Ultrasonographic characterization of hepatic cryolesions. An ex vivo study

To determine the physical basis for the ultrasonographic characteristics of the hepatic ice ball produced by cryotherapy and the size correlation between the actual hepatic ice ball and the ultrasonographic cryolesion.

Experimental ex vivo study involving controlled freezing with liquid nitrogen recirculating probes of fresh porcine livers immersed in various solutions at ambient temperatures (20.2 degrees C to 22.6 degrees C), together with measurements of the impedance of frozen and unfrozen liver.

First, the hyperechoic rim is caused by reflection of 34% of ultrasound waves at the interface between unfrozen and frozen liver as a consequence of an increased acoustic impedance of frozen liver that was calculated to be approximately 3.8 times that of unfrozen liver tissue. The increased acoustic impedance is due to the decrease in elasticity of hepatic tissue as it freezes. Second, the posterior acoustic shadowing is partly due to the attenuation of the incident ultrasound waves by reflection at the interface between unfrozen and frozen liver. It is also dependent on the crystalloid-protein content of hepatic parenchyma, which ensures a homogeneous lesion by preventing "shattering" within the cryolesion. This is in sharp contrast to the ultrasonographic appearance of an ice ball formed in ionized water, in which the hyperechoic rim overlies an area of posterior acoustic enhancement. Third, the correlation of the size between the ultrasonographic cyrolesion and the measured hepatic ice ball approached unity (r = .99), and the two measurements were identical for cryolesions less than 50 mm in diameter.

Ultrasound is an accurate method for depicting the actual diameter of frozen solid hepatic tissue in cryotherapy for liver tumors, but the present technology does not provide accurate assessments of the volume of frozen tissue.