Peripheral Microvascular Function Reflects Coronary Vascular Function

Abstract

Background

Coronary endothelial dysfunction is a precursor of atherosclerosis and adverse outcomes. Whether endothelial dysfunction is a localized or generalized phenomenon in humans remains uncertain.

Objectives

We simultaneously measured femoral and coronary vascular function with the hypothesis that peripheral vascular endothelial function will be reflective of coronary endothelial function.

Approach and Results

Eighty-five subjects underwent coronary angiography for evaluation of chest pain or abnormal stress tests. Endothelium-dependent and -independent vascular function were measured using intra-coronary and intra-femoral infusions of acetylcholine and sodium nitroprusside, respectively. Coronary flow reserve was assessed using intra-coronary adenosine infusion. Flow velocity was measured in each circulation using a Doppler wire (FloWire, EndoSonics). Coronary vascular resistance (CVR) and femoral vascular resistance (FVR) were calculated as mean arterial pressure (mmHg)/coronary blood flow (ml/min) and mean arterial pressure (mmHg)/femoral average peak velocity (cm/s), respectively. Mean age was 53±11 years, 37% female, 44% hypertensive, 12% diabetic and 38% had obstructive coronary artery disease. There was a correlation between the change in FVR with acetylcholine and acetylcholine-mediated changes in both the CVR (r=0.27; p=0.014) and in the epicardial coronary artery diameter (r=−0.25; p=0.021), indicating that subjects with normal endothelial function in the femoral circulation had normal endothelial function in the coronary epicardial and microcirculation, and vice versa. The coronary vasodilator response to adenosine also correlated with the femoral vasodilatation with acetylcholine (r=0.4; p=0.0002). There was no correlation between the coronary and femoral responses to sodium nitroprusside.

Conclusions

Endothelial functional changes in the peripheral and coronary circulations were modestly correlated. Thus, peripheral microvascular endothelial function reflects endothelium-dependent coronary epicardial and microvascular function, and the coronary flow reserve.

INTRODUCTION

The endothelium modulates vascular health by releasing a variety of paracrine factors that modulate smooth muscle vasomotion, platelet activation, and over the longer term influence endothelial layer integrity and susceptibility to atherosclerosis ^1–5. These endothelium-derived factors include nitric oxide (NO) that is synthesized in the endothelial cells by the constitutive action of endothelial nitric oxide synthase on L-arginine that is converted to L-citrulline and NO. NO then diffuses into the lumen and contributes to passivation of platelet aggregation and into the smooth muscle to stimulate cyclic guanosine monophosphate-dependent relaxation. Other endothelium-dependent vasodilators include endothelium-derived hyperpolarizing factors, prostaglandins, carbon monoxide and others ^6–9. Release of these endothelium-dependent agonists can be stimulated by pharmacologic probes such as acetylcholine, bradykinin, and substance P, and by physical stimuli including increased flow or shear ^10–14.

We and others have demonstrated that the endothelial reactivity in response to these pharmacologic probes indicative of vasomotion during physiologic stimulation as well. Coronary flow and epicardial responses to pacing, cold pressor testing and mental stress correlate with the responses to acetylcholine ^{7, 15–17}. We also have shown that forearm microvascular dilation in response to exercise mediated by NO release is greater in healthy subjects compared to those with hypercholesterolemia ¹⁸. Importantly, endothelial dysfunction in the coronary circulation is an independent predictor of worse long term prognosis in patients with and without significant coronary artery disease (CAD) ^19–21. Moreover, coronary microvacular endothelial dysfunction is commonly observed in patients with chest pain and non-obstructed coronary arteries ²². Finally, treatments associated with improvement in cardiovascular outcomes, including physical exercise, statins and angiotensin antagonists, at least partly, impart their beneficial effects by improving endothelial function ^23–26.

Endothelial function in the coronary and peripheral circulations can be measured using invasive techniques using infusions of endothelium-dependent agonists ^27–31. In subjects with normal coronary endothelial function, acetylcholine promotes vasodilation of both the coronary epicardial and microvascular circulations by stimulating endothelial muscarinic receptors that activates endothelial nitric oxide synthase to synthesize NO and other endothelium-dependent vasodilators. However, exposure to cardiovascular risk factors such as hypercholesterolemia, hypertension, aging and atherosclerosis leads to endothelial dysfunction and absence or reduction in NO synthesis. In these circumstances, the direct constrictor effects of acetylcholine on vascular smooth cells predominate, and the net effect is epicardial coronary vasoconstriction and reduced microvascular vasodilation ^{7, 32–36}

Non-invasive techniques that utilize increased shear as a stimulus for nitric oxide release are commonly used to test brachial arterial endothelial function. A significant correlation between flow-mediated brachial artery dilation and coronary epicardial endothelial function, both estimates conductance vessel function, has been reported previously ^37–40. However, whether endothelial function in both the coronary epicardial and microcirculation is similar to endothelial function in the peripheral circulation, assessed in detail using acetylcholine infusions, remains unknown, and was the aim of this study. Herein, we measured endothelium-dependent and -independent coronary epicardial and microvascular function and coronary flow reserve and compared these measures with endothelium-dependent and -independent function in the femoral circulation. Studies were performed simultaneously in patients with and without atherosclerotic risk factors and CAD. Our hypothesis was that endothelial function in different vascular beds is similar, and therefore endothelial dysfunction in the coronary epicardial arteries and microcirculation will be reflected in the femoral microcirculation.

MATERIAL AND METHODS

The data that support the findings of this study are available from the corresponding author upon reasonable request. We studied 85 subjects with chest pain or abnormal cardiac tests who underwent cardiac catheterization at the Cardiology branch of the National Institutes of Health. The study was approved by the Institutional Review Board, and all subjects gave written informed consent. Subjects with an acute coronary syndrome, myocardial infarction in the past 3 months, heart failure of New York Heart Association class III to IV, severe valvular disease, or those with left main or 3-vessel disease were excluded. CAD was defined as any angiographic evidence of plaque, luminal irregularity, or stenosis in any epicardial coronary artery. Non-obstructive CAD was defined as angiographic evidence of <50% stenosis in the epicardial coronary artery. Cardiac risk factors included age, body mass index (BMI), history of tobacco smoking (current or use within the past year), diabetes, hypertension, and hyperlipidemia. Diabetes was defined as a fasting blood glucose level >126 mg/dL or if the subject was receiving insulin or oral hypoglycemic therapy. Hypertension was defined as systolic blood pressure ≥140 mm Hg or diastolic pressure ≥90 mm Hg on at least 3 occasions, or the patient was being treated with medications. Hyperlipidemia was defined as a fasting serum low-density lipoprotein >130 mg/dL or if the subject was being treated with lipid-lowering medications or dietary modification.

Subjects were asked to refrain from smoking, drinking alcohol or caffeinated beverages for at least 24 hours before the studies. Additionally, all cardiac medications were withdrawn at least 48 hours before the study.

Coronary Vasomotor Function Assessment

After performing diagnostic coronary angiography, a 3F infusion catheter was introduced via a 7F guide catheter into a non-obstructed coronary artery (stenosis severity ≤50%). Blood flow velocity was measured, as previously described, with a 0.014- or 0.018-in Doppler wire (FloWire, EndoSonics) ¹⁹. Coronary blood flow (CBF) was calculated as arterial cross-sectional area X average flow velocity X 0.5 ⁴¹. The change in CBF was calculated as: (CBF^peak-CBF^rest/CBF^rest) x100. Coronary artery diameter measured in a 0.25- to 0.5-cm segment of vessel beginning 0.25 cm beyond the tip of the flow wire. Coronary vascular resistance (CVR) was calculated as: mean arterial pressure/CBF. Left coronary angiograms were obtained in the best projection to estimate epicardial coronary artery diameter at baseline and after each infusion to determine the percent change in diameter during drug administration compared to baseline and calculate the CBF in response to each vasoactive medication. Quantitative angiography was performed by one observer blinded to the identity of the study infusion with ARTREK software (Quantim 2001, StatView, ImageComm Systems, Inc) or the PIE medical CAAS system ³⁶ (Figure 1).

Figure 1: Coronary angiograms, intracoronary Doppler tracing and hemodynamic measurements during the study protocol.

The Doppler Wire is positioned in the mid-left anterior descending coronary artery. Coronary artery diameter is measured 5 mm distal to the Doppler wire tip at rest (A) and after acetylcholine infusion (B). Acetylcholine infusion resulted in severe coronary vasoconstriction. (C) Intracoronary Doppler tracing and hemodynamic measurements of the subject during the study protocol. LAD: left anterior descending artery; Ach: acetylcholine; APV: average peak velocity.

Endothelium-dependent vasodilation was assessed using intracoronary acetylcholine infusions for 2-minutes at 15 µg/min [estimated intracoronary concentrations of 10−⁶ mol/L). Ten minutes after acetylcholine infusion, endothelium-independent function was assessed using a 3-minute intracoronary infusion of 20 µg/min sodium nitroprusside, followed by measurement of coronary flow reserve with a 2-minute infusion of 2.2 mg/min of adenosine. After each medication infusion, cine image was obtained for quantitative coronary angiography.

Femoral Vascular Function Assessment

A 6F multipurpose catheter (CORDIS, Inc) was introduced through a 7F femoral arterial sheath to just above the sheath in the femoral artery. A 0.018-in Doppler wire was introduced and positioned 1 cm beyond the catheter tip. Because the Doppler Wire was positioned proximal to the delivery of the vasoactive medications, measurements of femoral artery diameter changes were not made. However, to exclude any significant changes in femoral artery diameter at the level of Doppler Wire in response to different vasoactive medications, we measured the changes in femoral artery diameter at the level of the flow wire in a preliminary study of 24 patients using serial angiography. Baseline diameter was 5.1±0.9 mm, while in response to 300 µg/min acetylcholine was 5.1±0.9 mm and 40 µg/min sodium nitroprusside was 5.1±0.9 mm. There was no constriction observed in femoral artery diameter with acetylcholine in patients with atherosclerosis or its risk factors. Hence, the femoral vascular resistance (FVR) was calculated as the mean arterial pressure/femoral average peak velocity ^{31, 42}.

After obtaining the resting femoral arterial average peak flow velocity, endothelium-dependent vasodilation in the femoral circulation was assessed using intra-femoral infusions of acetylcholine (300 µg/min) for 2-minutes. Following a 15-minute rest period and return to baseline flow velocity, intra-femoral sodium nitroprusside infusion (40 µg/min for 2-minutes) was initiated to assess endothelium-independent function and the femoral average peak velocity measurement.

Statistical analysis

Clinical variables are summarized as mean ± standard deviation for continuous variables and using count (%) for categorical variables. Data normality was assessed using the Kolmogorov-Smirnov test. Paired t-tests were used when no gross violation of normality was present and Wilcoxon signed-rank tests were used for non-normally distributed variables. All probability variables were two-tailed and a significance level of 0.05 was used for all analyses. Univariate and multivariate linear regression analyses were performed to determine the relationships between changes in CVR and FVR and clinical variables in both circulations in response to acetylcholine and sodium nitroprusside. Covariates entered in this regression model were age, BMI, sex, history of hypertension, diabetes, current smoking, hyperlipidemia and presence of CAD. Correlation analyses were performed using Spearman’s correlation coefficient. Analysis was performed using IBM SPSS version 25 (IBM, New York, NY).

RESULTS

Baseline characteristics of the enrolled subjects are summarized in Table 1. Mean age was 53±11 years and 54 (64%) were male subjects.

Table 1:

Characteristics of the study population

	All subjects (N=85)
Age (years), mean±SD	53±11
Female, N (%)	31 (37%)
BMI (kg/m2), mean±SD	29±5
White, N (%)	68 (80%)
Cholesterol (mg/dL), mean±SD	218±49
LDL (mg/dL), mean±SD	142±42
HDL (mg/dL), mean±SD	41±12
CAD, N (%)  - No CAD  - Non-obstructive CAD  - Obstructive CAD	47 (55%) 6 (7%) 32 (38%)
History of hypertension, N (%)	37 (44%)
History of diabetes, N (%)	10 (12%)
History of smoking, N (%)	52 (61%)

BMI: body mass index, LDL: Low-density lipoprotein, HDL: High-density lipoprotein, CAD: coronary artery disease. Obstructive CAD: coronary artery luminal stenosis ≥50%.

Coronary Vascular Function

There was no significant change in the mean arterial pressure in response to intra-coronary acetylcholine infusions (108±15 [rest] vs 108±15 [acetylcholine] mm Hg, p=0.87). The mean coronary epicardial artery diameter change in response to acetylcholine was 1.4±10% (range −25% to 30%), with 38 (45%) subjects demonstrating constriction (defined as % coronary diameter < 0%), indicating endothelial dysfunction. CBF increased by 122±98% (range −30 to 432%) and CVR decreased by −46±23% (range −80% to 43%) with acetylcholine (Table 2). In multivariate regression analyses, increased age was an independent predictor of the decrease in CBF in response to acetylcholine (Supplemental Table I).

Table 2:

Vasomotor parameters and changes in response to vasoactive medications

	Mean±SD	%Change
Coronary Study
Coronary blood flow (mL/min)  - Baseline  - Acetylcholine infusion  - Sodium nitroprusside infusion  - Adenosine infusion	42±30 94±70 99±68 172±110	122±98% 152±84% 361±169%
Coronary vascular resistance (mmHg.mL −1.min)  - Baseline  - Acetylcholine infusion  - Sodium nitroprusside infusion  - Adenosine infusion	3.6±2.3 2.0±1.6 2.0±1.3 0.98±0.7	−46±23% −59±15% −75±9%
Coronary artery diameter (mm)  - Baseline  - Acetylcholine infusion  - Sodium nitroprusside infusion	2.5±0.68 2.5±0.73 2.9±0.78	1.4±10% 19.4±16%
Femoral Study
Femoral blood flow velocity (cm/sec)  - Baseline  - Acetylcholine infusion  - Sodium nitroprusside infusion	19±8 46±26 42±16	139±108% 124±57%
Femoral artery resistance (mmHg.mL −1.min)  - Baseline  - Acetylcholine infusion  - Sodium nitroprusside infusion	5.9±2.3 2.8±1.6 2.3±0.9	−54±17% −59±10%

In response to sodium nitroprusside infusion, mean arterial pressure decreased from 108±15 mmHg [rest] to 99±15 mmHg; p<0.0001. Coronary artery diameter change was 19±16%; CBF increased by 152±84% (range 0.5% to 392%) and CVR decreased by −59±15% (range −83% to −8%). Presence of CAD was an independent predictor of reduced vasodilation in response to sodium nitroprusside. There was no correlation between the demographic and risk factor variables and the % changes in CBF or CVR in response to sodium nitroprusside infusion (Supplemental Table I).

CBF increased by 360±168% (range 103% to 880%) and CVR decreased by −75±9 (range −88% to −53%) with intracoronary adenosine infusion. Presence of CAD and diabetes were independent predictors of decreased CBF and vasodilator reserve in response to adenosine (Supplemental Table I).

Femoral Vascular Function

There was a significant reduction in mean arterial pressure in response to intra-femoral acetylcholine (101±14 [rest] vs 95±16 [acetylcholine] mmHg, p<0.0001 and intra-femoral sodium nitroprusside (100±14 [rest] vs 87±15 [sodium nitroprusside] mmHg; p<0.0001) infusions. In response to acetylcholine, the femoral average peak velocity increased by 139±108 % (range −5 to 511%) and the FVR decreased by −52±17% (range −93% to −10%). With sodium nitroprusside, femoral average peak velocity increased by 124±57% (range 33 to 273%) and FVR decreased by −59±10% (range −81% to −29%). In multivariate analyses adjusted for the aforementioned covariates, increasing age and male sex were independent predictors of reduced femoral vasodilation with acetylcholine infusion. Only male sex was an independent predictor of reduced vasodilation in response to sodium nitroprusside (Table 3).

Table 3:

Univariate and multivariate determinants of femoral vascular function

	Univariate Analysis		Multivariate Analysis
	β value	p value	β value	p value
% change in femoral vascular resistance in response to acetylcholine
Age	0.69	<0.001	0.55	0.002
BMI	0.4	0.3	0.17	0. 65
Male sex	7.8	0.047	7.5	0.045
Hypertension	4.2	0.27	2.1	0.56
Diabetes	12.2	0.037	5.6	0.32
Smoking	−2.5	0.53	−4.9	0.19
Hyperlipidemia	6.8	0.08	5.0	0.15
Presence of CAD	10.2	0.007	3.9	0.31
% change in femoral vascular resistance in response to sodium nitroprusside
Age	−0.15	0.18	−0.14	0.22
BMI	0.38	0.11	0.26	0.3
Male sex	4.1	0.09	4.9	0.042
Hypertension	4.1	0.08	2.7	0.3
Diabetes	5.8	0.12	6.1	0.11
Smoking	−3.0	0.21	−2.8	0.25
Hyperlipidemia	2.0	0.41	−1.8	0.44
Presence of CAD	−2.3	0.34	−2.4	0.37
% change in femoral blood flow velocity in response to acetylcholine
Age	−3.1	0.005	−1.8	0.072
BMI	−4.2	0.08	−0.9	0.36
Male sex	−52.5	0.03	−2.1	0.038
Hypertension	−35.2	0.14	−0.9	0.33
Diabetes	−70.3	0.053	−0.7	0.49
Smoking	20.3	0.40	1.6	0.11
Hyperlipidemia	−33.4	0.16	−0.9	0.33
Presence of CAD	−73.9	0.001	−2.0	0.049
% change in femoral blood flow velocity in response to sodium nitroprusside
Age	1.3	0.03	2.0	0.045
BMI	−1.5	0.27	−0.4	0.72
Male sex	−21.0	0.12	−2.1	0.037
Hypertension	−20.5	0.12	−1.4	0.18
Diabetes	−16.0	0.43	−0.9	0.37
Smoking	20.9	0.12	1.7	0.09
Hyperlipidemia	6.8	0.61	0.3	0.78
Presence of CAD	17.6	0.18	0.8	0.43

BMI: body mass index, CAD: coronary artery disease.

Relationship between the microvascular responses in the coronary and femoral circulations

Compared to the resting average peak velocities in the coronary circulation, resting average peak velocities in the femoral circulation were significantly lower (59.9±32 versus 46.1±26; p=0.0002). There was a significant correlation between the magnitude of coronary and femoral microvascular vasodilation in response to acetylcholine (Figure 2). Thus, the % change in FVR correlated with the % change in CVR (r=0.27; p=0.014) and the % change in femoral average peak velocity correlated with the % change in CBF in response to acetylcholine infusions (r=0.24; p=0.026). In contrast, there was no correlation between the endothelium-independent coronary and femoral microvascular vasodilation in response to sodium nitroprusside infusion (Supplemental Figure 1).

Figure 2: Relationship between coronary and femoral microvascular function in response to acetylcholine.

Relationship between the percent changes in (A) femoral blood flow velocity and coronary blood flow in response to acetylcholine infusions, (B) femoral vascular resistance and coronary vascular resistance in response to acetylcholine infusions.

Relationship between the epicardial coronary arterial and femoral microvascular responses

There was a significant correlation between changes in the epicardial coronary diameter and femoral microvascular dilation in response to acetylcholine (r=−0.25; p=0.021) (Figure 3). Subjects with epicardial coronary vasoconstriction in response to acetylcholine (N=38 subjects) had attenuated femoral vasodilation (FVR: −50±16.6 vs −57±17.6 mmHg.mL −¹.min; p=0.044) compared to those with epicardial coronary vasodilation (N=44 subjects). Similarly, the femoral average peak velocity in response to acetylcholine was significantly lower in subjects with epicardial coronary vasoconstriction compared to those with epicardial coronary vasodilation (114±94 vs 165±117 cm/sec; p=0.033). There was no association between in femoral microvascular dilation and the coronary diameter changes in response to sodium nitroprusside given in each circulation (Figure 3).

Figure 3: Relationship between coronary epicardial and femoral microvascular functions.

Relationship between the percent changes in (A) femoral vascular resistance and change in coronary diameter in response to acetylcholine infusion, (B) femoral vascular resistance and change in coronary diameter in response to sodium nitroprusside infusion. NS: not significant.

Relationship between the coronary flow reserve with adenosine and femoral microvascular responses

There was a significant correlation between coronary microvascular dilation in response to adenosine (coronary flow reserve) and femoral microvascular dilation in response to acetylcholine infusions (Figure 4). Thus, the change in CVR with adenosine correlated with the change in FVR with acetylcholine (r=0.4; p=0.0002). There was no association between the coronary flow reserve and femoral vasodilation with sodium nitroprusside.

Figure 4: Relationship between coronary flow reserve and femoral microvascular function.

Relationship between the percent changes in (A) femoral blood flow velocity and coronary blood flow in response to acetylcholine and adenosine infusions, respectively, (B) femoral vascular resistance and coronary vascular resistance in response to acetylcholine and adenosine infusions, respectively.

DISCUSSION

In a study that comprehensively examined the relationship between the coronary and peripheral endothelium-dependent and -independent vascular function in response to acetylcholine and sodium nitroprusside, respectively, performed simultaneously during the catheterization procedure, in a diverse population with and without coronary risk factors and CAD, we found that peripheral vascular endothelial responses to acetylcholine reflect (a) endothelial function in both the coronary epicardial arteries and microcirculation, and (b) the coronary flow reserve in response to adenosine (Graphic Abstract). There was no relationship between the peripheral and coronary endothelium-independent function assessed using sodium nitroprusside. Thus, subjects with peripheral vascular endothelial dysfunction had (a) coronary epicardial endothelial dysfunction, (b) coronary microvascular dysfunction, and c) abnormal coronary flow reserve in response to adenosine.

Previous studies have only investigated the relationship between conductance vessel endothelial function in the brachial and coronary epicardial arteries and reported similarity in endothelial function of the two circulations ^{37, 38}. To our knowledge this is the first study to assess endothelium-dependent and -independent microvascular function using acetylcholine and sodium nitroprusside infusions, respectively, in both the coronary and peripheral circulations. Conditions such as coronary microvascular dysfunction, where patients present with chest pain despite non-obstructive epicardial coronary arteries, are associated with impaired coronary microvascular endothelial function and reduced flow reserve ^{22, 43}. In this population and those with CAD, the magnitude of endothelial dysfunction and the magnitude of abnormality in flow reserve is predictive of adverse cardiovascular outcomes ^{20, 21, 44–46}. However, measurement of coronary endothelial function and flow reserve requires invasive testing that is expensive, time consuming, and not free of risk. There is thus a critical need for simpler tests that can accurately estimate coronary endothelial function and flow reserve. This would aid clinicians in both the diagnosis and prognostication of patients undergoing coronary angiography. Our study shows that assessment of peripheral endothelial function at the end of the diagnostic catheterization procedure, using a single infusion of acetylcholine, can provide valuable information regarding the coronary vascular endothelial function and coronary flow reserve. Importantly, a previous study has demonstrated the importance of acetylcholine-mediated vasodilator responses in the forearm microcirculation as an independent predictor of long term outcomes in CAD patients ⁴⁷.

Previous studies have suggested that coronary vascular dysfunction is associated with peripheral vascular structural changes. Also, subjects with coronary microvascular dysfunction appear to have narrower retinal arterioles ⁴⁸. Another study demonstrated that vasoreactivity in peripheral microvessels was abnormal in patients with vasospastic coronary syndromes and those with microvascular angina ⁴⁹. Whether the abnormal microvascular function in the coronary and peripheral circulation we observed is a functional abnormality or secondary to generalized structural changes in the microcirculation remains unknown. In a study examining the coronary microcirculation from ventricular biopsy tissue in patients with chest pain and normal coronary arteries, no structural abnormality was observed, except for mitochondrial changes ⁵⁰. Thus, it is likely that the observed microvascular changes are functional in nature due to endothelial dysfunction. This is further supported by the observation that smooth muscle function was not correlated with endothelial function in either circulation. We and others have reported previously that endothelial function, but not endothelium-independent function is impaired with increasing age, in men compared to women, and in the presence of cardiovascular risk factors in both the coronary and peripheral circulations ^{13, 29–31, 51, 52}. Finally, we have also previously shown that acetylcholine-mediated femoral microvascular responses, the technique used in this study, are abnormal in patients with cardiovascular risk factors or atherosclerosis compared to those without risk factors ³¹.

Interestingly, we also found that peripheral microvascular endothelial responses to acetylcholine correlated with the coronary flow reserve, measured as the vasodilator response to intracoronary adenosine. In a previous small study comparing patients with normal and abnormal coronary flow reserve, flow-mediated brachial arterial vasodilation was also noted to be lower in the latter group ⁵³. Another study demonstrated a correlation between the coronary flow reserve and microvascular endothelial function in the subcutaneous fat ⁴⁹. Our study is the first demonstration of the relationship between peripheral microvascular endothelial dysfunction and abnormalities in coronary flow reserve. The significant relationship between endothelial function in response to acetylcholine and coronary flow reserve in response to adenosine was perhaps due to the presence of adenosine receptors in human endothelial cells ⁵⁴. Additionally, it is known that vasodilation in response to adenosine leads to increased shear stress which triggers endothelial nitric oxide release which causes further vasodilation. Thus, the adenosine-induced vasodilation is at least partly due to nitric oxide release ⁵⁵.

The mechanisms regulating endothelium-independent smooth muscle dilatation of the coronary and peripheral microvasculature are complex and incompletely understood. Sodium nitroprusside causes vasodilation by donating nitric oxide directly to the smooth muscle. Recent study showed that vasodilation with sodium nitroprusside in the coronary and peripheral circulation were not correlated with each other, or with the acetylcholine responses, in either circulation ⁴⁹. One reason for the lack of correlation is because the effect of the NO donor on the endothelium-independent vasodilation is indeed different in the coronary compared to the peripheral circulation, while the effects of acetylcholine is similar. However, we used a single dose of sodium nitroprusside. A dose-response curve with multiple escalating doses of sodium nitroprusside could have shown a better correlation between the two vascular beds. However, the use of sodium nitroprusside infusions at higher doses is limited due to the systemic vasodilation that occurs at high doses.

Strengths and Limitations

This is the largest study to date that comprehensively examines coronary epicardial and microvascular endothelium-dependent and -independent function, and compares it to endothelium-dependent and -independent vasodilation in the peripheral circulation using the same agonists. Coronary vascular function was measured in a single unobstructed coronary artery and thus may not reflect endothelial function of obstructed coronary arteries. We also performed comparisons of single doses of these agonists, based on prior experience ²⁰. It is possible that a dose-response curve with multiple doses of each agonist would have provided better correlations than observed with a single dose. Finally, this study was not powered to examine the differences in vascular function based on racial and gender disparities. This is an area that should be further explored in future studies.

CONCLUSIONS

Herein, we demonstrate that coronary endothelium-dependent epicardial and microvascular function, and coronary flow reserve can be estimated by testing microvascular endothelial function, but not endothelium-independent function in the femoral circulation. Our findings suggest that endothelial dysfunction is a systemic disorder rather than localized one. A simple test that measures femoral vascular response to acetylcholine can provide important insights into the coronary vascular endothelial function and coronary flow reserve, tests that can assist in the diagnosis of coronary microvascular dysfunction and in prognostication. Whether strategies designed to improve peripheral endothelial dysfunction will also similarly improve coronary endothelial dysfunction need to be investigated.

HIGHLIGHTS.

• - In subjects with and without risk factors, coronary endothelium-dependent epicardial and microvascular function, and coronary flow reserve can be estimated by testing microvascular endothelial function in the femoral circulation.
• - Endothelial dysfunction is a systemic disorder rather than localized one.
• - Endothelial functional changes are generalized in the human circulation, and measures of peripheral endothelial function reflects coronary endothelial function.
• - A simple test that measures femoral vascular response to acetylcholine can provide important insights into the coronary vascular endothelial function and the flow reserve, tests that can assist in the diagnosis of coronary microvascular dysfunction and in prognostication.

Abbreviations

CAD

Coronary artery disease

BMI

body mass index

CBF

coronary blood flow

CVR

coronary vascular resistance

FVR

femoral vascular resistance