Table 2.

Analysis of the secondary end points.

Secondary end point	Analysis
Detect GLSa decrease of >15% from baseline, measured on ECHOb over the time window of 12 months.	The number of patients with a >15% decrease will be compared between groups with the Fisher exact test. The mean difference in proportions at 12 months and its 80% CI will be reported.
See if the changes in biomarkers will correlate with LVEFc measurements, assessed by ECHO and CMRd.	The association of changes in biomarkers and LVEF will be assessed with a linear regression model, while adjusting for oedema.
See if the changes in biomarkers will correlate with GLS measurements, assessed by ECHO.	The association of changes in biomarkers and GLS will be assessed with a linear regression model, while adjusting for oedema.
Compare the time to biomarkers’ positivity to the time to decrease in GLS >15% or decline in LVEF ≥10% in points with a final LVEF of <53% measured on ECHO.	The times will be compared with the Mann Whitney U test.
Find out if patients with increased baseline biomarkers will develop cardiotoxicity; identify predictors of cardiotoxicity by multivariable analysis.	A univariable and multivariable logistic model will be used.
Detect MACEe (defined as acute myocardial infarction, hospitalization due to heart failure, atrial flutter or fibrillation, and ventricular tachycardia) or death due cardiac problems during follow-up.	The rate of each overall MACE and that of each event will be computed per 100 person-year with 80% CI. Kaplan Meier curves will be plotted.
Assess the role of fibrosis on CMR (T1f mapping with evaluation of extracellular volume) after cardiotoxic radiation therapy or systemic therapy in predicting the incidence of cardiotoxicity.	A univariable and multivariable logistic model will be used.
Detect incidence of acute asymptomatic pericarditis after RTg, measured on CMR.	The proportion of patients with acute asymptomatic pericarditis and 80% CI will be computed.
Investigate if the area of the oedema on CMR correlates with RT dose distribution.	The Spearman correlation coefficient and 80% CI will be computed.
To assess the incidence of myocardial oedema on CMR (T2h mapping) after radiation therapy and cardiotoxic systemic therapy measured on CMR and ECHO over the time window of 12 months from the end of radiation therapy.	The proportion of patients with oedema and 80% CI will be computed.

^aGLS: global longitudinal strain.

^bECHO: echocardiography.

^cLVEF: left ventricular ejection fraction.

^dCMR: cardiac magnetic resonance.

^eMACE: major cardiovascular events.

^fT1: longitudinal relaxation time.

^gRT: radiotherapy.

^hT2: transverse relaxation time.