Number 45, 2009
Diabetes and cardiovascular disease

Nuclear magnetic resonance for the study of cardiac metabolism in diabetes

Back to the Summary

Giancarlo Todiere, Mario Marzilli
Cardiothoracic Department, University of Pisa, Pisa, Italy
Correspondence: Professor Mario Marzilli, Dipartimento Cardiotoracico, University of Pisa, Via Paradisa 2, 56100 Pisa, Italy.
Tel: +39 050 996751; e-mail: marzilli@med.unipi.it
Giancarlo Todiere, MD, Dipartimento Cardiotoracico, University of Pisa, Via Paradisa 2, 56100 Pisa, Italy.
Tel: +39 333 5071523; e-mail: todiere@alice.it

Nuclear magnetic resonance spectroscopy
Nuclear magnetic resonance (NMR) [1] is an imaging technique that uses the nuclear spin of protons in water and fat to provide anatomical and functional information, for example on the myocardium. Magnetic resonance spectroscopy (MRS) [2] is the only non invasive method, not involving exposure to radiation, that allows study of the metabolic components of the myocardium without the use of a tracer [3].
Proton (¹H) NMR is performed immediately before spectroscopy to enable selection of the anatomically most suitable “voxel” (the smallest distinguishable box-shaped part of a 3-dimensional space) for the subsequent metabolic study. This aim is achieved by analyzing signals from other nuclei possessing a nuclear spin [sodium-23 (²³Na), carbon-13 (¹³C), phosphorus-31 (³¹P)], providing information on the chemical and physical state of the tissue.
A current limitation of the technique, confining it to the field of research, is the low sensitivity of the signals acquired from these non proton nuclei. That low sensitivity is secondary to the lower polarization of the nuclei at thermal equilibrium. The level of polarization (P) is defined as:

1. P=(N+−N−)(N++N−)

where N⁺ and N⁻ are the number of spins in the (parallel) “up” and the (antiparallel) “down” directions, respectively; the net magnetization, and thus the available NMR signal, is proportional to the ratio of nuclei showing “up” spins to that of nuclei showing “down” spins. However, despite of the use of a high magnetic field [greater than 3 Tesla (T)] to increase the thermal equilibrium polarization (the energy difference between the two spin states), which contributes to the genesis of a useful signal, the polarization is very low.
In ¹H-MRS, the limit of low sensitivity is overcome by the high concentration of protons in organic tissues, but suboptimal spatial and temporal resolutions limit more widespread use of the technique in the clinical setting: using magnetic fields of 1.5 T, the voxels achieved are large (20–70 ml), compared with 8 ml obtained with 7 T, and acquisition takes about 30–40 min, these lengthy scans being associated with a corresponding increase in motion artifacts. In addition, MRS is technically demanding, and requires specialist expertise and additional magnetic resonance hardware and software [2].

Cardiac metabolism
In the past, most studies of cardiac metabolism were performed with ³¹P-MRS [4]. This technique mainly examines the ratio between phosphocreatine (PCr) and ATP, which represents a reliable index of the cardiac energy state. During myocardial ischemia, ATP demand exceeds its production, resulting in a decrease in PCr concentrations and, consequently, a decrease in the PCr/ATP ratio [5]. Using ¹H-MRS in 10 patients with a history of myocardial infarction, Bottomley and Weiss [6] measured the content of creatine in infarct zones with respect to those in control areas, and found it to be reduced in infarcted areas. With ²³Na-MRS it was also possible to show high levels of ²³Na signal both in areas of acute necrosis and in chronic myocardial scar tissue. In theory, combination of these three spectroscopic measurements could enable the identification of ischemic, non viable, and hibernating myocardial tissue.
Among these techniques, ¹H-MRS is now the easiest to use in the clinical setting because of its higher sensitivity compared with other MRS techniques. A large number of metabolites – including creatine, lactate, carnitine, and lipids – have indeed been studied. The major limitation of ¹H-MRS remains the need to suppress the signal deriving from water.

NMR and the diabetic heart
In the research setting, NMR (²³Na, ¹³C, ³¹P) has provided data on the metabolic changes associated with coronary artery disease, heart failure and its treatment (metabolic therapy with trimetazidine [7]), valvular disease (on a possible future role in the optimal timing for surgical treatment [8]), hypertrophic cardiomyopathy [9], and diabetes mellitus.
Given that diabetic patients surviving a myocardial infarction have higher mortality and higher prevalence of heart failure than non diabetic patients, it would be very useful to evaluate and monitor cardiac energy metabolism in diabetic individuals [10]. Indeed, the cardiac metabolic state has been identified as a major predictor of cardiovascular morbidity and mortality (stronger than left ventricular ejection fraction and New York Heart Association class [11]) in patients with heart failure.
The pathways of myocardial metabolism are summarized in Figure 1 [12]. In the diabetic heart, glucose and lactate oxidation decrease [13] and fatty acid oxidation is increased [14], increasing the oxygen requirement for ATP production. Studies with MRS in patients with type 1 and with type 2 diabetes have confirmed these observations. For example, 21 patients with type 2 diabetes and normal systolic left ventricular function [normal left ventricular ejection fraction (LVEF)] and normal myocardial left ventricular mass, free from other major cardiovascular risk factors, were studied with ³¹P-MRS [15]. The PCr/ATP ratio in the diabetes group (1.35) was decreased by 35% relative to that in the control group (2.35), and these figures correlated positively with fasting plasma glucose concentrations and negatively with plasma free fatty acid concentrations (Figures 2 and 3). Similar observations (a lower PCr/ATP ratio) were reported in patients with type 1 diabetes and a normal LVEF [16]. These findings are consistent with those from positron emission tomography (PET) studies demonstrating lower rates of uptake of [¹⁸F]2-fluoro-2-deoxyglucose [17] and decreased myocardial blood flow [18] in patients with diabetes compared with those in healthy volunteers.

Technological considerations
The use of spectroscopy in the clinical arena strongly depends on technological progress.
The transition from 1.5 T to 3 T MRI has already significantly improved the spatial resolution; further improvements are expected with 7 T magnetic fields, which could provide adequate coverage of the entire heart with a spatial resolution corresponding to a 17-segment model [11]. However, increasing the magnetic field gives rise to problems of cost, depth of penetration of the radiofrequency instrumentation, and the available sequences and tissue contrast.
An alternative strategy with which to improve the signal-to-noise ratio while limiting the increase in the magnetic field with all its attendant negative consequences, lies in a different means of increasing the polarization – a “hyperpolarization” technique [22]. This method can increase the signal by several orders of magnitude, using magnetic fields of relatively low strength (up to 3 T) (Figure 5). Referring to equation (1), and bearing in mind that the signal-to-noise ratio (SNR) is proportional to the gyromagnetic ratio (γ), the concentration of nuclear spin (c) and the polarization (P) (SNR = c γ P), it is clear that creating a non equilibrium distribution of the nuclei (N⁺ − N⁻) will achieve an increase in the signal-to-noise ratio. The main advantage of this approach is that the hyperpolarized nuclei create the signal themselves, rather than moderating the signal from surrounding protons.

REFERENCES