Coronary Microvascular Dysfunction in Ischaemic Heart Disease: Lessons From Large Animal Models

Authors: Oana Sorop, J. van de Wouw, Daphne Merkus, Dirk J. Duncker

Abstract

The coronary microvasculature is principally responsible for matching coronary blood flow to myocardial demand of oxygen and nutrients. Short‐term control of coronary blood flow is achieved via alterations in coronary microvascular tone, whereas long‐term control of coronary flow also involves remodelling of the coronary microvasculature, including adjustments in vascular structure, diameter and density. In the past 50 years, considerable research efforts have been directed at understanding the functional and structural coronary microvascular adaptations involved in matching myocardial oxygen supply to demand, and how these mechanisms are affected by various diseases. In this review article, we will discuss our current understanding of the mechanisms underlying the regulation of coronary microvascular tone under healthy physiological conditions and in ischaemic heart disease. We will specifically discuss the role of microvascular dysfunction in obstructive and non‐obstructive coronary artery disease, as studied in large animal models and confirmed in human studies. Future research should be directed at further unravelling the disease‐specific mechanisms of coronary microvascular dysfunction in order to identify therapeutic targets to improve microvascular function in patients with ischaemic heart disease.

Introduction

Adequate myocardial perfusion is critical for maintaining myocardial oxygen and energy balance in the heart, particularly during times of increased activity such as exercise. Because resting levels of myocardial oxygen extraction are already very high (65%–80%), any significant increase in myocardial oxygen demand must principally be met by a commensurate increase in coronary blood flow and oxygen delivery [1, 2, 3, 4].

The regulation of myocardial perfusion is a complex process involving a myriad of acute and chronic coronary microvascular adaptive mechanisms, which include the regulation of vascular smooth muscle tone for immediate adjustments and structural changes in microvascular diameter and density for more sustained responses. Understanding the basic principles and how these mechanisms are altered in various disease states has been the focus of numerous studies published in the past 50 years. Consequently, our understanding of coronary microvascular function in health and disease has significantly advanced, with important contributions from studies in large animal models, initially dogs and, in more recent years, swine, of which the cardiovascular anatomy and physiology closely resembles that of humans. Here, we will discuss the factors regulating coronary microvascular tone and structure under normal physiological conditions, as well as the impact of microvascular dysfunction in both obstructive and non‐obstructive coronary artery disease (CAD) as studied in large animal models and validated through human studies.

Physiological Control of Coronary Microvascular Tone

Because left ventricular myocardial oxygen extraction is already well above 65% at rest, any increase in oxygen demand must principally be met by an increase in myocardial oxygen delivery and hence in coronary blood flow. This is predominantly achieved by a reduction in basal vascular tone, leading to vasodilation and a reduction in vascular resistance. At rest, approximately 75% of total coronary resistance resides in the coronary small arteries and arterioles with a diameter of less than 200 μm, while only about 7% of the resistance resides in the veins [5]. Coronary vascular resistance is determined by various factors acting in an integrated fashion to maintain an optimal level of vascular tone. Such factors include (i) passive mechanical factors, e.g., vascular compression by the contracting myocardium in systole and the intraventricular pressures in diastole and distension due to changes in intravascular pressure, (ii) active adjustments in smooth muscle tone to mechanical stimuli, e.g., myogenic responses to the intravascular pressure and flow‐mediated dilation in response to changes in shear stress, as well as (iii) metabolic‐, (iv) neurohumoral‐, (v) endocrine‐ and (vi) endothelial factors. These factors exert diverse influences on different segments of the microvasculature [6, 7]. During increased metabolic demand, the smallest arterioles (< 100 μm diameter) appear to be most sensitive to metabolites; myogenic responses dominate the tone of more upstream arterioles (~100 μm), while larger arterioles and small arteries (100–400 μm) are most sensitive to flow‐mediated dilation. Figure 1 summarizes the most important factors influencing coronary microvascular tone. Circulating and endothelium‐derived factors as well as metabolic and neurohumoral influences act in concert to regulate coronary microvascular tone.

Endothelial Factors

Vascular endothelium releases a variety of vasoactive substances with potent effects on vascular tone. These include vasodilators such as nitric oxide (NO), cyclooxygenase (COX)‐derived vasodilators such as prostaglandins and endothelium‐derived hyperpolarizing factors (EDHFs) and vasoconstrictors such as endothelin‐1 (ET‐1) or COX‐derived constrictors.

NO, produced by endothelial NO synthase (eNOS) in response to a variety of agonists as well as shear stress, is released both luminally and abluminally. On the luminal side, NO can be taken up, as well as produced, stored, transported and released by erythrocytes [8]. When oxygen concentrations are low, nitrites (the predominant form of stored NO in RBCs) are converted back to NO and released, while RBC‐released ATP can also stimulate the release of NO by endothelial cells [9]. On the abluminal side, NO diffuses to the vascular smooth muscle cells (VSMCs), where it binds to soluble guanylyl cyclase, increasing cyclic GMP (cGMP) production and causing relaxation through the opening of KV channels and the subsequent reduction in intracellular calcium. Small arteries have been shown to exhibit greater responsiveness to NO as compared with arterioles [10]. Prostacyclin (PGI2) is the major active metabolite of arachidonic acid in the vascular endothelium and a potent vasodilator of coronary arteries via an increase in cAMP and the opening of KATP channels in the coronary smooth muscle cells. Inhibition of PG synthesis was shown to have little effect on the coronary blood flow regulation under normal healthy conditions in large animal models [11, 12], while findings in humans were unequivocal [13, 14, 15].

The third most important endothelium‐dependent vasodilator mechanism is endothelium‐dependent hyperpolarization (EDH). EDH factors (EDHFs) such as endothelium‐derived hydrogen peroxide (H2O2) derived from physiological redox signalling or in response to pulsatile flow, especially in the sub‐endocardial microvasculature [16] and cytochrome P‐450 epoxygenase metabolites EETs [17], are proposed to be released in response to certain agonists or to shear stress and act via hyperpolarization of vascular smooth muscle and opening of the calcium‐activated potassium channels (KCa). PGI2 and EDHFs are considered to be of minimal importance under healthy conditions but assume greater importance under conditions when NO bioavailability is reduced [18].

Conversely, ET‐1 is an extremely potent vasoconstrictor synthesized by cleavage via endothelin‐converting enzyme (ECE) of the precursor molecule (big–endothelin) and is the most abundant isoform of the endothelin family in the coronary vasculature [19]. ET‐1 results in vasoconstriction by binding to ETA and ETB receptors on coronary VSMCs, whereas binding to the ETB receptor on endothelial cells results in vasodilation mediated via NO production [19]. ET‐1‐induced vasoconstriction appears to be especially involved in microvascular angina associated with coronary microvascular dysfunction, as ET‐1 receptor blockade converted the coronary vasoconstrictor response to the cold‐pressor test to vasodilation [20]. Moreover, both ET‐1 production and vascular reactivity have been shown to be increased in patients with angina with no obstructive coronary arteries (INOCA) [21, 22]. Additionally, COX‐derived metabolites prostaglandin F2α and thromboxane A2 have been shown to cause coronary vasoconstriction [23], mainly during ischemia–reperfusion [24] or in the presence of endothelial injury causing coronary vasospasm [7, 23].

Importantly, all these mechanisms interact and influence each other. For example, NO is well known to inhibit the synthesis of ET‐1, thereby reducing its vasoconstrictor influence and maintaining the predominance of vasodilation in conditions where enhanced blood flow is required. Conversely, ET‐1 can both enhance NO production via the endothelial ETB receptor and inhibit NO bioavailability by promoting oxidative stress [25].

Neurohumoral Influences

In resting conditions, the sympathetic nervous system exerts minimal vasomotor influences on the coronary circulation in the healthy heart, while the parasympathetic nervous system exerts an influence mainly via lowering of heart rate. During exercise, the contribution of the autonomic nervous system to coronary resistance vessel tone increases. Studies in dogs, swine and humans have demonstrated that sympathetic activity contributes to exercise hyperaemia through beta‐adrenoceptor–mediated coronary vasodilation, which appears to outweigh the alpha‐adrenoceptor–mediated coronary vasoconstriction [3, 26]. The physiological role of vagal nerve control of coronary tone is uncertain. Coronary resistance arteries of humans and dogs with healthy endothelium are known to dilate to acetylcholine [18], while endothelial dysfunction results in an attenuation of acetylcholine‐induced dilation, even resulting in constriction [27]. Furthermore, there is an interaction between the sympathetic and parasympathetic regulation of coronary vascular tone, as data in swine indicate that during exercise, beta‐adrenergic vasodilation was supported by withdrawal of muscarinic receptor‐mediated inhibition of beta‐adrenergic coronary vasodilation [3]. Angiotensin (Ang) II has also been shown to exert vasomotor activity, depending on the concentration, in porcine coronary arterioles. At low concentration (1 nmol/L), Ang II causes activation of AT1 receptors, resulting in vasoconstriction, while higher concentrations (> 1 nmol/L) of Ang II induce vasodilation by activating AT2 receptors. Interestingly, a subvasomotor concentration of Ang II elicits superoxide production and inhibits NO production and endothelium‐dependent NO‐mediated dilation in response to adenosine [28].

Myocardial Metabolism

Metabolic vasodilation is probably the most important physiological mechanism through which blood flow to the myocardium is adjusted to meet the metabolic demands of the heart. This process ensures that the myocardium receives an adequate supply of oxygen and nutrients to match its energy consumption, particularly during periods of increased workload, such as during exercise or stress. The heart depends entirely on aerobic metabolism for its production of ATP. It is generally agreed that during periods of increased cardiac work, various metabolites produced by the myocardium play an important role in the regulation of coronary microvascular resistance. However, the exact nature of these factors and their mechanisms of action remain incompletely understood [4]. Among the mechanisms originally proposed were dissolved oxygen, CO2, H^+^, K^+^ and adenosine, but later studies failed to support a critical role for these molecules under normal physiological conditions [3, 4], although a role of adenosine in vascular tone regulation during myocardial ischemia has been demonstrated [3]. More recently, adenine nucleotides, released from erythrocytes at low oxygen levels (ATP), and H2O2 produced by mitochondria and acting via KV channels, have been proposed as key mediators coupling the coronary flow to metabolism in the heart [2, 29].

Intraluminal Forces: Myogenic Response and Flow‐Mediated Dilation

The myogenic response is triggered by the stretch of VSMCs in response to changes in intraluminal pressure, with vessels constricting in response to an increase in transmural distending pressure and dilating in response to a decrease in pressure. Its role is to maintain a fairly constant blood flow over a wide range of pressures, thereby contributing to coronary autoregulation. Species differences in the mechanisms modulating myogenic tone have been found in ex vivo thus, while in human and porcine small arteries myogenic tone has been shown to be endothelium‐independent and particularly important in coronary arterioles of ~100 μm in diameter [10, 16, 30, 31], in intramural rat coronary arteries, NO and prostaglandins modulate the myogenic tone [32]. Moreover, transmural differences in the magnitude of myogenic tone exist in the coronary circulation, with subendocardial arterioles exhibiting less myogenic response as compared with subepicardial arterioles, possibly associated with the lower transmural pressure in the subendocardial arterioles [30]. Mechanosensing of pressure may involve integrins that link the extracellular matrix to the cytoskeleton [33], while signalling mechanisms include stretch‐activated channels and voltage‐operated Ca^2+^ channels, resulting in an inward Na^+^ or Ca^2+^ current and activation of protein kinase C and/or Rho‐A kinase [34].

The second important mechanical factor acting on both conduit and resistance vessels is shear stress, the frictional force that the flowing blood exerts on the endothelium, that responds by releasing vasodilators in response to an increase in flow (i.e., laminar shear). Flow‐mediated dilation is one of the most important factors protecting vascular integrity and preventing endothelial damage, with plaque formation occurring at sites with turbulent, oscillatory or low flow. The mechanisms underlying the mechanosensing of shear by endothelial cells are still incompletely understood but are thought to involve multiple cellular components, such as the endothelial glycocalyx, integrins, cell–cell adhesions, caveolae, nuclear conformation changes and cellular‐ECM adhesion sites, as well as mechanosensitive ion channels, transient receptor potential (TRP) channels and G‐protein coupled receptors [35, 36, 37, 38]. The magnitude of endothelium‐dependent dilation and the underlying mechanism depends strongly on the level of flow, species, vascular bed, vessel size and age. For example, in normal conditions, while the conduit vessels rely primarily on NO, in the resistance vessels mediators such as NO, prostacyclin and EDHF have been shown to be involved. During exercise, when the levels of wall shear stress are increased, flow‐induced vasodilation was shown to be mediated mainly via endothelium‐derived NO and prostaglandins [39]. Moreover, in humans, factors mediating arteriolar flow‐induced dilation in the coronary circulation vary throughout life. Thus, prostacyclin is the principal vasodilator in young subjects; NO becomes dominant in adulthood, while H2O2 becomes important later in life or with the onset of CAD [40]. Overall, flow‐mediated dilation diminishes with age, as demonstrated in both humans [41, 42] and animal models [43].

Vascular Structure

Coronary vascular resistance as well as vasoconstrictor and vasodilator reserves are closely related to the biomechanical properties of the vasculature and the hemodynamic factors acting on the blood vessels. The process of vascular remodelling, for example, adaptations to regular aerobic exercise, is associated with increased vascular compliance, decreased microvascular elastic modulus and increased arteriolar and capillary density [39]. However, the internal vascular diameter and its adjustments to various influences are constrained by the thickness and composition of vascular layers, particularly the organization of VSMCs and the extracellular matrix. Chronic increases in active tone due to either myogenic autoregulation or biochemical factors induce inward remodelling of the vessel, either hypertrophic or eutrophic, hampering the vasodilator capacity of the particular vascular segment [44]. The mechanisms of the different types of coronary vascular remodelling have been studied and reviewed in detail elsewhere (Figure 2), and the reader is referred to these excellent articles [23, 38, 44].

Sex‐Differences in the Mechanisms of Coronary Blood Flow Regulation

Studies of sex differences in the mechanisms of coronary blood flow regulation often indicate different mechanisms controlling endothelial function in men and women. Such differences include sex hormone–dependent NO release in women and oxidative stress and impaired NO bioavailability in men. Similarly, differences exist in the role of renin angiotensin aldosterone system (RAAS) and ET‐1 signalling between pre‐menopausal women and men, women having decreased ET‐1‐mediated vasoconstriction and levels of circulating ET‐1 and lower RAAS activation [45]. For an in‐depth review of sex differences, the reader refers to several articles [46, 47].

Coronary Microvascular Dysfunction in Chronic Ischaemic Heart Disease

Ischaemic heart disease has long been viewed as a “large vessel” disease caused by obstructions in the epicardial coronary arteries. These obstructions usually result from atherosclerotic plaque buildup in a large coronary artery, which, when sufficiently severe, limits blood flow and oxygen delivery to the heart muscle. However, in recent years, the role of coronary microcirculation in ischaemic heart disease has gained significant attention, as nearly half of the patients undergoing coronary angiography for typical angina complaints or who have a positive stress test do not have an obstructive coronary artery lesion [48]. This form of ischaemia can be more challenging to diagnose because traditional coronary angiography and also fractional flow reserve (FFR) do not detect microvascular abnormalities.

In the healthy heart, microvascular dilation enables coronary blood flow to increase up to fivefold in response to elevated metabolic demands [49]. However, when atherosclerosis development leads to a stenosis in an epicardial artery, total coronary vascular resistance rises, and maximal coronary blood flow is consequently reduced [50]. This was thought to be mainly due to the stenosis‐associated reduction in distal perfusion pressure of the microvasculature [49]. However, there is evidence in patients undergoing catheterization for suspected CAD that alterations in the control of coronary microvascular tone and structure may also significantly contribute to the reductions in coronary flow reserve (CFR) [50]. This was shown in patients presenting with a wide spectrum of coronary obstructive disease (see Figure 3), where many patients with a hemodynamically significant coronary artery stenosis with FFR less than 0.8 (FFR < 0.8) had a CFR < 2.0 which was lower than expected based on the severity of the stenosis, suggesting the additional presence of diffuse microvascular disease distal to the stenosis (red plane in Figure 3). Importantly, the majority of patients had an FFR > 0.8 and a CFR < 3.0 (Micro, light blue plane), suggestive of diffuse microvascular disease in the absence of a flow‐limiting epicardial stenosis [50]. MVD distal to the stenosis but also in the absence of a proximal coronary obstruction will be discussed below.

Obstructive CAD

When the chronic coronary artery obstruction exceeds 75% reduction in vascular cross‐sectional area, the increased resistance results in a drop in perfusion pressure of the distal microvascular bed, leading to both structural and functional alterations [49]. It has been shown in swine that the low perfusion pressure of the distal (subendocardial) microcirculation may result in either inward hypertrophic or eutrophic remodelling of the resistance arteries (Figures 2 and 4) [51, 54, 55]. In addition, changes have been observed in the small artery (200–400 μm diameter) and/or capillary density in both subepicardium and subendocardium (Figures 4 and 5) [56]. This contrasts with the observations of Weil et al. where they observed an increase in arteriolar and capillary densities [51].

These alterations have been shown to be time‐dependent and can be influenced by both the presence/removal of a severe stenosis and pharmacological treatment of the patients. Indeed, in vitro, isolated arterioles cultured at low intraluminal pressure as seen distal to a severe stenosis showed eutrophic inward remodelling that could be prevented by incubation with the calcium antagonist amlodipine [57]. These observations were consistent with the findings of Verhoeff et al. [58] in patients with a coronary artery stenosis showing maintained or even slightly reduced minimal microvascular resistance upon revascularization. Interestingly, such findings were also consistent with in vivo studies in swine by Weil et al. [51], demonstrating microvascular remodelling (increased wall thickness‐to‐lumen ratio and decreased lumen area) in subendocardial arterioles distal to a chronic epicardial stenosis, that was compensated by increased subendocardial arteriolar and capillary densities. Strikingly, however, 1 month after revascularization, the inward microvascular remodelling persisted, while the compensatory increase in arteriolar and capillary density was no longer present, resulting in a reduction in subendocardial flow reserve [51]. Although capillaries are generally not considered to be an important site of coronary vascular resistance, their density in chronically ischemic myocardium may be a predictor of myocardial functional recovery following coronary revascularization [59]. Thus, myocardial biopsies in patients with a chronic coronary artery stenosis that underwent revascularization showed that segments with low capillary density displayed poor functional recovery, while segments with high capillary density showed good functional recovery [59]. Interestingly, areas with normal capillary density showed a similar functional recovery compared with low capillary density, which could be due to the fact that not only vascular integrity/density but also blood flow perturbations determine functional recovery after revascularization. It is possible that following revascularization persistent inward arteriolar remodelling in conjunction with the loss of arteriolar density can lead to sustained perturbations in flow responses, thereby contributing to delayed and incomplete myocardial functional recovery.

Arteriolar vascular remodelling is determined not only by the changes in haemodynamics distal to the stenosis but also by alterations in microvascular (endothelial) function. Increased circulating ET‐1 levels and an increased vasoconstrictor response to ET‐1 distal to a chronic stenosis of the left descending coronary artery, but not in the remote area, were observed in swine (Figure 5) [54]. Interestingly, the vasodilator response to bradykinin was preserved in subendocardial arterioles, although the contribution of NO to the bradykinin response was impaired, suggestive of preserved endothelial responsiveness, but with a shift in mediators from NO to prostanoids and/or EDHF [54]. Similarly, the contribution of prostanoids to the regulation of coronary microvascular tone in humans also depends on the progression of CAD, as inhibition of prostanoid production induced mild vasoconstriction distal to angiographically minimally diseased coronary arteries [60] whereas vasoconstriction was more pronounced in patients with ischemia at rest [61]. Importantly, abnormal coronary vasomotion, endothelial dysfunction and coronary vasospasm have been shown to predict adverse outcomes in patients with atherosclerosis and suspected ischemia [62]. In this group of patients, our observation of increased vasoconstrictor response to ET‐1 distal to a chronic stenosis may be of particular relevance, as increased levels of ET‐1 were measured not only in the circulating plasma of these patients but also in the pericardial fluid, and were capable of eliciting substantial arterial vasoconstriction [63]. In conclusion, there is increasing evidence that in patients with obstructive CAD, the distal microvascular bed undergoes functional and structural changes that likely contribute to the perturbations in myocardial perfusion and hamper functional recovery after revascularization.

Non‐Obstructive CAD

Patients with ischaemia with non‐obstructive CAD (INOCA) experience symptoms of myocardial ischaemia, but do not exhibit significant obstructive CAD on angiography. A high proportion of INOCA patients has features of metabolic dysregulation, such as obesity, hypertension, dyslipidaemia and insulin resistance. These conditions either alone or in combination increase the risk of different pathologies including coronary and peripheral artery disease, stroke and cancer [64]. Approximately 20%–25% of the world's adult population is estimated to have metabolic dysregulation, and its prevalence continues to rise [64]. In the United States, about 35% of all adults and nearly 50% of those over 60 years of age are estimated to have metabolic dysregulation and metabolic syndrome, with a higher prevalence in women than men [65]. Importantly, the rise in metabolic dysregulation is not limited to adults. A 2020 study estimated the global prevalence of metabolic syndrome at 2.8% for children and 4.8% for adolescents, equating to approximately 25.8 million children and 35.5 million adolescents affected worldwide, putting them at risk for future cardiovascular events [66].

Microvascular dysfunction has been postulated to be the major mechanism linking metabolic dysregulation to either obstructive or non‐obstructive ischaemic heart disease. Clinical and experimental studies show that perturbations in the control of coronary blood flow during increased metabolic demand in metabolic dysregulation and obesity are present well before overt CAD develops [49, 53, 67, 68, 69, 70]. Similarly, increasing evidence also supports the involvement of microvascular dysfunction in patients with INOCA [71, 72], as well as heart failure with preserved ejection fraction (HFpEF) [73, 74]. Indeed, metabolic syndrome and obesity are associated with impaired CFR, a common feature of INOCA and HFpEF [75]. These studies are supported by observations in dogs [76] and swine with comorbidities [77], demonstrating impairment of myocardial oxygen delivery during graded treadmill exercise [53, 77]. In swine with comorbidities, lower CFR and impaired oxygen delivery during exercise were associated with reduced myocardial lactate consumption, consistent with ischemia [53]. Also in humans, CFR worsens with the onset of type 2 diabetes [78]. In young subjects, acute hyperglycemia resulted in lower adenosine‐mediated increase in coronary blood flow and altered hyperemia in response to cold pressor testing, suggestive of sustained endothelial dysfunction [79]. Additionally, in asymptomatic individuals with isolated Lp(a) elevation or with familial hypercholesterolemia, the prevalence of coronary microvascular dysfunction and reduced CFR was higher than in healthy controls [80].

In farm pigs subjected to diabetes, a high‐fat/high‐fructose diet and chronic kidney disease for 6 months, the impaired CFR was not associated with alterations in arteriolar structure (Figure 4), stiffness or density, although capillary rarefaction was observed [53]. In contrast, 15 months of a high‐fat diet with or without diabetes did lead to stiffening of the small arteries, although hypertrophy of the vascular wall was not present [81]. Conversely, in Ossabaw swine, a well‐known model of metabolic syndrome, 16 weeks of a high‐fat/high‐fructose diet resulted in increased resting myocardial perfusion and a blunted response to adenosine associated with increased arteriolar density [82]. In the same model, 6 months of metabolic syndrome and a high‐fat diet resulted in impaired hyperaemic flow associated with augmented coronary myogenic tone, hypertrophic inward remodelling (Figure 4) of the coronary resistance arteries and capillary rarefaction [52].

While the alterations in vascular structure in the animals with comorbidities could not account for the impaired myocardial perfusion and reduced CFR, changes in coronary microvascular function, characterized by a reduction in coronary vasodilator responsiveness to various pharmacological agonists, were often seen. Our studies in swine with comorbidities clearly demonstrate the temporal changes in coronary microvascular dysfunction, thereby highlighting the importance of longitudinal studies when investigating perturbations in the control of coronary microvascular tone. Thus, at an early stage, in male swine subjected to 2.5 months of hyperglycaemia and hypercholesterolaemia, endothelium‐dependent vasodilation of isolated coronary small arteries to bradykinin was impaired due to loss of NO, while VSMC function, as measured by vasodilation to the exogenous NO‐donor SNAP (S‐nitroso‐N‐acetylpenicillamine) was maintained [70]. Interestingly, in this model, reduced ETA‐mediated vasoconstriction to ET‐1 was also observed, possibly as a compensatory mechanism. These alterations were not noted in hypercholesterolemic swine in the absence of hyperglycaemia, suggesting that both comorbidities were required to induce endothelial dysfunction. The progression of coronary microvascular dysfunction was studied in the same model after 15 months of hyperglycaemia and hypercholesterolaemia [81]. Unexpectedly, at this stage of the disease, when non‐obstructive atherosclerotic plaques were already observed in the coronary vascular tree, the endothelium‐dependent vasodilation to bradykinin was no longer reduced as compared with healthy controls, but the balance between the vasodilator and vasoconstrictor influences was dominated by an increased ETB—mediated vasoconstrictor response to ET‐1, while the levels of circulating ET‐1 were also elevated, similar to observations in patients with INOCA [21]. Although the endothelium‐dependent vasodilation to bradykinin was unaffected, altered NO/EDHF balance was documented in these animals, the EDHF contribution to the dilation being significantly reduced [81]. Similar to our observations in overt diabetes and hypercholesterolaemia, obesity has also been shown to increase coronary microvascular sensitivity to vasoconstrictors, including ET‐1, PGH2 or thromboxane A2, in both animal [68] and human [83] studies.

Because INOCA is more prevalent in females, in more recent studies, we investigated microvascular function in female pigs subjected to 6 months of diabetes, high‐fat diet and chronic kidney disease. In these animals, the vasodilation to bradykinin was impaired, suggestive of endothelial dysfunction, similar to the early findings in the male pigs; however, this was now mediated by a complete loss of NO, as measured both in vitro and in vivo in exercising swine [84]. The loss of NO bioavailability was associated with an increase in NOX expression, myocardial 8‐isoprostane levels and a decrease in antioxidant capacity, suggesting that oxidative stress was principally responsible for the reduced NO bioavailability. In vivo and in vitro ROS scavenging experiments indicated that the reduced NO bioavailability was partially compensated by increased H2O2‐mediated coronary vasodilation [85].

In addition to endothelial dysfunction, smooth muscle cell function has also been shown to be altered by metabolic dysregulation (Figure 6). Thus, 16 weeks of high‐fat diet in obese Ossabaw swine with metabolic syndrome resulted in increased coronary vascular tone mediated by altered electromechanical coupling between KV and CaV1.2 channels in smooth muscle cells [86]. Conversely, in swine with familial hypercholesterolemia, impaired endothelium‐dependent coronary vasodilation to ATP, measured in vivo, was paralleled by a compensatory increase in smooth muscle cell sensitivity to NO‐donor sodium nitroprusside (SNP). This increased VSMC sensitivity was confirmed in isolated small coronary arterioles ex vivo, together with impaired dilation to the endothelium‐dependent vasodilator bradykinin, and possibly acted as a compensatory mechanism for reduced NO bioavailability [77]. Similarly, in our swine subjected to 6 months of diabetes, hypercholesterolemia and chronic kidney disease, loss of NO bioavailability was also associated with an increased vasodilator response to SNP [84].

Although endothelial dysfunction is a key factor in ischaemic heart disease, chronic treatments aimed at alleviating endothelial dysfunction and, in particular, improving NO bioavailability have shown conflicting statin treatment in patients with inducible myocardial ischemia showed improved endothelial function and exercise‐induced perfusion [87], while nitrates, showing positive effects in acute settings of angina [87, 88], did not show improvements in patients with non‐obstructive CAD [89]. Conversely, long‐term oral nitrate therapy was associated with increased risk for adverse cardiovascular effects in patients with type II diabetes post PCI and in patients with vasospastic angina [90, 91]. Long‐term L‐arginine supplementation improved coronary small‐vessel endothelial function and decreased plasma endothelin concentrations in patients with non‐obstructive ischaemic heart disease; however, it did not improve the CFR [92]. Although conflicting, these findings highlight the complexity of endothelial dysfunction and the intricate interplay of mechanisms that contribute to it.

Increased sympathetic activity in metabolic dysregulation has also been documented by numerous studies both in patients and animal models, showing increased plasma catecholamines that may result in exaggerated alpha‐adrenergic coronary vasoconstriction [93]. This may be partially mediated by decreased coronary microvascular levels of the β‐adrenergic receptors with advancing age [94], diabetes [95] or cardiac hypertrophy [96], although a role for β‐adrenergic receptors in vasodilation has been shown in isolated vessels from human hypertrophied hearts [97]. Furthermore, there is substantial evidence for activation of the RAAS associated with adipose‐tissue–derived angiotensinogen, the major precursor of angiotensin, leading to increased Ang II–mediated vasoconstriction in the coronary circulation [98].

Additionally, adipocyte‐derived free fatty acids and leptin also lead to increased adrenergic tone [99]. In patients with metabolic dysregulation, the presence of adipose tissue was associated with microvascular dysfunction, as reflected in a lower CFR which correlated with the increased amounts of epicardial fat tissue. Moreover, epicardial fat tissue deposition was a predictor of worse CFR even after accounting for the presence or absence of metabolic syndrome [100]. Adipocytes and perivascular adipose tissue‐derived adipokines such as leptin, resistin, IL‐6 and TNF‐α are potent pro‐inflammatory molecules promoting oxidative stress in the endothelium and altering endothelial function and NO bioavailability, either directly or via increased ET‐1 production [101]. Furthermore, leptin derived from perivascular fat was also shown to promote coronary arterial vasoconstriction and smooth muscle cell proliferation via Rho kinase signalling [102]. Additionally, adipocyte‐derived circulating free fatty acids and hyperglycemia‐induced advanced glycation end products can lead to increased oxidative stress limiting NO bioavailability and increased production of vasoconstrictor factors such as thromboxane and ET‐1 [103].

While in the healthy heart, the venous circulation accounts for less than 10% of the coronary resistance, and increases in venous blood pressure in the healthy heart were shown to lead to myogenic vasoconstriction of precapillary vessels and capillary de‐recruitment [104], it has been shown that in patients with microvascular angina, an increase in coronary sinus pressure resulted in decreased microvascular resistance associated with small increases in capillary recruitment/diameter and an increase in blood flow [105], a mechanism proposed to have potential therapeutic implications for microvascular angina.

Sex‐Differences in the Mechanisms of Coronary Blood Flow Regulation in Ischaemic Heart Disease

In recent years, research efforts have been put into the understanding of sex differences in the pathophysiology of coronary microvascular dysfunction and ischaemic heart disease. Coronary microvascular dysfunction is thought to be the major aetiological factor for ischaemic heart disease and INOCA in women, especially after menopause when the protective effects of sex hormones are diminished [106]. Data from patients with coronary microvascular dysfunction indicated that despite similar coronary vascular resistance, females have a lower CFR than males, mainly due to higher basal flows [107]. Additionally, clinical data indicated that myocardial perfusion is more often related to hyperlipidemia and diabetes in men, but no association was found in women [108]. While coronary microvascular dysfunction in men was thus mainly driven by inflammation, among women, alterations in ventricular remodelling and fibrosis may play a more important role in the pathogenesis of coronary microvascular dysfunction [109]. Despite these sex differences, coronary microvascular dysfunction, and in particular endothelial dysfunction with loss of NO bioavailability, is an intricate part of ischaemic heart disease in both men and women.

Translational Value and Limitations

In this review article, we summarized various aspects of microvascular dysfunction and its role in both obstructive and non‐obstructive CAD, based on our experience with large animal models and comparing our findings with existing knowledge from human studies. However, it is imperative to acknowledge that no single animal model perfectly emulates the human disease, nor has a perfect translational capacity to the clinical setting. This stems mainly from the long‐term development of the disease in humans leading to a more complex pathology than in the animal models. Introduction of a coronary obstruction either acutely or over a period of months and the initiation of comorbidities may result in pathological processes similar to humans, but model limitations still remain. These factors should be carefully considered when selecting an animal model, and limitations such as young age and single‐sex cohorts should be avoided whenever possible, particularly in studies aiming at developing novel therapies for ischaemic heart disease [69].

Conclusions

Coronary microvascular dysfunction is an intricate part of ischaemic heart disease, with changes in both microvascular function and structure. An early consequence of exposure to cardiovascular risk factors is coronary microvascular dysfunction before overt macrovascular disease occurs. Coronary microvascular dysfunction encompasses endothelial dysfunction (loss of NO and increased ET‐1) in conjunction with arterial (inward) remodelling and reduced capillary densities, contributing to the observed perturbations in myocardial perfusion and decreases in CFR, thereby promoting myocardial ischaemia. Subsequently, impaired coronary flow and altered shear stress may promote macrovascular endothelial dysfunction and contribute to the development and progression of atherosclerosis and obstructive CAD, with the ensuing proximal coronary stenosis leading to further perfusion impediments and exacerbation of distal microvascular dysfunction and remodelling, thereby creating a feed‐forward mechanism of perturbations in coronary blood flow (Figure 7). Consequently, coronary microvascular dysfunction represents an important target for therapeutic interventions.

Conflicts of Interest

The authors declare no conflicts of interest.