How does alteration of hepatic blood flow affect liver perfusion and radiofrequency-induced thermal lesion size in rabbit liver?

The purpose of this study was to test the hypothesis that decreasing liver perfusion In rabbits results In an increase in thermal lesion size and that these effects can be accurately monitored using magnetic resonance Imaging (AMI). We additionally tested the hypothesis that the increase in thermal lesion size would depend on the particular vessel or vessels occluded (hepa tic artery, portal vein, or both). Using an Institutional Animal Care and U se Committee approved protocol, 20 New Zealand white rabbits were randomly assigned to four treatment groups (five in each group): control and ligatio n of portal vein (PV), hepatic artery (HA), or both PV and HA (HAPV). Surgi cal ligation of the appropriate vessel was performed under general anesthes ia. Immediately after ligation, the rabbits were placed in a 0.2-T open AM system, and an 18-G copper radiofrequency (RF) electrode with a 2-cm expose d tip was Inserted into the liver. RIP was applied for 10 minutes with the tip temperature maintained at 90 degrees h +/- 2 degreesC. Before and after ablation, perfusion data were obtained for 90 seconds using 30 3-second se quential single oblique-slice fast imaging with steady-state progression (F ISP) acquisitions after injection of gadolinium-diethylene triamine pentaac etic acid (Gd-DTPA) via the Inferior vena cava. Postablation scanning inclu ded axial and oblique turbo spin-echo (TSE) T2-weighted (T2w), STIR, and Gd -enhanced T1w sequences. Lesion size was determined perpendicular to the RF electrode using software calipers on the imager. The rabbits were sacrific ed after completion of the post-therapy scans, and their livers were harves ted for histologic analysis. The liver showed a mean increase In signal amp litude (SA) of 76% 24 seconds after Gd contrast injection in the control gr oup. After contrast injection, the SA increased to a mean of only 66% in th e group with ligated hepatic arteries, with no difference In the time to pe ak compared with the control group. No significant SA increase over baselin e could be found In the groups with ligated PV or ligated PV and HA. T2-wei ghted Images demonstrated the highest lesion-to-liver contrast-to-noise rat ios (CNRs; mean -5.5) on postprocedure images, followed by STIR images (mea n -2.2) in the control group. The lesions were poorly delineated on the Gd- enhanced Images. Average maximum lesion sizes (mean +/- 95% confidence inte rval) were 22 +/- 4.3 mm after ligation of PV, 22 +/- 2.6 mm after ligation of both PV and HA, 14 +/- 2.0 mm after ligation of HA, and 13 +/- 1.9 mm i n the control group. We accept the hypothesis that the diameter of the regi on of coagulation necrosis achieved by standardized RF ablation in the live r increases with reduced organ perfusion and that this effect can be accura tely monitored using MRI. The major factor influencing the size of the coag ulation area is the portal venous flow. Occlusion of the hepatic artery alo ne does not significantly increase lesion size. T2w sequences are best suit ed for postprocedure imaging due to the high lesion-to-liver CNR in rabbits with normal hepatic perfusion. (C) 2001 Wiley-Liss, Inc.