18fluorodeoxyglucose-positron emission tomography (FDG-PET) is a promising modality in imaging pancreatic tumors. We have previously reported that FDG-PET was useful in detecting pancreatic ductal cancer and the uptake of FDG in the tumor was largely dependent on the expression of GLUT-1 glucose transporter. The current study was undertaken to asses this technique from surgical points of view; 1) accuracy in differentiating pancreatic cancer from chronic pancreatitis, 2) detection of hepatic and peritoneal involvement, 3) efficacy in evaluating the response to chemoradiotherapy, and 4) detection of other indispensable pancreatic malignancies, such as cystic neoplasm and neuroendocrine tumors.
METHODS: 152 consecutive patients for suspected pancreatic tumor were underwent FDG-PET. Invasive ductal adenocarcinoma: 89 (resection: 40, non-resection: 49), neuroendocrine tumor: 18, cystic neoplasm: 11, other periampullary cancer: 9, chronic pancreatitis: 25. Diagnosis of chronic pancreatitis was confirmed by surgery or having no signs of malignancy during 12 months follow-up. PET was analyzed by calculation of a standard uptake value (SUV) 60 minutes after application of the tracer.
RESULTS: 1) With use of a 2.2 cutoff value for SUV, 88 ductal adenocarcinoma were positive, one was false negative, 20 chronic pancreatitis were true negative and 5 were false positive. Sensitivity of FDG-PET for ductal adenocarcinoma was 99%, specificity was 80% and diagnostic accuracy was 95%. Five false positives were attributed to acute inflammatory lesion. One false negative for pancreatic cancer and two for other periampullary cancer were less than 2 cm in size. 2) Sensitivity was 92% in detecting hepatic metastasis, specificity was 91% and accuracy was 91%; however, vales were not greater than those of CT scan. Peritoneal dissemination was not successfully detected by FDG-PET. 3) 14 patients received radiation therapy (IORT+EBRT) with or without chemotherapy, and post-radiation FDG-PET showed a significant decrease of SUV in primary tumor. 4) Sensitivity of FDG-PET for cystic neoplasm (malignant: 8, benign: 3) was 75%, specificity was 67% and diagnostic accuracy was 76%. 61% of neuroendocrine tumors (11/18) were detected by FDG-PET.
CONCLUSIONS: FDG-PET is a dependable imaging for the detection and differentiation of pancreatic ductal cancer. However, surgeons need to note that the method is not suitable to detect small cancer and hepatic or peritoneal involvement in advanced cancer. FDG-PET is not reliable enough to decide the surgical strategy for cystic malignancy.