A chronology of basic and clinical research in the coronary microcirculation

1. Introduction

The purpose of this review is to highlight some of historical landmarks that have provided the foundation for current basic and clinical research in the coronary microcirculation. In this brief review, we cannot possibly reference every important publication, but we have made every effort to cite those reports that have been important in the evolution of knowledge about the microcirculation of the heart. Fig. 1 illustrates an evolutionary tree crediting those studies the authors see as key in the myriad of advances in our understanding of the heart’s microcirculation in health and disease. Our brief review focusses on the coronary microvasculature and not lymphatics. This is not to dismiss the importance of lymphatics in the heart, but rather emphasize methods the lead to concepts about the regulation of coronary blood flow.

Fig. 1.

This figure includes reports focusing on morphometric parameters as well as studies directed at understanding the control of blood flow and vascular tone. A branch of the tree represents clinical investigations that provided insight into disease processes that appear related to coronary microvascular dysfunction. The end of each branch is depicted by lines that indicate that growth is dynamic and there will undoubtedly be advances made in the future.

2. The Base of the Tree and the Morphometrics Branch

At the base of the tree, we cite the classic work of Oskar Langendorff, who studied the isolated mammalian heart [1]. In this preparation, Langendorff documented the phasic nature of both coronary arterial inflow and coronary venous outflow. He likened the heart to a sponge that is squeezed during systole, where arterial inflow to myocardium is impeded, and venous outflow facilitated. During diastole the sponge fills due to arterial inflow and the cessation of venous outflow. Over three decades later, the first report of microvascular morphometric parameters in the coronary circulation appeared by Wearn [2]. Microscopic techniques were used by Wearn to evaluate various parameters (including capillary density) of cardiac tissue sections from cat, rabbit, and human hearts. Capillary densities in cardiac tissue were considerably higher than previously reported by Krogh [3] for skeletal muscle, thus leading to the conclusion that the heart has a “very rich blood supply.” Also, capillary densities in the ventricles were higher than in the atria. Although not shown in Fig. 1, Shipley et al. [4] observed a decrease in capillary density (compared with normal hearts) in cardiac tissue sections from hypertrophied ventricles. These authors speculated that due to the decrease in capillary density, there could be impairment in exchange of metabolic substrates and oxygen.

In the 1960s, Thomas James pioneered the study of human coronary arteries in many pathologic conditions; he published detailed photographic segments of vinylite casts of the small coronary arteries and histologic findings from myocardial necropsy specimens [5]. His work highlighted medial necrosis in the small arteries supplying the conduction system in various conditions, acute arteritis in polyarteritis nodosa, and various occlusive states due to abnormal platelet aggregation, amyloid deposition, and intimal proliferation in diabetes [6–8]. Although research in morphometric attributes, e.g., capillary and arteriolar density, of the coronary microcirculation under a variety of physiological and pathophysiological conditions in tissue sections continued for years [9–13], an important advance was implemented to involve a more comprehensive view of the entire coronary vasculature. In 1993 Kassab et al. provided a detailed morphometric analysis of the entire coronary arterial and capillary circulation from a vascular cast of a porcine heart [14]. This analysis revealed some striking findings, such as capillaries do not necessarily stem from the last order of arterioles, rather they can branch from vessels one and even two branch points from the pre-capillary vessel. This finding of heterogeneity in the architecture of the coronary microcirculation also led to some in vivo research involving fractal analyses that confirmed this concept [15,16]. In 1995, Kassab et al. used their analyses of a coronary vascular cast to show the morphometry of the coronary venous circulation [17]. However, the pinnacle on the morphometrics branch is the approach used by Spaan et al. [18]. This technique was even used to visualize and reconstruct the coronary microcirculation of the human heart, which was found to contain a significant number of subendocardial collateral vessels in heart failure [19].

3. The In Vivo Branch

The next branch we will discuss is the in vivo portion of the tree. The first paper that studied in vivo attributes of the coronary microcirculation was Yudilevich and Martindejulian [20]. These authors used indicator-dilution techniques to study of solute exchange in the heart. This seminal paper was the catalyst for investigations into exchange processes in the heart, transit times of solutes and gases, and capillary permeability-surface area products [21–28].

The next advance in in vivo involved direct visualization of the microcirculation of the beating heart. Martini and Honig used cinematography of the epicardial surface of the beating rat heart to measure intercapillary distances at varying arterial pO₂ tensions [29]. This involved a tedious frame-by-frame analysis until an image was found that was in focus. The results revealed that intercapillary distance was reduced during hypoxia, suggesting that not every capillary is being perfused at every moment and, in turn, suggestive of the idea of “capillary reserve.” The idea of capillary reserve was corroborated from results obtained from indicator dilution experiments that also revealed an increase of capillary surface area during vasodilation [26]. An advance in the evolution of in vivo research involved microscopic visualization of the beating heart in a preparation that constrained the motion (via insertion of pins) associated with the cardiac cycle [30,31]. High-speed cinematography was used to obtained images (films) of the epicardial surface to acquire measurements of phasic blood flow velocities in the atria Hellberg et al. [30] and left ventricle Tillmanns et al. [31]. A seminal advance was made by Nellis et al., who designed a system in which stroboscopic illumination was synchronized with the heart; such that a brief flash (15–20 μsec) occurred at the same point in each cardiac cycle [32]. When viewed through a microscope the heart would appear motionless (persistence of the image in one’s retina would retain the image until the next heartbeat). The authors also synchronized the frame capture of a camera to enable collection of high-resolution images in the beating heart. Nellis and colleagues also created a program in which the strobe flash would advance 1/100 of a cardiac cycle per beat, which then created an illusion that the heart was beating in slow motion when viewed through the microscope. One-hundred heartbeats then constituted a cardiac cycle. This slow-motion approach was used to synchronize a 3-dimensional electromagnetic micromanipulator to cardiac motion and via micropuncture techniques, pressures in coronary arterioles and venules were measured in the beating right ventricle. In 1984, Ashikawa and colleagues devised a “floating” microscope objective that moved in conjunction with the beating heart [33]. They used this approach to measure phasic flow velocities and diameters in epicardial vessels of the left ventricle.

Chilian et al. (1986) reported the distribution of microvascular pressures of the beating left ventricle under baseline conditions (pentobarbital anesthesia) and during coronary vasodilation with papaverine and dipyridamole [34,35]. These authors based their approach on the Nellis technique but adapted their stroboscopic system to video (rather than cinematography) and made measurements in the left ventricular coronary microcirculation. An important concept presented in this paper was that the bulk of coronary resistance resided in coronary arterioles and these vessels could decrease their resistance 12-fold during maximal dilation. Although small arteries only account for a small portion of total coronary resistance in the normal heart, during vasodilation as arterioles dilate, their contribution become more important. In 1993, two important advances were reported. The laboratory of Fumihiko Kajiya reported visualization of the endocardial coronary microcirculation of the beating heart [36]. Their preparation involved the use of an image-transferring conduit surrounded by a ring of fiberoptic illuminators and a silicone ring at the end of the conduit. When the ring was pressed against the endocardium and a small cannula allowed the flushing of blood from the area within the ring, the endocardial microcirculation could be visualized.

Trzeciakowski and Chilian used stroboscopic imaging of the epicardial surface to measure a parameter termed “flowmotion” [37], which was the “back and forth” flow occuring in an arterial-arterial anastomosis. Flowmotion was measured by following a “wavefront” of fluorescence. The authors reasoned that the “back and forth” motion is caused by changes in the pressure drop across the anastomosis, which in turn would be caused by fluctuations in upstream and downstream resistance. Modeling of this movement revealed a chaotic pattern that was affected by interventions such as antagonism of nitric oxide synthases, and adenosine receptor blockade. To provide some perspective, chaos is complexity, not randomness. The importance of this observation is that the factors that control blood flow in the coronary micro-circulation follow a non-linear control scheme. Thus, linear models will never accurately model the behavior of the microcirculation of the heart.

A criticism of the previous studies of the in vivo coronary microcirculation was that the measurements were confined to the epicardial surface, but this system circumvented that problem. Of note, the dynamics of flow velocity in microvessels from the subendocardium were very different than those reported in the subepicardium with a significant component of systolic retrograde flow in the subendocardium [38]. We would be remiss to not point out a caveat of all preparations involving in vivo analysis of the coronary microcirculation. Every approach employs some stabilization of the heart to prevent the “Z” motion, i.e., the “up and down” motion of the ventricle occurring during the normal cycles of contraction/relaxation. Undoubtedly, these procedures affect some attribute of mechanical forces and interpretations should always be made in the context of this issue.

Some recent techniques have been designed to non-invasively measure the diameter of coronary resistance vessels [39], and visualize diameters and flows in the beating heart via gated two-photon microscopy [40]. While the application of these techniques shows much potential, they are in their infancy.

4. The in vitro branch

In 1988 Kuo et al adapted an approach used for the study of isolated arterioles in peripheral organ systems to enable analysis of the functional attributes of the coronary microcirculation [41]. In their first report using this approach, the authors identified differences in myogenic responses between arterioles isolated from the subepicardial and subendocardial regions of pig hearts. In addition, the investigators used the isolated coronary arteriole preparation to study a variety of functional attributes of the coronary microcirculation including longitudinal gradients in arterial and arteriolar responses to metabolites, myogenic stimuli, and shear stress [42,43]. The concept presented by the longitudinal gradient concept was that different regions of the coronary microcirculation have different control mechanisms. Flow- or shear stress-dependent stimuli appear to dominate the control of small arteries, with myogenic mechanisms dominating intermediate sized arterioles and the smallest arterioles have the greatest sensitivity to metabolic factors. This type of control allows for an integrated response to the coronary microvascular network to any stimulus that alters cardiac work. In 1993 Yuan et al. adapted the isolated arteriole preparation to enable the study of isolated venues. This preparation was used to measure permeability of coronary post-capillary venules to macromolecules [44]. One key observation made using this approach was the discovery that increases in venular flow velocity, without changes in pressure, increase venular permeability [45] and also engendered the study of inflammatory cell-endothelial dynamics in a coronary exchange vessels [46].

Although the in vitro microvessel approach was used by many laboratories, in 1990s, the group of David Gutterman used it to study coronary vessels isolated from the human heart. Specifically, these investigators studied microvessels isolated from discarded atrial appendages following cardiopulmonary bypass [47]. Although there is a limitation of this approach concerning a suitable control group (patients undergoing valve replacements without coronary artery disease were used as controls for those with coronary disease), this group has made important contributions to our understanding of the pathophysiological consequences of vascular disease in the coronary microcirculation [48,49].

The isolated vessel preparation has also provided clues as to signaling mechanisms in the coronary microcirculation. In the late 1990s, Muller et al incorporated calcium imaging in conjunction with the video-microscopy of arterioles to reveal that dilation and production of nitric oxide in response to flow or shear stress does not require changes in cytosolic calcium [50]. This was demonstrated by “clamping” intracellular calcium with an intracellular chelator and showing that dilation to shear stress remained intact. Freed et al. (2014) further blended clinical pathophysiology with basic science to reveal some mechanisms underlying endothelial dysfunction in ischemic heart disease [49].

The next steps in the in vitro branch of the coronary microcirculation will likely involve experiments utilizing technology associated with induced pluripotent stem cells to study vascular cell biology of patients in a dish. The future holds promise with high-resolution microscopy and imaging techniques that create ‘Z stacks,’ enabling visualization of the endothelium and vascular smooth muscle to better understand the interactions among these cells.

5. Coronary microvascular dysfunction

According to the state of knowledge in the late 1960s, after the development of selective coronary angiography and coronary bypass surgery most cardiologists, and physicians in general, equated ischemic heart disease to epicardial coronary artery disease (CAD). The microcirculation was an afterthought. This concept was challenged by Marcus et al. in the early 1980’s, who measured coronary vasodilator reserve in patients with aortic stenosis [51]. Many of these patients had angina and little to no coronary vasodilator reserve—in the absence of large vessel coronary disease. This observation became one of the seminal observations showing that myocardial ischemia can occur in the absence of coronary disease. Similar observations were reported in patients with hypertrophic and other cardiomyopathies.

Cardiac syndrome X (CSX) was first suggested in 1973 by Harvey Kemp [52]. The “tongue-in-cheek” term was used to describe patients with chest pain and normal appearing coronary angiograms, which was a diagnostic combination that was previously unknown, in his editorial description of the work by Robert Arbogast and Martial Bourassa [53]. The term “syndrome X” was used to indicate the uncertainty of the cause of the chest pain in these patients [54]. The etiology was unclear with multiple hypotheses proposed including endothelial dysfunction, myocardial ischemia, abnormal pain perception, infection, and estrogen deficiency are among the most suggested. Also, it is possible that the defects may reside in the basement membrane, pericytes, smooth muscle cells, components of extracellular matrix, etc.

Due to advances in diagnostics, including identification of coronary microvascular dysfunction (CMD) with invasive flow reserve measurement in humans, predominantly female, the uncertainty of diagnosis is no longer an issue and the term Cardiac Syndrome X is no longer used [55]. Historically, risk factors such as hypertension, diabetes, smoking, chronic inflammation, etc. have all been associated with arteriolar remodeling and microvascular damage. More recently, anti-cancer therapeutics which can have adverse consequences in the microvasculature, are also recognized as important contributors to CMD. One of the research limitations in this field has been the confusion due to varying nomenclature and definitions of CMD. In 2018, the coronary vasomotor disorders international study group (COV ADIS) published standardized criteria for diagnosis of CMD [56]. Our ability to interrogate the microcirculation has now shown that CMD is prevalent and prognostic in both men and women [57,58]. However, there are notable sex-differences in this condition, where women are more impacted by CMD with more angina and a lower quality of life compared to men. [58] We briefly review the evolution in methods for human diagnosis of CMD that are clinically used in the United States.

In the 1990s, there were major advances in our clinical ability to investigate coronary physiology and flow. The initial work came from an assessment of coronary blood flow made from measuring x-ray contrast agent flow by counting its movement from cine frame to cine frame. Since this was done post-MI in a reperfused coronary by Robert Gibbons of the TIMI group, it was called the TIMI-Frame count [59]. Later others used Doppler techniques to enable quantification of coronary flow reserve (CFR) and fractional flow reserve (FFR) [60]. FFR provided information about hemodynamic significance of luminal narrowing, assessed by a pressure drop across the lesion. Invasive CFR reflected the ability of microvasculature to increase flow in response to substances such as adenosine (which was given intracoronary directly into the coronary artery but also via intravenous in some protocols) [56]. An attenuated response, characterized by inadequate increase in coronary flow velocity in response to hyperemia, was used to assess CFR, a measure of coronary microvascular dysfunction (CMD) in the absence of epicardial stenosis. Abnormal CFR independently predicts cardiac events in women as shown in the Women’s Ischemia Syndrome Evaluation (WISE) study [61]. Several studies have found that CFR is lower in women compared to men, and it is more prevalent in post-menopausal women, for reasons that are not entirely clear.

A more specific measure of CMD has come to the forefront in clinical care in recent years. In contrast to CFR, which evaluates flow through both the macro-and micro-circulation and is influenced by epicardial atherosclerosis and hemodynamics, the index of microcirculatory resistance (IMR) technique specifically tests the microvasculature and serves as a prognostic factor (Fig. 2). It is based on thermodilution technique that measures the mean transit time of blood flow; when IMR is high, flow is reduced at the microcirculatory level, and a cutoff of less than 25 is considered normal [62]. Catheter-based techniques to assess microvascular function have continued to evolve with the availability of pressure/temperature sensor-tipped guidewires. Recently, measurement of absolute coronary blood flow and microvascular resistance using continuous thermodilution technique was shown to be safe and reproducible in humans. [63] Microvascular resistance reserve (MRR) has been proposed as another new index of microvascular function, which is derived from continuous thermodilution, but is independent of epicardial disease and hemodynamic changes with hyperemia. [64–66] An important aspect of catheter-based assessment techniques in humans is safety and carefully weighing the risks vs. benefits of the procedure. In recent years, there have also been growing number of studies reporting that invasive coronary function testing is safe when done at experienced centers by trained interventionalists. [67–69]

Fig. 2. Endothelium-independent and endothelium-dependent angina mechanisms.

IMR, Index of Microcirculatory Resistance. *Coronary flow reserve (CFR), an integrated measure of flow through both epicardial and microvascular vessels, but in the absence of epicardial disease it serves as a measure of microvascular function. Figure made using biorender.com. Reprinted from Patel Net al. Climacteric 2023. DOI: https://doi.org/10.1080/13697137.2023.2281933

Simultaneous to advances in invasive flow detection methods, in the mid-1980s, non-invasive methods started to become available. With radioisotopes such as Rubidium-82 and Nitrogen-13 ammonia and cardiac positron emission tomography (PET) cameras, myocardial blood flow at rest and with maximal vasodilation could also be quantified as flow per minute per gram of tissue. Not only could regional perfusion be quantified based on these techniques, but also global myocardial flow reserve could be obtained. In the setting of no obstructive CAD, a low myocardial flow reserve reflects impaired microvasculature, and several studies over the years have noted that PET-derived flow reserve is prognostic of cardiovascular outcomes [57,70–72]. The evolution of non-invasive methods to assess perfusion reserve continued with cardiac magnetic resonance (CMR) imaging first-pass perfusion techniques in 2000s, a modality that had a distinct advantage of being a non-radiation exposure test and enabled superior myocardial tissue characterization compared to echocardiography [73]. In recent years, CMR based semi-quantitative methods to assess microvascular flow reserve and relationship to diastolic dysfunction has been further enhancing our understanding of the role of microcirculation in progression towards heart failure with preserved ejection fraction [74–76]. While these non-invasive methods have improved CMD detection, invasive coronary function testing remains the reference-standard to clinically assess vascular dysfunction (macro- and micro-vascular) [77].

5.1. INOCA

The term INOCA (ischemia but no obstructive coronary arteries) refers to patients who are suspected of ischemia but have no obstructive stenosis (defined as less than 50 % diameter reduction in any of the major epicardial coronary arteries) [78]. Seminal studies have demonstrated underlying endothelial dysfunction and CMD with associated adverse outcomes of myocardial infarction, heart failure, and repeated angina hospitalizations among those with INOCA [61,79–83] (Fig. 3). More recently, the clinical understanding has evolved, and refined to categorize CMD into two categories: structural CMD and functional CMD. Prior work in patients with myocardial hypertrophy due to aortic stenosis or hypertrophic cardiomyopathy had evidence of arteriolar remodeling and abnormal microvascular function. Hypertension and diabetes-related endothelial cell damage, arteriolar remodeling, and capillary rarefaction have also been noted in those with limited flow reserve. Microvascular spasm, an example of functional CMD, has been recognized as an important contributor to INOCA in the last two decades, diagnosed by simultaneous symptoms and electrographic changes during acetylcholine infusion (in the absence of visible epicardial vasospasm) [84]. One should note that these two CMD classifications share endothelial dysfunction and are not mutually exclusive and a patient can have both endotypes [85]. Myocardial blood flow is tightly regulated by intricate hormonal and metabolic factors through a complex network of various sized blood vessels. The large conduit vessels (i.e. epicardial coronary arteries) provide the primary route for blood to reach the myocardium but contribute relatively little to vascular resistance when open and not constricted. In contrast, the smaller sized arterioles are critical regulators of myocardial blood flow; they vasodilate and vasoconstrict in response to stimuli and to match local demand. The small arterioles are responsible for the greatest resistance due to their small diameter. Understanding the relative importance of these vessels is crucial in understanding the pathophysiology of CMD and for therapeutic targets. Heart rate changes are tightly coupled with coronary blood flow in healthy humans, such that a rise in heart rate increases coronary blood flow. Among those with CMD, the limited flow reserve during exercise can present with symptoms of chest pain and dyspnea with exertion despite lack of obstructive CAD.

Fig. 3. INOCA Key Points.

MI = myocardial infarction, CHF = congestive heart failure, SCD = sudden cardiac death, QoL = quality of Life, CAD = coronary artery disease. Reprinted from Patel N et al. Climacteric 2023. DOI: https://doi.org/10.1080/13697137.2023.2281933

Cardiometabolic risk factors such as insulin resistance/diabetes and obesity are associated with INOCA and CMD, and associated with heart failure with preserved ejection fraction (HFpEF). Treatment of these risk factors with weight loss medications such as glucagon-like peptide agonists is an exciting new frontier in CMD management. Liraglutide treatment was shown to improve coronary microcirculation in insulin-resistant Zucker obese rats on a high-salt diet. [86]

Pacanowski et al. found a genetic association with coronary micro-vascular dysfunction [87]. These investigators found polymorphisms in the beta1- and beta3-adrenergic receptors that were linked to increased risk in INOCA patients. In 2013, Fedele and colleagues completed a genetic analysis of polymorphisms associating with INOCA [88]. They reported that specific polymorphisms in eNOS, Kir6.2 and Nav1.5 were correlated with coronary microvascular dysfunction. Both of these studies led to the concept that genetic polymorphisms for ion channels appear to have a clinical impact in the susceptibility for coronary microvascular dysfunction, independent of the presence of classic cardiovascular risk factors.

5.2. MINOCA (acute coronary event)

One consequence of endothelial dysfunction and CMD is myocardial infarction with no obstructive coronary arteries (MINOCA), although the relative contribution of CMD is unknown. MINOCA occurs in approximately 5–15 % of acute MI cases, and predominates in younger women [89]. MINOCA tends to recur in 10–20 % of cases. The term MINOCA was first coined in 2013 and is increasingly used; however, it is considered a working diagnosis that requires further investigation to determine the potential pathophysiologic mechanism that triggered the acute event [90].

As intracoronary imaging techniques have evolved, our ability to identify elusive plaque and other vascular pathologies within the arterial wall has improved. Intravascular ultrasound technique in the late 1980s-early 1990s enabled coronary plaque imaging that is hidden within the arterial wall and not detected by traditional clinical coronary angiography which is a lumenography of the coronaries. IVUS studies have demonstrated diffuse epicardial plaque in a majority of cases of CMD. Around the early 2000s, optical coherence tomography (OCT) technique, originally used in ophthalmology to visualize the retina, became available for coronary imaging and provided much higher resolution images of coronary arteries. In a recent MINOCA study of women who underwent OCT, there was evidence of non-occlusive atherothrombotic plaque rupture in 85 % of cases [91]. It is clear that our understanding of pathophysiologic mechanisms has continued to evolve with improvement in diagnostic techniques and wider availability of testing.

6. Back to the Future

We start this section as “Back to the Future” because there are many unknowns in the coronary microcirculation, and filling in the gaps created by these unknowns will require studies to reveal mechanisms of fundamental processes. It is humbling to consider that science has known the human genome for over two decades, and yet, the precise mechanisms underlying processes such a coronary metabolic dilation, i.e., the connection of coronary blood flow to myocardial metabolism; and coronary autoregulation, the ability of the coronary circulation to maintain blood flow constant despite changes in perfusion pressure, are incompletely understood. These processes are fundamental to the ability of the heart to maintain pump function during physiological and pathophysiological challenges. The solutions to a better understanding will undoubtedly require further technical advancements as well as interdisciplinary science. Perhaps a more comprehensive vision of the coronary vasculature would be obtained with insights offered by single cell and spatial genomics approaches that are connected to physiological approaches.

In terms of clinical advancements, standard of care diagnostic algorithms and best practices for CMD diagnosis and treatment are evolving as CMD has come to the forefront as an important player in ischemic heart disease. While significant strides have been made in imaging and understanding the coronary microcirculation, many unknowns remain, particularly related to sex-differences with female predominance of angina and lower quality of life associated with CMD. Future research must delve deeper into molecular and cellular mechanisms that drive CMD. An improved understanding of the cross-talk between autoimmune dysregulation, inflammation, and sympathetic dysfunction in CMD is needed to help refine therapeutic approaches. Emerging technologies such as advanced microvascular flow imaging, artificial intelligence-driven analytics, and novel biomarkers may soon provide greater insights into microvascular pathophysiology. For example, uridine adenosine tetraphosphate is a novel vasodilator in the coronary microcirculation, which was shown to act through purinergic P1 receptors. [92] Precision-medicine based methods to identify at-risk CMD individuals and clinical trials designed to inform management and favorably alter the risk trajectory of CMD patients and improve angina and quality of life are ongoing research areas. In many ways, the future of CMD research takes us back to its origins—seeking answers to fundamental processes that continue to elude us.