Colorectal Cancer Liver Metastases: Biopsy of the Ablation Zone and Margins Can Be Used to Predict Oncologic Outcome

Abstract

Purpose To establish the prognostic value of biopsy of the central and marginal ablation zones for time to local tumor progression (LTP) after radiofrequency (RF) ablation of colorectal cancer liver metastasis (CLM). Materials and Methods A total of 47 patients with 67 CLMs were enrolled in this prospective institutional review board-approved and HIPAA-compliant study between November 2009 and August 2012. Mean tumor size was 2.1 cm (range, 0.6-4.3 cm). Biopsy of the center and margin of the ablation zone was performed immediately after RF ablation (mean number of biopsy samples per ablation zone, 1.9) and was evaluated for the presence of viable tumor cells. Samples containing tumor cells at morphologic evaluation were further interrogated with immunohistochemistry and were classified as either positive, viable tumor (V) or negative, necrotic (N). Minimal ablation margin size was evaluated in the first postablation CT study performed 4-8 weeks after ablation. Variables were evaluated as predictors of time to LTP with the competing-risks model (uni- and multivariate analyses). Results Technical effectiveness was evident in 66 of 67 (98%) ablated lesions on the first contrast material-enhanced CT images at 4-8-week follow-up. The cumulative incidence of LTP at 12-month follow-up was 22% (95% confidence interval [CI]: 12, 32). Samples from 16 (24%) of 67 ablation zones were classified as viable tumor. At univariate analysis, tumor size, minimal margin size, and biopsy results were significant in predicting LTP. When these variables were subsequently entered in a multivariate model, margin size of less than 5 mm (P < .001; hazard ratio [HR], 6.7) and positive biopsy results (P = .008; HR, 3.4) were significant. LTP within 12 months after RF ablation was noted in 3% (95% CI: 0, 9) of necrotic CLMs with margins of at least 5 mm. Conclusion Biopsy proof of complete tumor ablation and minimal ablation margins of at least 5 mm are independent predictors of LTP and yield the best oncologic outcomes. (©) RSNA, 2016.

Figure 1a:

Images in an 82-year-old female patient with one CLM who was undergoing treatment according to protocol. (a) Baseline contrast-enhanced portal phase CT image obtained 2 weeks before treatment shows a 1.5 × 1.7 cm CLM (arrow) in segment 5/6. (b) Intraprocedural PET image acquired for tumor targeting (tumor maximum standardized uptake value, 12.8) and fused with a subsequent CT image to ensure that tines are well positioned for ablation (Starburst RITA XLi; Angiodynamics) after intravenous injection of 4.4 mCi fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) and a 75-minute uptake period. (c) Core biopsies of the center and inferior margin of the ablation zone obtained with PET/CT guidance after RF ablation. Left: To sample the ablation zone where the tumor previously resided, the preablation PET image is fused with the immediate postablation CT image, overlapping the previous tumor with the ablation zone and targeting the previously PET-positive area within the ablation zone. Right: Biopsy of the presumed minimal margin of the ablation zone is performed in a similar manner, this time targeting the PET-negative area that overlaps the CT ablation zone. (d) PET/CT image of the liver obtained 60 minutes after intravenous administration of an additional 8.8 mCi ¹⁸F FDG given at the end of ablation. The FDG-avid tumor has been replaced by a photopenic area representing the ablation zone, which corresponds to treatment success. The tissue analyzed was classified as necrotic. LTP did not occur during the 9-month follow-up period.

Figure 1b:

Images in an 82-year-old female patient with one CLM who was undergoing treatment according to protocol. (a) Baseline contrast-enhanced portal phase CT image obtained 2 weeks before treatment shows a 1.5 × 1.7 cm CLM (arrow) in segment 5/6. (b) Intraprocedural PET image acquired for tumor targeting (tumor maximum standardized uptake value, 12.8) and fused with a subsequent CT image to ensure that tines are well positioned for ablation (Starburst RITA XLi; Angiodynamics) after intravenous injection of 4.4 mCi fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) and a 75-minute uptake period. (c) Core biopsies of the center and inferior margin of the ablation zone obtained with PET/CT guidance after RF ablation. Left: To sample the ablation zone where the tumor previously resided, the preablation PET image is fused with the immediate postablation CT image, overlapping the previous tumor with the ablation zone and targeting the previously PET-positive area within the ablation zone. Right: Biopsy of the presumed minimal margin of the ablation zone is performed in a similar manner, this time targeting the PET-negative area that overlaps the CT ablation zone. (d) PET/CT image of the liver obtained 60 minutes after intravenous administration of an additional 8.8 mCi ¹⁸F FDG given at the end of ablation. The FDG-avid tumor has been replaced by a photopenic area representing the ablation zone, which corresponds to treatment success. The tissue analyzed was classified as necrotic. LTP did not occur during the 9-month follow-up period.

Figure 1c:

Images in an 82-year-old female patient with one CLM who was undergoing treatment according to protocol. (a) Baseline contrast-enhanced portal phase CT image obtained 2 weeks before treatment shows a 1.5 × 1.7 cm CLM (arrow) in segment 5/6. (b) Intraprocedural PET image acquired for tumor targeting (tumor maximum standardized uptake value, 12.8) and fused with a subsequent CT image to ensure that tines are well positioned for ablation (Starburst RITA XLi; Angiodynamics) after intravenous injection of 4.4 mCi fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) and a 75-minute uptake period. (c) Core biopsies of the center and inferior margin of the ablation zone obtained with PET/CT guidance after RF ablation. Left: To sample the ablation zone where the tumor previously resided, the preablation PET image is fused with the immediate postablation CT image, overlapping the previous tumor with the ablation zone and targeting the previously PET-positive area within the ablation zone. Right: Biopsy of the presumed minimal margin of the ablation zone is performed in a similar manner, this time targeting the PET-negative area that overlaps the CT ablation zone. (d) PET/CT image of the liver obtained 60 minutes after intravenous administration of an additional 8.8 mCi ¹⁸F FDG given at the end of ablation. The FDG-avid tumor has been replaced by a photopenic area representing the ablation zone, which corresponds to treatment success. The tissue analyzed was classified as necrotic. LTP did not occur during the 9-month follow-up period.

Figure 1d:

Images in an 82-year-old female patient with one CLM who was undergoing treatment according to protocol. (a) Baseline contrast-enhanced portal phase CT image obtained 2 weeks before treatment shows a 1.5 × 1.7 cm CLM (arrow) in segment 5/6. (b) Intraprocedural PET image acquired for tumor targeting (tumor maximum standardized uptake value, 12.8) and fused with a subsequent CT image to ensure that tines are well positioned for ablation (Starburst RITA XLi; Angiodynamics) after intravenous injection of 4.4 mCi fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG) and a 75-minute uptake period. (c) Core biopsies of the center and inferior margin of the ablation zone obtained with PET/CT guidance after RF ablation. Left: To sample the ablation zone where the tumor previously resided, the preablation PET image is fused with the immediate postablation CT image, overlapping the previous tumor with the ablation zone and targeting the previously PET-positive area within the ablation zone. Right: Biopsy of the presumed minimal margin of the ablation zone is performed in a similar manner, this time targeting the PET-negative area that overlaps the CT ablation zone. (d) PET/CT image of the liver obtained 60 minutes after intravenous administration of an additional 8.8 mCi ¹⁸F FDG given at the end of ablation. The FDG-avid tumor has been replaced by a photopenic area representing the ablation zone, which corresponds to treatment success. The tissue analyzed was classified as necrotic. LTP did not occur during the 9-month follow-up period.

Figure 2:

Schematic diagram of tissue analysis. Samples positive for tumor cells at hematoxylin-eosin (H&E) staining were further interrogated with immunohistochemistry. A total of 16 ablation zones were considered positive for viable tumor cells.

Figure 3a:

Images in a 57-year-old man who underwent RF ablation followed by immediate postablation biopsy of the ablated tumor. (a) CT image shows a 3.0 × 2.0 cm CLM in hepatic segment 6 (arrow). (b) CT image obtained 6 weeks after ablation shows the ablation zone with expected postablation findings consistent with complete ablation. (c) Photomicrograph of the tissue obtained from the ablation zone and stained with hematoxylin-eosin shows tumor cells. (d, e) Further immunhistochemical analysis was positive for proliferative nuclear marker Ki-67 (d) and mitochondrial activity (e). (f) This patient developed LTP (arrow) 6 months after ablation, which was suspected at CT (left) and was confirmed at PET (right).

Figure 3b:

Images in a 57-year-old man who underwent RF ablation followed by immediate postablation biopsy of the ablated tumor. (a) CT image shows a 3.0 × 2.0 cm CLM in hepatic segment 6 (arrow). (b) CT image obtained 6 weeks after ablation shows the ablation zone with expected postablation findings consistent with complete ablation. (c) Photomicrograph of the tissue obtained from the ablation zone and stained with hematoxylin-eosin shows tumor cells. (d, e) Further immunhistochemical analysis was positive for proliferative nuclear marker Ki-67 (d) and mitochondrial activity (e). (f) This patient developed LTP (arrow) 6 months after ablation, which was suspected at CT (left) and was confirmed at PET (right).

Figure 3c:

Images in a 57-year-old man who underwent RF ablation followed by immediate postablation biopsy of the ablated tumor. (a) CT image shows a 3.0 × 2.0 cm CLM in hepatic segment 6 (arrow). (b) CT image obtained 6 weeks after ablation shows the ablation zone with expected postablation findings consistent with complete ablation. (c) Photomicrograph of the tissue obtained from the ablation zone and stained with hematoxylin-eosin shows tumor cells. (d, e) Further immunhistochemical analysis was positive for proliferative nuclear marker Ki-67 (d) and mitochondrial activity (e). (f) This patient developed LTP (arrow) 6 months after ablation, which was suspected at CT (left) and was confirmed at PET (right).

Figure 3d:

Images in a 57-year-old man who underwent RF ablation followed by immediate postablation biopsy of the ablated tumor. (a) CT image shows a 3.0 × 2.0 cm CLM in hepatic segment 6 (arrow). (b) CT image obtained 6 weeks after ablation shows the ablation zone with expected postablation findings consistent with complete ablation. (c) Photomicrograph of the tissue obtained from the ablation zone and stained with hematoxylin-eosin shows tumor cells. (d, e) Further immunhistochemical analysis was positive for proliferative nuclear marker Ki-67 (d) and mitochondrial activity (e). (f) This patient developed LTP (arrow) 6 months after ablation, which was suspected at CT (left) and was confirmed at PET (right).

Figure 3e:

Images in a 57-year-old man who underwent RF ablation followed by immediate postablation biopsy of the ablated tumor. (a) CT image shows a 3.0 × 2.0 cm CLM in hepatic segment 6 (arrow). (b) CT image obtained 6 weeks after ablation shows the ablation zone with expected postablation findings consistent with complete ablation. (c) Photomicrograph of the tissue obtained from the ablation zone and stained with hematoxylin-eosin shows tumor cells. (d, e) Further immunhistochemical analysis was positive for proliferative nuclear marker Ki-67 (d) and mitochondrial activity (e). (f) This patient developed LTP (arrow) 6 months after ablation, which was suspected at CT (left) and was confirmed at PET (right).

Figure 3f:

Images in a 57-year-old man who underwent RF ablation followed by immediate postablation biopsy of the ablated tumor. (a) CT image shows a 3.0 × 2.0 cm CLM in hepatic segment 6 (arrow). (b) CT image obtained 6 weeks after ablation shows the ablation zone with expected postablation findings consistent with complete ablation. (c) Photomicrograph of the tissue obtained from the ablation zone and stained with hematoxylin-eosin shows tumor cells. (d, e) Further immunhistochemical analysis was positive for proliferative nuclear marker Ki-67 (d) and mitochondrial activity (e). (f) This patient developed LTP (arrow) 6 months after ablation, which was suspected at CT (left) and was confirmed at PET (right).

Figure 4a:

(a) Cumulative incidence and (b) Kaplan-Meier survival curves for LTP. Cumulative incidence estimation (a) accounts for the competing event of death, whereas Kaplan-Meier analysis (b) treats patients who died before developing LTP as censored observations. The subgroups were created based on variables that were significant at multivariate analysis (ie, biopsy result, minimal ablation margin size). N = necrotic-negative biopsy, V = viable tumor-positive biopsy.

Figure 4b:

(a) Cumulative incidence and (b) Kaplan-Meier survival curves for LTP. Cumulative incidence estimation (a) accounts for the competing event of death, whereas Kaplan-Meier analysis (b) treats patients who died before developing LTP as censored observations. The subgroups were created based on variables that were significant at multivariate analysis (ie, biopsy result, minimal ablation margin size). N = necrotic-negative biopsy, V = viable tumor-positive biopsy.