Accuracy of Magnetic Resonance Cholangiopancreatography (MRCP) in the Diagnosis of Pancreas Divisum (PD)

Abstract

Background: Patients with PD may develop recurrent acute or chronic pancreatitis. ERCP is the gold standard for diagnosing PD. MRCP is a non invasive test reported to be highly accurate in diagnosing PD (Matos et al. GI Endosc 2001;53:728-33, Bret et al. Radiol 1996; 199:99-103). Aim: To evaluate the diagnostic accuracy of MRCP in detecting PD at our institution. Methods: During a 22-month period (10/02-07/04), 205 patients underwent ERCP and MRCP for various indications. 146 patients had diagnostic endoscopic pancreatograms (ERP) after MRCP and comprise our study population. MRCP was performed with the 1.5 T Signa Horizon LX scanner (GE Imaging Systems, Milwaukee, WI). Secretin 16 mg (SecreFlo™, ChiRhoClin, Burtonsville, MD) was given in 113/146 patients (S-MRCP). The remaining 33/146 had MRCP without secretin. Since 7/33 MRCP without secretin (21.2%) were non-diagnostic, further study focused on S-MRCP only. The S-MRCP was interpreted prior to ERP by 4 gastrointestinal radiologists as part of the routine daily workload and compared to subsequent ERP results. Results: 113 patients with S-MRCP and ERP (54 male, mean age 47.1 years) were evaluated. ERP identified PD in 19/113 (16.8%). S-MRCP identified 14/19 PD and was false positive in 3 cases (sensitivity 73.3%, specificity 96.8%, positive predictive value 82.4%, negative predictive value 94.8%). Five of 8 patients (63%) with inaccurate S-MRCP had changes of chronic pancreatitis by ERP. This differs from the frequency of chronic pancreatitis by ERP in 24/105 patients (25%) with accurate MRCP findings (p = 0.03). Summary: S-MRCP had a satisfactory specificity for detecting PD. However, the sensitivity of S-MRCP for diagnosis of PD was modest at 73.3%. This is low compared to previous smaller studies, which reported a sensitivity of MRCP up to 100%. Conclusions: 1) Our experience at a large referral center does not reflect the high diagnostic accuracy of MRCP for PD in previous studies. 2) MRCP without secretin results in a high percentage of non-diagnostic studies and should be obsolete if pancreatic assessment is desired. 3) It is more difficult to diagnose PD by MRCP in the setting of chronic pancreatitis. Background: Patients with PD may develop recurrent acute or chronic pancreatitis. ERCP is the gold standard for diagnosing PD. MRCP is a non invasive test reported to be highly accurate in diagnosing PD (Matos et al. GI Endosc 2001;53:728-33, Bret et al. Radiol 1996; 199:99-103). Aim: To evaluate the diagnostic accuracy of MRCP in detecting PD at our institution. Methods: During a 22-month period (10/02-07/04), 205 patients underwent ERCP and MRCP for various indications. 146 patients had diagnostic endoscopic pancreatograms (ERP) after MRCP and comprise our study population. MRCP was performed with the 1.5 T Signa Horizon LX scanner (GE Imaging Systems, Milwaukee, WI). Secretin 16 mg (SecreFlo™, ChiRhoClin, Burtonsville, MD) was given in 113/146 patients (S-MRCP). The remaining 33/146 had MRCP without secretin. Since 7/33 MRCP without secretin (21.2%) were non-diagnostic, further study focused on S-MRCP only. The S-MRCP was interpreted prior to ERP by 4 gastrointestinal radiologists as part of the routine daily workload and compared to subsequent ERP results. Results: 113 patients with S-MRCP and ERP (54 male, mean age 47.1 years) were evaluated. ERP identified PD in 19/113 (16.8%). S-MRCP identified 14/19 PD and was false positive in 3 cases (sensitivity 73.3%, specificity 96.8%, positive predictive value 82.4%, negative predictive value 94.8%). Five of 8 patients (63%) with inaccurate S-MRCP had changes of chronic pancreatitis by ERP. This differs from the frequency of chronic pancreatitis by ERP in 24/105 patients (25%) with accurate MRCP findings (p = 0.03). Summary: S-MRCP had a satisfactory specificity for detecting PD. However, the sensitivity of S-MRCP for diagnosis of PD was modest at 73.3%. This is low compared to previous smaller studies, which reported a sensitivity of MRCP up to 100%. Conclusions: 1) Our experience at a large referral center does not reflect the high diagnostic accuracy of MRCP for PD in previous studies. 2) MRCP without secretin results in a high percentage of non-diagnostic studies and should be obsolete if pancreatic assessment is desired. 3) It is more difficult to diagnose PD by MRCP in the setting of chronic pancreatitis. ErratumGastrointestinal EndoscopyVol. 62Issue 2PreviewIn the DDW 2005 Abstract issue (Volume 61, Number 5), two abstracts contained a typographical error in one author's name. For the abstracts “Accuracy of Magnetic Resonance Cholangiopancreatography (MRCP) in the Diagnosis of Pancreas Divisum (PD)” (Gastrointest Endosc 2005:61:AB100) and “ERCP Findings in Idiopathic Pancreatitis (IP): Cystic Fibrosis (CF) Gene Positive and Negative Patients” (Gastrointest Endosc 2005;61:AB186), the correct spelling of the author's name is Waleed Alazmi. Full-Text PDF