MRI of pancreas in patients with chronic pancreatitis and healthy volunteers: Can pancreatic signal intensity and contrast enhancement patterns be valuable predictors of early chronic pancreatitis?

Abstract

Background: Chronic pancreatitis is a progressive inflammatory disease of the pancreatic parenchyma and ductal structures. Typical MR imaging features of chronic pancreatitis include dilatation of the main pancreatic duct and irregularities of branch pancreatic ducts that can be evaluated with MRCP, loss of the normal high signal intensity of the pancreatic parenchyma on T1-weighted images, and decreased enhancement on dynamic contrast-enhanced sequences. Objective: The aim was to evaluate the parenchymal MRI findings of the pancreas in conjunction with MRCP in patients with chronic pancreatitis and normal healthy subjects. Methods: MRI and MRCP findings of 90 consecutive patients with chronic pancreatitis and 26 healthy volunteers were evaluated in this study. Patients were divided into five groups according to Cambridge classification system considering ERCP, MRCP, CT, US and EUS findings. On MRI, signal intensity ratios of the pancreas and the spleen on unenhanced T1 weighted fat saturated spoiled gradient echo images (SIR P/S) along with the enhancement ratio between the arterial phase and the portal venous phase (SIR A/V) were calculated and their frequency in each Cambridge score were documented. MRI findings in normal subjects were compared to patients with chronic pancreatitis. Results: MRCP findings were normal and pancreatic signal intensity was higher than spleen (SIR P/S > 1) in 26 healthy control subjects. Mean pancreas signal in control group (SIR P/S; 1.48 +/- 0.13) was significantly higher (P < 0,001) than mean signal intensity in patients with chronic pancreatitis (SIR P/S; 1.18 +/- 0.24). In the control group, the highest contrast enhancement occurred in arterial phase (SIR A; 1.7 +/- 0.32) and this was significantly higher (P < 0,001) than portal venous phase (SIR V; 1.45 +/- 0.28). In chronic pancreatitis group, the highest contrast enhancement occurred in portal venous phase (SIR V; 1.56 +/- 0.18) but there was no statistically significant difference (P = 0.06) compared to arterial phase (SIR A; 1.51 +/- 0.21). Mean SIR A/V values of control group were 1.18 +/- 0.08, and SIR A/V values of patients with chronic pancreatitis were 0.97 +/- 0.12 respectively. Mean SIR A/V value in control group was statistically higher than patients with chronic pancreatitis (P < 0.05). Between the control group and patients with chronic pancreatitis in terms of SIR statistically significant differences were found SIR P/S (Spearman correlation coefficient (rs) = -0.76, P < 0.001), SIR A (rs = -0.28, P = 0.003), SIR V (rs = 0.43, P < 0.001), SIR LV (rs = 0.54, P < 0.001) and SIR A/V (rs = -0.68, P < 0.001). Conclusion: In our study, MRI findings were significantly different in subjects with chronic pancreatitis compared to the control group. MRI findings correlated well with the ductal changes according to Cambridge classification. However, MRI findings may occur prior to ductal changes.