Disentangling metabolic impairment in the liver-heart axis: tissue-specific insulin sensitivity in type 2 diabetes

Aims/hypothesis The liver-heart axis in type 2 diabetes (T2D) reflects key metabolic interactions disrupted by insulin resistance (IR). Organ-specific effects of insulin remain unclear due to challenges in measuring tissue-level IR. This study aims to define liver-heart phenotypes and their associated metabolic impairments, which relate to hepatic fat accumulation linked to metabolic dysfunction-associated steatotic liver disease (MASLD) and coronary artery calcifications (CACs) tied to cardiovascular disease (CVD). Methods In this cross-sectional study, 41 individuals with controlled T2D underwent biochemical tests and [18F]FDG PET/CT scans before and after a hyperinsulinemic euglycemic clamp (HEC). Tissue-specific insulin-mediated glucose uptake was derived from PET imaging, while CT provided data on radiodensity, volume, fat, and calcifications. Results A strong inverse correlation was observed between myocardial and liver ΔSUV (r=-0.74, p=2×10 -7 ), thus suggesting the liver-heart metabolic axis in T2D. Three phenotypes were determined according to increased risks of T2D comorbidities including MASLD and CVD: HepGluc[+]+mIR (high risk of MASLD and CVD), HepGluc[−]+mIR (high risk of CVD, low risk of MASLD), and HepGluc[−]+mIS (low risk of MASLD and CVD). Moreover, HOMA-IR was only reflective of organ-level dysfunction in HepGluc[+]+mIR, which was the most at-risk phenotype in terms of systemic and tissue-specific metabolic impairment, including higher inflammation, IR, liver fat, CACs, and biomarkers of MASLD and CVD. Conclusions/interpretation This study explores a potential liver-heart metabolic axis in T2D, linked to insulin-mediated dysfunction that may originate in the heart and extend to the liver. The coexistence of organ-specific phenotypes suggests three possible risk profiles, with the HepGluc[+]+mIR phenotype appearing most consistent with advanced T2D progression. Careful identification of this phenotype could support improved monitoring and more personalized treatment strategies in T2D.