Non-invasive HCC diagnosis in daily practice

Studies on the performance of non-invasive hepatocellular carcinoma (HCC) diagnosis are mainly performed under controlled conditions, which might not be met in clinical practice.

We retrospectively analyzed all patients, in which a CT-guided liver biopsy was performed in a tertiary referral center within a 9 year period due to suspected HCC. The performance of non-invasive HCC diagnosis (incl. the Liver Imaging Reporting and Data System (LI-RADS)) was assessed.

There was no significant difference in diagnostic performance between CT, conventional MRI, and Gd-EOB-DTPA-MRI. Late arterial phase images, which had a significant higher sensitivity for HCC detection compared to early arterial phase images, were obtained in less than 50% of patients. The number of false positive non-invasive HCC diagnosis was higher in non-cirrhotic compared to cirrhotic patients (n = 13/19), while 19 out of 22 false-negative HCC diagnosis occurred in cirrhotic patients. Overall, the performance of LI-RADS to detect HCC was significantly better compared to AASLD and EASL-EORTC guidelines. Bleeding complications were observed in 3% of patients, all could be managed conservatively and biopsy-related mortality did not occur. Needle tract seeding was not reported in the files.

HCC biopsy is a reliable and safe procedure. Histological validation of non-invasive HCC diagnosis seems always warranted, when the radiological hallmarks are not fully met or the observation was made in a patient being low-risk for HCC development. Adjustment of examination protocols (e.g. bolus tracking for the timing of the arterial phase) may help to evaluate focal liver lesions in a standardized fashion. The use of LI-RADS is encouraged in clinical practice, as the differential diagnosis of HCC, ICC, and cHCC-CC has profound clinical implications and LI-RADS seems to have the power to separate these entities better than classical algorithms for non-invasive HCC diagnosis.