論文

© 2019 John Wiley & Sons, Ltd. Rationale, aims, and objectives: The aim of this observational study is to identify factors by which some pancreatic cystic lesions (PCLs) were undetectable at transabdominal ultrasonography (TAUS), using magnetic resonance imaging (MRI) as reference standard. Methods: The database for 781 consecutive subjects who underwent a health checkup including fat computed tomography and upper abdominal MRI as option was searched. The presence of fatty liver and fatty pancreas was diagnosed by TAUS, and atrophic pancreas was determined by reevaluating the image of the pancreas in the chest computed tomography for screening. Subjects with PCL detected and those undetected at TAUS were statistically compared in clinical characteristics. Results: The prevalence of PCL detected at MRI was 17.8% in the general population. Multivariate logistic regression analysis showed that fatty liver, body mass index, and the size of PCL were significantly associated with the factors influencing the visualization of PCL at TAUS (odds ratio [OR]: 0.337, 95% confidence interval [CI]: 0.154-0.734, P = 0.006; OR: 0.852, 95% CI: 0.737-0.985, P = 0.030; OR:1.120, 95% CI: 1.045-1.200, P =.001). Thirty-six PCLs (64.3%) in a total of 56 PCLs were undermeasured by TAUS. Additionally, nine (56%) out of 16 PCLs (≥ 15 mm) were undermeasured by 5 mm or more by TAUS, although a significantly higher detection rate was observed for PCLs (≥ 15 mm) in comparison with that for PCLs (< 15 mm) (80% vs 33.6%, P =.000). Conclusions: It should be noted that coexisting fatty liver may lower the detection of PCL, and its size may be underestimated by TAUS.