Pre-Clinical Research - Ablation and histology experiments in pigs


Healthy pigs were treated with RFA using RITA Medical Systems' RF generator. The animals were anesthetized according to the protocols of the Austrian animal care law and the position of the liver was localized using computer tomography (CT) according to the protocols. During the RFA continuous temperature at each RITA electrode, used power, and the efficiency of heating related to the tissue conductivity were recorded.

Pigs were sacrificed after 1 day / 7 days /14 days / 21 days /1 month/ 2 months and 3 month after the RFA. In a part of the treated animals, the livers were perfusion fixed and in the other half immersions fixed with formalin. Consequently, the livers were resected and trimmed for areas of interest and embedded in paraffin. Biopsies of the livers were also taken for storage in liquid nitrogen for further investigations. The perfusion fixed large liver blocks were scanned using a Micro-CT and then prepared for large scale sectioning and histological staining. To visualize the ablated area methylenblue-staining (MB) and chromotrope-aniline blue (CAB) staining were performed. Stained sections were scanned in high resolution (4800 dpi) for processing with the developed automatic image algorithms.

In further experiments healthy liver tissue was resected and heated for 10, 20 and 30 minutes at different temperatures to investigate cell death in liver tissue. Also the influence of temperature on hepatocytes and fibroblasts in cell-culture system was investigated.

To incur cirrhosis in pigs, N-Nitrosodiethylamine (DENA) injection was performed once a week. Each second week blood samples were taken to monitor liver enzymes and the pigs' general condition. These samples were stored for further analysis. Each 2 to 3 months a CT scan was performed in order to check the state of the liver. The injection of DENA to induce cirrhosis as well as tumours in pigs is going on. Unfortunately no tumours occurred. However, we hope to be able to deliver pigs suffering from HCC at the end of the following period; at least the induction of a broader range of different grades of cirrhosis as well as fibrosis of the liver tissue will be delivered to our partners in order to be able to evaluate the effects on RFA outcome.


The aim of the preclinical experiments on pig was to provide data for the modelling and validation activities. The macroscopic RFA-data collected on large animals (pigs) and microscopic data collected in cell-culture experiments were obtained. In addition MicroCT volumens and histological images of the ablated liver have been acquired. These had to provide ground truth information about the lesion volume incurred by RFA.

Tissue experiments:

The aim of our heating experiments was to investigate the impact of heat on liver tissue.

At first, healthy liver tissue was cut into sections and heated for 10, 20 and 30 minutes at different temperatures (55°C, 65°C, 75°C, 85°C and 100°C) in order to evaluate the influence of different temperatures and time-spans on liver tissue. PH-value was measured continuously during these heating experiments to document pH changes influenced by heating. Afterwards the tissues of the different experimental groups were embedded in paraffin and in liquid nitrogen for further immunohistochemical and molecular investigations. The immunohistochemical stained sections were negative against caspase-3. For that reason we switch to cell-culture experiments.

In cell-culture, we performed a set of heating experiments to investigate the sensitivity of hepatocytes and fibroblast cell lines to heat injury for different grades of simulated fibrosis in order to gain useful information on the impact of the liver tissue composition on the occurrence of cell death. Different amount of fibrosis was simulated by co-culture of hepatocytes and fibroblasts. These were exposed to increasing temperatures. Immediately, after 24 and 48 hours after exposure for different time spans the viability was tested using MTS-test.

Statistical analysis was performed to investigate the influence of the heating on cell survival. Our results show highly significant differences between the temperature capacitance of liver carcinoma cells (HepG2) and fibroblasts (MRC-5). HepG2 cells alone are most sensitive to heat-induced cell death and the sensitivity decreased with rising percentages of MRC-5 cells in the co-culture. Furthermore heating for 5 minutes for 55°C and 65°C does not cause cell death, but leads to a proliferation stimulus. Currently, experiments on a human hepatic stellate cell line (LX-1)

Changes in necrosis zone:

RFA is considered to be successful in patient, if the resulting necrosis zone encompasses the whole tumour and a safety margin without damaging too much surrounding healthy organ tissue. Furthermore, a follow up scan one month after RFA should not show a local recurrence of the tumour. It is known that the rapid and fixative effect of the RFA in combination with the coagulative effect of the vessel may hinder cells to induce an inflammatory cascade, while also minimising the dispersal of any pro-inflammatory cytokines produced by halting blood flow within the ablation. For that reason the strategy of healing is the encapsulation of the coagulated area without inflammatory response. A recent study found only a minimal apoptotic and heat shock activity in the transition zone in an immediately scarified pig. However, after a surviving time of 5 days the pig showed apoptosis and high HSP70 expression.

Based on these findings we investigated the processes going on in the RF affected tissue in the time frame between directly after RFA up to 3 month after treatment. Experiments were performed in healthy pigs with a surviving time after treatment of 0 day, 1 day, 2 days, 1 week, 2 weeks, 1 month, 2 months and 3 months. The livers of the pigs were cut, trimmed to the area of interest, immersion fixed in 4% formalin and embedded in paraffin according to standard procedure. We start to set up the protocols for histological and immunohistological staining to visualize and evaluate the changes in the coagulation area and the necrosis zone. Among others, the use of antibodies against heat shock protein 70 (HSP70), M30, Caspase 3 (apoptosis marker), Ki-67 (proliferation marker) is planned to evaluate the necrosis zone and the coagulation area over the time. Professional support will by provided by a pathologist.

Experiments on shape and geometry in large animals

38 healthy pigs were treated with RFA during the first and second year. The animals were anesthetized according to the protocols of the Austrian animal care law and the position of the liver was localized using computer tomography (CT) according to the protocols. Thereafter, one lesion in the peripheral parts of the liver and one central lesion in each pig were set using different RFA probes. During RFA several blood samples were taken continuously and aliquoted for further analysis. After RFA, the pigs were sacrificed and the livers were resected and trimmed for areas of interest. The dissected tissues were embedded in paraffin or stored in liquid nitrogen for further investigations. Besides, healthy tissue (liver, kidney, spleen, intestine, skin, etc) from one animal served as control sections in immunohistochemistry.
To generate CT images with different setups, we have tested two different RFA needles from RITA and Radionics. In each liver we applied two RFAs, one in the peripheral and one central near to big vessels. Additionally, we performed RFA in hypertonic and hypotonic pigs.
Future work will focus on the conductivity measurements in vivo during the RFA. These data are necessary for the bio-heat model development.

Histology preparation for image processing

Preparation of histology slices required a dedicated - non-standard - procedure to facilitate lesion reconstruction using semi-automatic image processing methods. The procedure is optimized to minimize tissue deformations inferred by sectioning as well as maximizing the contrast between RFA lesion and healthy tissue.

To minimize the effect of mechanical stress during processing we adapted different protocols for perfusion fixation. Hereby the anesthetised, heparinised pig is flushed with NaCl 0,9% to avoid blood trombi. The vessels were used to provide a fixative (4% formalin) to the organ (perfusion fixation). Afterwards the organ is explanted and trimmed to the area of interest. The tissue blocks were stored into 4% formalin until the fixation was complete (i.e. the whole tissue was saturated with the fixative). This additional immersion into the fixative guaranteed that no changes occurred in macroscopical morphology of the tissue. MicroCT scans of those blocks were taken to facilitate the recovery of the deformations due to slicing and staining.

To reduce the number of sections of ablation area, thus reducing the number of registration steps, we prepared the liver blocks for large scale sectioning (liver blocks up to 3x3x5 cm). Large section histology is a method of choice for cancer histopathology in gynaecological and obstetrics areas. This method is based on fixation and dehydration (alcohol serial, xylol and acetone incubation) of the tissue followed by its embedding into paraffin. We have adapted this method by optimizing the preparation of each sample to the special needs of histology imaging.

CAB and MB staining were selected for imaging the ablated area. Trichrome staining was not suitable for large scale sections as tissue sections got damaged during the long preparation time required by the staining protocol. The stained 1-2µm thick sections were uncovered (without coverslip), and placed directly on the scanner to avoid any optical artefacts. This procedure guarantees that the scans were in high resolution with smallest possible background. High resolution (4800dpi) scanning was performed to generate the histology images.

Establishing a tumour model in pigs:

Because the composition in a cirrhotic liver changes influences the effect of RFA, it is necessary to set up a large animal model to investigate the influence of fibrosis/cirrhosis on RFA outcome.

Guided by the literature, weekly injections of N-Nitrosodiethylamine (DENA) into pigs are beeing performed. We inject 15mg/kg DENA per pig while keeping injection during the whole investigation time. Each second week blood samples are taken to monitor liver enzymes and the pigs' general condition. Samples were stored for further analysis. Each 2 to 3 months a CT scan is performed in order to check the state of the liver.

Until now, no DENA-treated pig developed a HCC but different forms of changes in the liver tissue were observed ranging from steatosis and steatohepatitis as well as high grades of cirrhosis and angiosarcoma of the liver. Therefore, we will be able to get detailed insights in processes during RFA in livers of different tissue compositions despite the fact that up to now, the induction of a HCC in pigs was not successful in our laboratory. However, the HCC, the 5th most common malignant tumour in human, always occurs in combination with pathological changes in liver (fibrosis/cirrhosis) and therapeutic options highly depend on these changes.

Consequently to support our partners with data of RFAs performed in livers of different grades of cirrhosis we will focus in the next project period on different strategies of DENA application to generate a wider range of different liver damages. We plan to evaluate the grade of liver cirrhosis using the Child-Pugh score, additionally to pathological evaluation.

Copyright 2006 - Design by LS