291 - Quantitative Three-Dimensional (3D) Margin Analysis for Predicting Complete Pathologic Necrosis in Hepatocellular Carcinoma Treated with Percutaneous Thermal Ablation

Purpose: To evaluate the association between unablated tissue volume within a pre-defined tumor margin and the achievement of complete pathologic necrosis (CPN) in hepatocellular carcinoma (HCC) patients treated with percutaneous thermal ablation before liver transplant.

Materials and Methods: This HIPAA-compliant, IRB-approved retrospective study included consecutive adult HCC patients treated with thermal ablation before liver transplant at a single tertiary transplant center between 2001 and 2019. Lesions were segmented on the pre-ablation contrast-enhanced CT or MRI. The ablation zone was segmented on the intraprocedural contrast-enhanced post-ablation CT. The two images were then fused and registered using MIM software, with color-coded zero-to-10 minimum margin contours automatically generated around the segmented tumor. 3D volumes within a predefined margin that were not ablated were referred to as the volumes of insufficient coverage (VIC). Subcapsular tumors were defined as tumors within 5 mm of the liver capsule. In these tumors, all tumor voxels within 5 mm of the capsule were excluded from the analysis. Logistic regression and Receiver Operating Characteristic curve analysis with the Youden index were used to assess the ability of VIC to predict CPN.

Results: 101 patients (median age 59 years [IQR, 55-64], 80% males, 71% Child Pugh class A) with 128 tumors (median size, 2.3 cm [IQR, 1.9-2.7], 52% subcapsular) were included in the analysis. On histopathologic evaluation, CPN was achieved in 85/128 (66%) tumors. Tumors with CPN were significantly smaller in size than tumors with incomplete necrosis (2.1 cm vs 2.5 cm, p = 0.011). VICs for tumors with CPN were significantly lower than for tumors without CPN for all quantified minimum margins from 0- to 10 mm (P < 0.05). Univariable logistic regression analysis identified VICs for 6-, 7-, 8-, 9-, and 10-mm minimum margins as significant predictors of CPN, however, the highest accuracy for predicting CPN was observed with the intraprocedural 10-mm VIC (area under the curve, 0.7; 95% confidence interval, 0.6, 0.8).

Conclusion: A 10-mm margin should be covered circumferentially by the ablation zone on the intraprocedural postablation scan for optimal chances for CPN. Three-dimensional assessment of ablation margin provides reliable immediate intraprocedural metrics that predict CPN after percutaneous thermal ablation of HCC.