The aim of this study was to determine whether intraprocedural 18F-FDG PET/CT can be used as a predictor of local tumor progression after percutaneous ablation of colorectal liver metastases. Methods: In this institutional review board-approved prospective study, 39 patients (19 men and 20 women; median age, 56 y) underwent split-dose 18F-FDG PET/CT-guided ablation followed by immediate biopsy and contrast-enhanced CT imaging of the ablation zone. Binary categorization of biopsy tissues was performed on the basis of the presence of only nonviable coagulation necrosis or viable tumor cells. Minimum ablation margin measurements from contrast-enhanced CT imaging were categorized as 0 mm, 1-4 mm, 5-9 mm, or greater than or equal to 10 mm. SUVs were obtained from PET/CT imaging, and SUV ratios were calculated from 3-dimensional regions of interest located in the ablation zone and surrounding normal liver. All predictive variables (biopsy, minimum margin distance, and SUV ratio) were evaluated as predictors of time to local tumor progression identified on imaging using competing-risks regression models (uni- and multivariate analyses). Results: A total of 62 consecutive ablations were evaluated. The mean SUV ratio was significantly higher for viable tumor-positive immediate postablation biopsies (n = 10) than for tumor-negative biopsies (n = 52) (85.8 ± 92.2 vs. 42.3 ± 45.5) (P = 0.03) and for a minimum margin size of less than 5 mm (n = 15) than for a minimum margin size of greater than or equal to 5 mm (n = 47) (78.5 ± 99.1 vs. 38.3 ± 78.5) (P = 0.01). After a median follow-up period of 22.5 (range, 7-52) months, 23 of 62 ablated tumors showed local tumor progression (37.1%). The local tumor progression rate was significantly higher for viable tumor-positive biopsies (8/10) than for negative biopsies (15/52) (80% vs. 29%) (P = 0.001) and for a minimum margin size of less than 5 mm (9/15) than for a minimum margin size of greater than or equal to 10 mm (2/15) (60% vs. 13%) (P = 0.02) but not 5-9 mm (37.5%; 12/32) (P = 0.5). In a competing-risks analysis, biopsy results (P = 0.07) and the minimum margin size (P = 0.08) were borderline significant, but the SUV ratio was not (P = 0.22). However, for negative biopsy ablations, the minimum margin size and SUV ratio were predictive imaging factors for local tumor progression; subdistribution hazard ratios were 0.564 (0.325-0.978) (P = 0.04) and 1.005 (1.001-1.009) (P = 0.005), respectively. Conclusion: The SUV ratio and minimum margin size can independently predict colorectal metastasis local tumor progression after liver ablation when there are no viable tumor cells on immediate postablation biopsies.
Read full abstract