Magnetic resonance cholangiopancreatography reports in the evaluation of chronic pancreatitis: a need for quality improvement.

Background: Numerous publications from academic centers suggest that magnetic resonance cholangiopancreatography (MRCP) can diagnose early chronic pancreatitis (CP) and assess pancreatic secretory reserve/function. However, the rigorous composite interpretation methods and quantitative secretory dynamics reported in these studies are not routinely measured in clinical practice. Therefore, the utility of routine MRCP reports in the clinical setting is unknown.

Study Design: Cross-sectional study of patients referred to a tertiary center who underwent both MRCP and endoscopic pancreas function testing (ePFT) for assessment of chronic pancreatitis and abdominal pain.

Aims: To compare MRCP and sMRCP reports to a reference standard pancreas function test for diagnosis of chronic pancreatitis.

Methods: Source population: patients seen within a pancreas clinic at a tertiary referral center. MRCP and sMRCP reports were reviewed to record pancreas duct (dilation, side-branch changes), parenchyma enhancement (T(1), T(2) signal) and physiologic response (duodenal filling, pancreas duct response) to secretin. ePFT was categorized based on previously published data (normal peak bicarbonate >80 mEq/l). Referent values were calculated for MRCP and sMRCP using secretin ePFT as gold standard.

Results: A total of 69 patients were identified (mean age 43.5 +/- 12; 65.2% female). 28 (40.6%) patients had abnormal ePFT based on their peak bicarbonate level. The mean bicarbonate values in the abnormal PFT and normal PFT groups were 59 +/- 13.9 and 95.3 +/- 12.6 mEq/l, respectively. Peak bicarbonate decreased with severity of chronic pancreatitis on MRCP (p = 0.0016). There was fair agreement of MRCP and ePFT (kappa 0.335 [0.113, 0.557]). The pre-stimulation pancreas duct changes reported were found to be the only predictor of abnormal pancreas function (p = 0.002). The post-stimulation findings of duodenal filling (p = 0.47), T(2)enhancement (p = 0.21) or change in pancreas duct caliber (p = 0.3) reported did not improve MRCP agreement with ePFT. Overall diagnostic accuracy, sensitivity and specificity were 70, 85 and 46%, respectively, for MRCP reports using ePFT as the gold standard.

Conclusions: Pancreas ductal features described on routine MRCP reports correlate with abnormal pancreas function. Current MRCP reports should be standardized to include all radiologic information available in hopes of predicting early chronic pancreatitis.