Magnetic Resonance Imaging as a Valid Noninvasive Tool for the Assessment of Pancreatic Fibrosis

Abstract

Objectives The aim of this study was to evaluate the utility of magnetic resonance imaging (MRI) for the noninvasive assessment of pancreatic fibrosis (PF). Methods Fifty-two patients who underwent surgical resection of the pancreas, histological examination of resection margins, preoperative abdominal MRI, and fecal elastase-1 test were enrolled in the study. Pancreatic tissue was identified on the MRI T1-, T2-, and diffusion-weighted imaging sequences. Apparent diffusion coefficient (ADC) was measured at the expected resection margin of the pancreas. Results There was a significant negative correlation between the ADC mean and histologically determined PF (r = −0.752, P = 0.001). For equal to or greater than 25% of PF, the ADC cutoff value was 1.331 or less, with a sensitivity of 77% and specificity of 88%. The unenhanced T1-weighted signal intensity ratio (T1SI) cutoff value was 172.1 or less. For equal to or greater than 50% of PF, the ADC cutoff value was 1.316 or less with a sensitivity of 85% and specificity of 88%. The highest sensitivity was obtained by combining ADC and T1SI values. Conclusions Combining both the ADC and T1SI measurement allows the detection of early PF with good sensitivity and specificity. Magnetic resonance imaging has the advantage of being noninvasive and widely used in the clinical setting, thus making our results easily transferable to routine clinical practice.