Statistics from Altmetric.com
Regional myocardial blood flow can now be measured non-invasively with contrast echocardiography, CT, MRI, and positron emission tomography (PET). Measurements with PET have advanced furthest; their accuracy and reproducibility are well established, they can be and are employed in the clinical setting, and they are useful for probing the coronary circulation in humans. The ability to quantify myocardial blood flow rather than to only qualitatively evaluate its relative distribution expands current diagnostic capabilities. Flow measurements offer a means for functionally delineating more accurately the extent and severity of coronary artery disease and, importantly, for probing and measuring microcirculatory reactivity as a component of coronary circulatory function that thus far has escaped non-invasive examinations. This article briefly addresses methodological aspects of PET measurements of myocardial blood flow, and then examines how observations made with this new measurement tool relate to coronary circulatory function and how these measurements can uncover disease related alterations.
Methodological aspects of flow measurements
Several positron emitting radiotracers are available for PET measurements of myocardial blood flow. Oxygen-15 labelled water (O-15 water) is considered the ‘ideal flow tracer’ because its uptake in myocardium increases linearly with flow. Yet, its physical properties limit its clinical use: the 122 s physical half-life necessitates close coordination of clinical studies with the cyclotron production of the radiotracer. Further, the freely diffusible radiolabelled water equilibrates with the water space of the myocardium but also with those of blood and adjacent anatomical structures. The resulting low myocardium-to-blood activity ratio images require corrections for blood pool activity, either through additional labelled blood pool imaging or by mathematically separating the myocardial from blood pool radiotracer activity. In the clinical setting, most laboratories therefore prefer nitrogen-13 labelled ammonia (N-13 ammonia) or rubidium-82 chloride (Rb-82). Both are administered intravenously and promptly accumulate in myocardium while rapidly clearing from blood. The resulting high signal-to-noise myocardial activity images reflect the relative distribution of myocardial blood flow (figure 1).
N-13 ammonia, which is also cyclotron produced, is logistically easier to use than O-15 water because of its longer (9.97 min) physical half-life. Radiotracer administrations for rest and stress flow measurements can be repeated at 30–40 min intervals so that a rest-stress examination is completed within 60–80 min. Rb-82 is available through a strontium-82/Rb-82 generator system and does not require an on-site cyclotron. Its ultra-short physical half-life of only 1.27 min allows repeat imaging studies at 10–20 min intervals so that a stress-rest flow examination requires <1 h. Rb-82 is administered via a push-button operated delivery system; it elutes the radiotracer from the generator system and delivers radiotracer boluses at preselected activity doses and infusion rates. The generator can be eluted multiple times during the day for as long as 4–5 weeks, which is the clinical life of the generator.
Images of the myocardial N-13 ammonia or rubidium-82 uptake reflect the relative distribution of myocardial blood flow at the time of the radiotracer administration and are comparable to the standard single photon emission CT (SPECT) myocardial perfusion images. Estimates of myocardial blood flow are derived from the radiotracer tissue kinetics in myocardium and in arterial blood.1 2 The initial transit of the radiotracer bolus through the central circulation and its rate of radiotracer accumulation in the myocardium are recorded on serially acquired PET images. Image acquisition begins at the time of radiotracer administration and continues for 15–20 min with N-13 ammonia and for 9–10 min for Rb-82. After reorienting the transaxially acquired image data into short axis images of the myocardium, regions of interest are assigned to the left ventricular blood pool and myocardium. Time activity curves are obtained and reflect the arterial radiotracer input function and the myocardial tissue response. Operational equations, derived from tracer kinetic models that describe the radiotracer exchange between blood and myocardium as a function of flow, are applied to the time activity curves and yield estimates of myocardial blood flow in units of ml/min/g. These operational equations correct for the non-linear response of the myocardial radiotracer uptake to increasing flows so that the non-invasive flow estimates linearly track changes in myocardial blood flow.
With state-of-the art PET systems, all image data are acquired in list mode and are rebinned into serial image frames (as described above). Commercially available software analytical routines then reorient the image data and semi-automatically assign regions of interest to the left ventricular myocardium and the blood pool, extract the arterial radiotracer input function and the myocardial tissue response, and automatically display regional myocardial blood flows in ml/min/g. These software routines are similar to those employed for routine SPECT analysis of perfusion images so that flow estimates are available within 15–20 min after completion of data acquisition (figure 2).
Validated in animal experiments, the non-invasive flow estimates correlate accurately and linearly with estimates of flow by the invasive microsphere technique (as the ‘gold standard’ of flow measurements) from flows as low as 0.3 ml/min/g to as high as 6.0 ml/min/g.w1 w2 Flow estimates in normal volunteers at rest, during pharmacologic vasodilation or sympathetic stimulation are highly reproducible when repeated during the same study session or days or weeks later.3 w3–w6
Cardiac work as a determinant of myocardial blood flow
Myocardial blood flows in normal subjects and in patients with cardiovascular disease may differ considerably between individuals at rest. This variability largely reflects inter-individual differences in cardiac work—that is, in heart rate, systolic blood pressure and, thus, the rate–pressure product, which is an index of cardiac work. Accordingly, myocardial blood flows correlate linearly with the rate–pressure product4 at rest as well as during exercise stress or dobutamine stimulation (figure 3).
Several mechanisms regulate the flow response to changes in cardiac work and, thus, in metabolic needs of the myocardium.6 7 Flow across the myocardium is largely driven by the pressure gradient between the aortic root and the right atrium (coronary inflow and outflow), but is opposed by resistive forces. Most of the resistance resides at the level of the coronary microcirculation—that is, the pre-arterioles (200–500 μm diameter) and the arterioles (≤200 μm diameter). Vascular mechanisms regulate the resistance and adjust flows to the myocardium's metabolic needs. An increase in energy demand leads to a metabolically mediated vascular smooth muscle relaxation of the small arterioles (≤50 μm diameter), which in turn prompts a pressure related and myogenically controlled dilation of the immediate upstream arterioles. As resistance declines, flow velocities rise and cause a shear-stress mediated activation of the endothelial nitric oxide synthase (eNOS). This raises the bioavailability of nitric oxide (NO) and leads to further vascular smooth muscle relaxation and lower resistance to flow (figure 4).
The close interaction between endothelium and vascular smooth muscle adjusts and fine tunes rates of flow and thus of substrate delivery to changes in the myocardium's metabolic substrate needs. This interaction between endothelium related vasodilator and vascular smooth muscle related constrictive forces plays an even greater role during sympathetic stress and physical exercise. Local norepinephrine (NE) release in response to sympathetic activation leads to an α-adrenoreceptor stimulated vascular smooth muscle contraction and, thus, vasoconstriction that is normally offset by greater NO release from the endothelium, mediated by NE stimulation of endothelial adrenoreceptors and by shear-stress related increases in eNOS activity.
This ‘flow modulatory system’ also operates in the large epicardial coronary arteries. Under basal conditions, the conduit vessels exert little if any resistance to flow. Higher flow velocities in response to downstream reductions in microvascular resistance are met with greater resistance in the conduit vessel, due to the velocity of flow and, importantly, the vessel lumen. Through shear-stress mediated mechanisms, the vascular smooth muscle–endothelial interactive system recalibrates the diameter of the conduit vessel to the higher flow velocity, thereby minimising the resistance to flow. This ‘flow mediated dilation’ of the epicardial coronary arteries can be demonstrated on quantitative coronary angiography and serves as a measure of endothelial function.
Measurements of coronary responsiveness
PET based measurements of myocardial blood flow offer a non-invasive though quantitative tool for identifying disease related abnormalities of coronary function. At rest, myocardial blood flows even in patients with advanced non-coronary cardiac disease are almost invariably maintained. Flows at rest in patients with idiopathic dilated cardiomyopathy or with hypertrophic cardiomyopathy were found to be similar to those in normal volunteers.8–10 Detection of alterations in coronary circulatory function therefore requires measurements of flow responses and, thus, of the coronary responsiveness to well defined stimuli, including: (1) pharmacologic vasodilation; and (2) sympathetic stimulation through cold pressor testing.
Pharmacological vasodilation and integrated total vasodilator capacity
Vasodilator agents such as adenosine or dipyridamole cause vascular smooth muscle relaxation and maximally reduce coronary resistance. The resulting flow increase is thought to reflect predominantly vascular smooth muscle function. Flows achieved during hyperaemia then depend largely on the coronary driving pressure and on extravascular resistive forces. For example, myocardial blood flow may increase with higher aortic pressures when supplementing pharmacologic vasodilation with exercise stress. Yet, such potential flow increases can be opposed by higher extravascular compressive forces, related to increases in end-diastolic filling pressures, myocardial wall stress, contractility, and higher heart rates. Exercise supplementation of pharmacologic vasodilation may therefore fail to augment hyperaemic flows.w7 Whether such supplementation diminishes the diagnostic accuracy of stress perfusion imaging remains unknown, yet it may be offset by the effects of sympathetic activation during exercise with possible vasoconstrictive effects.
Apart from the coronary driving pressure and extravascular resistive forces, the endothelium contributes to the pharmacologically induced hyperaemia. For example, inhibition of eNOS activity with L-NAME in young normal subjects reduced hyperaemic flows by 21%,11 suggesting that endothelium related factors contribute as much as 30% to the adenosine induced flow response. This implies that pharmacologically stimulated hyperaemic flows do not selectively reflect the vascular smooth muscle dilatory function, but rather the combined response of vascular smooth muscle and endothelium, and thus the ‘total integrated coronary vasodilator capacity’. In normal individuals, vasodilator stress raises myocardial blood flow by as much as 300–600%. Accordingly, myocardial flow reserves as the ratio of hyperaemic to baseline blood flows range from 4–6.w8
Altered flow responsiveness in epicardial coronary artery disease
Consistent with stress induced defects on conventional myocardial perfusion images, fluid-dynamically significant coronary stenoses attenuate the flow response to pharmacologic vasodilation in the downstream myocardium. The degree of the attenuation depends on the functional severity of the upstream coronary stenoses; 50–70% diameter stenoses attenuate flow responses only moderately, whereas >85–90% stenoses severely compromise the flow response so that regional flow reserves may decline to <1.5.w9 12
Standard stress-rest myocardial perfusion imaging accurately detects coronary artery disease, but is limited in patients with balanced coronary artery disease or in identifying triple vessel disease. Balanced coronary artery disease (ie, left main and/or proximal left anterior descending and left circumflex coronary stenoses) can be associated with global reductions in myocardial flow reserves without inducing regional flow defects. Moreover, when all three myocardial coronary territories are subtended by vessels with disease, though different in severity, regions with the highest radiotracer activity on stress myocardial perfusion images serve as reference and are usually considered ‘normal’ (figure 5). However, quantitatively diminished flow responses or flow reserves in such ‘normal myocardium’ identify the presence of upstream coronary stenosis. Accordingly, though examined in only a small number of patients, standard myocardial perfusion imaging accurately detected triple vessel coronary artery disease in <50% of patients, whereas diminished flow responses by quantitative flow measurements correctly identified 92% of all patients with triple vessel disease.13
Even more challenging is the detection of transplant vasculopathy in patients with cardiac allografts. In most patients, the disease diffusely affects the coronary circulation so that myocardial perfusion is homogeneous at rest and during hyperaemic stress. Accordingly, the perfusion images characteristically demonstrate homogenous tracer distributions throughout the left ventricular myocardium and, thus, are considered normal. Importantly, however, the myocardial flow reserve as measured with PET may be substantially diminished in proportion to the severity of the vasculopathy (figure 6).14
Diminished flow responsiveness in microvascular disease
Cardiac risk factors including hypercholesterolaemia, smoking, chronic oestrogen withdrawal after menopause, and insulin resistance (including type 2 diabetes) adversely affect the coronary responsiveness. Most studies report moderate (about 30%) reductions in the flow response to vasodilator stress. For example, hyperaemic flows were found to be diminished in individuals with hypercholesterolaemia in patients with poorly controlled type 2 diabetes or in post-menopausal women.15–17 w10–w13 Aggressive risk factor modification such as lipid lowering or cardiovascular conditioning has been shown to improve the coronary vascular reactivity.w14–w16 Furthermore, hormone replacement therapy was found to normalise hyperaemic flows in post-menopausal women without other coronary risk factors.15 These treatment induced improvements implicate functional rather than structural alterations as the cause of diminished coronary responsiveness, which may largely be caused by endothelial dysfunction.w17–w19
Structural alterations of the microvessels may co-exist. Type 2 diabetes mellitus, for example, may be associated with morphologic microvascular disease that further reduces the vasodilator response. Diabetic patients with clinical evidence of microvascular disease, including stress induced angina, electrocardiographic signs of stress induced ischaemia but angiographically normal coronary arteries, demonstrated even lower hyperaemic flows and flow reserves than regular type 2 diabetes patients. Compared with normal subjects, myocardial flow reserves in type 2 diabetes patients were reduced by 32%, and by as much as 60% when microvascular disease was present.17 18
Lower limits of normal hyperaemic flows or of myocardial flow reserves remain to be established, possibly through long-term observational studies. Based on invasive coronary flow velocity measurements, some laboratories consider flow reserves <2.5 as abnormal. Others employ a value of 2.1 as derived from outcome data in patients with PET measured vasodilator responses.w20 Moreover, studies in patients with idiopathic dilated cardiomyopathy, hypertrophic cardiomyopathy as well as ischaemic cardiomyopathy and left ventricular dysfunction have consistently and convincingly identified hyperaemic myocardial blood flows of less than about 1.3 ml/min/g and myocardial flow reserves <1.5 as independent predictors of cardiac death and poor clinical outcome.8–10 Employing Gould's now classic relationship between coronary flow reserve and luminal stenosis severity,w21 a myocardial flow reserve of only 1.5 or less is equivalent to a ≥90% luminal stenosis.12 As increases in myocardial energy needs can no longer be adequately met by increases in substrate supply, repetitive ischaemic episodes or ischaemic events in cardiomyopathy patients possibly lead to progressive myocyte loss and replacement fibrosis, and thus to deterioration of left ventricular function.w22
Targeting coronary endothelial function
The interaction between vascular smooth muscle and endothelium can be tested through measurements of flow responses to sympathetic stimulation with cold pressor testing. Exposure to cold prompts an increase in heart rate and systolic blood pressure and, thus, in cardiac work, which under normal conditions is associated with commensurate increases in flow. In endothelial dysfunction and diminished NO bioavailability, however, adrenergically mediated vasoconstrictor forces are no longer fully offset by endothelium related vasodilator forces so that flow fails to increase appropriately with cardiac work. Flow responses may then be attenuated, be absent, or flows may actually decline.w23 w24 Cold pressor testing entails immersion of one hand in iced water, continuous monitoring of heart rate and blood pressure, and intravenous administration of the radiotracer of blood flow at 1 min, while the exposure to cold is maintained for another 1–2 min.
Reductions in coronary responsiveness to sympathetic stimulation contain predictive information on future cardiovascular events. For example, long term follow-up of individuals without angiographic evidence of coronary disease have demonstrated a statistically significant association between an abnormal cold pressor flow response and cardiac events.w25 Endothelial dysfunction as the likely reason for the diminished flow response to sympathetic stress may thus represent the substrate of future cardiac events and identify individuals at high coronary risk.
The flow responsiveness to cold pressor testing may already be diminished, even when the total integrated vasodilator capacity is maintained. For example, in individuals with milder forms of insulin resistance, hyperaemic flows were fully maintained while flow responses to cold pressor testing were already reduced (figure 7).19 Flow responses to cold pressor testing progressively declined with increasingly severe states of insulin resistance, whereas the total integrated vasodilator capacity decreased significantly only in patients with type 2 diabetes as the most severe state of insulin resistance. An abnormal flow response to cold pressor testing in a person with a fully preserved vasodilator flow response may thus identify endothelial dysfunction in the early phase of developing coronary atherosclerosis.
Clinical indications for quantitative blood flow measurements
From findings made with PET flow measurements, several clinical conditions emerge where such measurements may contribute to the diagnosis and risk stratification of patients with cardiovascular disease. PET flow measurements may prove clinically useful under the following conditions:
Balanced or triple vessel coronary artery disease. Suspected left main and/or balanced coronary artery disease where flow measurements can accurately delineate extent and severity of disease.
Functional characterisation of angiographic coronary stenosis for guiding percutaneous interventions (PCI). As shown in a randomised multicentre trial involving 1005 patients with multivessel coronary artery disease, PCI guided by functional rather than only angiographic criteria of the stenosis severity was associated with a significantly lower 1 year cardiac event rate (death, non-fatal myocardial infarction and repeat revascularisation: 13.2% vs 18.3%; p<0.02).20 PET measured hyperaemic flow responses may thus aid in identifying the target vessel for PCI.
Risk stratification in patients with and without obstructive coronary artery disease. The regional flow reserve in addition to myocardial perfusion imaging contains independent incremental predictive information for long term cardiovascular adverse events, even in patients with normal stress/rest myocardial perfusion images.
Cardiac transplant. Diminished flow reserve in cardiac transplant patients can identify the presence of transplant vasculopathy when stress/rest myocardial perfusion images are normal.
Microvascular dysfunction and microvascular angina. Coronary vasomotor abnormalities are frequently present in patients with chest pains but without obstructive coronary artery disease on invasive coronary angiography. In a retrospective analysis of 365 patients without obstructive coronary artery disease, the coronary responsiveness as assessed with intracoronary flow velocity probes was found to be diminished in 63% of all patients. Importantly, 59% of 233 patients with a normal stress/rest myocardial perfusion study had coronary vasomotor abnormalities.21 Such coronary vasomotor abnormalities may be present especially in women with chest pain as recently demonstrated in a WISE (Women's Ischaemia Syndrome Evaluation) substudy.22 In 189 women with chest pain followed for a mean of 5.4 years after invasive angiography, a flow reserve <2.32 was associated with a significantly increased rate of major cardiac events (27.0% vs 12.2%; p<0.01). Importantly, most of the women had either normal coronary vessels or only mild non-obstructive coronary artery disease and thus would likely have had normal stress/rest myocardial perfusion studies.
Microvascular disease in cardiomyopathies. In idiopathic dilated cardiomyopathy, hypertrophic cardiomyopathy, and ischaemic cardiomyopathy, measurements of the flow reserve can be predictive of the long term cardiac outcome.
Treatment responses. Measurements of blood flow can be useful for assessing effects of therapeutic interventions—for example, euglycaemic control in patients with type 2 diabetes mellitus, aggressive coronary risk modifications, and ß-blocker treatment in heart failure patients.
Several institutions now routinely combine flow measurements with standard stress/rest myocardial perfusion imaging. Findings forthcoming from these studies are expected to define further the clinical role and impact of PET flow measurements, to establish values of normal hyperaemic flows, and to identify those values that require clinical attention.
PET measurements of myocardial blood flow: key points
Flow responses measured with PET:
During pharmacologic vasodilation (adenosine, dipyridamole, regadenoson) reflect total integrated vasodilator capacity—that is, the combined effect of vascular smooth muscle and endothelial function.
Sympathetic stimulation with cold pressor testing primarily related to endothelial function.
PET flow measurements in microvascular disease:
Standard stress/rest myocardial perfusion imaging usually normal.
Associated with globally reduced hyperaemic myocardial flows and myocardial flow reserves.
May reflect: functional alterations of the coronary endothelium (ie, hypercholesterolaemia, smoking, chronic oestrogen withdrawal, diabetes): structural alterations (microangiopathies, loss of capillaries, increased myocardial fibrosis).
Observed in idiopathic dilated cardiomyopathy, hypertrophic cardiomyopathy, ischaemic cardiomyopathy, diabetic microvascular disease, vascular inflammation
You can get CPD/CME credits for Education in Heart
Education in Heart articles are accredited by both the UK Royal College of Physicians (London) and the European Board for Accreditation in Cardiology—you need to answer the accompanying multiple choice questions (MCQs). To access the questions, click on BMJ Learning: Take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to: http://heart.bmj.com/misc/education.dtl
RCP credits: Log your activity in your CPD diary online (http://www.rcplondon.ac.uk/members/CPDdiary/index.asp)—pass mark is 80%.
EBAC credits: Print out and retain the BMJ Learning certificate once you have completed the MCQs—pass mark is 60%. EBAC/ EACCME Credits can now be converted to AMA PRA Category 1 CME Credits and are recognised by all National Accreditation Authorities in Europe (http://www.ebac-cme.org/newsite/?hit=men02).
Please note: The MCQs are hosted on BMJ Learning—the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into Heart with their journal's username and password. All users must also complete a one-time registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.
- ↵The authors report the first application of non-invasive measurements of myocardial blood flow with N-13 ammonia and PET in normal human volunteers at rest and during supine bicycle exercise.
- ↵The authors demonstrate the feasibility of non-invasive measurements of myocardial blood flow with rubidium 8-82 and PET.
- ↵In a series of normal human volunteers the research described in this manuscript demonstrates the reproducibility of PET measured myocardial blood flows.
- ↵In a large series of normal human volunteers, the authors report on the heterogeneity of myocardial blood flow at rest and during vasodilator induced hyperaemia, indicating gender and age related differences in observed myocardial blood flows.
- ↵In normal individuals studied with N-13 ammonia flow measurements and C-11 acetate, the close correlation between cardiac work, oxidative metabolism and myocardial blood flow is demonstrated.
- ↵This review article provides a detailed account of the determinants of coronary blood flow and its response to physical stress.
- ↵This review article emphasises the importance of microvascular disease as a cause of myocardial ischaemia and reviews potential therapeutic strategies.
- ↵This study, similar to the immediate following two studies, emphasises the importance of assessing microvascular dysfunction and its prognostic value for assessing cardiac risk and long term clinical outcome.
- ↵This study demonstrates in normal human volunteers the contribution of the endothelium to the total vasodilator capacity.
- ↵The authors of this study report a statistically significant correlation between coronary stenosis severity as determined by quantitative angiography and myocardial flow reserve. This correlation observed in human coronary artery disease was similar to that reported previously in animal experimental studies.
- ↵This study demonstrates the value of quantitative measurements of myocardial blood flow for identifying left main equivalent or triple vessel disease.
- ↵The inverse correlation between the severity of vasculopathy and myocardial flow responses and allograft vasculopathy as reported in this research suggests the utility of flow measurements in patients with cardiac transplants.
- ↵This research assesses through quantitative blood flow measurements the coronary responsiveness in post-menopausal women. This responsiveness was determined by the total vasodilator response as well as by responses to cold pressor testing as a measure of endothelial function. The findings suggest that long term oestrogen administration in women without other coronary factors may restore impaired coronary vessel reactivity.
- ↵As demonstrated in this study, hyperlipidaemia in young normal volunteers is associated with impaired coronary vessel reactivity.
- ↵Elevated glucose levels in patients with poorly controlled type 2 diabetes are associated with a diminished myocardial flow reserve.
- ↵The findings of this study indicate a more severe impairment in the triple coronary vasodilator capacity when microvascular disease coexists with type 2 diabetes.
- ↵As shown in this study in insulin resistant individuals, the total integrated vasodilator capacity remains normal up to the most severe state of insulin resistance—that is, type 2 diabetes. However, there was evidence of endothelial dysfunction in individuals with milder forms of insulin resistance—for example, euglycaemic individuals and individuals with impaired glucose handling.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Download Supplementary Data (PDF) - Manuscript file of format pdf
Competing interests In compliance with EBAC/EACCME guidelines, all authors participating in Education in Heart have disclosed potential conflicts of interest that might cause a bias in the article. The author has no competing interests.
Provenance and peer review Commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.