[Myocardial microcirculation in humans--new approaches using MRI].

INTRODUCTION

One crucial goal of magnetic resonance imaging (MRI) in patients with coronary artery disease (CAD) is the characterization of myocardial microcirculation that reflects tissue supply much better than detection and quantification of a stenosis itself. PERFUSION: Myocardial perfusion is one important parameter of microcirculation and it is commonly detected by first-pass techniques using contrast agents (CA). Despite the quantification of perfusion it is an indispensable component of a comprehensive diagnosis to determine the perfusion reserve, which is believed a good indicator for viability of myocardium. However, most MRI techniques for perfusion imaging are Ca based and this implies a restricted reproducibility in humans. Beyond it, most first-pass techniques are qualitative and not quantitative. REGIONAL BLOOD VOLUME: Another parameter of microcirculation is the regional intracapillary myocardial blood volume (RBV) that almost represents the whole intramyocardial blood volume due to its dominating volume fraction. The RBV reflects the autoregulatory adaptation of microvessels, e.g., a severe stenosis may lead to an increase of the RBV by capillary recruitment, and the RBV is reduced in scar areas. The RBV may be quantified by first-pass techniques; however, this demands a definite relation between signal intensity and concentration of the CA, which is difficult to find for the range of concentrations present during the first pass. Until recently, no techniques existed for the exact and noninvasive assessment of the RBV. CAPILLARY RECRUITMENT: The evaluation of the relevance of a coronary artery stenosis is of paramount interest for the therapeutic decision. A severe stenosis implies the activation of compensation mechanisms, which includes poststenotic dilation of the microvascular system. This lowering of the vascular resistance aims to maintain sufficient blood supply at least under resting conditions. However, many obstacles hamper the noninvasive assessment of this autoregulatory response so far. Our laboratory recently developed different techniques for the assessment of myocardial perfusion, regional myocardial blood volume, and capillary recruitment. These techniques are based on theoretical and physiologic considerations and work mainly without CA. In this article, feasibility and reproducibility of these approaches are shown in volunteers and patients.

CLINICAL STUDIES

MR exams were performed on a 1.5-T whole body scanner (SIEMENS Vision) and a 2-T system (BRUKER Tomikon). Stress examinations were done repeatedly under pharmacologically induced stress (dipyridamole or adenosine, infusion rate: 0.56 mg/kg body weight over 4 min via an antecubital vein). Heart rate and blood pressure were continuously monitored during stress exams. T1

MEASUREMENTS

Spin labeling used in this work is based on T1 measurements after global and slice-selective spin preparation using a fast ECG-gated saturation recovery FLASH sequence. Due to the inflow of unsaturated proton spins, T1 in tissue is shortened after slice-selective preparation case compared to global saturation. We showed that, assuming a two compartment model with fast proton exchange between the compartments, the absolute perfusion P (in [ml/g/min]) can be calculated as P = lambda/T1(blood) ([T1(global)/T1(selective)] - 1), where the blood tissue partition coefficient lambda represents the quotient of water content of capillary blood and perfused tissue, which is approximately 0.9 ml/g in myocardial tissue. T1(blood) is the longitudinal relaxation time T1 of the arterial blood, measured in the left ventricle (LV). T1(global) and T1(selective) are the myocardial T1 calculated after the respective spin preparation. Perfusion reserve is evaluated as the quotient of perfusion under adenosine-induced stress and perfusion at rest. In volunteers quantitative perfusion was determined as 2.5 +/- 0.7 ml/g/min (rest), perfusion reserve was about 2.0. Absolute perfusion decreased to 1.6 +/- 0.6 ml/g/min under oxygen breathing. In patients with CAD, myocardial regions with decreased perfusion reserve could be identified. perfusion reserve could be identified. Performing the described spin-labeling technique with an intravascular CA facilitates the determination of the intra-extracapillary water proton exchange frequency and the RBV. In a patient study, the effect of the intravascular CA Feruglose (Amersham) on relaxation rate in myocardium (R1(myo)) in the steady state was investigated (Figure 1). The dependence of R1(myo) on R1(blood) was characterized and compared with a theoretical model which allowed determination of the intra-extracapillary water proton exchange frequency (f = 0.48 s(-1)) and the intracapillary blood volume (RBV = 12.9%). A linear response range of Delta R1(myo) on Delta R1(blood) was estimated which, in future studies, will allow the determination of RBV with intravascular CA (Figure 2). T2*

MEASUREMENTS

We anticipated that poststenotic vasodilatation implies a capillary recruitment. Almost all (i.e., > 90%) of intramyocardial blood residues in that type of vessel. Due to their large arteriovenous oxygenation difference, myocardial capillaries contain considerable amounts of deoxyhemoglobin (Figure 3). Hence, in regions with autoregulatory capillary recruitment the tissue concentration of deoxyhemoglobin should be elevated when compared to myocardium supplied by a normal vessel (Figure 5b). Due to its paramagnetic property and its intravascular confinement, the natural CA deoxyhemoglobin may be assessed by susceptibility sensitive, or also called blood oxygenation level-dependent (BOLD) MRI. For T2* measurements, a segmented gradient echo pulse sequence was used, which acquired ten successive gradient echoes per rf excitation in a single breathhold. In volunteers, there was an increase in T2* of about 10% under dipyridamole-induced stress (Figure 4). This means a decrease of the intracapillary deoxyhemoglobin concentration, whereas the oxygen consumption under increased perfusion did not change. In myocardial regions of patients, associated with the stenotic artery T2* was significantly lower than in residual myocardium (p < 0.01; Figure 5a). This difference in T2* increased after application of the vasodilator dipyridamole (p < 0.001). In patients being reinvestigated after therapeutic interventions, the microvascular dilation was partly removed (Figure 5c). For the first time we could show that myocardial BOLD MRI detects poststenotic capillary recruitment dependent on a coronary artery stenosis.