Histological characterization of lesions induced by perfused and non-perfused radiofrequency thermal ablation needles in liver on a swine in vivo model

RFA (Radiofrequency thermal ablation) is a technique currently applied in human oncology, based on tissue coagulative necrosis induced by means of several type of needles. The aim of this study was to characterize lesions induced by perfused and non-perfused RFA needles in liver on a swine in vivo model. The in vivo trial (aut. n. 885/2016-PR, 22/09/2016) involved eight 4-months-old Landrance x Large White sows, weighing about 45 kg. The animals underwent median celiotomy, from xiphoid cartilage to a point 5 cm caudal to the umbilicus, under general anaesthesia. The RFA procedure was carried out, under ultrasound guidance, inserting the electrode in one single liver lobe. Different RFA conditions were explored (six replicas for each), using 18G internally cooled needles (RF Medical Co. Ltd., Seoul, Korea), either perfused (P) or not (NP), selecting a fixed time of delivery of thermal energy (60 seconds). Four different saline solutions were used with P needle: saline 0.9% (P 0.9%), hypertonic saline 3% (P 3%), 7% (P 7%) and 10% (P 10%). The animals were euthanized at the end of RFA procedures, still under general anaesthesia, by an intravenous overdose of thiopental sodium. Samples from each RFA area were formalin fixed, routinely processed and the histological sections were H.E. stained and analysed to determine the total number of lobules, to investigate microscopical alterations and to characterize the type and amount of cell infiltrate. A semiquantitative score was attributed to the alterations observed in each examined lobule in the lesions and in the surrounding areas. A total score was calculated multiplying the mean score for the total number of lobules counted in the specific zone.Kruskal-Wallis test (followed by Dunnʼs post-test) was applied to compare the patterns of the lesions induced by the different treatments and to evaluate the differences in areas closer and more distant from the insertion site. P-value less than 0.05 was considered significant.Thermoablated areas showed histologically the presence of three different concentric zones: a central area involving coagulation necrosis (zone 1), an intermediate haemorrhagic area (zone 2), and a surrounding zone (zone 3), characterized by the presence of both cell injury and healthy tissue. Lymphocytes were the prevalent type of infiltrate detected in each examined section. The number of lobules involved in zone 1 was significantly higher in P 7%, compared to NP (p<0.05), whereas no significant differences were detected in the two more peripheral areas of the lesion among the treatments. No differences were observed comparing the scores obtained in the same areas by different treatments. The scores decreased along with the distance from the insertion site, and in the more distant areas they were negative, confirming the safety and effectiveness of RFA treatments.Moreover, the total score showed a significant reduction of lesions in zone 3, compared to zone 1 or 2, for each tested condition.The analyses confirm and support previous results obtained by morphometric investigations, showing that P 7% could represent the most promising choice to be further investigated, in order to obtain larger and more reproducible ablation effects.