Circulating beta-hydroxybutyrate levels in advanced HFrEF: the determinants and prognostic impact

Abstract Background Beta-hydroxybutyrate (BOHB) is the main ketone body and is generated mainly in the liver during states of exessive fatty acid (FA) utilisation, such as starvation or diabetic ketosis. BOHB serves as a metabolic substrate to heart. Mild ketosis or enahancement of BOHB utilisation may favorably affect cardiac function in preclinical HF models. Patients with HF have increased BOHB levels, but regulation of ketosis in HF is poorly understood. Purpose To investigate the determinants and prognostic implact of circulating BOHB levels in patients with advanced HFrEF. Methods We examined 847 pts with symptomatic HFrEF (age 57±11y, 50% CAD, 82% males, NYHA 2.7±0.6, LVEF 22±5%, BNP 802±850 pg/ml), referred to IKEM for TX or device (ICD/CRT/LVAD) implant. Patients underwent clinical and echocardiographic examination, MLHFQ questionnaire, circulating metabolite assessment, body composition (DEXA, n=150) and right heart catheterisation (n=383). BOHB in peripheral vein was measured in all subjects after overnight fasting by an enzymatic metod (Autokit 3-HB, Wako). Patients were followed for occurrence of an event (death/urgent Tx/LVAD). Results Median BOHB level was 64 (IQR: 33-161) mmol/l. BOHB was higher in DM patients (111 vs 163, <0.001) and inreased with age. Body composition, congestion status, insulin resistance (HOMA-IR), liver function tests, weight loss history, HF symptoms (NYHA, MLHFQ), gender, HF etiology or HF therapy (BB, ACE/ARB) were unrelated to BOHB levels. BOHB strongly correlated with markers of lipolysis (free FA, BNP) and GDF15, less strongly with pulmonary hypertension (PA systolic pressure, pulmonary vascular resistance-PVR), inflammation (CRP level, leukocytes%) and humoral regulators of ketogenesis in the liver (insulin/glucagon ratio). After median follow-up of 1126 days (IQR: 410; 1781), 60% patients experienced an event. In Cox analysis, Increased BOHB levels predicted worse survival (RR: 1.001, 95%CI: 1.0005-1.001, p<0.0001), particularly in the upper BOHB tercile. Impaired survival was also associated with elevated free FA (RR: 1.78 (1.23-2.51, p= 0.002). In bivariate Cox analysis of free FA and BOHB levels, only free FA, but not BOHB, remained significantly asccociated with adverse outcome. Conclusion Increased BOHB levels in HFrEF correlate with free FA levels, likely due to sympatetic activation-driven and BNP-driven lipolysis. The association of increased BOHB with adverse outcome dissapears after adjustment to free FA level. Other notable associations of BOHB were diabetic status, GDF15 level, systemic inflammation and pulmonary hypertension.