Background In patients with heart failure due to dilated cardiomyopathy, cardiac energy metabolism is impaired, as indicated by a reduction of the myocardial phosphocreatine-to-ATP ratio, measured noninvasively by ³¹P-MR spectroscopy. The purpose of this study was to test whether the phosphocreatine-to-ATP ratio also offers prognostic information in terms of mortality prediction as well as how this index compares with well-known mortality predictors such as left ventricular ejection fraction (LVEF) or New York Heart Association (NYHA) class.

Methods and Results Thirty-nine patients with dilated cardiomyopathy were followed up for 928±85 days (2.5 years). At study entry, LVEF and NYHA class were determined, and the cardiac phosphocreatine-to-ATP ratio was measured by localized ³¹P-MR spectroscopy of the anterior myocardium. During the study period, total mortality was 26%. Patients were divided into two groups, one with a normal phosphocreatine-to-ATP ratio (>1.60; mean±SE, 1.98±0.07; n=19; healthy volunteers: 1.94±0.11, n=30) and one with a reduced phosphocreatine-to-ATP ratio (<1.60; 1.30±0.05; n=20). At reevaluation (mean, 2.5 years), 8 of 20 patients with reduced phosphocreatine-to-ATP ratios had died, all of cardiovascular causes (total and cardiovascular mortality, 40%). Of the 19 patients with normal phosphocreatine-to-ATP ratios, 2 had died (total mortality, 11%), one of cardiovascular causes (cardiovascular mortality, 5%). Kaplan-Meier analysis showed significantly reduced total (P=.036) and cardiovascular (P=.016) mortality for patients with normal versus patients with low phosphocreatine-to-ATP ratios. A Cox model for multivariate analysis showed that the phosphocreatine-to-ATP ratio and NYHA class offered significant independent prognostic information on cardiovascular mortality.

Conclusions The myocardial phosphocreatine-to-ATP ratio, measured noninvasively with ³¹P-MR spectroscopy, is a predictor of both total and cardiovascular mortality in patients with dilated cardiomyopathy.