Aim: To determine whether the findings of microflow imaging (MFI), composed of a flash replenishment and a maximum intensity holding sequence, using contrast-enhanced sonography, correlate with the degree of histological differentiation of hepatocellular carcinoma (HCC).
Methods: This study was approved by the institutional review board; patients gave informed consent. The samples comprised of 61 nodules histologically diagnosed as HCC: 20 well-differentiated, 26 moderately-differentiated, and 15 poorly-differentiated HCC. SonoVue was used as the ultrasound (US) contrast agent. The US equipment used was a SSA-770 A with the imaging mode set at MFI. MFI is an imaging method combining flash replenishment imaging and maximum intensity holding. Two independent readers (readers 1 and 2) classified the microflow images into four patterns: (i) normal pattern; (ii) cotton pattern; (iii) vascular pattern; and (iv) dead wood pattern. The results were compared with the degree of histopathological differentiation of the HCC.
Results: In each of the 61 HCC, blood vessels in the tumor were clearly resolved down to their fine branches. With regard to the relationship between imaging patterns and thehistological findings, it was found (with high percentages) that the normal and cotton patterns were associated with well-differentiated HCC, that the vascular pattern was associated with moderately-differentiated HCC, and that the dead wood pattern was associated with poorly-differentiated HCC. If HCC with the normal and cotton patterns were assessed as well differentiated and those with the vascular or dead wood pattern were assessed as moderately or poorly differentiated, the sensitivity, specificity, and accuracy of these assessments were found to be 85%, 92.7%, and 90%, respectively, for reader 1, and 85%, 82.9%, and 83.6%, respectively, for reader 2.
Conclusion: The angioarchitecture and hemodynamics of HCC could be evaluated in detail using MFI. The results of this study demonstrate the feasibility of a non-invasive preoperative diagnosis of the histological differentiation of HCC using MFI.