The ability of 1-carboxymethyl-2-iminoimidazolidine (cyclocreatine), a synthetic creatine analog, to protect myocardium during global ischemia was assessed in isovolumic rat hearts using phosphorus-31 nuclear magnetic resonance spectroscopy. Wistar rats were fed a 1 % cyclocreatine diet. After 2 weeks, cyclocreatine-fed (n = 8) and control (n = 7) rats were anesthetized, the heart was excised and retrograde perfusion was begun at 10 ml/min per g with 37°C, phosphate-free buffer containing glucose and oxygen. Hemodynamic and spectroscopic data were obtained during baseline, ischemia and recovery periods (each 24 min). During ischemia, the heart of control rats developed a rigor-like increase in tonic pressure (ischemic contracture) not seen in the heart of cyclocreatine-fed rats (22 versus 1 mm Hg, p < 0.01). This change was associated with significantly more adenosine triphosphate (ATP) at endischemia in the cyclocreatine group (1.6 versus 0.6 μol/g, p < 0.01) and delayed development of acidosis (p < 0.001). With reperfusion, the heart of cyclocreatine-fed rats spontaneously defibrillated sooner than did the heart in control rats (178 versus 346 s, p < 0.03). Diastolic pressure remained significantly elevated throughout recovery in control hearts compared with treated hearts (p < 0.001).

An echocardiographic index of left ventricular (LV) short axis circularity can be defined by the equation: left ventricular systolic circularity index (LVSCI) = 4 pi(LV area) X 100/(LV perimeter). This index was measured from two-dimensional echocardiograms in 98 children (ages 1 day to 19 years) with congenital heart disease, and results were compared to right ventricular/left ventricular peak systolic pressure ratios (RVP/LVP) determined at cardiac catheterization. LVSCI was also computed in 50 children without cardiovascular or pulmonary disease to define the normal range. A short axis image of the left ventricle at the level of the papillary muscles was obtained from the left parasternal position. Area and perimeter were determined by computer planimetry of the LV endocardium at end systole. LVSCI was measured from three consecutive beats and averaged. In the normal group all values of LVSCI exceeded 93% (mean 96%). In the group with congenital heart disease RVP/LVP correlated exponentially with LVSCI: RVP/LVP = e2.6-0.04 LVSCI; with r = 0.88, SEE = 0.39, and p less than 0.001. If patients with suprasystemic right ventricular pressures (RVP/LVP greater than 1.2) are excluded, there is a linear correlation between RVP/LVP and LVSCI: RVP/LVP = 2.3-0.021 LVSCI; with r = 0.80, SEE = 0.14, and p less than 0.001. LVSCI could distinguish between patients with normal, mildly elevated, moderately elevated, and severely elevated RVP/LVP. We conclude that LVSCI is a readily determined parameter that is independent of age or body size and predicts RVP/LVP in children with congenital heart disease.

Twelve patients (pts), ages 7 mos. to 23 yrs. (mean 11 yrs.), with known univentricular atrioventricular (AV) connection (“Single ventricle”) were evaluated by ECG-gated magnetic resonance imaging (MRI). Multiple sections were obtained in transverse, coronal and oblique planes using a 0.3T or LOT magnet and SE 30 pulsing techniques. MRI studies were evaluated using a segmental approach prior to reviewing the echocardiograms and angiograms which were available on all pts. MRI accurately depicted: 1) cardiovisceral and atrial situs 2) presence of splenic tissue 3) systemic and pulmonary venous connections 4) the mode of AV connection (i.e. absent right, double inlet, etc.) 5) the morphology and spatial relationships of the main and rudimentary ventricular chambers (6) the ventriculoarterial (VA) connections and 7) surgical procedures (i.e. shunts, PA bands). Seven pts had absent right AV connection, 4 had double inlet anatomy and 1 had an absent left AV connection. The absent AV connection was true atresia in 6 with invagination to the cardiac crux of fat-containing sulcus tissue which appears bright on MRI and an imperforate valve in 1. In 11 pts, the rudimentary chamber could be visualized. VA discordance was demonstrated in 2 pts with absent right AV connection, in all 5 with double inlet LV's and the 1 with absent left AV connection. We conclude that MRI accurately depicts cardiac morphology in pts with univentricular AV connection, often providing information not available with other imaging techniques.

The metabolic effect of anoxia on hypertrophied myocardium was investigated using P-31 NMR spectroscopy. Hearts from 18 mo. SHR (hypertrophied, n=7) and age-matched WKY (control, n=6) rats were mounted on a modified Langendorff apparatus, paced at 240/min and perfused with 36°C phosphate-free, glucose-containing buffer bubbled with oxygen or nitrogen (anoxia). Left ventricular pressure (LVP), LV dP/dt and perfusion pressure were continuously recorded while consecutive 3-min spectra were collected. This allowed temporal assessment of myocardial phosphate levels [incl. inorganic phosphate (Pi), creatine phosphate (CP) and ATP] during baseline conditions, anoxia and recovery. Anoxia was maintained until a 70% fall in LVP occured. Compared to the WKY, SHR rats had higher in vivo BP (163 vs 104; p<.001) and cardiac hypertrophy (heart/ body weight = 5.3 vs 3.5 mg/g; p<.001). During baseline perfusion, SHR hearts had higher resistance (9.8 vs 5.9 mmHg/cc/min/g; p=.003) but no significant difference in LVP or dP/dt. SHR hearts demonstrated a faster fall in high-energy phosphates and LVP during anoxia (SHR fell to 30% baseline LVP in 8.5 vs 13.0 mins; p=.018). Throughout the protocol SHR hearts had lower CP (3.8 vs 5.6 umoles/g; p=.008), ATP (3.2 vs 3.9 umoles/g; p=.047) and CP/Pi ratio (1.5 vs 2.5; p=.043). We conclude that chronically hypertrophied hearts have: (1) less ATP/g and CP/g than age-matched controls, and (2) a faster decline in mechanical and metabolic function during anoxia.

MR was used to differentiate the classical form of tricuspid atresia, in which the atrioventricular connection is absent, from a rare type of right atrioventricular obstruction, in which the valve is imperforate, and from some Ebstein's malformations in which atrioventricular flow is obstructed by a displaced tricuspid valve. ECG-gated, T1-weighted spin-echo MR imaging was performed at 0.3 or 1.0 T in six patients with classical tricuspid atresia, in one patient with tricuspid atresia due to an imperforate tricuspid valve, and in two with Ebstein's malformation. In the patients with classical tricuspid atresia, epicardial fat within the atrioventricular sulcus that was interposed between the right atrium and the ventricular mass caused increased intensity. This finding was not present in the patients with imperforate tricuspid valve or Ebstein's malformation in whom the sulcus appeared normal. Postmortem examination of five other hearts with classical tricuspid atresia revealed epicardial fat extending into the atrioventricular sulcus, which separated the right atrium and the ventricular mass. In three hearts with Ebstein's malformation, the sulcus extended only to the level of the atrioventricular junction. The characteristic MR appearance of fat adjacent to the muscular floor of the right atrium, therefore, serves to identify the atrioventricular sulcus and allows differentiation of the types of right atrioventricular valve atresia.

ECG-gated MRI was performed at 0.3 T or 1.0 T in 19 patients, aged 6 years to 18 years, for suspected congenital abnormalities of the aorta (13 patients) or pulmonary artery (six patients). Seventeen of the patients were also evaluated by echocardiography, and 14 had angiograms. In 11 patients, MRI demonstrated lesions that echocardiography either failed to visualize or found inconclusive, including supravalvar aortic stenosis (one patient), coarctation of the aorta (three patients) hypoplastic aortic arch (one patient), and pulmonary artery hypoplasia or stenosis (six patients). MRI complemented echocardiographic diagnosis in four patients with Marfan's syndrome and in one with coarctation. One mild recurrent coarctation demonstrated angiographically was not visualized by MRI or echocardiography. The eight other angiographic studies of the aorta confirmed MRI findings. In all six MRI studies of the pulmonary arteries, obstructive lesions were revealed that had not been completely visualized on echocardiography or angiography. MRI is an excellent noninvasive method of depicting congenital abnormalities of the great arteries and may provide otherwise unobtainable information.

We quantified high-energy phosphate metabolites in hypertensive hypertrophied and normal myocardium and monitored temporal changes using the non-invasive 31P nuclear magnetic resonance (NMR) spectroscopy. Hearts from 18 month spontaneously hypertensive rats (SHR) and age-matched Wistar-Kyoto rats (WKY) were perfused with a phosphate-free buffer at 10 cc/min per g and paced at 240 beats/min on a modified Langendorff apparatus. Perfusion pressure, left ventricular pressure (LVP) and dP/dt were recorded and successive 31P NMR spectra were collected during a 24-min baseline period (oxygenated buffer), anoxia (N2-bubbled and glucose-free buffer) until a 70% fall in LVP occurred, and recovery. An aminomethylphosphonate standard, located within the LVP balloon, permitted absolute quantification of myocardial phosphate moieties (including inorganic phosphate (Pi), creatine phosphate (CP) and ATP). During perfusion, SHR hearts demonstrated higher coronary resistance but no significant differences in LVP or dP/dt. Spontaneously hypertensive rat hearts had lower CP, ATP and CP/Pi ratio and showed a faster fall in cardiac function during anoxia, associated with parallel rates of changes in the phosphate moieties.

Electrocardiographically gated magnetic resonance imaging (MRI) was successfully performed in 9 patients with atrioventricular (AV) septal defect: 6 had complete AV canal and 3 had partial AV canal. The defect was readily visualized in all patients on transverse scans taken at the level of the AV valve. The size and extent of the defect could be easily determined. All patients demonstrated a similar underlying morphologic pattern on MRI scans, consisting of deficiency of the primum atrial septum and inlet ventricular septum and a “common” AV valve ring with absence of the cardiac crux. The 3 patients with isolated atrial septal defect could be distinguished from the 6 with complete AV canal by the dense, fibromuscular bridging tissue, which coursed from the AV valve to the crest of the ventricular septum, obliterating the interventricular communication. Four patients had angiographically proved ventricular hypoplasia, which was also detected by MRI. AV valves and their patterns of chordal attachment were accurately imaged in 7 patients on systolic sections; accessory chordae were identified in 6 patients. MRI is a useful noninvasive imaging modality that can depict the underlying morphologic abnormalities in AV septal defect as well as important anatomic variations.

ECG-gated magnetic resonance imaging (MRI) has been shown to provide excellent tomographic images of congenital heart defects.1–4 Cardiac structures are especially well demonstrated because of the sharp distinction between relatively white myocardial walls or valves and the dark appearance of rapidly flowing blood within the heart. In particular, the clarity with which the endocardial and epicardial surfaces are resolved using MRI is superior to that produced by other noninvasive imaging modalities such as computed tomography (CT) or ultrasound and rivals the clarity produced by high quality cineangiograms. Thus, ventricular size, geometry, and wall thicknesses can be precisely defined. In this study, we examined the ability of MRI to depict the abnormal cardiac morphology in patients with known hypoplastic right heart syndrome (tricuspid atresia or pulmonic atresia with intact ventricular septum).

La prise en charge des cardiopathies congénitales est souvent un challenge diagnostique. Dans ce domaine, l’imagerie par résonance magnétique (IRM), du fait de sa complémentarité par rapport à l’échocardiographie et de son innocuité, occupe une place de choix dans la gamme des explorations cardiaques non invasives. L’IRM offre à la fois une représentation multiplanaire tridimensionnelle de l’anatomie du système cardiovasculaire avec un champ de vue large (sans interférence osseuse ou aérique) et une excellente résolution spatiale, mais également une approche fonctionnelle sur les flux. L’IRM permet ainsi d’accéder à des anomalies (notamment extracardiaques) qui échappent à l’échocardiographie mais aussi à l’angiographie. L’IRM s’est donc, dès la phase initiale de son développement, progressivement imposée comme un outil indispensable dans le bilan des cardiopathies congénitales. Sa place gagnée en deuxième intention après l’échocardiographie dans l’évaluation des cardiopathies congénitales répond en premier lieu au souci légitime d’éviter, dans la mesure du possible, le recours chez l’enfant ou le jeune adulte à des techniques d’exploration invasives et/ou génératrices de rayonnements ionisants.The incidence of congenital heart disease in live births is low; nevertheless, its early diagnosis and the establishment of appropriate presurgical management are critical. The significant progress in surgical treatment realized over the last two decades has resulted in the survival of many patients who would previously have died from their congenital malformation. Recently, a trend toward the use of new techniques such as magnetic resonance imaging (MRI) has been observed, in an attempt to obviate the need for more invasive or ionizing means. MRI has the potential of providing three-dimensional detailed anatomic and functional information on the cardiovascular system. It is not limited by intervening gas or bones, and provides a large view. Thus MRI clearly emerged as an excellent imaging modality complementary to sonography both during the preoperative and during the postoperative management. MRI bridges the gap between echocardiography and angiography, and consequently diagnostic angiographic procedures have become unnecessary. Indeed angiography has drawbacks; it is risky and invasive, exposes to radiation hazards and requires deeper sedation which is deleterious especially in evaluating a pediatric population.

Eleven patients with a total of 17 palliative systemic-pulmonary artery shunts underwent evaluation by electrocardiogram-gated magnetic resonance imaging (GMRI). GMRI successfully imaged 11 of 17 shunts (65%), including five of nine Blalock-Taussig shunts, four of six Glenn shunts, and both aortopulmonary shunts. All shunts except for the Waterston were imaged on coronal sections during end-systole. The single Waterston shunt was seen on sagittal and transverse scans. Shunt localization and identification were facilitated by obtaining multiple, contiguous sections through the body. Glenn shunts could be imaged entirely in one section, although multiple sections were required to locate the correct plane. Blalock-Taussig shunts generally required multiple sections to image different segments of the shunt. Both aortopulmonary shunts were seen as direct side-to-side connections of the aorta and pulmonary artery. GMRI permitted assessment of the size, course, patency, and distribution of systemic-pulmonary artery shunts as well as the size and morphology of the proximal pulmonary arteries. We conclude that GMRI is a useful, noninvasive method for imaging the anatomy of systemic-pulmonary artery shunts.