An Algorithm for the Diagnosis of Focal Liver Masses Using Microbubble Contrast-Enhanced Pulse-Inversion Sonography

The objective of this study was to develop an algorithm for liver mass diagnosis using microbubble contrast-enhanced pulse-inversion sonography. Ninety-six lesions in 92 patients were evaluated with DMP 115 (Definity)-enhanced pulse-inversion sonography, comprising 44 malignancies (29 hepatocellular carcinomas, 12 metastases, two peripheral cholangiocarcinomas, and one hepatic lymphoma) and 52 benign lesions (26 hemangiomas, 20 focal nodular hyperplasias, and six others). All had continuous low-mechanical-index imaging through the arterial and portal venous phase. A three-person blind review evaluated single images at baseline, early and peak arterial phases, and through the extended portal phases with a movie showing arterial phase wash-in. Reviewers assessed lesional vascularity and enhancement blindly but did not make a diagnosis. Combinations of answers were compared with independently determined final diagnoses to develop an algorithm for liver mass diagnosis. Portal phase enhancement comprises the first step of the algorithm, with positive or sustained enhancement identifying 48 (92%) of 52 benign lesions and negative enhancement or washout present in 41 (93%) of 44 malignancies. Sustained portal phase enhancement with arterial phase peripheral nodularity and centripetal progression predicted 24 (92%) of 26 of the hemangiomas; diffuse arterial phase enhancement greater than the liver identified 19 (95%) of 20 of the focal nodular hyperplasias. With negative portal phase enhancement, arterial phase information was less effective at differentiating hepatocellular carcinoma (25 [86%] of 29 cases) from another hepatic malignancy (11 [73%] of 15 cases). A simple diagnostic algorithm for interpretation of microbubble-enhanced sonography provides sensitive and accurate diagnosis of commonly encountered liver masses.