE.

PE is a hypertensive disorder in which predominance of placental anti-angiogenic factors like sFlt-1 over - angiogenic factors like PlGF exists [31]. PE increases 2- to 4-times maternal risk of CVD; although mechanisms linking PE and CVD are not fully understood [32], PlGF and sFlt-1 are regulators of angiogenesis, a mechanism which, if altered, constitutes the basis of several CVD [33].

Pregnancy promotes stressful changes on the maternal CV system, even in normal conditions. Maternal cardiac output and heart rate increase through the three trimesters of pregnancy, while blood pressure and vascular resistances decrease. Both the sympathetic system and renin-angiotensin-aldosterone axis are activated and contribute to an increased circulating volume. At a structural level, enlargement of the four cardiac chambers and an increase of the LV wall mass occur [11]. All these changes are worsened in PE, exceeding normal adaptation. Taken together, CV findings seen in PE resemble cardiac diastolic dysfunction, a condition promoting improper LV relaxation and increasing LV chamber stiffness, resulting in increased cardiac filling pressures [10].

The association between CVD and angiogenic factors has been investigated in non-pregnant populations. Increased concentrations of PlGF at admission are associated with worse outcomes in acute coronary syndromes, although PlGF stimulates angiogenesis and mediates recovery of cardiac pathologies [34]. Increased sFlt-1 concentrations are related to acute coronary occlusion and poor cardiac function in patients with stable coronary disease [35], and with development of endothelial vascular dysfunction, myocardial injury and adverse CV outcomes in patients with congestive heart failure [35]. As in PE, the use of the sFlt-1/PlGF ratio appears to be more predictive of adverse outcomes in coronary artery disease than either biomarker alone [36]. Finally, although angiogenic factors are critical players in PE endothelial dysfunction, there is no evidence supporting their role as maternal cardiotoxic factors, and CV risk later in life. A possible explanation could be that maternal cardiac damage in PE is so subtle that it cannot be detected by current imaging techniques; therefore, the use of cardiac biomarkers may have a role for early detection.

Some studies have shown that hs-cTn is elevated in pregnant women with hypertension or with PE, compared with healthy pregnancies. These are strong predictors for development of both conditions [37] and are related to mean peak arterial BP in asymptomatic PE patients [38]. Knowing hs-cTn can detect minor myocardial damage like that existing due to ischemia, we shall assess whether PE causes such damage. In addition to hs-cTn, copeptin and natriuretic peptides could be helpful biomarkers for cardiac dysfunction in PE. Copeptin concentrations may be increased in PE, because it is not only a regulator of circulating volume, but also responsive to the CV stress associated with the condition [39]. Volume changes and hemodynamic stress are characteristics of pregnancy that are particularly strong in PE. Santillan [20] was the first to report a high sensitivity and specificity of copeptin to predict PE in the first and second trimesters of pregnancy. A recent meta-analysis has confirmed these findings, extending them to the 3rd trimester, and showing a correlation between copeptin concentrations and PE severity [40]. On other hand, secretion of natriuretic peptides is increased in response to increases in BP, cardiac filling volume, cardiac chamber pressure and dilation [32]; consequently, natriuretic peptide concentrations are increased in PE [41]. We believe NT-proBNP concentrations in pregnant women may be an earlier and alternative tool to echocardiography for detecting cardiac function alterations existing in PE. Our study plans to use all information generated by these biomarkers, along with clinical and echocardiographic variables, to develop a CV risk score during a preeclamptic pregnancy, and at 12 months after delivery.

Studies conducted in patients with heart failure and in patients with PE have found a possible common pathophysiological pathway mediated by angiogenic/anti-angiogenic placental markers like PIGF and sFlt-1 [42, 43]. Our group has recently shown that 12 years after delivery, lower levels of PlGF in pregnancy affected by PE were are associated with worse maternal hemodynamics (higher BP) and lower high-density lipoprotein cholesterol concentrations and with subclinical myocardial dysfunction (worse global longitudinal strain) and increased carotid intima-media thickness [16]. The results of this study corroborate that angiogenic imbalance in pregnancies complicated by PE is related to a negative impact in future CV risk, long term after pregnancy. These findings compel to further study causes of PE and the possibility of prediction and prevention. In the present study, we wish to assess if an anti-angiogenic profile from the beginning of pregnancy determines CV remodeling, that in turn, promotes PE development and associated CV injury that can persist after pregnancy, conditioning future CV health.

In addition to maternal assessment, we plan to evaluate fetal cardiac function at the third trimester of pregnancy, to analyze if there are potential associations between maternal CV status and angiogenic factors and occurrence of fetal cardiac programming. Recently, various studies have demonstrated the occurrence of a fetal cardiac programming in the growth-restricted fetus, which persists into childhood [19]. Furthermore, this fetal cardiac remodeling is also present in an independent fashion, in pregnancies affected by preeclampsia [18]. Given that fetal growth-restriction and PE are clinical expressions of underlying placental disease, we hypothesize and plan to analyze whether angiogenic factors will be associated with the existence of cardiac remodeling.

There are potential limitations and considerations of the present project. We would only observe subclinical CV changes due to the short length of postpartum follow-up (12 months). Clinically evident CV outcomes will probably require much longer follow-up given that the studied population will only include young, pregnant women. However, our aim is to initiate a cohort study of patients with high-risk for PE or that develop the disease, in order to adequately follow-up in the long-term, as we have previously done with other cohorts. Development of a predictive model of future CVD risk or complications, including clinical and echocardiographic variables, CV and placental biomarkers, will require further validation in prospective well-powered studies. Finally, in our study we will assess associations between CV findings and angiogenic factors but finding of potential causal links between both will still require further studies.

To conclude, this study will provide, for the first time, evidence characterizing the association between angiogenic imbalance and CV dysfunction/remodeling in women at risk for or developing PE and, in their offspring, both during and after pregnancy. We expect to provide data supporting the association between abnormal angiogenesis and CV dysfunction/remodeling during pregnancy and fetal life, therefore identifying new therapeutic targets for future studies aiming to prevent CV morbidity associated with PE. The goal of this research is to obtain insight in the particular risk factors that link pregnancy and CV health, and thus, reduce the burden of the disease.