Assessment of therapeutic response in hepatocellular carcinoma treated with percutaneous radio frequency ablation: comparison of multiphase helical computed tomography and power Doppler ultrasonography with a microbubble contrast agent.

OBJECTIVE: To compare the results of multiphase helical computed tomography and power Doppler ultrasonography with a microbubble contrast agent in the assessment of the therapeutic response to radio frequency ablation in hepatocellular carcinoma. METHODS: In 66 patients with 73 nodular hepatocellular carcinomas ranging from 1.0 to 4.0 cm (mean, 2.6 cm) in diameter, contrast-enhanced power Doppler ultrasonography was performed after intravenous bolus injection of a galactose-based microbubble contrast agent before and after radio frequency ablation. The results of the studies were compared with the findings of follow-up 3-phase helical computed tomography. All patients were regularly followed up with computed tomography for more than 1 year (range, 13-19 months). RESULTS: In 8 (11%) of 73 hepatocellular carcinomas, immediate follow-up computed tomography obtained within 2 hours after radio frequency ablation showed focal enhancing portions within the treated lesions, suggesting residual non-necrotic tumors. All 8 of these tumors had intratumoral flow signals on contrast-enhanced power Doppler ultrasonography. The diagnostic agreement between computed tomography and contrast-enhanced power Doppler ultrasonography was achieved in 100%. Among the remaining 65 hepatocellular carcinomas with the absence of residual tumors at both immediate follow-up computed tomography and contrast-enhanced power Doppler ultrasonography, subsequent follow-up computed tomography showed local regrowth at the margins of 10 lesions (15%). CONCLUSIONS: The results of contrast-enhanced power Doppler ultrasonography closely correlated with those of immediate follow-up computed tomography for detecting residual tumors in hepatocellular carcinomas treated with radio frequency ablation. Both techniques, however, showed a limitation in detecting small or microscopic residual tumors and in predicting local regrowth in the treated lesions.