Hepatology. 2022 Mar 29. doi: 10.1002/hep.32494. Online ahead of print.
ABSTRACT
BACKGROUND AND AIMS: The diagnostic accuracy of Liver Imaging Reporting and Data System (LI-RADS) v.2018 and European Association for the Study of the Liver (EASL) criteria for the diagnosis of hepatocellular carcinoma has been widely evaluated, but their reliability has not. We aimed to assess and compare the reliability of LI-RADS v.2018 and EASL criteria for the diagnosis of hepatocellular carcinoma using MRI with extracellular contrast agents (ECA) and gadoxetic acid (GA) and to determine the effect of ancillary features on LI-RADS reliability. Approach and Results Ten readers reviewed MRI studies of 92 focal liver lesions measuring <3 cm acquired with ECA and GA less than 1 month apart from two prospective trials, assessing EASL criteria, LI-RADS major and ancillary features, and LI-RADS categorization with and without including ancillary features. Inter-reader agreement for definite HCC diagnosis were substantial and similar for the two contrasts for both EASL and LI-RADS criteria. For ECA-MRI and GA-MRI, respectively, inter-reader agreement was k=0.72 (95%CI:0.63-0.81) and k=0.72 (95%CI:0.63-0.8) for nonrim hyperenhancement, k=0.63 (95%CI:0.54-0.72) and 0.57 (95%CI:0.48-0.66) for nonperipheral washout, and k=0.49 (95%CI:0.4-0.59) and k=0.48 (95%CI:0.37-0.58) for enhancing capsule. The inter-reader agreement for LI-RADS after applying ancillary features remained in the same range of agreement.
CONCLUSION: Agreement for definite hepatocellular carcinoma was substantial and similar for both scoring systems and the two contrast agents in small focal liver lesions. The agreement for LI-RADS categorization was lower for both contrast agents, and including LI-RADS ancillary features did not improve the agreement.
PMID:35349760 | DOI:10.1002/hep.32494
