The Use of Dual Energy Computerized Tomography to Detect Residual Viable Hepatocellular Carcinoma after Transarterial Chemoembolization

Abstract

Objective: To determine the value of the dual energy computerized tomography (DECT) for detection of residual viable hepatocellular carcinoma (HCC) after transarterial chemoembolization (TACE). 
Methods: Single-source (ss) DECT of liver was performed in adult patients who were diagnosed as HCC and treated with TACE at Siriraj Hospital during October 1st, 2013- December 31st, 2014. The diagnostic 5-point performance score of conventional liver CT imaging set (CCTI) and iodinated material density imaging set (IMDI)
obtained simultaneously by using DECT, were evaluated by two radiologists. The follow up imaging at 6 months was regarded as gold standard. The sensitivity and specificity were calculated by assigned score 4 or 5 lesions as positive for the presence of HCC, assigned score 1 or 2 lesions as negative for viable tumor and assigned score 3 lesions as uncertain diagnosis. McNemar’s test was used to compare the sensitivity and specificity between CCTI and IMDI. The reading time of both technique and radiation dose were recorded and the mean reading time were compared using a paired t-test.
Results: Out of total 21 patients with 66 lesions, 81% were male and 19% were female with mean age 61.8 ± 10.2 years old. After monitoring for 6 months, 35 of the total 66 lesions were still viable HCCs and 31 lesions became non-viable HCCs. CCTI had excellent inter-observer agreement while IMDI had moderate agreement (K = 0.931
and 0.534, respectively). The sensitivity of CCTI and IMDI for detection of viable tumor were 88.6% and 100%, respectively (p-value cannot be computed). The specificity of CCTI and IMDI were 96.8% and 93.5%, respectively (p-value = 1.000). The mean reading time of two radiologists for CCTI was 151.2 ± 134.7 seconds and 123.2 ±
126.8 seconds for IMDI (p-value = 0.048). Total radiation dose of dynamic liver CT was 1194.22 ± 179.44 mGy cm.
Conclusion: IMDI has higher sensitivity for detection of viable HCCs after TACE and consumes less reading time than CCTI.