Clinical utility and limitations of FDG PET in detecting recurrent hepatocellular carcinoma in postoperative patients.

Background: The clinical usefulness of positron emission tomography (PET) with (18)F-fluorodeoxyglucose (FDG) for the detection of recurrent hepatocellular carcinoma (HCC) is controversial because HCC displays varying FDG avidity. The purposes of this study were to re-evaluate the utility of FDG PET for the detection of recurrent HCC, and to assess its prognostic value in a large series of postoperative patients.

Methods: We retrospectively reviewed 113 scans in 86 patients undergoing FDG PET after curative surgery for HCC. These scans were performed for suspected recurrence on radiologic imaging (group A: n = 44) because of an elevated tumor marker level with negative prior imaging results (group B: n = 32) or with no suspicion of recurrence (group C: n = 37). FDG PET's accuracy for recurrence detection and its value as a predictor of survival were assessed.

Results: The sensitivity, specificity, and diagnostic accuracy were 53, 100, and 55 % for group A; 34, 100, and 41 % for group B; and 11, 100, and 78 % for group C, respectively. A change in therapy resulted from the scan results in 7, 9, and 8 % in groups A, B, and C, respectively. The combined sensitivities for intra- and extrahepatic recurrence were 30 and 42 %, respectively. Histopathological features at initial surgery did not affect the sensitivity. The overall survival of patients with positive scans was significantly poorer than that of patients with negative scans (P = 0.008).

Conclusions: The sensitivity of FDG PET for recurrent HCC was low, with little change in treatment resulting. However, it can predict prognosis in postoperative patients.