Comparison of LI-RADS v.2017 and ESGAR Guidelines Imaging Criteria in HCC Diagnosis Using MRI with Hepatobiliary Contrast Agents

Abstract

Purpose The purpose of this study was to assess and compare diagnostic ability of LI-RADS (LR) v. 2017 and ESGAR guidelines in hepatocellular carcinoma (HCC) diagnosis using MRI with hepatobiliary contrast agents. Methods Seventy pathologically confirmed lesions in 32 patients (24 males and 8 females) who had MRI with hepatobiliary contrast done before surgery or biopsy were reviewed retrospectively. Six lesions were <10mm, 31 lesions 10-19mm, and 33 lesions ≥20mm. Two readers assessed all lesions according to LI-RADS v.2017 criteria and ESGAR consensus statement on liver MR imaging and clinical use of liver-specific contrast agents. Statistical analysis was performed to compare diagnostic ability of both guidelines including receiver operative curves (ROC) and area under curve (AUC). Results For LR ≥ 4 sensitivity, specificity, accuracy, and AUC were 96%, 75%, 88.6%, and 85.5, respectively. For LR5 they were 74%, 95%, 80%, and 84.5, respectively. For ESGAR criteria with major and additional features, they were 88%, 75%, 84.3%, and 81.5, respectively. For ESGAR criteria only with major features they were 78%, 80%, 78.6%, and 79, respectively. AUC analysis revealed that overall diagnostic ability of LI-RADS was higher than ESGAR but the results did not show statistical significance. Conclusions Both LI-RADS and ESGAR guidelines presented high diagnostic ability in HCC diagnosis of MRI studies with hepatobiliary contrast agents. More complex LI-RADS criteria performed better than ESGAR guidelines and it may justify extra effort that needs to be put in the report. However, the results were not statistically different and the simplicity of the ESGAR guidelines should also be taken into consideration.

Diagnostic Performance of CT/MRI Liver Imaging Reporting and Data System v2017 for Hepatocellular Carcinoma: A Systematic Review and Meta‐Analysis

“…In that study, five of 10 studies had only used pathology as the reference standard, which might have led to verification bias and overestimation of the probability of HCC for LR‐2 5 . In our meta‐analysis using CT/MRI LI‐RADS v2017, only one of six studies that reported the proportion of HCC for LR‐2 used pathology 13 and the other five studies used pathology and CCRS as the reference standard 12,16,18,21,24 . In addition, the definition of the LR‐2 distinctive nodule without malignant features (solid nodule <20 mm distinctive in imaging appearance compared to background nodules and with no major feature of HCC, no feature of LR‐M, and no ancillary feature of malignancy) was clearer in CT/MRI LI‐RADS v2017 4 compared to the LR‐2 cirrhosis‐associated nodule in LI‐RADS v2014, 29 which may refine the LR‐2 category so that it achieves ≤5% probability of HCC.…”

Performance of initial LI-RADS 2018 treatment response in predicting survival of patients with hepatocellular carcinoma following TACE: a retrospective, single-center cohort study

“…The LI-RADS is currently the most popular system of lesion classification in CT and MRI studies in patients with increased risk of HCC (Chernyak et al 2018 ; Kim et al 2019 ; Aslam et al 2020a ; Rosiak et al 2018b ). There are many studies showing the high specificity and sensitivity of LI-RADS for HCC reporting (Rosiak et al 2018a ; Furlan 2019 ).…”

Diagnostic performance of LI-RADS for MRI and CT detection of HCC: A systematic review and diagnostic meta-analysis

“…The 18 studies comprised a total of 3386 patients (age range, 20-88 years). Three studies were prospective [17][18][19] and 14 were retrospective [20][21][22][23][24][25][26][27][28][29][30][31][32][33]. The design of another study was unclear [34].…”

“…Nine studies enrolled subjects consecutively [17,[22][23][24]27,28,30,31,34]; the remaining nine studies did not describe how subjects were enrolled [18][19][20][21]25,26,29,32,33]. Seven studies used only histopathology as a reference standard [20,[26][27][28]30,33,34] and 11 studies used both histopathology and follow-up imaging as the reference standard [17][18][19][21][22][23][24][25]29,31,32]. Ten studies had fewer than three readers [18,[20][21][22][23][24][25][26]28,32], while eight studies had three or more readers [17,19,27,…”