MR elastography-based prediction of development of hepatocellular carcinoma in patients with chronic hepatitis B with sustained virological response.

[OBJECTIVES] The goal of this study was to develop a risk score based on MR elastography (MRE) to predict hepatocellular carcinoma (HCC) development in patients with chronic hepatitis B (CHB).

[MATERIALS AND METHODS] A total of 345 patients with CHB who underwent 2D/3D MRE between July 2015 and December 2018 were enrolled and then randomly assigned to training (n = 243) and validation (n = 102) cohorts. An MRE-based HCC risk score was developed for the prediction of HCC development based on a multivariable Cox model and compared with previous clinical scores. The predictive performance was evaluated using the C-index and time-dependent ROC.

[RESULTS] The 2D/3D MRE-based risk scores incorporating age, platelet count, albumin, and liver stiffness provided better predictive performance in HCC development than three existing clinical risk scores [CAMD (cirrhosis, age, male sex, and diabetes mellitus), PAGE-B (platelet age gender-B), and mPAGE-B (modified platelets, age, gender-hepatitis B)] in the training (C-index: 0.859 and 0.872, respectively, vs 0.762, 0.754 and 0.818, all p value < 0.05) and validation cohorts (C-index: 0.878 and 0.887, respectively, vs 0.815, 0.709 and 0.810, all p value < 0.05). The 2D and 3D MRE-based risk scores provided high negative predictive values in the training and validation cohorts at 3 years (97.8-100.0%) and 5 years (94.6-100.0%) with the two optimal cut-off values of 43.2 and 69.2, respectively, and 56.5 and 71.6, respectively.

[CONCLUSIONS] Both 2D and 3D MRE-based risk scores may serve as a valuable tool for predication of HCC development for CHB patients and provided superior predictive performance compared to existing clinical scores.

[KEY POINTS] Question Can MRE be a useful component of a risk score to predict HCC development in CHB patients with sustained virological response? Findings 2D/3D MRE-based stiffness values were independent predictors for HCC development, and the 2D/3D MRE-based risk scores demonstrated better performance than three existing clinical risk scores. Clinical relevance The 2D/3D MRE-based risk scores provide high negative predictive values for HCC development, which may become a noninvasive tool for clinicians to stratify CHB patients for HCC surveillance and to improve early HCC detection and reduce mortality.