Methods: A retrospective analysis was done on 47 patients with a history of pancreatic carcinoma (n=24) or cholangiocarcinoma (n=23), all of whom were referred for restaging with a FDG-PET/non-ceCT and a ceCT study during the follow- up period after first-line therapy. Histological and radiological follow-up were used to determine the accuracy of the imaging findings.
Results: A lesion-based analysis showed, when equivocal lesions were considered negative, that the sensitivity and specificity of PET/ceCT were 89.1% and 76.9% , respectively, whereas those of PET/non-ceCT were 71.4% and 80.8% , respectively, and those of ceCT were 63.9% and 84.6% , respectively. PET/ceCT had significantly higher sensitivity compared to PET/non-ceCT (p =0.0003) and ceCT alone (p=0.0003). When equivocal lesions were considered positive, the sensitivity and specificity of PET/ceCT were 93.3% and 69.2% , respectively, whereas those of PET/non-ceCT were 77.3% and 80.8% , respectively, and those of ceCT were 76.5% and 55.1% , respectively. Within the PET/ceCT group, negative reading increased specificity from 69.2% to 76.9% (p= 0.04), at no significant cost to sensitivity (decrease from 93.3% to 89.1% , p= 0.07).
Conclusions: PET/non-enhanced CT when interpreted together with ceCT, significantly improves sensitivity of restaging of pancreatic and cholangiocarcinomas at a minor expense to specificity. Our results provide evidence that ceCT and PET/CT are complementary tests that enhance the diagnostic accuracy of restaging in patients with pancreatic and cholangiocarcinomas.