Myocardial fibrosis identification in patients with anthracycline-induced cardiotoxicity: a biomarkers and cardiac magnetic resonance correlation

Abstract Background Anthracycline-induced cardiotoxicity (AIC) remains a significant challenge in Oncology. Early detection is crucial for optimizing therapeutic strategies. Emerging biomarkers of fibrosis, including soluble suppression of tumorigenicity-2 receptor (sST2) and galectin-3 (Gal-3), show promise for earlier diagnosis. Recent advances in cardiac magnetic resonance with gadolinium (CMR) offer a comprehensive assessment (late gadolinium enhancement, LGE, T1 mapping and extracellular volume fraction, ECV) to detect fibrosis. Purpose We investigated the correlation between novel biochemical markers of fibrosis and CMR findings in patients with AIC, evaluating their relationship with established biomarkers such as high sensibility troponin I (cTnI) and brain natriuretic peptide (BNP). Methods We prospectively enrolled 90 patients with a history of AIC (i.e., 10% decline in left ventricular ejection fraction, LVEF, from baseline to 50%). Patients were categorized in two groups: those with LVEF recovery after therapy (5% increase and final LVEF 50%, LVEF-rec group) and those without (LVEF-norec group). CMR assessments included LV strain (GLS), LGE, and T1 mapping. The biomarker panel included clinically established biomarkers (cTnI and BNP), emerging fibrosis biomarkers (sST2) and research-oriented biomarkers (Gal-3, Enhanced Liver Fibrosis, ELF, Score, procollagen III, PIII, amino terminal peptide, and tissue inhibitor of metalloproteinase 1, TIMP-1). Multivariable logistic regression and receiver operating characteristic (ROC) analysis were performed to assess predictive value. Results In our population, 69 patients were included in the LVEF-rec group and 21 in the LVEF-norec group. LVEF-norec group patients were significantly older (mean age 63 vs. 55 years, p=0.04) and experienced a more severe LV systolic dysfunction after therapy with anthracyclines (LV EF nadir 36% vs. 43%, p 0.0001). There were no significant differences between the two groups in terms of use of different cardioprotective drugs, anthracycline dose, therapy duration or use of HER2 inhibitors. LVEF-norec group exhibited significantly greater levels of TnI, BNP, Gal-3, and PIIINP (p0.05), while ELF score and TIMP-1 showed no differences between the two groups. LGE mass was comparable between groups, but T1 relaxation time was prolonged in LVEF-norec group (p=0.013), indicating diffuse fibrosis. GLS was significantly lower in the LVEF no-rec group, suggesting greater impairment in myocardial contractility. A biomarker-based risk score incorporating Gal-3, BNP, TnI, and PIIINP demonstrated good predictive value for persistent dysfunction (AUC=0.74). Conclusion Fibrosis-related biomarkers, particularly Gal-3 and PIIINP, strongly correlate with structural myocardial changes in AIC. The integration of serum biomarkers with CMR parameters enhances risk-stratification and may improve early identification of patients at risk of poor cardiac recovery.