The impact on patient management of whole-body FDG-PET scanning in patients with liver metastases from colorectal carcinoma

Abstract

3569 Background: The accuracy and utility of FDG-PET (PET) scanning in patients with colorectal carcinoma metastatic to the liver was evaluated. Methods: Whole-body PET scans performed on 100 consecutive patients with liver metastases from colorectal carcinoma between July 2001 and January 2003 were reviewed and correlated with spiral triple-phase computed tomography (CT), serum carcinoembryonic antigen (CEA) level, and final histopathologic diagnosis. Results: The indication for PET was a rise in CEA in 19% of patients, surveillance or restaging in 37%, to exclude additional disease prior to metastasectomy in 33%, or to evaluate “indeterminate” findings on CT in 11%. The sensitivity, specificity, and accuracy of CT were 81%, 100%, and 84%, respectively. The sensitivity, specificity, and accuracy of PET were 84%, 89%, and 85%, respectively. The PET false-positive (FP) rate was 2%: liver & wound (1), mediastinum (1). Sixteen false-negative (FN) sites were observed in 13% of PET scans: liver (5), lung (5), porta hepatis lymph node (2), abdomen/pelvis (4). Two patients underwent unnecessary laparotomy based on FN PET results, whereas CT correctly suggested extrahepatic disease in these two. There was no difference in CEA level between the PET true-positive (TP) and FN groups. In the six cases where CT was “indeterminate,” PET correctly diagnosed patients. PET directly altered patient management in 23% of cases and this was independent of PET indication. PET detected liver metastasis not seen on CT in 8%, confirmed liver metastasis read as “indeterminate” on CT in 2%, detected extrahepatic metastasis not seen on CT in 6%, detected metabolic response in metastasis which led to a change in management in 4%, and excluded disease read as “indeterminate” on CT in 2%. When PET was combined with CT, the overall sensitivity, specificity, and accuracy of diagnosis were 96%*, 88%*, and 95%*, respectively (*p<0.05 vs CT alone). Conclusions: PET, when used selectively in patients with liver metastases from colorectal carcinoma, marginally improved the accuracy of diagnosis but directly altered management in 23% of patients. No significant financial relationships to disclose.