Myocardial ATP depletion detected noninvasively predicts sudden cardiac death risk in patients with heart failure

Abstract

BACKGROUNDSudden cardiac death (SCD) remains a worldwide public health problem in need of better noninvasive predictive tools. Current guidelines for primary preventive SCD therapies, such as implantable cardioverter defibrillators (ICDs), are based on left ventricular ejection fraction (LVEF), but these guidelines are imprecise: fewer than 5% of ICDs deliver lifesaving therapy per year. Impaired cardiac metabolism and ATP depletion cause arrhythmias in experimental models, but to our knowledge a link between arrhythmias and cardiac energetic abnormalities in people has not been explored, nor has the potential for metabolically predicting clinical SCD risk.METHODSWe prospectively measured myocardial energy metabolism noninvasively with phosphorus magnetic resonance spectroscopy in patients with no history of significant arrhythmias prior to scheduled ICD implantation for primary prevention in the setting of reduced LVEF (≤35%).RESULTSBy 2 different analyses, low myocardial ATP significantly predicted the composite of subsequent appropriate ICD firings for life-threatening arrhythmias and cardiac death over approximately 10 years. Life-threatening arrhythmia risk was approximately 3-fold higher in patients with low ATP and independent of established risk factors, including LVEF. In patients with normal ATP, rates of appropriate ICD firings were several-fold lower than reported rates of ICD complications and inappropriate firings.CONCLUSIONTo the best of our knowledge, these are the first data linking in vivo myocardial ATP depletion and subsequent significant arrhythmic events in people, suggesting an energetic component to clinical life-threatening ventricular arrhythmogenesis. The findings support investigation of metabolic strategies that limit ATP loss to treat or prevent life-threatening cardiac arrhythmias and herald noninvasive metabolic imaging as a complementary SCD risk stratification tool.TRIAL REGISTRATIONClinicalTrials.gov NCT00181233.FUNDINGThis work was supported by the DW Reynolds Foundation, the NIH (grants HL61912, HL056882, HL103812, HL132181, HL140034), and Russell H. Morgan and Clarence Doodeman endowments at Johns Hopkins.

Keywords: Arrhythmias; Bioenergetics; Cardiology; Heart failure.

Conflict of interest statement

Conflict of interest: The authors have declared that no conflict of interest exists.

Figures

Figure 1. Representative CMR image and 31P MR spectra.

Representative axial MRI (right) of the chest of a 45-year-old woman (lying prone) who subsequently experienced an arrhythmic event. Yellow lines denote localized volumes from which 31P MRS spectra (left) were derived (arrows). The upper spectrum from the heart shows phosphocreatine (PCr), diphosphoglycerate (DPG), and the 3 phosphate resonances of ATP. The myocardial ATP was low (2.5 μmol/g wet wt) but PCr/ATP was normal (1.7) in this individual. The lower spectrum includes chest muscle and is shown for comparison but was not used in analysis.

Figure 2. Event-free survival in patients with low versus normal myocardial energetics.

Kaplan-Meier curves depicting the proportion of event-free survival across a median follow-up period of 10.7 years (range: 3.2–14.7 years) in individuals with low (solid line) versus individuals with normal (dashed line) myocardial high-energy phosphates. We defined low myocardial energetics as 2 standard deviations below the grouped mean of healthy individuals previously reported by our group (ATP: A, C, E, and G) Event-free survival is shown where the event was defined as the composite endpoint of appropriate ICD firing or cardiac death. (B, D, F, and H) Event-free survival is shown where the event was defined as appropriate ICD firing only. Low myocardial ATP concentration was associated with lower event-free survival when considering the composite endpoint (log-rank, P = 0.0079) or appropriate ICD firing alone (log-rank, P = 0.024). Neither low PCr or low CK flux were significant predictors of event-free survival (log rank, P > 0.05). (A and B) n = 32 versus 14 (normal versus low ATP); (C and D) n = 17 versus 29 (normal versus low PCr); (E and F) n = 28 versus 13 (normal versus low CK flux); (G and H) n = 30 versus 16 (normal versus low PCr/ATP).

Figure 3. Event-free survival based on cardiac ATP and LVEF.

Kaplan-Meier curves depicting the proportion of event-free survival across a median follow-up period of 10.7 years (range: 3.2–14.7 years) in individuals grouped by myocardial ATP (normal versus low ATP, P = 0.024). The lowest arrhythmic risk subgroup had normal ATP and less depressed LVEF and exhibited nearly a 90% event-free survival at 10–15 years. The highest arrhythmic risk subgroup had both low ATP and the lowest LVEF, with a 5-year event-free survival of only approximately 35%. Low ATP, low ejection fraction n = 6; low ATP, lower ejection fraction n = 8; normal ATP, low ejection fraction n = 17; normal ATP, lower ejection fraction n = 15. The log-rank P value reflects the overall comparison of the 4 groups, and a significant P value rejects the null hypothesis that the groups are identical.