Heart and Metabolism

Metabolic imaging of heart failure

F.C. Visser
Department of Cardiology, Vrije Universiteit Medical Center, Amsterdam, The Netherlands
Correspondence: Professor F.C. Visser, Department of Cardiology, Vrije Universiteit Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands
Tel: +31 20 4440123, fax: +31 20 4442446, e-mail: fc.visser@vumc.nl

Introduction
Cardiac metabolism is the basis for all cardiac processes such as contraction and maintenance of cellular structure and integrity. Therefore, abnormalities of metabolism may result in abnormalities of contractile function. In recent years, new data have become available giving insight into the metabolic abnormalities in patients with heart failure. In this overview, some of the aspects of the metabolic abnormalities in heart failure will be discussed with emphasis on energy metabolism and its noninvasive assessment using radiolabeled nutrients.

Exploring cardiac energy metabolism
Under normal conditions the heart uses a variety of substrates such as carbohydrates, lipids, and proteins. Of these nutrients, glucose, fatty acids, and acetate contribute most to the production of energy-rich phosphates, whilst other carbohydrates and proteins play a minor role. Availability (measured in plasma levels) strongly determines which substrate is oxidized. For example, during fasting there is a high blood level of fatty acids, and fatty acid oxidation provides 60 to 70% of the energy production,[1] whereas after a carbohydrate-rich meal, when glucose and insulin levels are increased, fatty acid contribution falls to approximately 30%, and 60% of the energy production is derived from glucose oxidation. During exercise the skeletal muscles produce lactate and this substrate is readily extracted by the heart, leading to a lactate energy production contribution of approximately 60%.
Under ischemic conditions with low flow and oxygen deprivation, the heart switches to anaerobic glycolysis, leading to production of lactate and a limited amount of ATP.[1]

Exploring energy metabolism of the heart
Cardiac energy metabolism can be studied in patients either by invasive procedures in which the arterial system and the coronary sinus are cannulated, or by using radiolabeled nutrients. The invasive approach can only determine global extraction and uptake of nutrients. The advantage is, however, that lactate extraction and release can be measured. Radiolabeled glucose and structurally different labeled fatty acids have all successfully been employed in patient studies. The main advantage is that it is a noninvasive procedure and it can measure regional substrate uptake. No studies have yet been performed using labeled lactate.
Although glucose can be labeled with C-11, it is nowadays exclusively labeled with F-18 by replacing an OH group. After uptake, F-18 deoxyglucose (FDG) is converted to FDG-6-phosphate and is not further metabolized. Myocardial FDG uptake can be detected by both PET and SPECT.
Its main clinical application is the detection of dysfunctional but viable tissue that is able to recover in function either spontaneously (stunning after infarction) or after revascularization (hibernation or repetitive stunning). This has been discussed in detail in previous issues of Heart and Metabolism by Senior and Lahiri[2] and Bax et al.[3]
Free fatty acids can be labeled with C-11 (C-11 palmitate), F-18 thioheptadecanoic acid (FTHA), and I-123 (I-123 b-methyl-phenyl pentadecanoic acid, BMIPPA; heptadecanoic acid, IHDA; and ortho- and paraphenylpentadecanoic acid, oPPA and pPPA, respectively). C-11 palmitate undergoes b-oxidation and enters the Krebs cycle, releasing the radioactivity as C-11 CO2. The rate of C-11 CO2 release which is monitored during dynamic PET imaging is related to the oxidation rate of fatty acids. However, several factors have limited its application. First, uptake and turnover of C-11 palmitate is dependent on substrate availability: for example, during high glucose/ insulin levels, b-oxidation of palmitate is low because the heart switches to glucose oxidation. Therefore, metabolic imaging conditions need to be standardized. Second, part of the fatty acids do not undergo b-oxidation but are stored in triglycerides. Therefore, for accurate assessment of the oxidation rate, one needs to correct for the amount of stored fatty acids in triglycerides. Third, during conditions of ischemia, a variable amount of the fatty acids that are taken up by the myocardium are again released without undergoing metabolism. As this release cannot be distinguished from C-11 CO2 release by the PET camera, accurate determination of the oxidation rate of fatty acids under ischemic conditions is difficult. Similar kinetics (and limitations) can be observed with the iodinated fatty acids IHDA and pPPA. In contrast to C-11 palmitate, the modified fatty acids FTHA, BMIPPA, and oPPA undergo only limited b-oxidation and are trapped in the myocardium, similarly to FDG.
A distinct tracer of myocardial metabolism is C-11 acetate. After uptake in the myocardium, acetate is converted to acetyl-CoA and immediately enters the Krebs cycle, releasing the radiolabel as C-11 CO2. Previous studies[4,5] have demonstrated a direct relation between the rate of C-11 CO2 release from the myocardium and oxygen consumption. In contrast to C-11 palmitate, turnover of C-11 acetate is not substrate-dependent and there is no back-diffusion under ischemic conditions. C-11 acetate imaging is therefore a unique technique to determine regional oxygen consumption of the heart (see reference 6 for an extensive overview).
These labeled fatty acids and acetate have been explored to characterize physiological and pathophysiological conditions in humans, but there are no established clinical indications in routine practice at present.

Metabolic changes in heart failure
During fetal development, the heart mainly uses glycolytic pathways for its energy metabolism. After birth the heart switches from glycolysis primarily to fatty acid oxidation. There is increasing evidence that in the failing heart, energy substrate utilization is changed: fatty acid oxidation decreases and the contribution of glycolysis increases. This pattern resembles the fetal situation.[7–10] This changed pattern has been demonstrated both in rats and in humans. The reduction in fatty acid oxidation is not due to changes in substrate availability in the blood, since the reduced oxidation rate has also been demonstrated in isolated cardiomyocytes.[11] Moreover, Sack et al[12,13] demonstrated in the failing human heart that the expression of genes that encode the cardiac enzymes for fatty acid oxidation is depressed. Thus, enzymes that are necessary for the metabolic handling of fatty acids are depressed, leading to a reduced rate of fatty acid oxidation. The main question is whether a reduction of the involved enzymes and the fatty acid oxidation rate can be a cause of (the transition to) heart failure, or whether it is only a secondary phenomenon.[14] It is clear that genetic defects in fatty acid oxidation enzymes lead to childhood cardiomyopathies. However, in the adult situation the cause-effect relation is not clear and needs further study.

Metabolic imaging in heart failure
Heart failure can be caused by any cardiac disease (coronary artery disease, valvular disease, tachyarrhythmias, hypertrophic and idiopathic dilated cardiomyopathies) and also by a number of noncardiac diseases (hyperthyroidism, hypertension). Most data in the literature on metabolic imaging in heart failure are from cases of ischemic cardiomyopathy. This is not surprising given that ischemic cardiomyopathy has the highest incidence and prevalence among all cardiac diseases. However, idiopathic dilated cardiomyopathy (IDCM) is an interesting disease because coronary artery disease is not present, and coronary underperfusion is in principle not the primary cause of the metabolic abnormalities. Moreover, the data of Sack et al[12,13] suggest that the abnormalities in gene expression of enzymes involved in fatty acid oxidation are present in both ischemic and nonischemic dilated cardiomyopathy.

Metabolic imaging in ischemic cardiomyopathy
From a clinical point of view it is important to distinguish irreversible necrosis from potentially reversible dysfunction (viable tissue). In necrotic/scar tissue no or little metabolic activity is present and no improvement in regional or global function after revascularization is to be expected. Necrotic tissue is characterized by a severe reduction in flow with a concomitant reduction in FDG or fatty acid uptake (flow-metabolism match). Also in these areas C-11 acetate clearance is slow.
In contrast, in dysfunctional but viable tissue, uptake of FDG is preserved relative to flow, indicating that the myocardium preferentially utilizes glucose (Figure 1).

Figure 1. Short-axis views of glucose uptake (FDG) and resting perfusion (resting thallium-201) in a patient with left ventricular dysfunction. Depicted are mid-ventricular short-axis slices with the anterior wall at the top, the lateral wall on the right, the inferior wall at the bottom, and the septum on the left. A clear FDG mismatch (increased FDG relative to perfusion) is observed in the septal and inferior area. Reproduced from reference 15 with permission.

Fatty acid uptake is decreased in viable tissue, as has been shown particularly with BMIPPA. It is more depressed than the concomitant reduced flow. Although confusing, in BMIPPA imaging the term mismatch is also used, but in this instance it indicates a more severe reduction of fatty acid uptake relative to flow, being the opposite to the flow-FDG mismatch. C-11 acetate imaging shows that in the areas of viable tissue the clearance rate of the radioactivity is reduced in comparison to that of normal myocardium. However, it is faster in comparison to the rate in infarcted tissue. These different patterns of FDG, fatty acid and C-11 acetate imaging have been extensively reviewed by Bax et al3 in Heart and Metabolism.

Metabolic imaging in IDCM
Although idiopathic cardiomyopathy has been studied with radiolabeled fatty acids, FDG, and C-11 acetate, the number of studies is very limited.
Geltman et al[16] studied patients using C-11 palmitate and compared the cardiac fatty acid uptake with that of normal subjects and patients with myocardial infarction. These authors found that IDCM patients had, throughout the myocardium, multiple discrete regions with depressed fatty acid uptake, giving the myocardial images a “moth-eaten” appearance. This marked heterogeneity of fatty acid uptake was independent of perfusion and contractile abnormalities, and of the degree of myocardial thinning. Similar findings were observed by Höck et al,[17] Ugolini et al,[18] and Yazaki et al[19] using various radioiodinated fatty acids.
Endocardial biopsy findings corroborate these observations: both patchy fibrosis and patchy metabolic derangements, including depression of oxidative phosphorylation and altered myocardial enzyme content, have been reported.[20,21] Thus, abnormal myocardial fibrosis and regional depression of fatty acid oxidation may explain the observed scintigraphic abnormalities.
In contrast to the previous studies, Feinendegen et al[22] did not observe heterogeneous fatty acid uptake in patients with IDCM: in their study the fatty acid uptake was homogeneous, possibly related to the imaging technique. Using a combination of two fatty acids, oPPA (which is retained in myocardium) and pPPA (undergoing oxidation), they focused on the fatty acid oxidation rate. In their patients a highly variable rate of oxidation (higher, equal, and lower compared with normal subjects) was found. As the authors did not correlate their findings with hemodynamic and perfusion data, this approach needs further validation.
An interesting study has recently been reported by Yazaki et al,[19] who correlated fatty acid (BMIPPA) uptake patterns with perfusion and follow-up data. First, the extent of fatty acid abnormality correlated with echocardiographic, hemodynamic, and biopsy data in the patients, suggesting that there is an association between the severity of disease and the extent of fatty acid abnormality. Second, patients with a more severe reduction of fatty acid uptake relative to perfusion, had an unfavorable follow-up in comparison to
patients with relatively preserved fatty acid metabolism (Figure 2).

Figure 2. Short axis views of fatty acid uptake (BMIPPA) and perfusion in a patient with a cardiomyopathy. Depicted are mid-ventricular short-axis slices with the anterior wall at the top, the lateral wall on the right, the inferior wall at the bottom, and the septum on the left. Both matches (decrease of perfusion and fatty acid uptake) and mismatches (fatty acid uptake more decreased than perfusion) are observed in the septal area. Reproduced from reference 19 with permission.

Thus, most of the available evidence suggests that fatty acid metabolism is decreased in patients with IDCM. Preliminary results show that the degree of abnormality is correlated with the hemodynamic status and is associated with a poor prognosis.
IDCM is presumed to involve the myocardium diffusely and to have a homogeneous effect on function. However, heterogeneity in regional function has long been noted.[23,24] Bach et al[25] correlated regional function with regional myocardial oxygen consumption using C-11 acetate and found that regions with relatively preserved function (although the mean ejection fraction was 21%) also had a higher oxygen consumption than regions with an average depressed function. The heterogeneity of metabolic abnormalities in this disease was also studied by van den Heuvel et al.[26] albeit in a different manner. These authors investigated C-11 acetate in areas with increased FDG uptake and found that C-11 acetate clearance (oxygen consumption) in this region was lower. As FDG is a tracer of glucose consumption (which can be metabolized aerobically and anaerobically), the data may indicate that a larger fraction of glucose is metabolized by anaerobic glycolysis. These observations are in line with animal experimental data showing that glucose oxidation was impaired despite acceleration of glycolysis.[27,28] This uncoupling of increased glycolysis and unchanged or decreased glucose oxidation may lead to the production of lactate and protons. This is unfavorable since it leads to acidosis, a major cause of contractile dysfunction.[14] An example of C-11 acetate images in patients with IDCM compared with normal subjects is presented in Figure 3.[29]

Figure 3. C-11 acetate images in a healthy individual (a) and in a patient with IDCM (b). Depicted are mid-ventricular short-axis slices with orientation similar to that shown in Figures 1 and 2. Soon after injection, activity is present in the blood pool with subsequent uptake and later washout by the yocardium. Reproduced from reference 29 with permission.

The same group[26] also compared FDG uptake in these patients with that in normal subjects. Similar to uptake of fatty acids and acetate, uptake of FDG was inhomogeneous, comprising areas of normal, increased, and decreased uptake. This heterogeneity was also found by Yokoyama et al.[30] Interestingly, they divided IDCM patients into a group with and without events during follow-up. Although global FDG uptake was not different between the two groups and not different compared with normal subjects, the patients with events (worsening of heart failure, hospital admissions, and sudden death) had a significantly higher heterogeneity than patients with a favorable prognosis. Homogeneous FDG uptake also predicted improvement of left ventricular function on medical therapy with 85% accuracy. This accuracy was higher than that of clinical and hemodynamic parameters.
To summarize the available metabolic imaging data in IDCM, the most prominent feature is a strongly heterogeneous uptake and turnover of the different nutrients. In patients there is a reduction of fatty acid metabolism: the degree of abnormality is correlated with the hemodynamic status and is associated with an unfavorable prognosis. Similarly, regional function and regional oxygen consumption are highly variable, and there are indications that part of the increased glucose uptake is metabolized anaerobically. Finally, abnormal glucose metabolism is also an indicator of poor prognosis. It is a challenging concept that the oxidative metabolic abnormalities may contribute to the development of heart failure or to the transition from compensated to decompensated failure. If so, metabolic imaging may be powerful tool in clinical practice to discriminate between high- and low-risk patients.

REFERENCES
1. Opie LH. The Heart. Physiology and metabolism. New York: Raven Press; 1991.
2. Senior R, Lahiri A. Metabolic imaging: predicting recovery of function in heart failure. Heart Metab. 1999;6:12–17.
3. Bax JJ, Klein LJ, Sloof GW, Visser FC. Metabolic imaging in the evaluation of myocardial ischemia and viability. Heart Metab. 2000;10:11–22.

4: J Am Coll Cardiol 1988 Oct;12(4):1054-63