Circulating Ceramide 16:0 in Heart Failure with Preserved Ejection Fraction

Ceramides are a class of bioactive sphingolipids increasingly recognized in metabolic disease, heart failure (HF), atherosclerosis, and cardiovascular death (1). Despite accumulating evidence linking ceramides to HF risk, little is known about circulating ceramides and clinical outcomes in HF.

We performed targeted sphingolipidomics to test whether circulating ceramides are associated with death or heart failure admission (DHFA) among participants with heart failure with preserved ejection fraction (HFpEF) in the TOPCAT trial. TOPCAT data and samples were obtained from the National Heart, Lung, and Blood Institute Biolincc repository. Serum concentrations of 76 sphingolipids were determined for all participants with available samples (n = 433) using liquid chromatography-tandem mass spectrometry at the Virginia Commonwealth University Lipidomics Core Facility (Richmond, VA). Analyses were conducted in a blinded manner.

Compared to participants without available samples, the prevalence of myocardial infarction (25.1% vs 31.9%), coronary artery bypass surgery (12.1% vs 18.2%), atrial fibrillation (34.3% vs 42%), and a history of smoking (40.2% vs 47.9%) was higher in the subset with available samples (all P<0.05). Subjects in this subset also had higher body mass index (30.8 kg/m² vs 32 kg/m²) BNP levels (mean 397 pg/ml vs 508 pg/ml) and were more frequently from the Americas (49.4% vs 44.1%; all P<0.05). Amongst the 433 subjects, 103 reached the endpoint of DHFA. In univariate Cox models adjusted for multiple comparisons by a modified Bonferroni correction based on the principal components explaining >95% of the variability in measured species, increased concentrations of ceramides 16:0, 18:0 and 18:1 were associated with increased risk of DHFA (Figure 1A). When further adjusted for the MAGGIC risk score (the time of heart failure diagnosis was not available, and excluded from the score), increased concentrations of ceramides 16:0 and 18:0 were associated with the risk of DHFA (Figure 1B). Amongst these, ceramide 16:0 exhibited the most significant association (HR 1.35 95% CI 1.13–1.63, P=0.0011). Both ceramide 16:0 and 18:0 were also associated with all-cause death. When participants were stratified by tertiles of ceramide 16:0, we found no significant differences in clinical variables including age, gender, race, history of myocardial infarction, coronary revascularization, chronic obstructive pulmonary disease, hypertension, diabetes, atrial fibrillation, smoking, body mass index, use of insulin, aspirin, statins, beta blockers, renin-angiotensin-aldosterone system inhibitors, estimated glomerular filtration rate, hematocrit, or natriuretic peptide levels. We did not observe significant interactions between baseline ceramide 16:0 levels and randomization to spironolactone therapy (P = 0.60), statin therapy, or BMI for the outcome of DHFA.

Figure 1. Associations between sphingolipids and outcomes in TOPCAT.

A) Univariate standardized hazard ratios and B) MAGGIC risk score adjusted standardized hazard ratios (x-axis) versus p-values (y-axis). The red-dotted lines indicate p = 0.05, while the blue-dotted lines indicate statistical significance after correction for multiple comparisons (p = 0.0025).

Abbreviations: MH-ceramide- monohexosylceramide

Although increased ceramide 24:0 levels, the most abundant very long-chain ceramide species, were previously associated with reductions in all-cause mortality (3), we found no association between ceramide 24:0 and DHFA. Moreover, increased ceramide 16:0 was associated with DHFA in a model adjusted for ceramide 24:0 (Standardized HR = 1.35; 95%CI = 1.12–1.61; P = 0.0012).

In this study, we identify that increased circulating concentrations ceramide 16:0 and 18:0 are associated with DHFA in HFpEF, consistent with the growing body of literature that long-chain ceramides are associated with cardiovascular outcomes (3,4). In contradistinction to prior reports suggesting very-long chain ceramides are associated with reduced risk(2), we did not observe any associations between increased ceramide species and reduced DHFA risk. Our findings are consistent with the recent report by Lemaitre et al. who found that increased ceramide 16:0 is associated with incident HF (4).

Because palmitate is the most abundant fatty acid in the Western diet and we have measured predominantly palmitate-derived ceramides, our studies are consistent with the possibility that that long-chain ceramides are potential mediator of the link between poor diet and poor health outcomes. Our samples were not necessarily obtained in the fasting state, which could have confounded our results. Regardless of diet, carbon chain lengths of ceramide species are also dependent on particular ceramide synthases. Ceramide synthase 5 (CerS5), for example, is involved in the synthesis of ceramide 16:0 species. In murine models of high fat feeding, genetic ablation of CerS5 leads to decrease in ceramide 16:0 and 18:0 species in multiple tissues, decreased weight gain, and reduced white adipose tissue inflammation (5). In summary, ceramide 16:0 concentrations are associated with risk of incident heart failure, DHFA in HFpEF, and metabolic dysfunction and inflammation in animal models.

Abbreviations:

HF

heart failure

HFpEF

heart failure with preserved ejection fraction

BNP

B-type natriuretic peptide

DHFA

death or heart failure admission