Perturbed pathways of lipid metabolism in patients with chronic hepatitis B infection after cessation of long term nucleos (t)ide analogue

Abstract

<p>Background and aims: The ‘stop-to-cure’ approach is a potential strategy to achieve finite duration of nucleoside analogue (NUC) in patients with chronic hepatitis B (CHB) infection. We examined the serial serum metabolomic profiles among flare compared to nonflare CHB patients who had stopped NUCs, followed by in silico data mining to identify relevant pathways. Method: Patients with HBeAg-negative non-cirrhotic CHB on ≥3 years of NUC were recruited. NUCs were stopped and they were prospectively followed up every 6 weeks for virological relapse (VR: serum HBV DNA rise from undetectable to >2000 IU/ml) ± clinical flare (VR plus alanine aminotransferase [ALT] >2 times upper limit of normal), with serial blood sampling for metabolomic profiling at week 0, at NUC resumption (for flare patients, or non-flare patients with VR or HBeAg sero-reversion), and at ALT normalization following NUC resumption among flare patients. A total of 8 patients (4 flare [among which 2 developed subsequent HBsAg seroclearance, SC], 4 VR or HBeAg sero-reversion) were included in this interim analysis. Serum metabolomics data was acquired with 1D 1H-Nuclear magnetic resonance spectroscopy. Metabolomics data was analysed with MetaboAnalyst 6.0. GEO datasets was analyzed with GEO2R. Results: Metabolites involved in lipid metabolism (triglyceride and total fatty acid) were significantly downregulated in flare patients compared to non-flare patients at NUC resumption. Among flare patients, succinic acid is significantly downregulated in those who subsequently achieved SC compared to flare-no-SC patients upon ALT normalization. Dimethylsulfone involved in triacylglycerol degradation is upregulated in flare patients with SC, compared to flare-noSC and non-flare patients, at all timepoints. Compartment specific in silico data mining based on GEO DataSets showed genes involved in fatty acid synthesis (FAR1, FASN), fatty acid elongation in mitochondria (NDUFAB1), fatty acid elongation in endoplasmic reticulum (ELOVL1, ELOVL5) were significantly increased in the liver of CHB patients (GSE230397) compared to control, but not in peripheral blood CD8 T cells (GSE217838), indicating a compartment specific imbalance of lipid metabolism. In CHB patients with severe flare, fatty acid beta-oxidation associated genes ACAT1, ACSL1 and ACSL4 were significantly downregulated in the PBMCs of the survival group compared to patients who died (GSE168049). Conclusion: We demonstrated alterations of lipid metabolism pathways in CHB patients with post-NUC cessation flare that were associated with clinical outcomes. Fatty acid metabolism is perturbed in patients with CHB and is inferred to be associated with survival in those with severe HBV flare based on in silico datasets analysis. Our work reveals an imbalance in lipid metabolism in CHB both in spatial location and molecular signature.</p>