Accurate differentiation of focal nodular hyperplasia from hepatic adenoma at gadobenate dimeglumine-enhanced MR imaging: prospective study

Abstract

Purpose: To prospectively determine the accuracy of differentiating benign focal nodular hyperplasia (FNH) from hepatic adenoma (HA) and liver adenomatosis (LA) by using gadobenate dimeglumine-enhanced magnetic resonance (MR) imaging.

Materials and methods: The ethics committee at each center approved the study, and all patients provided informed consent. Seventy-three patients with confirmed FNH and 35 patients with confirmed HA (n = 27) or LA (n = 8) underwent MR imaging before (T2-weighted half-Fourier rapid acquisition with relaxation enhancement or T2-weighted fast spin-echo and T1-weighted gradient-echo [GRE] sequences) and at 25-30 seconds (arterial phase), 70-90 seconds (portal venous phase), 3-5 minutes (equilibrium phase), and 1-3 hours (delayed phase) after (T1-weighted GRE sequences only, with or without fat suppression) bolus administration of 0.1 mmol per kilogram of body weight gadobenate dimeglumine. The enhancement of 235 lesions (128 FNH, 32 HA, and 75 LA lesions) relative to the normal liver parenchyma was assessed. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy for the differentiation of FNH from HA and LA were determined.

Results: Hyper- and isointensity on T2-weighted and iso- and hypointensity on T1-weighted GRE images were noted for 177 (88.9%) of 199 lesions visible on unenhanced images. On dynamic phase images after contrast material administration, 231 (98.3%) of 235 lesions showed rapid strong enhancement during the arterial phase and appeared hyper- to isointense during portal venous and equilibrium phases. Accurate differentiation of FNH from HA and LA was not possible on the basis of precontrast or dynamic phase images alone. At 1-3 hours after contrast material enhancement, 124 (96.9%) of 128 FNHs appeared hyper- or isointense, while 107 (100%) HA and LA lesions appeared hypointense. The sensitivity, specificity, PPV, NPV, and overall accuracy for the differentiation of FNH from HA and LA were 96.9%, 100%, 100%, 96.4%, and 98.3%, respectively.

Conclusion: Accurate differentiation of FNH from HA and LA is achievable on delayed T1-weighted GRE images after administration of gadobenate dimeglumine.