Differentiation of mucinous cysts and simple cysts of the liver using preoperative imaging

Abdominal Radiology(2022)

引用 0|浏览8
暂无评分
摘要
Purpose Preoperative radiographic differentiation of mucinous cystic neoplasms (MCN) and simple cysts (SLC) of the liver is challenging. Previous data have demonstrated that the finding of septations arising from the cyst wall without indentation on cross-sectional imaging is associated with MCN. We aim to assess whether this radiographic feature is diagnostic of MCN. Methods A prospectively maintained database was queried for patients with a preoperative diagnosis of a cystic liver lesion who subsequently underwent operative intervention. The feature of septations without indentation of the cyst wall was evaluated on cross-sectional imaging obtained within 3 months of operation. Imaging was independently evaluated by three radiologists blinded to pathology and interobserver agreement was compared to assess the diagnostic accuracy of this feature as well as the overall likelihood of the lesion representing a MCN. Results There were 95 patients who met inclusion criteria; 80 (84%) had SLC on pathology, while 15 (16%) had MCN. Presence of septa without indentation of cyst wall had high sensitivity (range 80–87%), but low specificity (range 48–66%). Interobserver percent agreement (PA) was 51% [ κ = 0.35 (95% CI 0.22–0.47)]. Sensitivity among the three radiologists ranged between 20 and 80% and specificity between 71 and 91% for the likelihood of the lesion representing MCN versus SLC, with an area under the curve (AUC) of 0.67–0.79; however, interobserver agreement was fair [ κ = 0.40 (95% CI 0.25–0.55), PA = 67%]. Conclusion The presence of septations without indentation of cyst wall demonstrates adequate sensitivity to differentiate MCN and SLC. However, there is variability for detection of this feature and therefore, it alone is of limited clinical value.
更多
查看译文
关键词
Simple liver cyst, Mucinous cystic neoplasm, Biliary cystadenoma, Biliary cystadenocarcinoma, Radiographic imaging
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要