Immediate 3D Imaging Assessment is the Most Accurate Predictor of Local Tumor Progression after Microwave Ablation of Colorectal Cancer Liver Metastases

According to the guidelines, assessments of the ablation zone (AZ) is routinely performed at 3-8 weeks post ablation of colorectal liver metastases (CLMs). Information regarding the intraprocedural 3D/stereotactic evaluation of the AZ and treatment efficacy is lacking. This study aimed to compare the prognostic accuracy of intraprocedural and 4-6-week post-microwave ablation imaging assessments of the AZ using a 3D software assessment methodology.
Materials and Methods:

From a prospective IRB-approved clinical trial, 52 CLMs treated with real time split-dose FDG PET/CT-guided microwave thermal ablation were collected. The following steps were taken for the 3D assessments:

Intraprocedural:

Using the immediate pre-ablation real-time PET/CT obtained within 60 minutes of FDG injection, we segmented the CLMs and desired surrounding 5- and 10-mm margins. To increase the registration accuracy, the immediate pre-ablation non-contrast CT and the post-ablation contrast enhanced (portal venous) CT (CECT) were also registered in a 3D software. The tumor and margin contours were then superimposed on the registered post-ablation CT. The volumes of insufficient coverage (VICs) were defined as the tumor and margin volumes that were not covered by the AZ, depicted on the immediate post- ablation CECT.

4-6 weeks:

The immediate pre-ablation non-contrast CT and 4-6 weeks CECT or MRI were registered. The tumor and margins' contours from the pre-ablation CT and PET were transferred to the 4-6 weeks registered images and the VICs were assessed in the same manner and using the same 3D software used on the day of the ablation.

The area under curves (AUCs) and 95 % CI for prediction of local tumor progression (LTP) were computed.

Results:

AZ area significantly decreased from intraprocedural to 4-6 weeks post-ablation imaging (related-samples Wilcoxon signed rank test < 0.001). The cases were followed up for the median of 18 months (first and third quartiles: 10-26). AUCs were higher in the intraprocedural compared to the 4-6 week post-ablation assessments. The highest AUC for prediction of LTP was for evaluation of 5 mm VIC intraprocedurally [AUC: .705 (95% CI: .539-.871) compared to the 4-6 weeks assessments (AUC: .591 (.418-.764)].

Conclusion:

Higher AUCs for prediction of LTP were achieved for intraprocedural assessment compared to the 4-6 post-ablation evaluation of AZ. The highest AUC for prediction of

LTP   was for intraprocedural assessment of 5 mm VIC.