In 8 pigs, super-selective catheterizations of the right hepatic artery were achieved (Fig. 1). Then, the conventional IRE in the left liver lobe (control group; C-IRE) and subsequently IRE in the right liver lobe with simultaneous continuous lidocaine injection through a catheter positioned selectively in the right liver artery (study group; L-IRE) was performed. Two different IRE protocols were used with the intention to create different sizes of IRE zones (tip exposure and applicator spacing Protocol-1/Protocol-2: 2.0 and 2.0 cm/2.5 and 2.5 cm) (Fig. 2). Then animals were killed and livers explanted 2 hours after IRE. Applying CT and non-enhanced MRI, size and shape of the IRE zones were analyzed by using a semi-automated software (MITK freeware “Geometric evaluation of ablations”; www.mitk.org; German Cancer Research Center, Heidelberg, Germany) [3-5]. Two independent readers segmented all IRE zones twice. The averages were involved in the further statistical analysis, and intra-observer and inter-observer agreements of the measurements were calculated applying the Bland-Altman analysis with bias and 95% limits of agreement [6]. IRE generator protocols were compared between groups.

The early gross and histopathological changes of the central and peripheral IRE zones were analyzed and compared applying different and specific staining techniques [e.g. hematoxylin and eosin (HE) and TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling)] [3]. All pathologic examinations were carried out by pathologists under blind estimations.