Interreader reliability of liver imaging reporting and data system treatment response: A systematic review and meta-analysis

Abstract

Background: For a proper management strategy in patients with locoregionally treated hepatocellular carcinoma (HCC), it is essential that the Liver Imaging Reporting and Data System (LI-RADS) treatment response algorithm (LR-TR) has high interreader reliability. We aimed to systematically evaluate the interreader reliability of LR-TR and sources of any study heterogeneity. Methods: Original studies reporting the interreader reliability of LR-TR were identified in MEDLINE and EMBASE up to 20 September 2020. The pooled kappa coefficient (κ) was calculated using the DerSimonian–Laird random effects model. Subgroup analyses were performed according to imaging modality (magnetic resonance imaging (MRI) or computed tomography (CT)). Meta-regression analyses were performed to explore study heterogeneity. Results: Eight studies with 851 HCCs were finally included. Pooled κ was 0.70 (95% CI, 0.58–0.82) for CT/MRI LR-TR, and those of MRI and CT were 0.71 (95% CI, 0.53–0.89) and 0.71 (95% CI, 0.65–0.78), respectively. Study design (p < 0.001) and type of treatment (p = 0.02) were significantly associated with substantial study heterogeneity. Conclusion: LR-TR showed substantial interreader reliability regardless of the imaging modality. Because of substantial study heterogeneity, which was significantly associated with study design and type of treatment, published values for the interreader reliability of LR-TR should be interpreted with care.