Comparison of Image Quality and Diagnostic Performance of Cone-Beam CT during Drug-Eluting Embolic Transarterial Chemoembolization and Multidetector CT in the Detection of Hepatocellular Carcinoma.

PURPOSE: To compare image quality and diagnostic performance of cone-beam computed tomography (CT) and multidetector CT in the detection of hypervascular hepatocellular carcinoma (HCC) in patients with cirrhosis undergoing transarterial chemoembolization with drug-eluting embolic agents.

MATERIALS AND METHODS: Fifty-five consecutive patients referred for chemoembolization of hypervascular HCC were prospectively enrolled. Imaging included preprocedural multidetector CT within 1 month before planned treatment, intraprocedural cone-beam CT, and 1-month follow-up multidetector CT. Analysis of image quality was performed with calculations of lesion-to-liver contrast-to-noise ratio (LLCNR) and lesion-to-liver signal-to-noise-ratio (LLSNR). One-month follow-up multidetector CT was considered the reference standard for the detection of HCC nodules.

RESULTS: Median LLCNR values were 3.94 (95% confidence interval [CI], 3.06-5.05) for preprocedural multidetector CT and 6.90 (95% CI, 5.17-7.77) for intraprocedural cone-beam CT (P < .0001). Median LLSNR values were 11.53 (95% CI, 9.51-12.44) for preprocedural multidetector CT and 9.36 (95% CI, 8.12-10.39) for intraprocedural cone-beam CT (P < .0104). Preprocedural multidetector CT detected 115 hypervascular nodules with typical HCC behavior, and cone-beam CT detected 15 additional hypervascular nodules that were also visible on 1-month follow-up multidetector CT.

CONCLUSIONS: Cone-beam CT has a significantly higher diagnostic performance compared with preprocedural multidetector CT in the detection of HCCs and can influence management of patients with cirrhosis by identifying particularly aggressive tumors.