Distinct metabolipidomic signatures differentiate de novo acute heart failure from decompensated chronic heart failure

Background In the broad population of patients hospitalized due to acute heart failure (AHF), two major subgroups should be distinguished: patients with de novo AHF, in whom the index hospitalization is the first manifestation of the disease, and patients with previously diagnosed chronic heart failure (HF) experiencing a worsening of heart failure (WHF) episode. Objectives To investigate whether circulating metabolomic and lipidomic profiles differ between de novo AHF and decompensated pre-existing HF. Methods We performed omics data analysis in venous blood samples collected from patients hospitalized for AHF in a single tertiary cardiology center. Results Data from 274 patients with AHF were analyzed ("wet and warm" haemodynamic profile: 96%), of whom 33% were classified as having de novo AHF. Patients with pre-existing HF were older, more frequently had coronary artery disease as the primary HF etiology, and presented with slightly higher LVEF despite comparable left ventricular end-diastolic diameters. Regarding cardiac biomarkers, including those reflecting neurohormonal activation, both groups were comparable. Patients with pre-existing HF, however, had a significantly higher burden of major comorbidities, including chronic kidney disease and diabetes. De novo AHF was characterized by higher levels of sphingolipids, angiotensin (1–7), and purine-related metabolites, indicating a pro-inflammatory and pro-apoptotic state and oxidative stress, with metabolomic signals of neurohormonal activation despite no differences in conventional neurohormonal biomarkers. WHF was associated with elevated levels of amino acid derivatives, glucuronidated compounds, and organic acids, which reflect alterations in energy metabolism and detoxification pathways. Lipidomic analysis revealed higher levels of glycerophospholipids (GPEtn, GPGro, GPA, SQDG) and lower levels of sphingomyelins and selected GPEtn species in WHF, which may be related to impaired cellular membrane integrity or cellular injury. Biomarkers of catabolism and inflammation were also associated with pre-existing disease. In exploratory multivariable models, several metabolites demonstrated consistent associations with 12-month all-cause mortality across modeling strategies, notably positive associations for glucose–lactate–acetate and nitrosyl acetate and a strong inverse association for 1,2-ethanedisulfonic acid. While additional metabolites emerged in strategy-specific analyses, only a subset retained signals after accounting for clinical covariates, supporting their potential independent prognostic relevance in this hypothesis-generating setting. Conclusion De novo AHF was characterized by a metabolomic profile indicative of a pro-inflammatory state and neurohormonal activation, while patients hospitalized for exacerbated pre-existing disease presented with metabolomic and lipidomic signatures of abnormal energy metabolism, a catabolic state, and cellular injury.