Proteomics-informed reconstruction of myocardial metabolism reveals profound bioenergetic failure in patients with anthracycline-induced cardiotoxicity

Background

Anthracyclines remain a cornerstone of contemporary anti-cancer therapy, yet their clinical utility is limited by severe cardiotoxicity. Experimental data implicate impaired myocardial energetics as a driver of anthracycline-induced cardiotoxicity (AIC). However, assessment of myocardial metabolism in patients is challenging, and the bioenergetic phenotype of AIC remains poorly characterised.

Purpose

To investigate impaired myocardial metabolism in patients with AIC using proteomics-based, patient-specific computational metabolic modelling.

Methods

Left ventricular biopsies were obtained from patients with advanced AIC (n=17; 53±8 years) and age-matched non-failing donor hearts (n=15). Myocardial proteins were quantified by unbiased high-resolution mass spectrometry. Differentially expressed proteins were functionally annotated through Gene Ontology enrichment analyses. Based on the individual proteomics-derived enzyme abundance profile, patient-specific computational models of myocardial metabolism were generated, encompassing 296 biochemical reactions across all main energy-generating pathways. Proteomic and metabolic alterations were correlated with clinical characteristics, cardiac imaging parameters, and outcomes.

Results

Patients with AIC exhibited severely reduced left ventricular systolic function (LVEF 20±5%) and marked symptom burden (76.5% NYHA class IV) despite guideline-directed medical therapy. Most had received anthracyclines as part of combination chemotherapy for non-Hodgkin lymphoma or breast cancer (76.7%).

Among 6,940 identified proteins, 527 (7.6%) were significantly upregulated and 305 (4.4%) downregulated in AIC (Figure 1A). Upregulated proteins were mainly associated with inflammatory and adaptive immune pathways, whereas downregulated proteins were enriched for mitochondrial respiration and ATP synthesis (all FDR <0.01).

Integration of patient-specific proteomic profiles into computational metabolic models revealed a markedly reduced maximal ATP production capacity in AIC compared with controls, evident under both fasting and post-prandial conditions (Figure 1B-C). AIC was characterised by a shift in myocardial substrate utilisation from fatty acid oxidation towards greater glucose use (Figure 1D). Myocardial uptake capacity for ketone bodies and branched-chain amino acids was significantly impaired (p<0.01), whereas maximal fatty acid uptake capacity was not different.

Conclusions

AIC is characterised by profound bioenergetic impairment driven by downregulation of mitochondrial oxidative pathways and accompanied by a maladaptive shift in substrate utilisation. Proteomics-based metabolic network reconstruction enables quantitative delineation of cardiac bioenergetics in patients and identifies impaired ATP generation as a central pathomechanism. Further research is warranted to evaluate targeted metabolic interventions as novel cardioprotective strategies in cancer survivors.Figure 1