부산대학교 도서관

Abstract Objective Endoscopic ultrasound (US)‐guided radiofrequency ablation (RFA) has been investigated for pancreatic ductal adenocarcinoma (PDAC) but studies are limited and heterogeneous. Computed tomography (CT) scan features may predict RFA response after chemotherapy but their role is unexplored. The primary aim was to investigate the efficacy of ex‐vivo application of a dedicated RFA system at three power on surgically resected PDAC in patients who underwent neoadjuvant chemotherapy. The secondary aim was to explore the association between pre‐treatment CT‐based quantitative features and RFA response. Methods Fifteen ex‐vivo PDAC samples were treated by RFA under US control at three power groups (10, 30, and 50 W). Short axis necrosis diameter was measured by two expert blinded pathologists as the primary outcome. Two radiologists independently reviewed preoperative CT images. Results Eighty percent of specimens showed coagulative necrosis consisting of few millimeters: 5.7 ± 3.9 mm at 10 W, 3.7 ± 2.2 mm at 30 W, and 3.5 ± 2.4 mm at 50 W (p = 0.3), without a significant correlation between power setting and mean necrosis short axis (rho = –0.28; p = 0.30). Good agreement was seen between pathologists (k = 0.76; 95% confidence interval 0.55–0.98). Logistic regression analysis did not show associations between CT features and RFA response. Conclusions RFA causes histologically evident damage with coagulative necrosis of a few millimeters in 80% of ex‐vivo PDAC samples after chemotherapy and no clinical or pre‐operative CT features can predict efficacy. Power settings do not correlate with the histological ablation area. These results are of relevance when employing RFA in vivo and planning clinical trials on its role in PDAC patients.