Plasma GLP-1 and metabolic dynamics during human liver regeneration and their association with posthepatectomy liver failure

Metabolic regulation is critical during liver regeneration in rodents, but human data are limited. We investigated perioperative dynamics of circulating metabolites and plasma levels of glucagon-like peptide-1 (GLP-1) and GLP-2, in patients undergoing liver resections, exploring their associations with the histological phenotype of metabolic dysfunction-associated steatotic liver disease (MASLD) and posthepatectomy liver failure (PHLF).

Significant metabolic changes occur during human liver resection, particularly in phospholipid metabolism, along with distinct perioperative dynamics of GLP-1 and GLP-2, closely linked to PHLF and independent of the histological phenotype of MASLD. Additionally, we provide exploratory results on the predictive value of GLP-1 for PHLF, emphasizing a holistic model of liver function assessment highlighting the metabolic component of human liver regeneration.

Eighty-one and 75 patients from two centers between 2012 and 2023 were studied. Targeted quantitative metabolomic assay of 180 circulating metabolites, perioperative GLP-1, GLP-2, and standard lipid parameter level evaluation was employed. An exploratory PHLF prediction model was developed, including GLP-1 as a metabolic parameter.

Significant alterations of 44 metabolites by postoperative day (POD) 1 and 40 by POD5 were observed, mainly among phospholipid species. Unsupervised clustering identified two metabolic clusters, with one encompassing 93% of PHLF patients by POD5 (P<0.001). Standard plasma lipid parameters displayed consistent decrease after hepatectomy, independent from MASLD phenotype, with the lowest levels in PHLF patients. Postoperative GLP-1 and GLP-2 dynamics displayed a reciprocal pattern, indicating adaptive change in secretion. Preoperative GLP-1 levels were significantly increased in PHLF (P=0.02). Furthermore, incorporation of GLP-1 into the established aspartate aminotransferase to platelet ratio index (APRI) + albumin-bilirubin (ALBI) score, improved PHLF prediction [area under the curve (AUC): 0.833, 95% confidence interval (CI): 0.660-0.964]. Conclusions: Significant metabolic changes occur during human liver resection, particularly in phospholipid metabolism, along with distinct perioperative dynamics of GLP-1 and GLP-2, closely linked to PHLF and independent of the histological phenotype of MASLD. Additionally, we provide exploratory results on the predictive value of GLP-1 for PHLF, emphasizing a holistic model of liver function assessment highlighting the metabolic component of human liver regeneration.