Variability of CT Contrast Enhancement in the Pancreas: A Cause for Concern?

Background: Multidetector CT is a valuable technique for diagnosis/staging in several pancreatic pathologies. Diagnosis is usually based on tissue density measurements. Recently, newer functional CT techniques have been introduced. The aim of this study was to assess variability in perfusion and dual-energy CT data, and to compare these data with density measurements in the pancreas of a healthy population. Methods: Two groups were included: 20 patients underwent perfusion CT imaging, and 10 patients were scanned using a dual-energy protocol. In both groups, tissue density [Hounsfield units (HU)] was measured in the pancreatic head, body and tail. Functional data were calculated (blood flow/blood volume in the perfusion CT group, iodine concentration in the dual-energy group), and variability was assessed. Results: Density measurements were comparable for the perfusion and dual-energy CT groups, and ranged from 14 to 60 HU. Maximal enhancement differences between the head/body/tail of the pancreas ranged between 2 and 21 HU. Considerable variability was observed, both in density measurements (ranging from 3 to 34%) and in functional parameters (mean variability in perfusion CT parameters blood flow and blood volume was 21.3 and 10% respectively; mean variability in dual-energy iodine-mapping results was 24.4%). Conclusion: This study demonstrated the presence of important intraindividual variability in pancreatic tissue contrast enhancement, regardless of the CT technique used. Considering the variability observed in this study, the use of cut-off values to characterize pancreatic pathologies seems troublesome, and morphologic primary and secondary changes will remain important, even when using novel functional imaging techniques.