Nuclear magnetic resonance-based lipid metabolite profiles for differentiation of patients with liver cirrhosis with and without hepatocellular carcinoma

Background: Hepatocellular carcinoma is frequently unrecognized in its early stage limiting the access to the first therapeutic steps resulting in a low cure rate. Therefore, an early diagnosis is crucial. In this scenario the analysis of lipidome and metabolome emerged as a promising tool for early detection. Aims: Aim of the study was to characterize metabolomic profiles as novel markers of early hepatocellular carcinoma. Methods: Serum basal levels of metabolites, isolated from a cohort of 90 patients (n = 30 early stage; n = 30 advanced stage; n = 30 liver cirrhosis) were analysed using a nuclear magnetic resonance spectroscopy platform. To assess the predictive value of nuclear magnetic resonance profiles, we included the magnetic resonance imaging follow up of control patients with liver cirrhosis. Results: Significant differences were observed in the levels of individual parameters that included total cholesterol, LDL and HDL subclasses, Isoleucine, Valine, Triglycerides, Lactate, Alanine, Albumin, alpha Fetoprotein, Dimethylamine, Glycerol, and total Bilirubin levels in cancer compared to liver cirrhosis (p < 0.05). Furthermore, a significant difference in glycerol levels (p < 0.05) and a decreasing trend in dimethylamine were observed in cirrhotic patients who later developed HCC (16%, n = 5). Retrospective MRI analysis revealed precursor lesions in 3/5 patients, initially not classified as HCC due to their size and hemodynamic features. Conclusion: Nuclear magnetic resonance based assessment of lipidomic and metabolomic profiles permit the differentiation of cancer from liver cirrhosis. The data obtained suggests a possible role of lipidomic based serum profiles for early detection.