Comparison of bipolar radiofrequency ablation zones in an in vivo porcine model: Correlation of histology and gross pathological findings

2.1 Animals

The study was performed with six female domestic pigs (body weight 69±2 kg, age 149±19 days). All animals were housed at 15–24°C at the Department of Experimental Medicine (certified by ISO 9000 and ISO 8000), Charité - Universitätsmedizin Berlin, Germany using a 12:12-hrs light:dark cycle. Before RFA was started animals were anaesthetized by intramuscular (i.m.) injection of ketamine 10% (24 mg/kg, Ursotamin^TM, Serumwerk Bernburg, Bernburg, Germany) and azaperone (180 mg/animal, Stresnil^TM, Janssen Animal Health, Beerse, Belgium) in a mixing syringe and separate i.m. injection of xylazine (2.7 mg/kg, Rompun^TM, Bayer Vital GmbH, Leverkusen, Germany) and atropine (0.01 mg/kg, Atropinsulfat B. Braun, 0.5 ml/ml, B. Braun, Melsungen, Germany). Anesthesia was maintained by constant intravenous (i.v.) administration of ketamine and xylazine (8.4 ml/h, ratio ketamine/xylazine: 2.8:1). According to physiological requirements electrolyte solution(Sterofundin, B. Braun Melsungen AG, Melsungen, Germany) and lactated Ringer’s solution (Fresenius Kabi Deutschland GmbH, Bad Homburg, Germany) were applied intravenously. Each animal received additional i.m. injection of 2 μg/kg fentanyl (Fentanyl 0.5 MG, Rotexmedica GmbH, Trittau, Germany) for analgesia.

All in vivo experiments have been approved by the regional office for health and social welfare (LAGeSo, Berlin, Germany, G0281/12). All principles of laboratory animal care were followed, according to the guidelines of the European Society of Laboratory Animal.

2.2 Radiofrequency ablation

The surgical approach to the liver was ensured by a median laparotomy and an incision along the right costal arch. RFA was performed in the periphery of the liver using a bipolar ablation system (CelonPOWER System, Olympus Surgical Technologies Europe, Hamburg, Germany) and an internally cooled bipolar RFA applicator with an active length of 20 mm (CelonProSurge T-20, Olympus Surgical Technologies Europe, Hamburg, Germany). Internal cooling of the applicator was maintained by a peristaltic pump (CelonAquaflowIII, Olympus Surgical Technologies Europe, Hamburg, Germany).

For planning and exact positioning of RFA applicators, CT scans of the porcine liver were performed using a 80-slice MDCT scanner (Aquilion PRIME, Toshiba Medical Systems) with a multiphasic protocol with standardized parameters (voltage 120 kV, maximum 400 mA, mean mAs 260; collimated slice thickness: 80×0.5 mm, 40 mm total detector width; 0.5 s rotation speed, pitch-factor 0.813).

According to manufacturer’s recommendations the starting power for the RFA was set to 20 J/s (W). Power feed was regulated by a preinstalled resistance controlled automatic power mode (RCAP). Each ablation was stopped after an energy input of 10 kJ.

After RFA, the center of the ablated liver volume was marked by a cut-to-size plastic tube, which was advanced along the applicator up to the applicator’s isolator. This position of this marker corresponded to the largest cross sectional area perpendicular to the applicator.

The animals were euthanized in deep anesthesia by i.v. injection of 15 ml T 61 (Intervet Deutschland GmbH, Unterschleißheim, Germany) directly after ablation.

2.3 Evaluation of ablation zones

Not later than 20 min after euthanasia the liver segment including the RF ablation area was explanted. The liver segment with the inserted plastic tube was placed in a dedicated cutting device and was cut into halves, orthogonally to the applicator on the height of the largest cross sectional area. The RFA zones within these two halves were photographed besides millimeter scaling. Thereafter the specimens were snap-frozen with liquid nitrogen and stored at –80°C. 6–8 μm thick cryosections were cut from each sample for histological workup. Each cryosection was stained with a solution of reduced α-NADH (nicotinamide adenine dinucleotide) and NBTC (nitro blue tetrazolium chloride) for detecting viable and non-viable cells as described by Neumann et al. [19]. Additionally hematoxylin and eosin (HE) staining was performed for histopathologic characterization of thespecimens.

Qualitative analysis of the RFA zones was conducted by a board-certified pathologist (R.K.) and based on the following cellular characteristics: a) complete loss of cell structure, b) cellular degeneration with remaining cell borders and loss of NADH-diaphorase stain, c) hepatocellular separation, d) intercellular edema and reduced NADH-diaphorase stain intensity, and e) unscathed hepatic parenchyma.

Quantitative analysis of histological sections was performed using image analysis software (eSlide Manager / Genie, Leica Biosystems Imaging Inc, DB Maarn, The Netherlands) which was trained to recognize areas with similar histologic patterns regarding to the predefined cellular characteristic, and to display these areas using different colors (color-coded histological images).

The histological and macroscopic images were aligned to each other using landmarks. The applicator was marked as the ablation’s center point. The radius was defined as distance between the ablation center and the ablation border. The ablation area and the median radius (measured every 45°) of all ablation zones were determined independently by two investigators (a specialist in veterinary anatomy and histology (O.G.) and a surgeon (F.P.)). The results of both investigators were correlated to each other to rule out inter-observer differences. The median values of the results were used for further statistical analysis.

2.4 Statistical analysis

Results are presented as medians (minimum-maximum). The Wilcoxon signed-rank test was used to compare histological and macroscopic ablation areas and radii. The Pearson coefficient was computed for the correlation of histological and macroscopic ablation areas and radii of the ablation zones. Statistical analysis was performed with IBM^® SPSS^® Statistics (Version 23). Statistical significance was assumed at p≤0.05 (two sided).

3.1 Ablation zones

Eight bipolar ablations were performed on open cavity in six pigs. Median ablation time was 11 minutes and 43 seconds (10:01–13:33 min). Ablations were round and confluent. One ablation appeared to be asymmetric due to a major hepatic vessel adjacent to the applicator.

3.2 Qualitative analysis

Histopathological analysis revealed different ablation zones within the ablated liver tissue. Centrally, a “white zone” (WZ) was observed, which was characterized by a complete loss of cell structure and cell viability according to the literature [1] (see Figs. 1 and 2). This white zone was surrounded by a red zone, which could histologically be divided into two distinctive zones: an inner red zone 1 (RZ 1) and an outer red zone 2 (RZ 2). Red zone 1 typically showed cellular degeneration with remaining cell borders but loss of NADH-diaphorase staining (severely damaged but equivocal viable or nonviable cells). Red zone 2 was characterized by hepatocellular separation, intercellular edema and reduced cell viability (see Figs. 1 and 2). When comparing NADH-diaphorase staining and HE staining regarding cell viability, a significant number of cells seemed to be viable in HE staining (e.g. undamaged nuclei and remaining cell borders), but proved to be non-viable according to NADH-diaphorase staining.

Also rare vital cells were found in the ablation center in some of the histological samples. These cells were most likely displaced by retracting the applicator after RFA without back track ablation.

On macro-morphological images, a beige gray and homogenous area could be observed in the center of the ablations, according to the white zone. Within the white zone, sometimes a narrow dark brown area was visible adjacent to the applicator channel, which most likely was carbonized tissue. The white zone shaded off gradually into a rosy gray area (red zone). Due to different shades of gray color and the results of the histopathological examination, the red zone could be subdivided into the inner red zone 1 and the outer red zone. The red zone was sharply circumscribed by a thin fringe of dark red tissue, which was surrounded by apparently native and macroscopically undamaged liver tissue (see Figs. 1 and 3).

3.3 Quantitative analysis

Quantitative analysis was performed by means of the color-coded histological images (WZ: red, RZ 1: pink, RZ 2: yellow, and viable hepatic parenchyma: blue) and the gross pathological cross sections (Fig. 3). The values of the ablation areas and radii for the white and red zones are summarized in Table 1. The red zone was divided into red zone 1 and red zone 2 according to the results of the qualitative analysis. A strong correlation was obvious between macroscopic and histological measurement of the ablation zone areas. The highest correlation was found for the white zone (WZ: r = 0.95, p < 0.01; RZ 1: r = 0.85, p < 0.01; RZ 2: r = 0.89, p < 0.01). Comparison of macroscopic and histological findings revealed no significant differences for all three ablation zone areas (WZ: p = 0.33, RZ 1: p = 0.40, RZ 2: p = 0.89) and radii (WZ: p = 0.36, RZ 1: p = 0.61, RZ 2: p = 0.77).

In one case, a major hepatic vessel was located inside the ablation. The vessel caused an unambiguous heat sink effect and led to an asymmetric kidney-shaped ablation area. This resulted in a small minimal ablation radius (see Table 1).

The correlation analysis between the results of the two investigators showed a strong correlation for the histological data (WZ: r = 1.0, RZ 1: r = 0.99, RZ 2: r = 1.0; p < 0.01), and also for the macro-morphological analysis (WZ: r = 0.83, RZ 1: r = 0.90, RZ 2: r = 0.81; p < 0.02).