Prospective evaluation of the diagnostic accuracy of secretin-enhanced magnetic resonance cholangiopancreaticography in suspected chronic pancreatitis.

The purpose of this study was to determine the improvement in sensitivity and specificity of magnetic resonance cholangiopancreaticography (MRCP) in the diagnosis of chronic pancreatitis due to the application of secretin. Sixty-two consecutive patients with suspected chronic pancreatitis were prospectively included in our study. All investigations were carried out in a clinical 1.0-T MR-scanner using heavily T2-weighted pre- and dynamic post secretin measurements. Dynamic images were acquired every 30 s for a total imaging time of 10 min. Two experienced radiologists blinded for all other results evaluated the images in random order in two sessions: one for the native images and one for the stimulated MRCP series. The gold standard for the final diagnosis was the summary of all clinical and radiological results. Sensitivity and specificity for the diagnosis of chronic pancreatitis and the kappa coefficient for the agreement between both observers were calculated. Prior to secretin application, investigators achieved sensitivities of 78% and 56%, respectively, and specificities of 89% and 96%, respectively. After secretin application, the sensitivity for the first observer remained at 78%; for the second observer it increased to 67%, whereas specificities improved to 94% and 98%, respectively. Following secretin application, agreement determined between the two investigators improved from 0.73 to 0.84. While MRCP already provides high sensitivity and specificity, application of secretin increases both values and the reliability of the diagnosis further.