[18 F-fluorodeoxyglucose positron emission tomography for characterization and initial staging of pancreatic tumors]

To evaluate positron emission tomography (PET) with (18)F-fluorodeoxyglucose (FDG) for characterizing and initial staging of pancreatic tumors and to determine its impact on therapeutic strategy.

This study included 24 patients with pancreatic tumor who underwent PET before treatment. Twenty-two patients had a malignant tumor and two had a benign tumor. The PET scan was performed after intravenous injection of FDG with a gamma camera. PET findings were compared with histology of the pancreatic tumor (n=24), liver metastases (n=5), peritoneal metastases (n=5), and lymph node metastases (n=5). Absence of metastasis or lymph node involvement was determined by surgery or by CT scan, ultrasonography, magnetic resonance imaging and at least two months follow-up.

The sensitivity of the PET scans to identify pancreatic carcinoma was 64% (95% confidence interval 44-84%). PET scans could not be interpreted for lymph node involvement adjacent to the tumor. For liver metastases, the PET scan was positive in 3 out of 5 patients. For peritoneal metastases, the PET scan was positive in 4 out of 5 patients but was doubtful in one. There were two false positives. Among the 4 cystic tumors, the PET scan was positive for 2 malignant tumors and negative for 2 benign tumors. Surgical strategy was modified in only one of the 24 patients on the basis of PET findings.

The sensitivity of PET for the diagnosis of primary malignant pancreatic tumor was found to be low. The contribution of FDG-PET to the surgical decision appears to be limited to the detection of metastases or lymph node involvement distant from the tumor, contraindicating surgery. Nevertheless, the sensitivity of FDG-PET is lower than that of laparoscopy for peritoneal metastases. Indications for PET should be included in an evaluation of therapeutic decision making and cost analysis.