Utility of non-contrast-enhanced CT for improved detection of arterial phase hyperenhancement in hepatocellular carcinoma.

Purpose: To determine the utility of non-contrast-enhanced CT (NC-CT) for improved detection of hyperenhancement in hepatocellular carcinomas (HCC).

Methods: We performed a retrospective analysis on CT studies of 104 consecutive patients with 124 HCCs between October 2006 and December 2012. The enhancement of HCC was evaluated on quadriphasic CT: non-contrast (NC), arterial phase (AP), portal-venous phase (PVP), and delayed phases (DP) were performed. For determination of enhancement, mean attenuation of the HCC and surrounding non-tumor bearing liver in the four phases were recorded. Hyperenhancement was defined by either HCC (AP-NC) representing absolute lesion enhancement or AP (HCC-liver), representing relative enhancement. An increase of ≥10 Hounsfield units (HU) was considered as enhancement. Two radiologists (R1 and R2) independently performed a conventional qualitative assessment for hyperenhancement of HCC. The accuracy of detection by proposed absolute lesion enhancement HCC (AP-NC) was compared to relative enhancement AP (HCC-liver) and qualitative assessment by readers with the McNemar test.

Results: The proposed criteria of absolute lesion enhancement, HCC (AP-NC) identified more hypervascular nodules compared to qualitative analysis (92% vs. 62% for R1 and 69% for R2), equating to an improvement of 23%-31%. In contrast, relative enhancement AP (HCC-liver) was found to be inferior to conventional qualitative analysis (50% vs. 62%-69% for readers).

Conclusion: NC-CT is useful to improve the detection of hyperenhancement in HCCs for imaging diagnosis.