The objective of the study was to determine the performance of Acoustic Radiation Force Impulse (ARFI) imaging to differentiate benign from malignant liver masses, both of hepatocellular origin and metastases, by quantification of their stiffness.Methods
This study has institutional review board approval and informed consent. Eighty-nine patients (42 female and 47 male patients) with 105 liver masses had ARFI evaluation on ultrasound, S2000 (Siemens, Mountain View, Calif). Mean age of the patients was 53.67 years (range, 27–83 years). Mean diameter of the masses was 2.77 cm (range, 1.0–13.0 cm). Final diagnoses, confirmed by imaging on contrast-enhanced computed tomography, magnetic resonance, or ultrasound or biopsy, include hepatocellular carcinoma (n = 28), metastasis (n = 13), hemangioma (n = 35), focal nodular hyperplasia (n = 15), focal fat sparing (n = 8), focal fat deposit (n = 4), and adenoma (n = 2). Receiver operating characteristic analysis was performed to evaluate the diagnostic accuracy of the ARFI measurement and to extract the optimal cutoff values in the differentiation of benign from malignant disease.Results
Acoustic Radiation Force Impulse values showed a statistically significant difference between benign (1.73 [SD, 0.8] m/sec) and malignant masses (2.57 [SD, 1.01] m/sec) (P < 0.001). However, the area under the receiver operating characteristic curve was 0.744, suggesting only fair accuracy. For differentiation of malignant from benign masses, the sensitivity, specificity, positive predictive value, and negative predictive value were 68% (28/41), 69% (44/64), 58% (28/48), and 77% (44/57), respectively, when 1.9 m/sec was chosen as a cutoff value, reflective of a wide variation of ARFI values in each diagnosis. For differentiation of metastasis from benign masses, sensitivity, specificity, positive predictive value, and NPV were 69% (9/13), 89% (57/64), 56% (9/16), and 93% (57/61), respectively, when 2.72 m/sec was chosen as a cutoff value.Conclusions
Acoustic Radiation Force Impulse measurement may be helpful to differentiate benign masses from metastases, in particular. Otherwise, ARFI measurements alone do not differentiate benign and malignant masses because of variations in stiffness of all types of masses.