We evaluated the prognostic accuracy of established PET response criteria in patients with liver metastases from pancreatic cancer after treatment with (90)Y microspheres.

Methods: Seventeen patients underwent (18)F-FDG PET/CT before and 3 mo after radioembolization for liver metastases from pancreatic cancer. Overall survival, progression-free survival, and time to intrahepatic progression were among other factors correlated with metabolic response as revealed by PERCIST 1.0-defined declining SUVpeak and total-lesion glycolysis.

Results: Metabolic response by change in SUVpeak (7/17) and change in total-lesion glycolysis (7/17) was a predictor for overall survival (P = 0.039; hazard ratio [HR], 0.24; 95% confidence interval [CI], 0.06-0.93), progression-free survival (P = 0.016; HR, 0.15; 95% CI, 0.03-0.69), and time to intrahepatic progression (P = 0.010; HR, 0.16; 95% CI, 0.04-0.65). A summed baseline CT diameter of less than 8 cm for the 2 largest liver metastases predicted time to intrahepatic progression (P = 0.013; HR, 0.21; 95% CI, 0.06-0.72) but did not predict overall or progression-free survival. Patient outcome was not predicted by other parameters, including baseline SUVpeak, baseline total-lesion glycolysis, or change in serum level of carcinoembryonic antigen or carbohydrate antigen 19-9 from baseline to follow-up (each, P > 0.05).

Conclusions: Metabolic response by (18)F-FDG PET/CT predicts overall survival, progression-free survival, and time to intrahepatic progression after radioembolization for liver metastases from pancreatic cancer.