Myocardial extracellular volume derived from contrast-enhanced chest computed tomography in longitudinal evaluation of cardiac toxicity in patients treated with immune checkpoint inhibitors

Background: The immune-related adverse effects after immune checkpoint inhibitors (ICIs) treatment have always been a hot topic. Although the incidence of myocarditis is not high among the related adverse effects, the mortality rate is extremely high once it occurs. In the past, the risk of cancer therapy-related cardiac dysfunction (CTRCD) after drug treatment was evaluated based on imaging examinations, but this evaluation still had certain limitations. Currently, the extracellular volume (ECV) score measurement calculated using cardiac magnetic resonance T1 mapping has become a reliable method for evaluating myocardial toxicity and computed tomography (CT) examination may become an alternative. This study aimed to longitudinally evaluate the cardiac toxicity of patients treated with ICIs using myocardial ECV derived from contrast-enhanced chest CT. Methods: A total of 500 patients with III-IV lung cancer and esophageal cancer treated with ICIs were evaluated. Participants underwent baseline examination and at least 1 follow-up examination after treatment. Contrast-enhanced chest CT-ECV, left ventricular ejection fraction (LVEF), and measurement of cardiac troponin T (cTnT) were conducted before the first treatment, 3-6 months after the first treatment, and about 12 months after the first treatment, respectively. The ECV value of the middle part of the left ventricular septum was evaluated on CT venography and plain scan, the LVEF value was evaluated by color Doppler ultrasound, and the quantity of cTnT was detected by chemiluminescence. Cancer therapy-related cardiac dysfunction was recorded. Results: The mean baseline LVEF value was 68.51%+/- 4.81% (N-0=500), and those of LVEF1, LVEF2, and LVEF3 were 68.77%+/- 4.30%, 68.16%+/- 3.59%, and 66.23%+/- 4.20%, respectively (N-1=500, N-2=467, and N-3=361, respectively). There was no significant difference between LVEF1, LVEF2, and LVEF0 (P-1=0.095, P-2=0.062), whereas LVEF3 was significantly lower than LVEF0 (P<0.001). The average baseline cTnT(0) value was 7.42 +/- 3.95 (N-0=500). The values of cTnT(1), cTnT(2), and cTnT(3) were 10.05 +/- 11.40, 12.24 +/- 13.59, and 14.54 +/- 14.49, respectively (N-1=500, N-2=467, N-3=361). The values of cTnT(1), cTnT(2), and cTnT(3) were significantly higher than cTnT(0) (P-1<0.001, P-2<0.001, P-3<0.001). The average ECV0 was 47.14%+/- 7.48% (N-0=500). ECV1, ECV2, and ECV(3 )were 50.85%+/- 6.79%, 53.44%+/- 6.96% and 52.64%+/- 7.58% respectively (N-1=500, N-2=467 and N-3=361). ECV1, ECV2, and ECV3 were significantly higher than ECV0 (P-1<0.001, P-2<0.001, P-3<0.001). CTRCD occurred in 49 patients. There were significant differences between the CTRCD (+) group and the CTRCD (-) group in cTnT(1), cTnT(2), and cTnT(3) (P-1<0.001, P-2<0.001, and P-3<0.001, respectively) and in ECV1, ECV2, and ECV3 (P-1=0.039, P-2=0.041, and P-3=0.013, respectively). Conclusions: CT-ECV began to increase at the early stage after the treatment of ICIs. CT-ECV is a potential biomarker for dynamically monitoring the cardiac toxicity of tumor patients after receiving ICIs. ECV may be used to speculate the CTRCD caused by the treatment of ICIs.