(LLC), distant liver control (DLC), and distant metastasis and is reported using the Kaplan-Meier method via log-rank. Hazard ratios were calculated using Cox regression. Results: In the 34 patients we followed, 32 (94.1%) patients were not surgical candidates. Thirty one (91.2%) patients had Child-Pugh class A and 3 (8.8%) patients had class B. Extrahepatic disease was present in 3 (8.8%) patients. Theraspheres were injected to 48 lobes and segments involved with the disease. Twenty-five (54%) lesions were PET avid with SUV max of 3-20. Median volume treated was 945cc with a median dose of 130Gy. Significant predictors for 18 F-FDG avidity on univariate analysis included female gender (OR 1.22, 95% CI 1.046-1.427, pZ0.025) and age (OR 0.019, 0.001-0.445, pZ0.003). Median follow up for our patients was 11.6 months (1.2-62.8). PET avid disease was associated with lower median time for LLC (6 vs. 27 months, pZ0.001) and decreased 1 year rate of LLC (14% vs 63%). Median times for DLC and DM were not reached in non-PET-avid disease, and for PET avid disease they were 12.3 and 16.3 months, respectively. On multivariate analysis (MVA), PET avidity remained a significant predictor for LLC (HR 5.06, 1.89-13.57, pZ 0.001) and for DLC (HR 7.5, 1.8-30.9, pZ0.005). Predictors for DM were PET avid disease (HR 79.7, 6.2-1020.6, pZ0.001), age (HR 0.93, 0.86-1, pZ0.05), and AFP response 6 months post treatment (HR 0.29, 0.09-0.88, pZ0.03). Age was the only predictor for survival on MVA (HR 0.91, 0.84-0.99, pZ0.027). Conclusion: PET avid unresectable HCC tumors undergoing radioembolization have worse local and distant outcomes. More studies are needed to validate the potential for 18F-FDG PET/CT as an imaging biomarker in HCC.
