Nitric Oxide-mediated Coronary Endothelial Function is Impaired in Patients with Heart Failure with Preserved Ejection Fraction

The pathophysiology of heart failure with preserved ejection fraction (HFpEF) is incompletely understood but is likely multifactorial and critical to the development of effective therapeutic interventions. Abnormal coronary endothelial-dependent function was proposed as a driver of HFpEF pathophysiology¹, but the few prior reports of endothelial-dependent function in HFpEF patients used invasive catheter-based methods that limit HFpEF participants to those referred for catheterization and precluded studies in healthy controls. Therefore we exploited a noninvasive MRI method to directly measure coronary endothelial-dependent function (CEF) by assessing changes in coronary vasodilatation and blood flow in response to isometric hand-grip exercise (IHE), a demonstrated endothelial-dependent stressor.^2,3 We tested the hypothesis that NO-mediated CEF is impaired in outpatients with stable HFpEF as compared to the responses of age-, sex-, and race-comparable control subjects by performing 3T coronary MRI without contrast administration before and during IHE using previously described methods.^2,3 After written informed consent approved by the Institutional Review Board, MRI was performed in three groups: control normotensive subjects with no HF history, control hypertensive subjects with no HF history, and HFpEF patients with a clinical diagnosis of HF (NYHA II-III) symptoms for ≥1 month and LVEF ≥50%, but no history of systolic HF (LVEF <40%), infiltrative cardiomyopathy, constrictive pericarditis, significant valvular disease, or cor pulmonale. Subjects with a history of clinical CAD, including acute coronary syndrome, myocardial infarction or coronary revascularization were excluded. If coronary angiography or CTA was available, those with >50% stenosis or revascularization were excluded. Data are available upon reasonable request.

There were no significant differences in age, sex, race, smoking status, baseline/exercise BP, RPP, coronary area, flow, or resistance among the three groups. Body mass index (BMI) was significantly higher in the HFpEF cohort, (mean±SD, HFpEF: 37.6±4.6 kg/m², p=0.017), than in normotensive or hypertensive controls (31.3±4.7 and 32.1±5.5 kg/m²), however all cohorts met criteria for obesity. Importantly, the IHE-induced coronary cross-sectional area change (CSA, vasodilation) and the response of coronary blood flow (CBF), another metric of endothelial-dependent function, were lower in HFpEF subjects than in control subjects (Figure 1). When abnormal CEF was defined, as previously reported, as a CSA change of ≤0% or a CBF change of ≤0%, reflecting what is sometimes referred to as “macrovascular” and “microvascular” CEF, respectively, then abnormal CEF was present in 60% of HFpEF subjects in comparison to 0% of age-comparable normotensive and hypertensive control subjects (p<0.001). 30% of HFpEF participants had abnormal CSA change alone, 10% had abnormal CBF change alone, and 20% had both abnormal CSA and CBF change in response to IHE stress.

Figure 1:

Coronary Endothelial Function is Impaired in Patients with Heart Failure with Preserved Ejection Fraction (HFpEF). HFpEF patients (n=10, mean± standard deviation (SD), 63.2±8.7 years, 50% Women, 50% Black adults), and two groups of age-, sex-, and race-comparable controls: Normotensive Control (n=10, 55.3±18.4 years, 50% Women, 30% Black adults) and Hypertensive Control (n=12, 57.3±11.8 years, 75% Women, 67% Black adults) subjects underwent coronary magnetic resonance imaging (MRI) at rest and during isometric handgrip exercise (IHE) stress. The diagnosis of HFpEF was based on 1) clinical history of congestive heart failure (HF) requiring hospitalization, outpatient intravenous (IV) diuretics, or outpatient oral diuretics to control symptoms, 2) echocardiographic evidence of diastolic dysfunction (abnormal diastology, elevated left atrial pressure by E/E’, or left atrial dilation) within prior 4 months of enrollment, 3) left ventricular ejection fraction (LVEF) ≥ 50%. Coronary cross-sectional area (CSA) significantly increases in normotensive controls (A) and hypertensive controls (B) with IHE stress, but not in HFpEF subjects (C). Similarly, coronary blood flow (CBF) significantly augments in normotensive and hypertensive control subjects (D,E), but not in HFpEF subjects (F). CSA is measured in millimeters-squared (mm²) and CBF in milliliters per minute (ml/min). Summary data (G) shown for HFpEF subjects (red bars) demonstrate decreased coronary endothelial function (CEF) as indicated by significantly impaired cross-sectional area change and coronary blood flow augmentation in response to IHE stress, as compared to that of age-comparable, normotensive (gray bars) and hypertensive controls (blue bars). The change in coronary resistance (CR), calculated as mean blood pressure (BP) divided by coronary blood flow (CBF), during IHE was greater in HFpEF than in hypertensive control participants. Each box represents mean±95% confidence interval (CI). Analysis of Variance (ANOVA) with Tukey were used for comparison of the three groups with correction for multiple comparisons. Statistical significance was considered a two-tailed p-value <0.05 or confidence interval of 95%.

Most prior noninvasive imaging studies in HFpEF patients measured left ventricular flow reserve changes to endothelial-independent stressors without specifically measuring endothelial-dependent coronary vasomotor responses. In two prior large invasive studies of endothelial-dependent responses, one a retrospective, cross-sectional study between 1993–2015 in patients referred for cardiac catheterization and the other in hospitalized HFpEF referred for catheterization, roughly 24%–40% of HFpEF patients exhibited abnormal endothelial-dependent CEF.^4,5 There were no healthy control subjects for comparison in either catheterization-based study.

In the current study, we assessed endothelial-dependent CEF by previously validated noninvasive MRI obtained before and during IHE to directly visualize and measure coronary vasomotor changes. We prospectively enrolled a contemporary population of stable, outpatient HFpEF patients who were not referred for cardiac catheterization, thus avoiding the inherent referral and selection bias of previous catheterization-based CEF studies. This also allowed enrollment of age-comparable controls without known heart disease. We report a significant reduction in CEF in HFpEF patients as compared to age-comparable normotensive and hypertensive control subjects, as measured by both macrovascular (CSA) and microvascular (CBF) indices (Figure 1). We also observe a high prevalence of coronary endothelial-dependent dysfunction by blood flow criteria in 30% of these HFpEF patients, similar to prior invasive CEF studies (~24%–40%). If we also include impaired vasodilatation in the criteria for abnormal CEF, then the prevalence of abnormal CEF is even higher, at 60%, in these stable, outpatient HFpEF subjects, while none of the controls exhibited abnormal CEF. The other strengths of this prospective study include the use of a reproducible, safely administered CEF test that avoids intravenous or intracoronary infusions and that is well suited to serial studies over time. Given the prevalence of obesity in the control cohort, the relative impairment in CEF in HFpEF reported here may be underestimated if the results are compared to those of a cohort of non-obese, control subjects.

In summary, CEF is significantly impaired in stable HFpEF outpatients without obstructive coronary artery disease, as compared to age-comparable, obese normotensive and hypertensive controls. A high proportion (~60%) of contemporary, stable HFpEF outpatients exhibit abnormal coronary endothelial function which now can be safely and serially assessed noninvasively. Coronary endothelial function is an appealing target for further mechanistic studies and HFpEF therapeutics.

Non-standard Abbreviations and Acronyms:

BMI

body mass index

CAD

coronary artery disease

CEF

coronary endothelial function

CTA

computed tomographic angiography

CBF

coronary blood flow

CFV

coronary flow velocity

CSA

cross-sectional area

ECG

electrocardiogram

eNOS

nitric oxide synthase

HF

heart failure

HFpEF

heart failure with preserved ejection fraction

IHE

isometric handgrip exercise

L-NMMA

N-Methylarginine (nitric oxide inhibitor)

LVEF

left ventricular ejection fraction

MRI

magnetic resonance imaging

NO

nitric oxide

RPP

rate pressure product

SD

standard deviation

3T

three Tesla