Use of diffusion-weighted MRI to differentiate chronic pancreatitis from pancreatic cancer

OBJECTIVE. The purpose of this study was to compare diffusion-weighted MRI (DWI) and conventional (non-DWI) MRI sequences in differentiating mass-forming chronic pancreatitis from pancreatic cancer. MATERIALS AND METHODS. A retrospective cohort study included 36 patients who underwent pancreatic resection for pancreatic cancer (n = 13) and chronic pancreatitis (n = 23) after preoperative MRI with DWI. Two independent reviewers assessed the DW images for signal intensity and apparent diffusion coefficient (ADC) values. Four weeks later, they reviewed the other MR images for size of mass, double-duct sign, pancreatic duct cutoff, and perivascular soft-tissue cuffing. A score for conventional MRI was given with 1 meaning definitely benign and 5 meaning definitely malignant. Univariate and multivariate analyses and receiver operating characteristic (ROC) curve analysis were performed with surgical pathologic examination as the reference standard. RESULTS. The only finding that differentiated the two groups was the presence of a well-defined mass, favoring the diagnosis of cancer (p = 0.02, p < 0.01). There was no significant difference between the two groups in signal intensity on DW images (p = 0.82, p = 0.85) or ADC (p = 0.51, p = 0.76). Double-duct sign, pancreatic duct cutoff, and perivascular soft-tissue cuffing were not useful in differentiating the two groups. The areas under the ROC curve were 0.873 and 0.878 for the conventional MRI scores, compared with 0.602 and 0.552 for ADC measurements (p = 0.02, p = 0.008). CONCLUSION. The addition of DWI to conventional MRI does not facilitate differentiation of pancreatic cancer from chronic pancreatitis.