Objectives Current guidelines recommend magnetic resonance imaging (MRI)/magnetic resonance cholangiopancreatography (MRCP) for risk stratification of intraductal papillary mucinous neoplasms (IPMNs). We assessed the interobserver agreement among radiologists in evaluating and… Click to show full abstract
Objectives Current guidelines recommend magnetic resonance imaging (MRI)/magnetic resonance cholangiopancreatography (MRCP) for risk stratification of intraductal papillary mucinous neoplasms (IPMNs). We assessed the interobserver agreement among radiologists in evaluating and risk stratifying IPMNs. Methods This single-center study evaluated 30 patients with IPMNs who had undergone MRI/MRCP, endoscopic ultrasound, and/or surgical resection. Six abdominal radiologists evaluated the MRI/MRCPs to document multiple parameters. The analysis applied Landis and Koch κ interpretation for categorical variables and intraclass correlation coefficient (r) for continuous variables. Results Radiologists demonstrated almost perfect agreement for location (κ = 0.81, 95% confidence interval [CI], 0.74–0.87), size (r = 0.95; 95% CI, 0.89–0.98), and main pancreatic duct diameter (r = 0.98; 95% CI, 0.96–0.99). Substantial agreement was observed for communication with the main pancreatic duct (κ = 0.66; 95% CI, 0.57–0.75) and classification of IPMN subtype (κ = 0.77; 95% CI, 0.67–0.86). Presence of intracystic nodules (κ = 0.31; 95% CI, 0.21–0.42) and wall thickening (κ = 0.09; 95% CI, −0.01 to 0.18) reached only fair and slight agreement, respectively. Conclusions Although MRI/MRCP is excellent in the evaluation of spatial aspects, there is lower reliability for nondimensional characteristics of IPMNs. These data support guideline-recommended complementary evaluation of IPMNs with MRI/MRCP and endoscopic ultrasound.
               
Click one of the above tabs to view related content.