Stephan W Anderson1, Jorge A Soto. 1. Department of Radiology, Boston University Medical Center, Boston, MA 02215, USA. Stephan.Anderson@bmc.org
Abstract
BACKGROUND: Many patients presenting with nonspecific signs and symptoms often receive CT scans using general protocols, not optimized to evaluate for pancreatic pathology. Therefore the purpose of this study was to evaluate portal venous phase 64 multi-row detector CT (MDCT) scans for detecting pancreatic duct strictures, stones, pancreas divisum, and communication between pancreatic ducts and cystic pancreatic lesions. METHODS: Institutional review board approval with waived informed consent was obtained for this HIPAA-compliant study. We included all patients that underwent abdominal, portal venous phase, intravenous contrast-enhanced 64 MDCT scans between 6/7/05 and 5/01/07 and MR cholangiopancreatography (MRCP) or endoscopic retrograde pancreatography (ERCP) within 2 months of the CT. This yielded 93 patients (42 males, 51 females) with a mean age of 59 years. In addition to CT, 75 patients underwent MRCP and 37 patients underwent ERCP. Two radiologists independently evaluated the CT images, including multiplanar and minimum intensity pixel projection reformations, for pancreatic duct strictures, stones, pancreas divisum, or cystic pancreatic lesions. The latter were classified as communicating or not communicating with the pancreatic ducts. Findings on ERCP or MRCP were used to calculate diagnostic performance parameters. RESULTS: On standard of reference examinations, 15 (16%) of the 93 patients had a pancreatic duct stricture. The sensitivity and the specificity for Observer 1 were 87% and 100%, respectively; for Observer 2, 100% and 100%, respectively. Six (6%) of the 93 patients had main pancreatic duct stones. The sensitivity and the specificity for Observer 1 were 83% and 100%, respectively; for Observer 2, 100% and 99%, respectively. Five (5%) patients had pancreas divisum; Observer 1 correctly identified four and Observer 2 correctly identified three cases. Eleven (12%) patients had cystic pancreatic lesions. Observer 1 correctly determined whether or not there was communication between the cystic pancreatic lesion and the pancreatic duct in ten cases; Observer 2 correctly made this determination in nine cases. CONCLUSION: Portal venous phase 64 MDCT images are moderately sensitive and highly specific for detecting pancreatic duct stricture, stones, and pancreas divisum and moderately accurate for detecting communication between pancreatic ducts and cystic pancreatic lesions.
BACKGROUND: Many patients presenting with nonspecific signs and symptoms often receive CT scans using general protocols, not optimized to evaluate for pancreatic pathology. Therefore the purpose of this study was to evaluate portal venous phase 64 multi-row detector CT (MDCT) scans for detecting pancreatic duct strictures, stones, pancreas divisum, and communication between pancreatic ducts and cystic pancreatic lesions. METHODS: Institutional review board approval with waived informed consent was obtained for this HIPAA-compliant study. We included all patients that underwent abdominal, portal venous phase, intravenous contrast-enhanced 64 MDCT scans between 6/7/05 and 5/01/07 and MR cholangiopancreatography (MRCP) or endoscopic retrograde pancreatography (ERCP) within 2 months of the CT. This yielded 93 patients (42 males, 51 females) with a mean age of 59 years. In addition to CT, 75 patients underwent MRCP and 37 patients underwent ERCP. Two radiologists independently evaluated the CT images, including multiplanar and minimum intensity pixel projection reformations, for pancreatic duct strictures, stones, pancreas divisum, or cystic pancreatic lesions. The latter were classified as communicating or not communicating with the pancreatic ducts. Findings on ERCP or MRCP were used to calculate diagnostic performance parameters. RESULTS: On standard of reference examinations, 15 (16%) of the 93 patients had a pancreatic duct stricture. The sensitivity and the specificity for Observer 1 were 87% and 100%, respectively; for Observer 2, 100% and 100%, respectively. Six (6%) of the 93 patients had main pancreatic duct stones. The sensitivity and the specificity for Observer 1 were 83% and 100%, respectively; for Observer 2, 100% and 99%, respectively. Five (5%) patients had pancreas divisum; Observer 1 correctly identified four and Observer 2 correctly identified three cases. Eleven (12%) patients had cystic pancreatic lesions. Observer 1 correctly determined whether or not there was communication between the cystic pancreatic lesion and the pancreatic duct in ten cases; Observer 2 correctly made this determination in nine cases. CONCLUSION: Portal venous phase 64 MDCT images are moderately sensitive and highly specific for detecting pancreatic duct stricture, stones, and pancreas divisum and moderately accurate for detecting communication between pancreatic ducts and cystic pancreatic lesions.
Authors: Vladimir M Kushnir; Sachin B Wani; Kathryn Fowler; Christine Menias; Rakesh Varma; Vamsi Narra; Christine Hovis; Faris M Murad; Daniel K Mullady; Sreenivasa S Jonnalagadda; Dayna S Early; Steven A Edmundowicz; Riad R Azar Journal: Pancreas Date: 2013-04 Impact factor: 3.327