Evaluation of the accuracy of gadobenate dimeglumine-enhanced MR imaging in the detection and characterization of focal liver lesions

Objective: We evaluated the extent to which hepatic lesion characterization and detection is improved by using gadobenate dimeglumine for enhancement of MR images. Materials and Methods: Eighty-six patients were imaged before gadobenate dimeglumine administration, immediately after the 2 mL/sec bolus administration of a 0.05 mmol/kg dose (dynamic imaging), and at 60-120 min after the IV infusion at 10 mL/min of a further 0.05 mmol/kg dose (delayed imaging). The accuracy for lesion characterization was assessed for a total of 107 lesions. Sensitivity for lesion detection was assessed for a total of 149 lesions detected on either intraoperative sonography, iodized oil CT, CT during arterial portography, or follow-up contrast-enhanced CT as the gold standard. Results: The accuracy in differentiating benign from malignant liver lesions increased from 75% and 82% (the findings of two observers) on unenhanced images alone, to 89% and 80% on dynamic images alone (p < 0.001, p = 0.8), and to 90.7% when combining the unenhanced and dynamic image sets (p < 0.001, p = 0.023). Delayed images did not further improve accuracy (90% and 91%; p = 0.002, p < 0.05). A similar trend was apparent in terms of accuracy for specific diagnosis: values ranged from 49% and 62% on unenhanced images alone, to 76% and 70% on combined unenhanced and dynamic images (p < 0.001, p = 0.06), and to 75% and 70% on inclusion of delayed images (p < 0.001, p = 0.12). The sensitivity for lesion detection increased from 77% and 81% on unenhanced images alone, to 87% and 85% on combined unenhanced and dynamic images (p = 0.001, p = 0.267), and to 92% and 89% when all images were considered (p < 0.001, p = 0.01). Conclusion: Contrast-enhanced dynamic MR imaging with gadobenate dimeglumine significantly increases sensitivity and accuracy over unenhanced imaging for the characterization of focal hepatic lesions, and delayed MR imaging contributes to the improved detection of lesions.