Purpose: Despite recent advances in clinical imaging modalities, differentiation of pancreatic masses remains difficult. Here, we tested the diagnostic accuracy of molecular-based imaging including 3'-deoxy-3'-[(18)F]fluorothymidine (FLT) positron emission tomography (PET) and [(18)F]fluorodeoxyglucose (FDG) PET/CT in patients with suspected pancreatic masses scheduled to undergo surgery.
Methods: A total of 46 patients with pancreatic tumours suspicious for malignancy and scheduled for resective surgery were recruited prospectively. In 41 patients, FLT PET and FDG PET/CT scans were performed. A diagnostic CT performed on a routine basis was available in 31 patients. FLT PET and FDG PET/CT emission images were acquired according to standard protocols. Tracer uptake in the tumour [FDG and FLT standardized uptake value (SUV)] was quantified by the region of interest (ROI) technique. For FDG PET/CT analysis, correct ROI placement was ensured via side-by-side reading of corresponding CT images.
Results: Of 41 patients, 33 had malignancy, whereas 8 patients had benign disease. Visual analysis of FDG and FLT PET resulted in sensitivity values of 91% (30/33) and 70% (23/33), respectively. Corresponding specificities were 50% (4/8) for FDG PET and 75% (6/8) for FLT PET. In the subgroup of patients with contrast-enhanced CT (n = 31), sensitivities were 96% (PET/CT), 88% (CT alone), 92% (FDG PET) and 72% (FLT PET), respectively. Mean FLT uptake in all malignant tumours was 3.0 (range SUV(max) 1.1-6.5; mean FDG SUV(max) 7.9, range 3.3-17.8; p < 0.001).
Conclusion: For differentiation of pancreatic tumours, FDG PET and FDG PET/CT showed a higher sensitivity but lower specificity than FLT PET. Interestingly, visual analysis of FLT PET led to two false-positive findings by misinterpreting physiological bowel uptake as pathological FLT uptake in the pancreas. Due to the limited number of patients, the clinical value of adding FLT PET to the diagnostic workup of pancreatic tumours remains to be determined.