Kathryn Lindley, M.D.
Project Summary:
Preeclampsia (PE) during pregnancy is associated with the development of future cardiovascular disease (CVD). Recent epidemiologic studies suggest that rather than being a marker of CVD, PE is mechanistically linked to the development of CVD. A potential mechanism for this link is through the development of endothelial dysfunction resulting in left ventricular diastolic dysfunction (LVDD) and the subsequent development of Heart Failure with Preserved Ejection Fraction (HFpEF). The objective of this pilot study was to identify PE-related antiangiogenic biomarkers associated with LVDD. The central hypothesis is that increased levels of antiangiogenic biomarkers (sFLT1, sFLT1/PlGF ratio) is associated with postpartum LVDD.
We prospectively enrolled 48 women with severe PE during their hospitalization for delivery. Within 24 hours of delivery, subjects underwent an echocardiogram (to evaluate for LV structure and LVDD) and measurement of sFLT1 and PlGF. At the 12-week post-partum follow-up they underwent a repeat echocardiogram. Overall, the cohort had multiple risk factors for preeclampsia including Black race (56%), obesity (mean BMI 34.2, 63% obese), and chronic hypertension (20%). Diagnosis of preterm PE occurred in 59% of women. Echocardiographic findings were consistent with age-adjusted indices of LV diastolic dysfunction (Septal e’: 9.4±2.5; Lateral e’: 11.9±3; E/e’: 9.3±2.6; LAVI: 33.1±8.7); 50% had LV remodeling (strong predictor of CVD & HF risk); and those with abnormal diastolic function exhibited a trend towards elevated angiogenic biomarkers (sFLT1/PlGF: 2,446±5,059 vs. 490±1,362, p=0.09), consistent with our hypothesis of an association between angiogenic biomarkers and development of LVDD. Given the outstanding recruitment (despite COVID-19) and significant novel findings suggestive of a clinical path towards development of HFpEF, and the high likelihood of subsequent funding based on these and additional preliminary data supported by LLF, continuation of this study with further characterization of exercise tolerance and invasive hemodynamics is warranted.
Progress Report
Recruitment in year one of this study has been outstanding, despite limitations secondary to COVID-19. Our initial echo data identifies findings consistent with significant left ventricular diastolic dysfunction in the majority of women with severe PE. This is consistent with prior studies showing that during a pre-eclamptic pregnancy, significant LV concentric remodeling occurs, with 50% of women developing LVDD, documented in small studies to persist for several years post-partum. Asymptomatic LV concentric remodeling (a pattern that has been shown to progress to LV enlargement and/or dysfunction) and LVDD have also been identified at increased rates in previously pre-eclamptic women years after pregnancy. Patients with abnormal cardiac remodeling, particularly concentric hypertrophy, are at increased risk for adverse cardiovascular events including heart failure.
In the current study, we identified an association between LVDD and elevated sFLT1/PlGF ratio, indicating a dose-response relationship between angiogenic imbalance and cardiac dysfunction. Our group also recently identified a possible dose-response effect related to duration of expectant management of preterm preeclampsia, yielding an increased risk of cardiovascular hospital admission or death. In light of the important role of sFLT1 in angiogenesis, we hypothesize that elevated levels of sFLT1 in PE lead to decreased capillary density, resulting in diffuse myocardial ischemia. Diastolic dysfunction is a sensitive marker of myocardial ischemia and is apparent before systolic dysfunction is present. There is a high incidence of diastolic dysfunction in patients with endothelial dysfunction and microvascular ischemia.
Dysregulated angiogenesis may occur in the setting of placental hypoxia, leading to excessive production of antiangiogenic factors such as sFLT1 and proinflammatory cytokines such as TNF-α, as well as reduced levels of proangiogenic factors such as PlGF. sFLT1 is an anti-angiogenic protein produced predominantly by the placenta, but also by monocytes and endothelial cells. sFLT1 is a soluble vascular endothelial growth factor (VEGF) inhibitor, which binds to VEGF and PlGF in the circulation, preventing their interaction with their endothelial receptors. VEGF is important in both angiogenesis and in maintenance of endothelial cell health. PlGF works synergistically with VEGF to promote angiogenesis. sFLT1 has been shown in vitro to induce vasoconstriction and endothelial dysfunction, and induces a syndrome resembling PE in rats, with the development of hypertension, proteinuria, and glomerular endotheliosis. In the mouse model, exposure to sFLT1 leads to decreased capillary density (and thus ischemia/hypoxia) and impaired cardiac function. Thus, sFLT1 and PlGF appear to play an important pathophysiologic role in the development of PE, and possibly also in the subsequent increased prevalence of post-partum hypertension, cardiovascular dysfunction including progression to heart failure, and/or CAD.
HFpEF, a disease that predominantly affects women with a history of hypertension, exhibits significant overlap with PE in terms of clinical presentation and abnormal serum biomarkers, supporting the hypothesis that these two condition share common pathophysiologic pathways. After a pre-eclamptic pregnancy, pre-existing CV risk factors and irreversible endothelial and myocardial damage from PE contribute to the development of excess CV morbidity and mortality due to hypertension, heart failure (HFpEF and HFrEF), and CAD. Recently we described the spectrum of echocardiographic phenotypes of peripartum heart failure, showing that 81% of women with HFpEF had a history of PE, vs. 45% of those with HFrEF (P<0.001). A recent epidemiologic analysis from our group identified that PE was independently associated with hospitalization for HFpEF in the decade following pregnancy (aHR 2.09), after adjusting for traditional HFpEF risk factors including hypertension and diabetes.
These data provide evidence in support of PE and the associated LVDD being important echocardiographic biomarkers of later HFpEF. Increased sFLT1 levels have previously been associated with adverse outcomes including increased mortality in patients with heart failure. sFLT1 levels rise during the third trimester of all pregnancies but are markedly elevated in pregnancies complicated by PE. sFLT1 levels then rapidly decline following delivery; however, circulating levels of sFLT1 in PE patients remain persistently elevated above post-partum levels compared to patients without PE. The source of the persistent sFLT1 secretion has not been well-characterized, but placental remnants, mononuclear cells, and shed syncytial microparticles have been suggested. Higher levels of sFLT1 are associated with more severe PE phenotypes. Our group recently identified that women with PE who were treated with prophylactic aspirin, which has been shown in vitro to reduce hypoxia-induced sFLT1 production, had reduced sFLT1 and sFLT1/PlGF levels, and lower incidence of postpartum hypertension. Thus, ongoing endothelial dysfunction and myocardial ischemia in the setting of persistently elevated postpartum sFLT1 levels may contribute to subsequent exercise intolerance due to HFpEF after a pre-eclamptic pregnancy.
Pilot data from our group has shown wide variability in exercise tolerance among age- and BMI-matched women 3-12 months postpartum. In the absence of underlying chronic hypertension or diabetes mellitus to account for these differences, angiogenic insult due to recent PE contributing to ongoing diastolic dysfunction is suspected. Impaired exercise capacity is a hallmark of HFpEF, and pulmonary capillary wedge pressure (PCWP) at peak stress has previously been identified as an independent variable consistent with HFpEF. Exercise PCWP ≥25 has previously been classified as consistent with HFpEF in invasive hemodynamic studies. If invasive RHC hemodynamic study identifies the presence of exercise-induced HFpEF in women with recent PE, post-partum exercise stress testing could be used as a predictor of future HFpEF, representing an opportunity to risk-stratify a postpartum population in whom aggressive risk modification and/or treatment strategies can be evaluated to delay and/or prevent development of clinically-relevant CVD. Based on the preliminary findings of our study, and these pilot exercise data, continuation of the study is warranted.
