The differential diagnosis between pancreatic carcinoma (PC) and mass-forming focal pancreatitis (FP) is extremely important because their prognosis and management are different [1]. However, such differentiation is invariably difficult because of their similar clinical presentations, imaging features and, on some occasions, inconclusive biopsy results [2, 3]. This study was conducted to assess DCE-MRI combined with DWI for improving the differentiation between FP and PC, based on qualitative and quantitative analysis of TIC and ADC values.

The institutional review board approved this retrospective study and waived the requirement for informed consent. This study included 32 patients with surgically confirmed PC and 18 patients with surgically or histologically proven FP who underwent five phases DCE-MRI and DWI at 3.0 T. The pancreatic TIC was generated as a percentage increase in the signal intensity (SI), according to the following enhancement formula: (SIpost-SIpre)/SIpre × 100, where SIpre and SIpost represent the pre- and post-contrast SIs, respectively. The patterns of the TICs of ROI in the pancreatic tumor and non-tumor adjacent pancreatic parenchyma were classified into 5 types according to the time of a peak (18s, 45s, 75s, 2.5min, 4min after bolus injection of contrast material), namely, type-Ⅰ, Ⅱ, Ⅲ, Ⅳ, Ⅴ, respectively (Figure 1). Then, according to the part of the TIC profile after the peak time, the type of the masses were classified into two subtypes, subtype-a (slow decline) and subtype-b (plateau) (Figure 2)..The apparent diffusion coefficient (ADC) value between PC and FP on DW images obtained with a b value of 600 sec/mm2 were compared. Also, the tumor-to-pancreas contrast ratio of ADC value was calculated as follows: | (SI_T – SI_P) | / SI_S, where SI_T is signal intensity of pancreatic tumor, SI_P is signal intensity of NAP, and SI_S is signal intensity of spleen, which was used for the normalization of ADC values. The sensitivity and specificity of DCE-MRI and DWI alone and a combined image set was evaluated using receiver operating characteristic curve (ROC) analysis.Results: The prevalent TIC profiles differed in the PC and FP groups in that PC showed the type-Ⅴ or subtype-b profile (25 [80%] of 32) and FP showed the type-Ⅰ, type-Ⅱor subtype-a profile (14 [78%] of 18). The type-Ⅴ TIC was only recognized in PC group (P<0.01). Type-Ⅳb (P=0.036) were more frequently observed in PC (Figure 2), while type-Ⅱa (P<0.01), type-Ⅰa(P=0.037) in FP (Figure 3) . The mean ADC value ± standard deviations (×10-3mm2/s) of tumors are lower in PC than FP (1.17 ± 0.24, 1.47 ± 0.18, respectively, and P<0.01). We also found a significant difference in the mean tumor-to-pancreas contrast ratio of ADC value ± standard deviations between PC and FP (0.30 ± 0.16, 0.48 ±0.21, respectively, and P<0.01). The combined image set of DCE-MRI and DWI (96.9%, 83.3%) yielded better sensitivity than the DCE-MRI or DWI alone (93.8%, 66.7%; 81.3%, 61.1%) (Figure 4,5).Pancreatic pancreatic TIC from DCE-MRI and ADC value and tumor-to-pancreas contrast ratio of ADC value from DWI were found to provide reliable information for differentiating PC from FP, while the combination DCE-MRI and DW imaging can achieve higher sensitivity and specificity. These imaging techniques may therefore make it possible to eliminate unnecessary major pancreatic surgery and delays in making a correct diagnosis of pancreatic carcinoma.