Sweat the small stuff: The human microvasculature and heart disease

Abstract

Traditionally thought of primarily as the predominant regulator of myocardial perfusion, it is becoming more accepted that the human coronary microvasculature also exerts a more direct influence on the surrounding myocardium. Coronary microvascular dysfunction (CMD) not only precedes large artery atherosclerosis, but is associated with other cardiovascular diseases such as heart failure with preserved ejection fraction and hypertrophic cardiomyopathy. It is also highly predictive of cardiovascular events in patients with or without atherosclerotic cardiovascular disease. This review focuses on this recent paradigm shift and delves into the clinical consequences of CMD. Concepts of how resistance arterioles contribute to disease will be discussed, highlighting how the microvasculature may serve as a potential target for novel therapies and interventions. Finally, both invasive and non-invasive methods with which to assess the coronary microvasculature both for diagnostic and risk stratification purposes will be reviewed.

1 |. INTRODUCTION

For decades, heart disease has been responsible for the majority of deaths in the United States. Traditionally, the term heart disease has been used interchangeably with coronary artery disease (CAD) implying that obstructive atherosclerosis within an epicardial artery is to blame for signs and symptoms of myocardial ischemia. Although “heart disease” remains the number one killer of Americans, the clinical presentation has changed. It is estimated that 50% of patients with classic complaints of dyspnea and/or angina do not have significant obstructive atherosclerotic disease.¹ Interestingly, while the incidence of ST-segment elevation myocardial infarction (STEMI) has decreased steadily for the last 20 years,² hospitalization rates for patients diagnosed with heart failure with preserved ejection fraction (HFpEF) have risen dramatically.³ Epidemiological changes including the aging population, as well as increased rates of obesity and diabetes, likely contribute to these observed findings.^4,5 As the definition of heart disease evolves, a commonality exists between the old and the new: the coronary microcirculation, small blood vessels that make up approximately 60% of the total coronary vascular resistance⁶ is both structurally, and functionally, abnormal.

Despite this recognition, many challenges associated with studying the role of microvessels in disease exist. For one, confusion remains regarding appropriate nomenclature to be used in both the clinical and laboratory setting. Coronary microvascular disease (CMD) may be considered by some as a diagnosis of exclusion in patients presenting with angina who show reduced coronary flow reserve (CFR), without evidence of obstructive atherosclerosis during cardiac catheterization.⁷ McCabe et al reported that roughly a third of all patients who presented with clinical markers of STEMI in the emergency room did not have obstructive CAD requiring percutaneous coronary intervention.⁸ This clinical scenario led the development of the term MINCA (Myocardial Infarction with Normal Coronary Arteries)⁹; however, some researchers refer to this presentation as MINOCA (Myocardial Infarction with Non-Obstructed Coronary Arteries) which includes patients with stenotic lesions of less than 50%.¹⁰ The term microvascular dysfunction may be used to describe a range of observations that include decreased myocardial perfusion in response to adenosine during cardiac catheterization¹¹ to reduced endothelium-dependent dilation in isolated human resistance vessels.¹²

Compared to obstructive conduit coronary disease, the microcirculation is invisible by traditional angiography, making assessment of the coronary microvasculature and providing a clinical diagnosis of CMD challenging and indirect in vivo. Coronary angiography utilizing catheter-based techniques to measure CFR has traditionally served as the primary approach to diagnose CMD.¹³ The invasiveness of this technique is a major limitation; however, other non-invasive diagnostic imaging modalities including cardiac magnetic resonance (CMR) and positron emission tomography (PET) have been used to interrogate coronary vasomotor function and quantify myocardial perfusion.¹⁴ Non-invasive imaging devices that utilize side-stream or incident-stream darkfield technology (eg, Microscan, CytoCam, respectively) as well as application of optical coherence tomography (OCT) may allow for assessment of the peripheral microvasculature, acting as a window to the coronary circulation.

Aside from the connection between diseases of the coronary macrocirculation (atherosclerosis) with that of the microcirculation, CMD is prevalent in other cardiac conditions including HFpEF¹⁵ and non-ischemic cardiomyopathies.^16–18 Here, we summarize what is known about CMD and cardiovascular disease. Translational approaches used to understand the link between microvascular function and the prevention or development of cardiac disease will be the central thesis of the review. The role of the coronary microcirculation in regulating myocardial perfusion as well as how in vitro analysis of human arterioles can help gain insight into these pathophysiological mechanisms will be discussed. Both invasive and non-invasive methods used to assess the coronary microcirculation, or technologies that examine systemic microvascular beds that may help predict future cardiac risk, will also be reviewed.

2 |. BEYOND PERFUSION: THE EXPANDING ROLE OF THE CORONARY MICROVASCULATURE

Conduit arteries comprise only ~7% of total coronary resistance. During normal physiological conditions, the coronary microcirculation, resistance arterioles ~50–250 μm in diameter, is responsible for the majority of vascular tone.¹⁹ The function of the arterioles is to ensure efficient delivery of nutrients while removing metabolites, a process that requires both access to essentially every cell, and the ability to rapidly respond to changes in oxygen demand with appropriate adjustments in flow to maintain myocardial oxygen supply. These critical adjustments in perfusion are made possible by both direct and indirect vascular influences. Myogenic tone, endothelial-derived mediators, autoregulation, and pharmacological agents are considered as direct influences, while indirect regulators include humoral, metabolic, and neural factors. Aside from this traditional role of tightly matching metabolic supply and demand, a new paradigm has emerged: factors derived from the microcirculation which elicit changes in tone can also evoke paracrine effects on local surrounding tissue.

2.1 |. Flow-induced dilation

While much less is known about the regulation of the coronary microvasculature in the intact human heart, a large amount of information has been extrapolated from the studying isolated microvessels in vitro. Small arterioles (<30 μm in diameter) are primarily responsible for metabolic dilation. Slightly upstream, the larger arterioles (30–60 μm) respond predominantly to changes in pressure and provide myogenic tone and serve to regulate pressure in smaller downstream vessels, thus providing protection to capillaries and post-capillary venules from acute elevations in pressure. Coronary vessels with diameters of approximately 100–150 μm are the most responsive to changes in flow.²⁰ This longitudinal gradient within the coronary microvasculature allows for rapid dilation in metabolically active myocardium while controlling the amount of total flow and pressure within the vascular bed.

The endothelial lining of most vessels, especially resistance arterioles mechanically sense an increase in flow as an elevation in shear stress along the vessel wall resulting in the generation and release of vasoactive mediators. These compounds are capable of traversing the basement membrane and eliciting relaxation of the underlying smooth muscle. Under normal physiological circumstances in adult humans, a substantial amount of dilation due to increased flow relies on endothelial-derived nitric oxide (NO) in conduit arteries (brachial artery²¹), peripheral arteries (radial artery²²), and isolated human arterioles collected from cardiac tissue (atrial²³), as well as from systemic adipose tissue.²⁴ The vast majority of studies have attributed the shear stress-induced increase in NO due to activation of endothelial nitric oxide synthase (eNOS) (also known as nitric oxide synthase 3; NOS3 or constitutive nitric oxide synthase; cNOS).

2.2 |. Plasticity in FID signaling

In addition to NO, many other compounds (endothelial-derived hyperpolarizing factors; EDHFs) have been identified that contribute to coronary microvascular tone. These include, but are not limited to, prostacyclin (PGI₂), epoxyeicosatrienoic acids, hydrogen sulfide, and hydrogen peroxide (H₂O₂). One may ask, why do so many mediators exist? Data from our group as well as others suggest that this redundancy of vasoactive compounds reflects the pathophysiological importance of endothelium-dependent dilation, allowing loss of one pathway to be compensated by another. These compensatory mechanisms are crucial in order to maintain FID and adequate perfusion even during times of stress and/or disease.

Videomicroscopy allows in vitro examination of human microvessel reactivity to explore mechanisms of FID during both health and disease. Collection of discarded surgical specimens from cardiac surgery has provided a means to study the behavior of this vascular bed during health and disease, as well as how mechanisms change throughout life. For instance, children may lack compensatory pathways as evidenced the inability of their arteriole to dilate in response to flow during exposure to the cyclooxygenase inhibitor indomethacin, whereas exposure to the NOS inhibitor N-ω-nitro-L-arginine methyl ester hydrochloride (L-NAME) had no effect on FID.²⁵ Interestingly, although NO contributes to the majority of FID in younger adults (18–55 years old), PGI₂ also participates as a mediator. Over the age of 55, FID is strictly mediated by NO. This may be a disadvantage during times of increased oxidative stress and decreased NO bioavailability. Such is the case when human arterioles from healthy adults are subjected to elevated intraluminal pressure. Following a 30-min exposure to high pressure, the mediator changes from NO to H₂O₂,²⁶ the same mediator that is responsible for eliciting dilation in human coronary arterioles from patients with known CAD.²⁴

2.3 |. FID mediators and tissue homeostasis

Even in times of disease, FID for the most part is maintained, but relies on H₂O₂ as the endothelial-derived relaxing agent. During CAD, shear stress leads to an increase in mitochondrial superoxide which then is oxidized to H₂O₂ via superoxide dismutase.²⁷ Although formation of H₂O₂ allows microvessels to maintain this important physiological dilator mechanism during disease, both NO and H₂O₂ can migrate through the vascular wall and exert different effects on surrounding parenchymal tissue. In contrast to the anti-inflammatory and anti-thrombotic effects of NO, H₂O₂ promotes vascular inflammation, thrombosis, and apoptosis of coronary endothelial cells.^28,29 It is important to note that although this transition in FID mediator to H₂O₂ is associated with the disease state, the formation of basal levels of H₂O₂ has been shown to be essential for vascular function³⁰ and is also critical for coupling cardiomyocyte metabolism with coronary vascular tone.³¹

This suggests that beyond the dilation itself, other properties of a dilator substance may be involved in promoting or preventing disease. Administration of lower amounts of NO results in a positive inotropic response in isolated cardiomyocytes³² which has been attributed to many mechanisms including stimulation of adenylyl cyclase, activation of L-type calcium current, and binding of superoxide anions to form peroxynitrite, which not only removes the harmful ROS, but also activates K_ATP channels.³³ NOS3 knockout mice are not only hypertensive, but exhibit ventricular remodeling and hypertrophy³⁴ as well as accelerated development of atherosclerosis.³⁵

More recently, it has been suggested that aquaporin-1 mediates the transfer of H₂O₂ into cardiomyocytes. Aquaporin-1-deficient mice do not develop cardiac hypertrophy and fibrosis in response to angiotensin II treatment.³⁶ However, it is important to note that low levels of H₂O₂ may be required for proper vascular function. It has been reported that H₂O₂ is formed immediately in cultured endothelial cells exposed to laminar shear stress. The study also showed that this lower physiological level (μmol/L range) of H₂O₂ activated eNOS, which was inhibitable by blocking the activity of NADPH oxidase 4 (NOX4).³⁷ It remains unclear whether it is the absolute amount, the subcellular localization or the source of H₂O₂, that determines whether the ROS promotes cellular homeostasis or disease-related vascular changes.

Human resistance vessels from adipose tissue react in a similar manner to flow as coronary arterioles.²⁴ Collection of human adipose tissue, typically more abundant than cardiac samples, has allowed us to advance our understanding of how this transition in mediator occurs, and has provided insights for restoring NO-dependent FID. One critical signaling pathway involved in the transition to H₂O₂ during disease is sphingolipid metabolism. The balance of S1P and ceramide-related sphingolipids, referred to commonly as the “sphingolipid rheostat,” is a major controller of cellular reactive oxygen species (ROS).³⁸ Increased intracellular ceramide, a sphingolipid elevated in plasma of patients with CAD and recognized as an independent risk factor for MACE, triggers the transition in FID mediator from NO to H₂O₂ in arterioles from patients with no known risk factors for cardiovascular disease.³⁹ To the contrary, exposure to the sphingolipid sphingosine-1-phosphate (S1P) restores NO as the FID mediator in vessels from patients with CAD.⁴⁰ Immunohistochemistry analysis shows that the expression of neutral ceramidase (NCdase), an enzyme that hydrolyzes ceramide to sphingosine before being phosphorylated to S1P, is decreased in human CAD arterioles. Inhibition of ceramide formation via neutral sphingomyelinase (NSmase) in diseased vessels also restores NO-dependent FID³⁹ while inhibition of NSmase in arterioles from healthy individuals promotes FID mediated by H₂O₂. (unpublished observations). Together these data suggest that the ability to properly metabolize sphingolipids may prove more useful than simply eliminating ceramide or balancing the effect of ceramide with S1P.

The transition in FID mediator at times may not be definite, as we have observed scenarios that result in both NO and H₂O₂ contributing to dilation, or the “dual mediator phenomenon.” Using a lentiviral approach, decreasing expression of peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α), an important regulator of mitochondrial ROS, promotes H₂O₂-dependent FID in healthy arterioles, yet when PGC-1α expression is increased in diseased vessels using α-lipoic acid or ZLN005, dilation was only partly decreased in the presence of either PEG-catalase or L-NAME. It was only when both inhibitors were used together that dilation to flow was abolished.⁴¹ We have observed a similar phenotype when microvessels from healthy patients are pre-treated with S1P prior to ceramide exposure, as FID is only decreased during simultaneous treatment with both inhibitors (unpublished observations). It remains unclear whether there is dual release of these two mediators during flow, or whether one pathway compensates for loss of the other. It has been postulated that the two mediators can act synergistically, as H₂O₂ is capable of activating eNOS.⁴² The source of H₂O₂ may also prove critical in this scenario in that mitochondria ROS tend to exert pro-inflammatory actions on parenchymal tissue⁴³ whereas H₂O₂ derived from NADPH oxidase 4 (NOX4) is anti-inflammatory.⁴⁴

Albeit complex, regulation of coronary microvascular tone is a highly orchestrated process that heavily involves the generation and release of endothelial-derived vasoactive compounds. The mechanism by which resistance arteries dilate to flow changes throughout one’s life may rely on a sole mediator or the contribution of several, allows for compensatory pathways to emerge during times of stress or disease, and has a significant impact on surrounding parenchyma, which may contribute to the acquisition or prevention of heart disease.

3 |. THE LITTLE THINGS MATTER: THE LINK BETWEEN MICROVASCULAR DYSFUNCTION AND CARDIOVASCULAR DISEASE

Microvascular dysfunction precedes large artery atherosclerosis. Rubinshtein et al showed in patients without obstructive CAD that the Framingham risk score and CAD risk factors (eg, diabetes mellitus, hyperlipidemia, hypertension, congestive heart disease, smoking history) correlate with microvascular function measured by invasive CFR.⁴⁵ The WISE (Women Ischemia Syndrome Evaluation) trial concluded that women with evidence of microvascular dysfunction (reduced CFR), despite a lack of obstructive coronary disease, was predictive of future cardiac events.⁴⁶ With the development of CAD, the vascular endothelium also becomes damaged as evidenced by a decrease in endothelial-dependent dilation to pharmacological agents (eg, acetylcholine). Microvascular endothelial dysfunction, or an impaired response to the endothelium-dependent vasodilator acetylcholine, has also been shown to be a risk factor for MACE independent of CAD.⁴⁷ The role of the endothelium was further highlighted by Reriani et al who reported that human microvascular coronary endothelial function, reported as percent-change of coronary blood flow in response to acetylcholine, adds power for risk stratification in patients without obstructive disease.⁴⁸

However, a new paradigm has emerged suggesting that the coronary microvasculature also contributes to an array of other cardiovascular diseases including cardiomyopathies (hypertrophic,¹⁶ restrictive,⁴⁹ and stress-related⁵⁰) and heart failure (HFpEF).^51–55 The PRevalence Of Microvascular dySfunction in Heart Failure with Preserved Ejection Fraction (PROMIS-HFpEF), a multicenter prospective study, found that 75% of HFpEF patients had evidence of CMD.⁵³ The concept that CMD contributes to HFpEF was first postulated by Paulus and Tschope, suggesting that the damaged microvascular endothelium promotes inflammation within the underlying myocardium.⁵⁶ They propose that co-morbidities common to HFpEF (eg, diabetes mellitus, obesity, hypertension) promote an inflammatory state, induce microcirculatory endothelial dysfunction, and trigger a cascade that ultimately causes fibrosis and hypertrophy. This hypothesis is strengthened by data showing that cardiomyocytes collected via myocardial biopsy from HFpEF patients contain more collagen and oxidative stress markers.⁵⁷ This has also been shown in animal models where pigs with multiple co-morbidities (eg, hyperlipidemia, hypertension, hyperglycemia) have increased myocardial ROS production, increased NOX activity, dysfunctional eNOS, reduced NO levels, and impaired endothelium-dependent vasodilation in the small coronary arteries. These changes were associated with increased cardiomyocyte and LV end-diastolic stiffness, despite a normal ejection fraction.⁵⁸ More recently, it was reported in mice that a high-fat diet in combination with the NOS inhibitor L-NAME reduced expression of myocardial X-box binding protein 1 (XBP1), also known to be decreased in the myocardium from humans with HFpEF.⁵⁹

Cardiomyopathies are clinically divided into ischemic and non-ischemic types, though impaired microcirculatory function has been identified in both.⁶⁰ In fact, nearly 82% of patients diagnosed with either type of cardiomyopathy have an abnormal CFR and is associated with an increased risk of MACE.⁶⁰ Arterioles from patients with hypertrophic cardiomyopathy (HCM) exhibit both medial hypertrophy and intimal hyperplasia, creating a thick vascular wall with a reduced luminal area.⁶¹ The anatomical changes within the microvessels may account for the myocardial ischemia and arrhythmias commonly associated with HCM.⁶² Cardiac PET imaging studies have shown that CMD occurs both in the hypertrophied septum, as well as the free wall of the ventricle, with substantial microvascular disease in the left ventricle subendocardium.⁶³

As we shift our thinking to this novel concept that the microvasculature may be the nidus for both ischemic and non-ischemic heart disease (Figure 1), efforts need to be focused on understanding how such small vessels can impact myocardial function.

FIGURE 1.

The microcirculation as a driver of disease. Coronary endothelial dysfunction, here defined as the release of H₂O₂ in response to an increase in blood flow, precedes the development of CAD, and may contribute to other cardiac pathologies including HFpEF and HCM. CAD, coronary artery disease; FID, flow-induced dilation; HFpEF, heart failure with preserved ejection fraction; HCM, hypertrophic cardiomyopathy

4 |. INSIDE-OUT AND OUTSIDE-IN: ASSESSMENT OF HUMAN CORONARY MICROVASCULAR FUNCTION

Clinical signs and symptoms of CMD parallel those observed with obstructive ischemic heart disease including exertional dyspnea and angina. Although patients may present with heart failure symptoms (eg, venous distention, pulmonary rales, pedal edema), the physical examination and electrocardiogram (ECG) are typically normal. Unfortunately, exercise stress testing, with or without stress imaging (eg, stress echocardiography or nuclear scintigraphy) is neither sensitive nor specific for diagnosing CMD.⁵ Approximately 50% of patients with CMD experience angina that is identical in nature to those who suffer from obstructive CAD.⁶⁴ Both invasive and noninvasive techniques exist that allow for interrogation of in vivo coronary microvascular function; however, significant limitations exist. Ideally, a non-expensive, non-invasive approach that does not require significant training would be ideal to study the regulation of human microvascular function in vivo and allow for early diagnosis of CMD.

4.1 |. Invasive microvascular assessment

Coronary angiography, the gold standard for diagnosing obstructive CAD, can be used to diagnose CMD by both excluding obstructive atherosclerosis as a diagnosis, and by allowing for the use of catheter-based techniques to examine microvascular coronary physiology. Use of Doppler-tipped guidewires allows for measurement of blood flow velocity at baseline and after administration of adenosine or acetylcholine (endothelial-independent and endothelial-dependent, respectively). CFR is a dimensionless index calculated by the ratio of hyperemic to baseline coronary blood flow or flow velocity. Abnormal CFR (CFR < 2) is indicative of CMD.⁶⁵ Clinicians can also assess coronary flow following reperfusion therapy by measuring the wash-in speed of a radiopaque dye (TIMI flow grade), or the number of angiographic frames needed for the dye to traverse the coronary artery (TIMI frame count).⁶⁶ Fractional flow reserve (FFR) and instantaneous wave-free ratio (iFR) are indirect measures of microvascular resistance calculated before and after restoring flow through a stenotic artery. The Index of Microcirculatory Resistance (IMR) however is considered a more direct measurement of microcirculatory resistance independent of epicardial stenoses,⁶⁷ is highly reproducible,^68,69 and serves as an independent predictor of adverse outcomes such as microvascular obstructions⁷⁰ and death.⁷¹ While these methods may prove useful in the diagnosis of CMD, they suffer from the same limitations as other invasive measures, most notably that these are all indirect measures of microvascular function.

4.2 |. Non-invasive microvascular assessment

4.2.1 |. Myocardial imaging

Positron emission tomography is arguably the gold standard of non-invasive measurement of microvascular blood flow.⁷² Both validated and reproducible, quantitative microvascular flow measurements are now performed during routine PET stress testing.¹⁴ Global myocardial blood flow is measured using a blood flow radiotracer following administration of a vasodilator. While PET offers the advantage of measuring absolute myocardial blood flow, the poor spatial resolution does not allow for quantification of regional blood flow to a particular myocardial region. CFR values derived from PET can identify subjects at high risk for MACE and cardiac death.⁷³

Cardiac magnetic resonance uses a gadolinium-based contrast agent to quantify myocardial perfusion. The protocol is similar to that of PET, measuring flow at both rest and during stress. A major advantage of CMR compared to PET is the lack of ionizing radiation required for measurement of myocardial perfusion. Liu et al recently showed that CMR using T₁ mapping as opposed to gadolinium can be used to diagnose myocardial ischemia in those with and without obstructive CAD.⁷⁴ The use of CMR has allowed for a decreased rate of unnecessary coronary angiographies in patients with angina suggestive of CAD.⁷⁵ Major drawbacks to both PET and CMR for diagnosis of CMD include the cost associated with acquiring these images and the inability to directly visualize the coronary microvessels.

Myocardial contrast echocardiography (MCE) assessment of microvascular function uses high-energy ultrasound to burst injected microbubbles and measures the rate of replenishment as an index of perfusion. The echo contrast used is safer than radiotracers and measurement of coronary microvascular flow has been verified against both PET⁷⁶ and CFR.⁷⁷ The use of echo contrast as opposed to radiotracers, and the relatively low cost of this technique, makes it a desirable tool to assess for CMD. However, an experienced operator is required.

Other modalities exist to non-invasively quantify coronary blood flow including but not limited to doppler echocardiography and dynamic myocardial perfusion computed tomography (CT); however, studies have yet to validate these approaches for the assessment of coronary microvascular flow.

4.2.2 |. Assessment of the peripheral microvasculature

Although not used clinically, the systemic microvasculature, namely the skin, retina, limb microvasculature, and mucosal tissues (eg, sublingual, buccal, rectal) can be visualized using a variety of techniques. Although some animal studies have shown that co-morbidities associated with cardiovascular disease result in a distinct remodeling pattern specific to the coronary microvasculature,⁷⁸ our work using isolated human arterioles suggests that peripheral microvessels from subcutaneous, visceral, and peritoneal fat depots have a similar response to shear as those collected from right atrial appendages. The case for assessing peripheral microvessels as a surrogate to coronary arterioles is strengthened by studies showing that decreased vessel density in the sublingual microvasculature is predictive of complications in patients presenting with acute myocardial infarction.⁷⁹ No only do structural and functional changes correlate with atherosclerotic risk,⁸⁰ but decline in retinal artery function is much more pronounced in patients with CAD and reduced cardiac function, compared to those with CAD alone, suggesting that a continuum of microvascular damage exists.⁸¹

Due to the technical challenges of interrogating the human coronary microvasculature, use of the peripheral microvasculature as a window into coronary microvascular function may improve risk stratification. As early as the 1990s, videomicroscopy has been used to examine the human peripheral microvasculature, predominantly the sublingual and/or buccal vascular beds. Using orthogonal polarization spectral imaging (OPS), De Backer and colleagues pioneered the process of performing human in vivo imaging studies on critically ill patients.^82–84 OPS technology has since been replaced with side-stream dark field (SDF) and most recently, incident dark field (IDF) imaging, both of which use visible green light at wavelengths that correspond to the isosbestic point for hemoglobin, allowing for visualization of red blood cells. This technique allows to measure flow in a given area (microvascular flow index) or measure the total and perfused vessel density, respectively, either through a manual or automated analysis. Multiple lightweight, handheld devices have been created for this purpose. The Microscan (MicroVision Medical, Amsterdam, the Netherlands) utilizes SDF technology and records analog video, requiring the images to be digitalized prior to analysis. The Capiscope HVCS and HVCS-HR (KK Technology) are also handheld SDF cameras, but allow for digital recording. These devices are also compatible with GlycoCheck, a software system used to analyze the perfused boundary region of blood vessels. The CytoCam (Braedius Medical BV) is the newest generation of hand-held IDF devices. It records digital video at a higher resolution and increases the field of view, increasing the number of vessels visualized by twofold.⁸⁵ A limited number of studies using SDF devices have shown that a thin glycocalyx, and thus an increase in the perfused boundary region in the peripheral microcirculation, correlates with the presence of ischemic heart disease.^86,87 While the use of SDF/IDF technology to visualize the peripheral microvasculature is still in its infancy, these devices hold great promise as a means to detect changes in the microcirculation and allow for early diagnosis of cardiovascular disease.

Optical coherence tomography is another non-invasive imaging modality that operates similar to ultrasound. Briefly, by quantifying the time delay in reflected light waves, a 3D image can be generated by applying a lateral beam of light along the sample surface. The use of OCT, more specifically OCT angiography (OCT-A), has predominantly been used to effectively diagnose disorders of the human eye⁸⁸ but its use for cardiovascular disease is an active area of investigation. If combined with intravascular ultrasound, OCT can be used to examine arterial plaques in larger conduit arteries.⁸⁹ Using optical microangiography (OMAG), Wang and colleagues were the first to visualize the intact coronary microvasculature from an explanted rat heart.⁹⁰ Although OMAG has been used to visualize the density of the coronary vasa vasorum during cardiac catheterization, which negatively correlates with endothelium-dependent microvascular function, this technology has primarily been used to link CAD to changes in the peripheral microcirculation within the eye.⁹¹ The high resolution of OMAG is offset by the lack of penetration through multiple layers of biological tissue, limiting examination to superficial microvascular beds (eg, eye) compared to other areas such as the forearm. The high cost associated with OCT/OMAG also currently limits its use as a means to assess the microvasculature.

Laser Doppler has been used extensively to study cutaneous microvascular beds. Laser Doppler flowmetry (LDF), laser Doppler perfusion imaging (LDPI) and laser speckle contrast imaging (LCSI) all optically measure relative flow in a small and shallow region of the skin. Commonly used in tandem with iontophoresis and intradermal microdialysis, these techniques can be used to study microvascular reactivity. Laser Doppler is both non-invasive and safe; however, only changes in absolute flow with a confined region are measured as opposed to examining changes in the microvasculature directly through visualization. Microvascular dilation in response to Ach, local heating and cuff occlusion are all diminished in people with known CAD.³⁷ Using LDF, Sheikh et al showed in patients presenting with angina that epicardial artery endothelial dysfunction positively correlated with decreased dermal microvascular reactivity to Ach, suggesting that the changes in the peripheral microcirculation may offer clues regarding the health of our coronary system.⁹² A summary of the methods and techniques in which to assess the microvasculature are presented in Table 1.

TABLE 1.

Comparison of both invasive and non-invasive techniques to clinically assess the coronary microvasculature

	Method	Pros	Cons
Invasive	CFR	Commonly used, relatively inexpensive	Does not distinguish micro- from macrocirculation, high variability, questionable during ACS
	Doppler wire	Highly reproducible	Possible issues with calibration, signal loss, and potential difficulties in placement
	IMR	Dedicated to microcirculation, low variability, possible during PCI	Debated corrections in the presence of stenosis
Non - invasive	PET	Highly sensitive, accurate, reproducible, quantifiable	Extremely expensive, needs CT for structural assessment
	SPECT	More prevalent and with up to 10 times more studies performed than PET	Higher tracer radiation exposure and worse spatial and temporal resolution than PET
	CMR	High resolution, no radiation, clinical use increasing	Very expensive, image artifacts hinder absolute quantification of flow, not feasible with some implants
	CTP	High spatial resolution, provides anatomical and functional measurements, cost effective	High radiation exposure, requires slow heart rate
	TTDE	Cost effective, highly feasible	Requires an expert operator, modest agreement with PET
	MCE	Clinically studied, correlates well with PET, cost effective, widely available	Complex, requires an expert operator, less feasible in obesity and some lung diseases

Abbreviations: ACS, acute coronary syndrome; CFR, coronary flow reserve; CMR, cardiac magnetic resonance; CT, computed tomography; CTP, computed tomography myocardial perfusion; IMR, index of microcirculatory resistance; MCE, myocardial contrast echocardiography; PCI, percutaneous coronary intervention; PET, positron emission tomography; SPECT, single photon emission computed tomography; TTDE, transthoracic Doppler echocardiography.

5 |. SUMMARY AND CONCLUSIONS

It is becoming increasingly clear that in addition to its role as a critical modulator of myocardial perfusion, the coronary microvasculature also acts in a paracrine fashion, exerting influence on surrounding parenchymal tissue. A vast array of mediators generated from the microvascular endothelium can limit or intensify the degree of inflammation within the myocardium. Here, we have reviewed concepts regarding the role of CMD in heart disease, including current and up-coming methods in which to interrogate the microvasculature for diagnostic and research purposes. As data continue to emerge suggesting that coronary resistance vessels contribute to the pathophysiology of atherosclerosis, HFpEF, and HCM, efforts need to focus on mechanisms by which changes in endothelial-derived mediators occur, an achievable goal through use of isolated arterioles from patients who suffer from these disorders. Thorough understanding of these mechanisms will allow for the development of potential therapeutics that aim to improve coronary perfusion, limit ischemia-induced structural changes, and prevent large artery atherosclerosis from developing. The ability to intervene early when endothelial dysfunction is occurring, prior to the formation of obstructive coronary disease or HFpEF, may drastically alter an individual’s disease trajectory and outcome. Improvements in the ability to non-invasively analyze the microvasculature, to determine not only that one is dilating but what specific mediators are eliciting the dilation, should be an area of future investigation. After all, the devil is in the details.