Assessments of microvascular function in organ systems

Abstract

Microvascular disease plays critical roles in the dysfunction of all organ systems, and there are many methods available to assess the microvasculature. These methods can either assess the target organ directly or assess an easily accessible organ such as the skin or retina so that inferences can be extrapolated to the other systems and/or related diseases. Despite the abundance of exploratory research on some of these modalities and their possible applications, there is a general lack of clinical use. This deficiency is likely due to two main reasons: the need for standardization of protocols to establish a role in clinical practice or the lack of therapies targeted toward microvascular dysfunction. Also, there remain some questions to be answered about the coronary microvasculature, as it is complex, heterogeneous, and difficult to visualize in vivo even with advanced imaging technology. This review will discuss novel approaches that are being used to assess microvasculature health in several key organ systems, and evaluate their clinical utility and scope for further development.

INTRODUCTION

The importance of the microvasculature cannot be understated. Throughout the body in all organ systems, the microvasculature is present to transfer oxygen and nutrients, to remove waste products, and to regulate blood flow. Consequently, microvascular dysfunction is well known to be implicated in numerous disease processes such as cardiovascular and renovascular pathologies, and diabetes (1). Besides these more commonly recognized conditions, the neurological, pulmonary, and gastrointestinal organ systems have many diseases that are rooted in microvascular dysfunction, and pathologies such as severe sepsis and coronavirus disease 2019 (COVID-19) can result in multiorgan failure from reduced microvascular perfusion (2–5). Pregnancy is also associated with microvascular-related complications (6).

Several imaging tools have been developed to study microvascular disease, either through indirect or direct assessment. The indirect methods look at the microvasculature of external and easily accessible structures, in which the findings can be correlated to multiple organ systems and disease processes. With indirect methods, however, care must be taken to extrapolate to other organ systems given differences in organ-specific microvascular physiology. Direct methods examine the microvascular system of the organ itself.

Despite the variety of methods available to study microvascular dysfunction, these techniques have largely not been incorporated into routine clinical practice for either screening or treatment purposes, because of the insufficient number of studies showing a benefit from the use of these techniques and the lack of evidence-based treatment guidelines for microvascular disease. In addition, the study of the heart’s microvasculature has shown to be extremely difficult, given certain inherent anatomic and complex physiological properties. Fortunately, there are some ongoing studies to address these deficiencies by exploring the diagnostic capabilities of these techniques in both experimental and clinical settings. This review will discuss techniques used to evaluate the microvasculature in different organ systems and disease processes, and the current research to determine the clinical utility of these techniques. If these tools can be effectively instituted in patient care to look for microvascular disease, there could be tremendous implications for prevention and/or earlier intervention of many pathologies, as endothelial dysfunction and microvascular disease often precede clinical illness (7).

INDIRECT METHODS

Skin

Many techniques have been developed to look at the microvasculature of skin, the body’s most superficial structure. These techniques include capillary microscopy, laser Doppler flowmetry (LDF), and transcutaneous oxygen measurements (Fig. 1). These methods provide valuable data about the reactivity and function of the skin microvasculature that have been correlated with diseases such as hypertension (HTN), chronic kidney disease (CKD), and peripheral vascular disease (PVD), and have been able to detect microvascular changes before patient disease is detected clinically (7). Also, given the noninvasive nature of these techniques, they are more accessible for real-life applications and will help elucidate important clinical questions more easily and cost-effectively.

Figure 1.

Indirect methods of assessing the microvasculature. Created with BioRender.com and published with permission.

Capillary microscopy is a technique that looks at the capillaries at the finger and toe nailfold, where the vessels are parallel to the skin surface. In contrast, capillaries elsewhere are perpendicular to the skin surface, so only the capillary density can be measured. The capillaries are then examined using a videomicroscope at ×100 magnification with blue or green light. From this, baseline capillary density, functional capillary recruitment and maximum capillary density can be measured as well (8). Its applications have been studied in HTN since the 1950s, and patients with HTN were found to have capillary rarefaction, abnormal vasoconstriction, and impaired capillary recruitment (9, 10). Then in the late 1990s and 2000s, work was published on the treatment of HTN and the effects on the capillaries observed through microscopy, but these studies were conflicting and results differed depending on the antihypertensive agent, but overall it was determined that capillary density increased with hypertension treatment (11–15).

The use of capillary microscopy in CKD is more recent. In CKD, capillary rarefaction and microvascular dysfunction were found to possibly precede end-organ dysfunction and can be independently correlated with the abnormalities in mineral metabolism, especially hyperphosphatemia, and vascular calcification (16, 17). Also, these microvasculature abnormalities appeared to evolve in different stages of CKD. In a study of pediatric patients, capillary rarefaction was strongly correlated with the severity of hyperparathyroidism and high serum phosphorus, but capillary functionality was preserved (17). In patients with advanced CKD, there was a loss of capillary functionality as well (16). Like HTN, routine screening has not been implemented to detect early capillary disease. More studies are needed to determine the benefit of treating preclinical abnormalities based on capillary abnormalities in this patient population.

In contrast, capillary microscopy has been used in dermatology practice for a while now for the diagnosis of many skin disorders including connective tissue disease (18). For example, nailfold microscopy is used to distinguish between primary and secondary Reynaud’s, and general practitioners are proficient in this technique (19). Current research in this field focuses on the microvascular damage present from autoimmune and connective tissue disorders and the effects of treatment, including correlations with distant organ involvement and disease progression, and the effects of anticoagulant therapy for antiphospholipid syndrome on microscopy findings (20, 21). This research will hopefully shed further light on the changes in the microvascular endothelium during these disease processes and subsequent treatment.

Finally, the relationship between peripheral microangiopathy identified on capillary microscopy and pulmonary arterial hypertension (PAH) is being studied. This relationship was first identified in patients with systemic sclerosis with resulting PAH (22). In a small prospective case-control study of patients with either idiopathic PAH or chronic thromboembolic PAH, nailfold capillary microscopy showed that these patients had a significantly reduced capillary density and increased capillary loop width compared with healthy controls (23). Future larger studies are needed to confirm these results and determine clinical utility.

LDF is another technique that looks at the tissue blood flow by measuring the Doppler shift of laser light from the moving red blood cells in the skin microvasculature to quantify the flowmotion. Flowmotion is a consequence of the constriction and relaxation of blood vessels, or vasomotion, and it ensures adequate tissue perfusion (24). Decreased flowmotion is found in obesity, HTN, peripheral arterial disease, diabetes, CKD, and hypercholesterolemia (25). Functional tests can be done as well, such as postocclusion hyperemia to determine maximal response capacity and reactivity to acetylcholine (endothelium-dependent vasodilator) or sodium nitroprusside (endothelium-independent vasodilator). These tests have been found to be very sensitive and are able to identify reductions in capillary response in adults with obesity, HTN, poor glucose control, diabetes, and advanced CKD (1).

LDF is able to not only detect microvascular changes in established disease but also detect early changes when there is an increased risk for cardiovascular disease. A cross-sectional study in the 2000s demonstrated that in patients with end-stage renal disease, abnormal LDF parameters were correlated with increased cardiovascular mortality (26). More recently, in a retrospective study of hemodialysis patients, LDF was able to detect early-stage PVD (27). A pilot study in 2021 comparing healthy patients with patients with risk factors for cardiovascular disease demonstrated that reduced reactivity found on LDF may be used as a biomarker for cardiovascular disease and a potential risk factor for cardiovascular events (28). And to further demonstrate the sensitivity of this technique, after only 7 days of a high-salt diet in young healthy normotensive patients, LDF demonstrated reduced microvascular reactivity (29). However, there are conflicting data showing some limitations of LDF; a retrospective study showed that microvascular changes were found in patients with stage 2 but not stage 1 HTN (30).

LDF has also potential applications in surgical patients. LDF has been used to evaluate vascular patency, and both preclinical and investigative studies have been done. In a rat model, simultaneous LDF and indocyanine green-videoangiography have been used to successfully show vascular stenosis in free flaps (31). Also, a prospective trial of 42 patients who underwent tibial bypass grafts and serial LDF measurements showed that patients who experienced graft occlusion had significantly poorer microperfusion postoperatively, demonstrating that LDF could be used to determine the risk for future graft occlusion and the need for preventative therapies (32). Further research investigating the utility of LDF in these approaches is needed to firmly establish this modality in current clinical practice.

Also, LDF has been used in conjunction with skin microdialysis, a technique that infuses substances into the superficial tissue via a thin tubular dialysis membrane to induce a microvascular response, which then can be measured by LDF (33). With this combination of both LDF and skin microdialysis, physiological changes in response to active agents can be easily studied in vivo and safely in human patients, and precise mechanisms can be determined. One prospective study of 10 hypertensive adults was able to quantify endothelial-related changes in response to hypertension treatment and uncover mechanisms of action for the antihypertensive agent. It was found that captopril (a sulfhydryl-donating angiotensin-converting enzyme inhibitor) improved acetylcholine-induced vasodilation and increased hydrogen-sulfide-related vasodilation. This study demonstrated that captopril was able to restore cutaneous microvascular reactivity in adults with hypertension, possibly through hydrogen-sulfide-dependent mechanisms (34). Second, in a study of healthy young adults who had recovered from mild to moderate COVID-19, LDF with skin microdialysis demonstrated that there was no persistent COVID-19-related nitric oxide-mediated vascular dysfunction (35). Finally, the cutaneous microvascular reactivity was studied in young non-Hispanic black women, a population in which there is a high prevalence of cardiovascular disease and microvascular dysfunction, but is rarely ever the focus of clinical studies. With LDF and skin microdialysis, it was found that the reduced microvascular function in these women may be related to endothelin-1 (36). More studies are needed to clarify these mechanisms of vascular dysfunction in this population, but this preliminary study is valuable, especially to close the gender and racial disparities in scientific research.

Finally, transcutaneous oxygen monitoring (TcPo₂) is a simple noninvasive technique to obtain the partial pressure of oxygen at the skin surface. TcPo₂ monitoring is done by placing an electrode on the clean skin of a supine or semirecumbent patient, and the TcPo₂ value is obtained on the monitor. It has been used in premature infants, adults in the intensive care unit, for wound assessments in wound clinics, and in the determination of amputation levels in PVD. In PVD, this technique has been studied since the 1980s, but has not been a part of routine practice (37). Part of this has been due to the lack of standardization of TcPo₂ protocols and consequently conflicting data (38). Environmental factors such as room temperature, dermal conditions such as hardened skin or inadequate skin prep, physiological factors such as decreased cardiac output or hypoxia, and mechanical factors such as monitor calibration or improper electrode application all affect the TcPo₂ measurements (39).

Recently, there has been an increase in studies implementing TcPo₂ as a part of the evaluation of lower extremity critical limb ischemia. Currently, the main methods for evaluation of vascular disease (ankle-brachial indices and angiography) look primarily at macrovascular disease. Only toe blood pressures are used, mostly in diabetic patients, to evaluate the degree of microvascular ischemia. Toe blood pressures used in conjunction with TcPo₂ could further aid in the classification of limb ischemia to decrease the risk of underestimation of the severity of distal limb ischemia and could identify patients without diabetes with microvascular dysfunction, where microvascular dysfunction is just as common (40). A feasibility study using photooptical oxygen tension measurement (pTcPo₂), a more practical variation of the standard TcPo₂, was used in 10 patients undergoing revascularization. This study checked for concordance between pTcPo₂ measurements, ankle/brachial indices, and angiography. They found that pTcPo₂ was certainly a feasible tool in the operating room with a strong correlation with the before and after ankle-brachial indices (41). In addition, a paramagnetic skin adhesive film has been developed to facilitate easy and reproducible TcPo₂ measurements and the preliminary results show that this superficial perfusion oxygen tension chip is reliable with reproducible results (42). Hopefully, with these tools and new research available, TcPo₂ could be a valuable data point in the complex decision-making needed for PVD, but protocols will need to be standardized before it becomes a part of regular clinical application.

Retina

Retinal imaging has been used for the diagnosis and examination of eye diseases for the past century (Fig. 1). With this technique, both the structure and function of the retinal blood vessels can be examined, including arteriolar vessel diameters, response to vasodilators and constrictors, and HTN-related changes in the vessel wall. It is well established that cardiovascular pathologies can be gleamed from retinal microvascular studies. HTN is associated with reduced arteriolar diameters, and interventions such as fish oil consumption, blood pressure treatment, and lifestyle changes can normalize arteriolar diameters (1). The data on renal function are not as strong but have demonstrated that smaller retinal arteriolar diameters are associated with CKD, and patients with smaller diameters were more likely to have renal function loss or start dialysis (43).

A newer area of study is the use of retinal imaging as a marker of cerebrovascular and neurodegenerative health. Currently, magnetic resonance imaging (MRI) is the most common imaging to assess minute pathology in the brain, where risk factors for dementia and stroke (white matter lesions, lacunar infarcts, microbleeds, and atrophy) can be seen. However, MRI is not capable of visualizing minute microvascular changes, thus it may not be as effective in the diagnosis of early disease. The retina has embryological and physiological (i.e., barrier function, autoregulation, low-flow, and high-oxygen extraction) similarities with the brain, and the study of its vasculature could shed light on the health of the microvasculature of the brain (44–46). There is an ongoing Phase II clinical trial to assess whether the vasodilator medication cilostazol will effectively slow the progression of white matter dementia and induce changes in retinal vasoreactivity in patients with small vessel cerebral vessel disease (47). If this trial proves successful, it could be an important step toward both early prevention, diagnosis, and treatment of dementia and stroke.

Sublingual Mucosa

Sublingual microcirculation evaluation is typically done at bedside with a side-stream darkfield camera, which uses green light to detect the passing red blood cells (Fig. 1). This modality has been used primarily in the critically ill population with sepsis, a condition that can lead to multiorgan dysfunction and/or failure because of derangements in the microvasculature with increased endothelial permeability, endothelial-leukocyte adhesion, and heterogeneity in blood flow with associated tissue hypoxia (48).

This imaging technique has been used to study the microvascular response in patients with sepsis undergoing early goal-directed therapy for septic shock or with evidence of end-organ dysfunction. A prospective observational study found that increased microcirculatory flow during resuscitation was associated with decreased rate of organ failure at 24 h, even though hemodynamic parameters were not significantly changed (49). There are two ongoing prospective observational longitudinal studies to look at endothelial glycocalyx damage in septic patients and their sublingual microcirculation, in both the emergency department (Early Detection of Glycocalyx Damage in the Emergency Room Patients) and intensive care unit (Analysis of Sublingual Glycocalyx Damage at Intensive Care Unit Admission to Predict Risk of Death) (50, 51). The glycocalyx of the endothelium is a protective layer within the lumen of the vessel that plays a critical role in microcirculatory homeostasis, including the reduction of endothelium-leukocyte interactions, decreased endothelial permeability, and regulation of the shear-stress mechanisms (52–54). With these data, it is possible to hypothesize that goal-directed therapy with microcirculatory flow in mind could help prevent organ failure and become a key part of clinical practice.

Specialized In Vitro Techniques

Microvessels-on-a-chip is a fascinating technology that simulates microvasculature in vitro, and recent advancements have been made to assess the microvessel permeability or microvascular exchange by using avidin-biotin complex formation. The microvessels are perfused with fluorescent-labeled avidin. The avidin molecules that pass through the vessel endothelium bind to biotin, which can be recorded with confocal imaging. A baseline permeability coefficient can be determined, and interventions on the microvessels can be performed to monitor for changes in the permeability coefficient (55). With this new capability, the microvessels-on-a-chip becomes a better physiological model with more similarities to in vivo experiments and can be used to simulate preclinical studies in microvascular research.

DIRECT METHODS

Neurological

MRI is the most sensitive and commonly used noninvasive method to image the brain and can be used to look for and follow cerebral microvascular lesions and neurodegeneration loads to predict dementia and cognitive decline (56). The Atherosclerosis Risk in Communities Study looked at the risk of stroke in the setting of brain microvascular disease progression; this was a prospective study of 907 stroke-free participants who underwent serial brain MRI and the patients were followed for stroke incidence for 4 years. It was found that participants with increased progression of microvascular brain disease on MRI (white matter hyperintensities/lacunar brain infarcts) had an increased risk of stroke (3). Based on these findings, future studies should look at both early identification and intervention in patients at high risk of stroke. However, the main limitation of MRI is that it may capture microvascular disease only when it is more advanced.

In contrast, ultrahigh-resolution computed tomography imaging is an advanced type of computed tomography (CT) with higher resolution obtained via a 0.25-mm detector, in which 40-μm vessels (within arteriolar size) have been clearly visualized. With this modality and the use of long-circulating CT contrast agents, the complete rodent cerebrovasculature was able to be imaged in vivo and signs of cerebrovascular disease were able to be detected in a rodent model of human cerebrovascular arteriopathy. Minute changes impossible to be seen using the usual histological methods, such as the width of the Circle of Willis, straightening of cerebral arteries and arterial stenosis, were detected on ultrahigh resolution CT (57). Another study used this technology to image the lenticulostriate arteries of 13 patients with aneurysmal disease, and compared with conventional CT, significantly more lenticulostriate arteries were able to be identified (58). With the ability to better visualize the cerebrovasculature in vivo, easily, and noninvasively, this imaging technique will facilitate future longitudinal studies that aim to develop therapeutic agents for cerebral microvascular disease.

More invasive methods of the brain’s microcirculation have been studied in animal models to look at cerebromicrovascular disease as they allow for elucidation of precise mechanisms that are not able to be visualized by imaging alone. One such study investigated the importance of the adaptor protein p66Shc, a protein that regulates levels of reactive oxygen species, on the middle cerebral artery of transgenic rats with several types of mutations in p66Shc as well as varied genetic backgrounds. The rats had hypertension induced with a nitric oxide inhibitor for 4 wk, and afterward the middle cerebral artery was harvested and placed on microglass cannulas within an organ bath. From this, changes in vessel diameter and myogenic tone were measured, and it was determined that p66Shc had a functional impact on myogenic tone in cerebrovascular function in a low renin model of hypertension though not other phenotypes (59). This study demonstrated the complexity of the cerebrovascular regulation and may help guide future personalized medicine in this field.

To address the need for longitudinal studies on age-related microvascular disruption and vascular cognitive impairment, a model of aged mice with a chronic cranial window was developed to allow for intravital two-photon imaging and optical coherence tomography. With these imaging techniques, both cortical microvascular density and the blood-brain barrier integrity (via fluorescent tracer dye) measurements were taken over 36 wk, and it was demonstrated that microvascular rarefaction and increased blood-brain barrier permeability were present in aged mice (60). With this innovative technology, the same area of brain can be monitored in longitudinal studies to evaluate the effects of targeted therapeutics in preclinical studies.

Cardiac

Myocardial blood flow (MBF) is complex, heterogeneous, and tightly regulated, and there are intrinsic difficulties in imaging and directly studying the microvasculature of the heart given the constant movement of the heart and the low transparency of the tissue. Thus, there is currently no way to directly visualize the human coronary microcirculation in vivo. Given these limitations, assessments of the coronary vasculature are largely assessed by quantifying the MBF or the coronary flow reserve (CFR) and measuring coronary reactivity to acetylcholine (61). The in vivo imaging technology available is improving, and there are ex vivo techniques to study the more minute physiological properties of the coronary microvasculature.

First, it is important to discuss the structure and function of the components of the coronary vasculature. The epicardial arteries, such as the right and left coronary arteries, serve as the conductance vessels; they provide minimal resistance, have a diameter greater than 400 µm, and can be seen on standard imaging such as coronary angiography. However, they make up only a small portion of the coronary vasculature. The microvasculature is composed of the prearterioles, arterioles, capillaries, postcapillary venules, and venules and has key regulatory roles of the MBF (Fig. 2). The prearterioles are 100–400 µm in diameter and along these vessels, the pressure drops significantly. The arterioles are 10–100 µm; they respond to myocardial oxygen demand by dilation/constriction and contribute to most of the coronary vascular resistance. The capillaries, the smallest vessel (<10 µm), are made up of a single layer of endothelial cells and a thin basement membrane, allowing it to be the primary site of gas, solute, and nutrient exchange (62). Also, some of the capillary beds lie dormant and can respond to changes in the coronary circulation, such as metabolic demand, pharmacological agents, ischemia, and reperfusion (63). The capillaries drain into the postcapillary venules (20–50 µm), which are the site of most reactive inflammation and have focally open intercellular junctions that allow for plasma protein permeability and leukocyte diapedesis (64–66). The venules or muscular venules (40–120 µm) drain the postcapillary venules and have smooth muscle (unlike the capillaries and postcapillary venules), but with less than that of arterioles (65, 67). Both the postcapillary venules and muscular venules serve as capacitance vessels given the passive and distensible nature of the venule wall.

Figure 2.

Anatomy and physiology of the coronary microvasculature. A: terminal arterioles lead into capillaries, which then drain into postcapillary venules and venules. Arterioles can constrict or dilate to direct the blood flow to capillary beds. Capillaries are the main site of gas, solute, and nutrient exchange. Postcapillary venules make up the portion of the microvasculature that is the most reactive to inflammation and permeable to both plasma proteins and leukocytes. Venules, or muscular venules, have smooth muscle fibers present, though not as much as that of the arterioles. B: summary of the size and function of each vessel type found in the microvasculature. *Vessel diameters are approximate values and vary between publications and literature sources. Created with BioRender.com and published with permission.

Second, there are multiple mechanisms to regulate the coronary blood flow (CBF) so that the flow meets the myocardial oxygen demand. At rest, already 75% of the oxygen from the blood is extracted; so if there is an increase in oxygen consumption, there is little reserve to increase the oxygen extraction and therefore MBF must increase by changes in the coronary vascular resistance (68). In addition, most of the regulation of the CBF is intrinsic to the heart and is tightly linked to cardiac metabolism, allowing the CBF to remain constant over a wide range of coronary perfusion pressures (68, 69). The change in vascular resistance is primarily accomplished via the change in diameter of the prearterioles and arterioles, which is thought to happen by many mechanisms, including myogenic control, flow-mediated dilation/constriction, metabolic, neural, and endothelial-dependent pathways (68, 70–77). In the setting of myocardial ischemia or extremely low myocardial Po₂, Po₂ coupled with adenosine appears to play an active role in increasing myocardial perfusion by arteriolar dilation, where low myocardial Po₂ induces adenosine (potent vasodilator) release (69, 78–83).

Coronary microvascular disease (CMD) occurs when there are abnormalities in the microcirculation that result in decreased dilation/constriction responses, causing an oxygen supply/demand mismatch and myocardial ischemia. In contrast to traditional coronary artery disease (CAD), CMD may have a nonobstructive pattern seen on coronary angiography, which is also called nonobstructive coronary artery disease (NOCAD). CMD in patients with angina and NOCAD appears to affect more women than men, and these patients have an increased risk of major adverse cardiovascular events and may have decreased intramyocardial blood volume (68, 84).

There are several noninvasive imaging modalities that quantify the MBF to characterize the heart’s microcirculation; they are positron emission tomography (PET), single-photon emission computerized tomography (SPECT), CT, MRI, and myocardial contrast echocardiography (68). PET is the most commonly used of these modalities. Coronary flow reserve (CFR) is another important parameter that can be quantified, and it is a measure of capacity to increase blood flow to maximal capacity from baseline coronary flow. However, a decreased CFR could indicate either epicardial artery stenosis in CAD or microvascular dysfunction in the setting of normal angiography.

PET is used with radiotracers such as ¹⁵O-water, ¹³N-ammonia, and rubidium-82, and compartmental modeling of radiotracer kinetics is used to quantify the MBF (Fig. 3). The ideal radiotracer has a measureable concentration that increases linearly with coronary flow over a wide range and has high first-pass extraction without significant recirculation and low extracardiac uptake. ¹⁵O-water fulfils all these requirements but is limited by its short half-life, leading to images that cannot be subjected to visual analysis. MBF is determined through radiotracer extraction from the arterial blood into the myocardium, or from the blood into the tissue compartment. With PET, the rate of uptake describing the transfer from blood to tissue is calculated to estimate the MBF (68). PET however is relatively high cost and so has limited clinical use, but the software package for the kinetic modeling is widely available.

Figure 3.

Compartmental modeling of radiotracer kinetics to calculate myocardial blood flow (MBF) in positron emission tomography (PET). There are two compartments depicted here: the blood compartment and the tissue compartment. The radiotracer is extracted from the blood to the tissue, and the rate at which this happens is k1. The rate that the radiotracer is washed out back into the blood is k2, and with early imaging, it is assumed that tracer washout and tracer metabolism is negligible. Myocardial blood flow (MBF) is calculated by k1 multiplied by the tracer extraction (E).

One major limitation of using MBF to characterize microvascular disease is that stenotic epicardial disease can result in a low MBF, and MBF is a value that represents the global perfusion of the heart, which can be difficult to apply to local areas of microvascular disease given the heterogeneity of the cardiac circulation. Oxygen-sensitive cardiovascular magnetic resonance (CMR) can noninvasively image myocardial oxygenation by using the magnetic properties of deoxygenated hemoglobin, and therefore can reflect the oxygenation of the blood. A decrease in signal of T2*-weighted images is found with an increase of deoxygenated hemoglobin, and an in increase in signal in T2*-weighted images is found with a decrease of deoxygenated hemoglobin (85). This modality has been demonstrated in both animal and human studies in effectively assess myocardial oxygenation independent of blood flow (85–88). In a recent exploratory study to apply this technology to microvascular disease, women ages 40–65 yr old with ischemic chest discomfort without obstructive coronary artery disease (INOCA) and healthy volunteers were imaged with CMR to evaluate global and regional oxygenation variability. It was found that women with INOCA had increased regional heterogeneity compared with healthy controls, though there were no significant differences in global changes in oxygenation, or CMR-measured heart volume, function, or mass; this imaging modality may become a useful tool in the future for diagnosis in patients with INOCA (89).

To see the vessels themselves, ultrahigh-resolution CT, as mentioned before for brain imaging, has been applied to the cardiovascular system as well. In a rodent model of coronary ischemia-reperfusion, the infarcted area was imaged with ultrahigh resolution SPECT/CT and was compared with high-resolution autoradiography. This study demonstrated that SPECT/CT could evaluate the severity of the infarction with similar efficacy compared with autoradiography, a tracer deposition method (90). In addition, visualization of the coronary arteries is improved with this technology: a study of 79 patients underwent high-resolution CT, and of those 59 underwent invasive coronary angiography and 19 underwent conventional CT. The degree of luminal stenosis was graded, and it was found that ultrahigh-resolution CT had better image quality than conventional CT, was able to visualize smaller calcified plaques, and luminal stenosis was graded more precisely (91). These capabilities will hopefully evolve further and allow for improved noninvasive diagnosis and prognostication.

Novel ex vivo techniques have been developed to study the microvasculature from a more microscopic view. MicroangioCT is an ex vivo imaging technique that uses a polymer-based contrast agent, μAngiofil, and allows for three-dimensional visualization of the vasculature, including the capillaries (92). The major limitation of this technique and ex vivo techniques, in general, is the inability to use this in longitudinal studies.

Another method available uses human coronary microvessels from living patients in experimental settings; from a portion of the right atrial appendage, skeletal muscle or other microvessels can be harvested during cardiovascular surgery before and after cardiopulmonary bypass (CPB) or during other operations. Arterioles can be dissected from the tissue and secured on glass microcannulas and pressurized in a no-flow state in an organ bath (93). With this method, electronic dimension analysis can be used to monitor arterial dimensions (dilation or vasoconstriction), and the response to different drugs, neurohumoral agents, stimuli, and conditions can be directly assessed (93–98). This technique appears to be valuable in assessing the effects of diabetes, hypertension, hypercholesterolemia, or other clinical conditions on microvascular function directly such that autoregulation and other factors do not interfere with the assessment (95, 97–99). Receptor blockers such as β-blockers and ion channel blockers can be added to the organ chamber to more carefully examine specific pathways of altered vasomotor regulation (100, 101). Also, at least in the case of cardiac surgery or other clinical interventions, the effects of myocardial ischemia, cardioplegia or cardiopulmonary bypass can be assessed (95, 97, 98, 101). The limiting factor of using this method is that the influence of the surrounding tissue, autoregulation, and metabolic influences are not taken into account. Furthermore, human patients are a diverse group of subjects with various medical conditions taking a variety of medications. These factors may make the assessment and comparison of the microvascular change problematic, as opposed to the assessment of microvascular function in a homogeneous cohort of transgenic mice. However, all things considered, the direct in vitro assessment of the human coronary, skeletal muscle, or other organ microcirculation is a very valuable tool in the overall investigation of microvascular function.

Altogether, there have been many techniques developed to study the coronary microvasculature, but there are still no strong evidence-based guidelines for the screening or treatment of CMD. If CMD is diagnosed, the treatment currently recommended by the European Society of Cardiology is based on data not quite relevant to CMD, does not treat the primary pathophysiology, and consists only of reassurance, symptom relief, and the traditional antianginal therapy (nitrate, β-blockers, or calcium channel blocker) (102). Fortunately, there have been ongoing studies hoping to address this deficiency.

The MICORDIS Study is an ongoing single-center observational cross-sectional study that aims to see if myocardial blood volume is decreased in NOCAD at baseline, hyperinsulinemia, or stress. Patient enrollment will include patients with long-standing angina with NOCAD and inadequate response to medical therapy, and these patients will be compared with healthy age-matched controls (103). Another study looked at ranolazine, a chronic angina medication that has been investigated for the treatment of coronary microvascular dysfunction. A systemic review and meta-analysis of the randomized trials were performed and showed that ranolazine may cause an increase in the CFR and improvement in some quality of life measurements, but it does not seem to impact the frequency of symptoms or treatment satisfaction. Improvement in prognosis is not known at this time (104). Also, coronary microvascular dysfunction distal to epicardial stenosis has become an increasingly important topic in recent years; though clinical trials have not been successful thus far, animal models testing new therapeutic strategies are promising and focus on improving lipid and myocardial metabolism and reducing oxidant stress (105). These kinds of studies will hopefully soon fill the current gaps in knowledge in cardiology practice for CMD.

Pulmonary

The pulmonary system is the largest microvascular network in the body, and it receives the entirety of the cardiac output from the right heart. Significant microvascular diseases of the lung are acute respiratory distress syndrome (ARDS) and coronavirus disease-19 (COVID-19).

ARDS is thought to be from microvascular inflammation and can be caused from a variety of conditions, including sepsis, acute pancreatitis, and aspiration, which trigger an inflammatory response. Plain chest radiographs and CT scans are more commonly performed in ARDS, and advanced imaging is not routinely done. PET, however, is a functional imaging technique that has been increasingly used to look at structure and function relations and metabolic and physiological processes, including the quantification of regional perfusion, vascular permeability, endothelial receptor and enzyme function, and the metabolic activity of inflammatory cells (106, 107). PET is able to study all of these processes depending on the radiotracer given; pulmonary vascular permeability can be measured with ⁶⁸Ga-transferrin or ¹¹C-methylalbumin, endothelial binding can be checked with radiotracer ligands such as ¹¹CGP-12177 (for the β-adrenergic receptor system) and ¹⁸F-fluorocaptopril (for the angiotensin-converting enzyme system), and inflammation can be measured with 2-[¹⁸F]fluoro-2-deoxy-d-glucose (as activated neutrophils have high glucose metabolism) (106, 108). Although PET can provide a lot of valuable information, it is unlikely that it will be a part of routine practice because of its cost and low availability. Most of its value may be in clinical studies to determine the efficacy of treatments targeted toward the pulmonary microvascular endothelium.

COVID-19, a widely known respiratory illness, leads to a microvascular angiopathy. This is a sequelae of severe inflammation from the overactivated host immune system and the release of proinflammatory mediators that affect the endothelium to cause acute cellular dysfunction, loss of endothelial barrier function, increased vascular permeability, leukocyte infiltration, and higher risk of microvascular thrombosis (Fig. 4) (5, 109). COVID-19 also infects the host using the angiotensin-converting enzyme 2 receptor, which is expressed on vascular endothelial cells as well as many other organs in the body. In autopsies of patients who succumbed to COVID-19, pathology demonstrated the presence of many inflammatory cells associated with the endothelium, and viral particles within the vascular endothelial cells, consistent with endotheliitis (110).

Figure 4.

Lung endothelial dysfunction in coronavirus disease-19 (COVID-19). In physiological conditions, nitric oxide (NO) levels are adequate to mediate vasodilation, prevent platelet aggregation, and dampen inflammatory response. There are intact endothelial cell (EC) tight junctions and a protective layer of pericytes around the vasculature. In COVID-19, there is acute endothelial dysfunction: ECs produce large amounts of chemoattractants, cytokines, and adhesion molecules. This leads to leukocyte activation, adhesion, migration through the disrupted tight junctions, and increased vascular permeability. There is also an insufficient production of NO, with resulting increased reactive oxygen species (ROS) levels and vasoconstriction. Viral particles can infect endothelial cells via the angiotensin-converting enzyme 2 (ACE-2) receptor. D-dimer (protein fragments of dissolved blood clots) levels have also been found to be elevated in severe COVID-19 and are associated with worse prognosis.

Dual-energy CT (DECT) has been used to study the degree of microvascular involvement in vivo. DECT is also known as spectral CT, and it uses two separate X-ray photon energy spectra so that the tissue/lesion characterization is superior. In a retrospective study of 85 patients with severe COVID-19 who underwent DECT for suspicion of pulmonary artery thrombosis, it was found that while only 34% of patients were diagnosed with pulmonary artery thrombosis, 68% had parenchymal ischemia (111). This suggests that there may be microthrombosis contributing to the patients who had parenchymal ischemia without detectable pulmonary artery thrombosis. In another retrospective study of patients hospitalized for COVID-19 pneumonia, DECT was used to look at pulmonary microvascular involvement in patients and microvascular changes over time. These patients underwent DECT initially for clinically worsening symptoms or suspicion of pulmonary embolus. Five male patients were studied, all with an elevated D-dimer but no pulmonary embolus. There appeared to be two different radiological patterns seen on DECT. The early clinical phase was associated with diffuse bilateral ground-glass opacities with associated increased perfusion, where hypoxemia in these patients may be due to pulmonary shunting (112). Intussusceptive angiogenesis, found in the lungs of deceased COVID-19 patients, was ∼2.7 times higher than lungs of patients with influenza; this may be contributive to the shunting effect (112, 113). The later phase was associated with bilateral alveolar consolidation and was associated with decreased perfusion in the affected lobes, which may be due to the microthrombi characteristic of COVID-19 (112). Thus, DECT will be a useful tool in studying the natural history of this rampant disease and may be used in monitoring future therapeutic interventions if endothelial-targeted novel therapies are developed.

Finally, in animal models, intravital microscopy (IVM) is a high-resolution technology that has been used to look at the intricacies of dynamic pulmonary microvasculature regulation in live tissue. This technique is performed through a thoracic window, where visual access is surgically gained and the animal is mechanically ventilated. A fluorescent plasma marker is injected so that the pulmonary arterioles can be better visualized, and vasoactive responses can be seen in response to exposure to different conditions or drugs (114). In a rodent study looking at the effects of type 1 diabetes on the pulmonary vasculature, decreased microvascular dilation was seen in the setting of decreased nitric oxide bioavailability, with associated endothelial injury (115). While using IVM, the actions and interactions of individual cells can be pictured using cell labeling strategies such as fluorescent probes/dyes, fluorescent antibodies, florescent reporter mice, and knockout mice. The movements of inflammatory cells have been tracked within the pulmonary microvasculature, and with this much has been learned about the behavior and movement of neutrophils in diseases such as pneumonia, viral infections, cystic fibrosis, and sepsis-induced lung injury (116). IVM has been used to study the effects of intravenous administration of mesenchymal stem cells in the pulmonary vasculature, including the cells’ kinetics and site of arrest in the vasculature. It was found that the cells were lodged at microvascular bifurcations, where they became deformed and appeared to release extracellular vesicles (117). With this fascinating technology, much can be uncovered about the intimate physiology of the microvasculature and can guide therapeutic development. The limitations of this technique are its ability to only visualize the vessels of the organ surface (which are not necessarily representative of the inner vessels), limited access to certain organ systems, the unknown effects of positive pressure ventilation, and the inability to do longitudinal studies given the invasiveness of the procedure.

Gastrointestinal

The gut blood supply originates from three major arteries that branch off the aorta: the celiac, superior mesenteric, and inferior mesenteric arteries. Between these vessels, there are many anastomoses to allow for continuous flow throughout the gut. The microvasculature of the gut within the intestinal villi is of particular interest; the arteriole and postcapillary venule run parallel to each other so that the tips of the villi receive the lowest amount of oxygen. This makes the villi tips physiologically hypoxic and the gut especially susceptible to ischemia (118). Also, the blood flow of the gut is regulated tightly. During stress, catecholamines are released and the vasculature constricts; after meals, there is postprandial hyperemia; and infections can result in acute inflammation. Novel techniques that have been used to assess the gut mucosa and its associated pathologies are confocal laser endomicroscopy (CLE) and the balloon-tipped transpyloric probe. In surgical procedures, dynamic fluorescence videoangiography has become an increasingly used to assess adequacy of blood supply to bowel anastomoses.

CLE is an advanced endoscopic technique that allows for real-time assessment of the gut mucosa at the subcellular level, with microscopic power comparable to that of white light microscopy with the option for further image enhancement with the use of a contrast agent. It has been used for numerous applications, including distinguishing nonneoplastic and neoplastic tissues, grading tumor vasculature, surveillance of nondysplastic Barrett’s esophagus and inflammatory bowel disease, and monitoring the effects of sepsis and treatment-related changes (119). For example, this modality has been used in colonoscopies for “in vivo histology” where the vessel and crypt architecture is graded to determine the presence of neoplasia. The accuracy of this method was acceptable, around 84%, but became much higher at 94% when all interpreters of the data agreed (120). Though this application is useful, it is very unlikely to replace pathological histology for final diagnosis and may be more useful for monitoring treatment response rather than diagnosis. CLE has been used to evaluate tumor angiogenesis in gastric and rectal cancer, where detailed vascular assessments are made (vessel shape and size, vessel permeability, flow), and to evaluate changes in the angiogenic score in response to cancer therapy (121). This assessment of the tumor vasculature can provide relatively noninvasive prognostic data, and both guide and personalize cancer therapy without needing histology of the tumor itself. Finally, CLE has been used in both porcine and human sepsis models to look for changes in the gut microcirculation and the response to fluid resuscitation, as a part of early goal-directed therapy. The functional capillary density appeared to be markedly decreased in sepsis, but with volume therapy, there was a significant improvement in the capillary density, showing the efficacy of the intervention (122). Like with the use of sublingual assessments of the microcirculation, CLE would be useful to determine the resuscitation endpoints of sepsis management. It is unknown how this would compare with measurements taken from the sublingual microcirculation.

The balloon-tipped transpyloric probe, developed very recently, is a balloon-tipped triluminal intestinal probe that is able to assess gut perfusion directly. This device consists of an enteral feeding tube along with a photoplethysmography (PPG) sensor that lies within the duodenal mucosa. The PPG sensor is a versatile technology that is not only able to detect blood volume changes in the microvascular bed but is also capable of assessing the vascular age of larger blood vessels; increases in arterial stiffness with age results in increased pulse wave velocity (123). With the balloon-tipped transpyloric tube, the PPG sensor looks at blood flow in both the arteries and veins and the signal measures the nonpulsatile and pulsatile components of the vasculature (124). While also establishing enteral access, there is continuous direct visualization of the gut mucosa, an area highly susceptible to ischemia in septic shock. In a porcine model of septic shock, the success of this technique has been demonstrated: 14 piglets were anesthetized and ventilated, and in half of the pigs, sepsis was induced with Pseudomonas aeruginosa. A balloon-tipped transpyloric probe was placed, and there was a significant decrease in the gut microcirculation and pulsatile part that correlated with the lactate value (125). In a human patient under advanced hemodynamic monitoring, a PPG probe was placed for both enteral access and gut monitoring, and improvements in the gut circulation were seen after a fluid challenge (126). Overall, this technology could help with decision making in the management of enteral feeding in septic shock, a historically controversial topic. However, larger clinical trials will be needed to fully establish its efficacy and role in clinical practice.

Finally, in general and colorectal surgical procedures, bowel anastomosis viability is one of the prime concerns, as inadequate blood flow can lead to an anastomosis breakdown or stomal necrosis. Dynamic fluorescence videoangiography with indocyanine green is already semiroutinely used by some surgeons in clinical practice. With this technology, the patient is given intravenous indocyanine green, and a fluorescent picture of the target area (e.g., anastomosis, stoma) is taken with a camera to evaluate the amount of fluorescent dye prevalent within the target area allowing for both a qualitative and quantitative analysis (127). There is now an ongoing clinical trial (Evaluation of Microcirculation in Colon Wall and Bowel Anastomosis by Laser-Induced Fluorescence Video Angiography) to assess the efficacy of this modality in patients with colon cancer undergoing resection, and this will hopefully more firmly establish the role of this technology (128).

Renal

The evaluation of the renal vasculature has largely been indirect, such as with functional methods (e.g., venous occlusion plethysmography, flow-mediated dilation, LDF) and biochemical markers to assess endothelial function (such as nitric oxide synthase inhibitors, cell adhesion molecules, coagulation pathway molecules, etc.). These methods, however, have not become a part of clinical practice as studies on novel noninvasive imaging of microvascular and endothelial dysfunction in patients with CKD are not commonly done (129). Endothelial dysfunction, decreased microvascular density, and increased cardiovascular risk are well established in CKD, but there is also evidence that endothelial dysfunction and chronic vascular inflammation are present even in the early stages of renal insufficiency, and it is associated with increased cardiovascular mortality (130). Thus, it has become apparent that microvascular dysfunction in the setting of early renovascular disease is important to address. Endothelial dysfunction is detected primarily by biomarker testing, partially because of many limitations of older imaging techniques, and it is unclear at this time what role imaging of the renal microvasculature plays in routine clinical practice despite the high prevalence of renal insufficiency. However, with development of superb microvascular imaging (SMI) for better visualization of the renal cortical microvasculature, the role of imaging may change in the management of renovascular disease.

SMI is an ultrasound image processing technique that results in finer characterization of the smaller vessels and less motion artifact. To further define the vasculature, intravascular microbubbles can be used. When compared with conventional color and power Doppler ultrasonography, SMI has appeared to be more sensitive in showing the renal cortical microvasculature with significantly increased color pixel intensity and decreased distance of cortical end vessel to the renal capsular surface (1.06 ± 0.43 mm) (131). A rodent model of acute renal ischemia demonstrated that just based on SMI images alone, it was possible for the assessors to qualitatively distinguish healthy renal microvasculature from the microvasculature postischemia (132). In a study of 47 patients with CKD stage 2–4, SMI was performed and it demonstrated a strong correlation between the SMI index and estimated glomerular filtration rate (133). With these capabilities, SMI may provide a way to monitor acute or chronic morphological changes in patients with renal dysfunction. This modality may have a place in future studies for the early diagnosis of kidney injury or CKD, or longitudinal studies following novel treatments for endothelial dysfunction.

SMI has a role in renal malignancy evaluation as well. Using the Bosniak classification of renal cystic masses, SMI has been able to correctly identify malignancy with improved evaluation of blood flow in the septa and solid structures compared to conventional color Doppler ultrasonography (134). Another method of microvascular assessment in the malignancy is through tumor microvascular density (MVD) assessment of pathological slides. In a study of 57 cases of clear cell renal carcinoma, after surgical removal the sections were immunostained for CD34 (marker of vascular endothelial progenitor cells) and vascular endothelial growth factor (VEGF). A computerized image analysis method was created to count CD34-positive cells and measure the intensity of VEGF staining. Increased adjusted MVD values of the malignant tissue (normalized by MVD values of normal tissue) were significantly associated with shorter disease-free survival (135). This technique could provide important prognostication of malignancies and potentially help tailor cancer medical therapies after surgical resection.

PREGNANCY AND MICROVASCULAR DISEASE ASSESSMENT

Microvascular complications can occur in pregnancy as well, such as placental vascular dysfunction with consequent preeclampsia or fetal growth restriction. Risk factors for these conditions are ones that are shared with cardiovascular disease, such maternal hypertension, hyperlipidemia, hyperglycemia, and obesity (6). Even when preeclampsia has resolved after pregnancy, there is a significantly increased risk of cardiovascular disease, hypertension, and stroke long-term (136–138). Also, subclinical insulin resistance can worsen to gestational diabetes, and in women with type 1 diabetes, there is a transient increased in the risk of retinopathy during pregnancy with the need for postpartum surveillance (139, 140).

Current screening for preeclampsia consists of a review of risk factors, blood pressure measurement at every prenatal visit, urine checks for proteinuria, and treatment with aspirin. With Doppler ultrasound assessments of the umbilical and uterine artery blood flow, signs of preeclampsia can be detected by changes in the waveform with an increased impedance in the fetoplacental vascular bed (141). However, this assessment has a low sensitivity for detecting preeclampsia or related pathologies, ranging from 15% to 75%, and can be affected by benign factors such as fetal movement or gestational age (142–144).

Recent research has been performed to search for better methods to assess placental health and vascular endothelial dysfunction following preeclampsia. One such study used a mouse model of fetal growth restriction to measure the reflected pressure waves that are propagated in the opposite direction of the flow of the umbilical artery, and the presence of these reflections was demonstrated to be sensitive in detecting placental vascular abnormalities and characterizing the fetoplacental tree (145). This represents an important step toward improved interpretation of hemodynamic measures and may be beneficial in future clinical screening. Another study looked at the microvascular reactivity in the forearm of women 6 mo to 5 yr postpartum from pregnancies complicated by preeclampsia. With laser speckle contrast imaging, iontophoresis, and tests for endothelial-dependent and endothelial-independent vasodilation, it was found that severe preeclampsia was associated with increased endothelial-dependent and -independent vasodilation compared with that of healthy controls (146). This finding sheds light on some of the possible mechanisms of vascular dysfunction after preeclampsia, and future research should be targeted to addressing the root causes of the microvascular changes to reduce the risk of maternal cardiovascular disease.

Conclusions

Many tools have been developed to study the microvasculature as microvascular disease has been increasingly recognized as a major player in very common and relevant disease processes. As of now, there largely are no screening guidelines to use imaging to assess for microvascular dysfunction, but fortunately there are a few ongoing studies to assess its clinical value. The information gleamed from these studies will be valuable for learning more about the natural history of these diseases, prognostication, inspiring therapeutic interventions, and longitudinal studies to assess efficacy of treatment. A major limitation is a need for therapies targeted toward endothelial dysfunction; there are none effective currently that is widely known or used. Once the microvascular pathologies can be targeted with effective treatments, it is likely that these novel diagnostic approaches will be further established in future clinical practice.

GRANTS

This study was funded by the National Institutes of Health (NIH) Grants 1R01HL133624 (to M.R.A.) and 1HL46716 and R01HL128831-01A1 (to F.W.S.), and C.X. was supported by NIH Grant T32 GM065085-10.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

C.X. and M.R.A. conceived and designed research; C.X. drafted manuscript; C.X., F.W.S., and M.R.A. edited and revised manuscript; C.X., F.W.S., and M.R.A. approved final version of manuscript.