Evaluation of small (<or=2 cm) dysplastic nodules and well-differentiated hepatocellular carcinomas with ferucarbotran-enhanced MRI in a 1.0-T MRI unit: utility of T2*-weighted gradient echo sequences with an intermediate-echo time.

Objective: To evaluate the detectability and signal intensities of small (

Methods: Pathologically confirmed DNs (n=13) and w-HCCs (n=31) with a median largest dimension of 1.1cm were scanned using ferucarbotran-enhanced MRI. Conventional T2*-weighted GRE sequences (conventional-T2*-GRE: repetition time, 280 ms; echo time, 14 ms; flip angle, 60 degrees) and specific T2*-weighted GRE sequences using an intermediate-TE (specific-T2*-GRE: repetition time, 140 ms; echo time, 8 ms; flip angle, 30 degrees) were obtained before and after ferucarbotran administration. Two independent observers scored all nodules for visibility and assigned confidence level scores to their observations. To assess the effect of ferucarbotran, the tumor-liver signal contrast-to-noise ratio (tumor-liver-CNR) was also calculated for detected nodules by the same two observers with consensus.

Results: There was good interobserver agreement regarding the presence of nodules for both sequence types. Qualitative and quantitative analyses indicated that specific-T2*GRE sequences were superior to conventional-T2*-GRE sequences for detecting DNs and w-HCCs with hypointense signals. The tumor-liver-CNR of DNs was significantly different between specific-T2*-GRE sequences and conventional-T2*-GRE sequences (Mann-Whitney test, P<0.001). Both qualitative and quantitative analyses indicated that conventional-T2*-GRE sequences were superior to specific-T2*-GRE sequences for detecting w-HCCs with heterogeneous and hyperintense signals.

Conclusions: Specific-T2*-GRE sequences with ferucarbotran are useful for detecting DNs and w-HCCs that produce hypointense signals on a 1.0-T MRI unit.