Doxorubicin-induced cardiovascular toxicity: a longitudinal evaluation of functional and molecular markers

Aims Apart from cardiotoxicity, the chemotherapeutic doxorubicin (DOX) induces vascular toxicity, represented by arterial stiffness and endothelial dysfunction. Both parameters are of interest for cardiovascular risk stratification as they are independent predictors of future cardiovascular events in the general population. However, the time course of DOX-induced cardiovascular toxicity remains unclear. Moreover, current biomarkers for cardiovascular toxicity prove insufficient. Here, we longitudinally evaluated functional and molecular markers of DOX-induced cardiovascular toxicity in a murine model. Molecular markers were further validated in patient plasma. Methods and results DOX (4 mg/kg) or saline (vehicle) was administered intra-peritoneally to young, male mice weekly for 6 weeks. In vivo cardiovascular function and ex vivo arterial stiffness and vascular reactivity were evaluated at baseline, during DOX therapy (Weeks 2 and 4) and after therapy cessation (Weeks 6, 9, and 15). Left ventricular ejection fraction (LVEF) declined from Week 4 in the DOX group. DOX increased arterial stiffness in vivo and ex vivo at Week 2, which reverted thereafter. Importantly, DOX-induced arterial stiffness preceded reduced LVEF. Further, DOX impaired endothelium-dependent vasodilation at Weeks 2 and 6, which recovered at Weeks 9 and 15. Conversely, contraction with phenyl-ephrine was consistently higher in the DOX-treated group. Furthermore, proteomic analysis on aortic tissue identified increased thrombospondin-1 (THBS1) and alpha-1-antichymotrypsin (SERPINA3) at Weeks 2 and 6. Up-regulated THBS1 and SERPINA3 persisted during follow-up. Finally, THBS1 and SERPINA3 were quantified in plasma of patients. Cancer survivors with anthracycline-induced cardio-toxicity (AICT; LVEF < 50%) showed elevated THBS1 and SERPINA3 levels compared with age-matched control patients (LVEF ≥ 60%). Conclusions DOX increased arterial stiffness and impaired endothelial function, which both preceded reduced LVEF. Vascular dysfunction restored after DOX therapy cessation, whereas cardiac dysfunction persisted. Further, we identified SERPINA3 and THBS1 as promising biomarkers of DOX-induced cardiovascular toxicity, which were confirmed in AICT patients. Translational perspective DOX induced arterial stiffness and endothelial dysfunction, which preceded impairment of left ventricular systolic function. Hence, arterial stiffness and endothelial dysfunction represent potential early, functional markers of future cardiovascular dysfunction in patients receiving DOX. Of note, DOX-induced arterial stiffness and endothelial dysfunction were transient, highlighting the importance of timing for evaluating these vascular parameters in patients. Furthermore, anthracycline-induced cardiotoxicity patients showed elevated SERPINA3 and THBS1 in plasma, raising awareness for a role of these proteins in cardiovascular toxicity with possible diagnostic value in DOX-treated patients.