Dynamic Study Of Diffusion-weighted Imaging In Rabbit Models Of Hepatic VX2 Carcinoma After Radio-frequency Ablation And Pathologic Correlation

Objective: To explore the feasibility and value of dynamic diffusionweighted imaging in the assessment of the treatment response of hepaticcarcinoma by studying the characteristics of the rabbit models of hepatic VX2carcinoma in DWI at different times after Radio-frequency ablationMaterials and Methods: 20 rabbit models of hepatic VX2 carcinomawere constructed successfully，and were randomly divided into 4 groups（1week,2 week, 3 week, 4 week）, 4 normal rabbits were used as controls andrandomly assigned to each group. CT-guided RFA was performed in the liverof 24 rabbits. MR scan was performed on 1, 2, 3, and 4 weeks respectively.After MR scan, a group of corresponding experimental rabbits and controlrabbit were immediately killed, and correlation was made withhistopathological- findings. The appearances in MR T1WI, T2WI, DWI andthe apparent diffusion coefficient （ADC） of necrosis area induced by RFA,viable tumor, inflammatory hyperemic reaction area, and normal liverparenchyma at different times were analyzed according to histopathologicalfinding. The statistical differences between different groups were analyzed byANOVA.Results: 20 rabbit models of hepatic VX2 carcinoma treated with RFA,but 2 rabbits died of anesthesia overdose during operation and 3 rabbits diedof peritonitis in the follow-up process. Ultimately, MR imaging- histopathologic correlation was possible in only 15 of these. Macroscopically,a oval or round necrotic lesion is centered about a central track of charringalong the electrode tract, and a central necrosis is surrounded by thin-walledinflammatory hyperemic reaction rim. In addition, peripheral viable tumortissues are seen in some specimens. The Central necrosis was graduallygetting absorption liquid. The findings of peripheral inflammatory reactionzone early were mainly hyperemic response and inflammatory cell infiltration,and then capsule was gradually formed in peripheral ablation tissue from 1 to4 weeks. The mean ADC value of tumor and normal liver parenchyma beforeRFA were 1.31±0.275(×10^-3mm²/s), 1.80±0.272(×10^-3mm²/s)respectively,which were statistical significant （P< 0.05）. The ADC values in viable tumortissue, necrosis tissue, inflammatory hyperemic reaction rim and normal liverparenchyma after RFA 1 week were 1.30±0.223, 0.77±0.144, 2.72±0.029,1.871±0.016(×10^-3mm²/s)respectively. The ADC values in viable tumortissue, necrosis tissue and normal liver parenchyma after RFA 2 ,3, and 4week were 1.359±0.018, 0.871±0.011, 1.887±0.011(×10^-3mm²/s)（afterRFA2weeks）; 1.462±0.14, 2.242±0.107, 1.843±0.151(×10^-3mm²/s)（afterRFA 3 weeks）; 1.43±0.018, 2.543±0.129, 1.927±0.163(×10^-3mm²/s)（afterRFA 4 weeks）. The ADC values of ablation necrosis tissues, which wasminimum value in 1 week and was maximum value in the 4th week, was arising trend over time. The differences among 1, 2，3 and 4 weeks werestatistical significance respectively （P< 0.05）. The ADC value in viable tumortissue at different times showed no statistical significance （P＞0.05）.Conclusions: DWI can dynamically monitor treatment response ofrabbit hepatic VX2 carcinoma after RFA. The manifestations of ablation zoneon DWI are basically matched with pathologic findings. ADC is possible to be biomarker in the assessment of treatment response of hepatic carcinomaafter RFA.