18F-FDG PET/MRI and clinical risk factors for early cardiotoxicity induced by anthracycline in patients with malignant tumor

Introduction: To investigate the availability of ¹⁸F-FDG PET/MRI for early myocardial damage and the risk factors of early cardiotoxicity induced by anthracycline in patients with malignant tumor.

Methods: Cancer patients after anthracycline chemotherapy and controls without chemotherapy were prospectively recruited. All patients fasted for at least 12 hours and had their blood glucose level measured prior to injection of ¹⁸F-FDG. Cardiac magnetic resonance examinations including left ventricular ejection fraction (LVEF), mass, T1 and T2mapping was explored. Myocardial uptake of ¹⁸F-FDG were analyzed by semi-quantitative (maximum standardized uptake value, SUV_max) methods. SUV_max-heart/SUV_{max-background (scapularmuscles)} ratios were calculated. Receiver operator characteristics (ROC) curve analysis was performed to determine optimal cut-off values of those PET/MRI imaging criteria for evaluating early anthracycline cardiotoxicity, taking ECG-positive as the end point. A binomial logistic regression was performed to evaluate PET/MRI and clinical parameters in a multivariate analysis.

Results: ¹⁸F-FDG FDG PET/MRI was performed in thirty-one patients (mean age 49±15 years) and 17 controls (mean age 28±7 years) in control group (CG) were recruited. Chemotherapy group had higher native T1 value (1194.4±41.5 versus 1167.4±39.7ms, P=0.030) than control group. LVEF (60.50±8.20 versus 61.74±6.55%, P=0.625), T2 values (51.9±6.7 versus 51.1±5.4 ms, P=0.321) were similar between chemotherapy group and CG.

T2 values (55.8±4.6 versus 51.1±5.4 ms, P=0.040) from the patients after 3-4 cycle chemotherapy were higher than CG. There was no significant difference in LVEF (P=0.435), native T1 value of left ventricular in septal (P=0.230) and lateral wall (P=0.099) between the two groups.

ROC curves showed optimal thresholds of T1 value (lateral wall) and SUV_max-heart/SUV_{max-background} ratio were 1182ms and 10.2 respectively. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of T1 value (lateral wall) and SUV_max-heart/SUV_{max-background} ratio were 57.14% (4/7), 66.67% (14/21), 36.36% (4/11), 82.35% (14/17), 64.29% (18/28); 100.00%(8/8), 72.73%(8/11), 72.73%(8/11), 100.00%(8/8), 84.21%(16/19). When T1 value greater than 1185ms and SUV_max-heart/SUV_{max-background} ratio greater than 10.2 were combined, the specificity and accuracy achieved 77.78% and 85.7%.

Univariate analyses indicated that the SUV_max-heart/SUV_{max-background} ratio (β =0.462, P=0.022, OR=1.588), age (β =0.113, P=0.038, OR=1.119), weight (β =0.213, P=0.034, OR=1.237) and height (β =0.192, P=0.027, OR=1.211) were risk factors of early anthracycline cardiotoxicity. Binary logistic regression found only SUV_max-heart/SUV_{max-background} ratio (β =0.633, P=0.035, OR=1.883) was significantly associated with early anthracycline cardiotoxicity with height (β =0.330, P=0.054, OR=1.391) in the equation.

Conclusions: ¹⁸F-FDG PET/MRI could assess early cardiotoxicity induced by anthracycline by using 10.2 as the threshold of after treatment SUV_max-heart/SUV_{max-background} ratio and T1 value higher than 1185ms. After chemotherapy, patients with increased FDG uptake are exposed to a higher risk of cardiotoxicity. Clinical predictors need large studies to explore.

• Download figure
• Open in new tab
• Download powerpoint