High‐grade atypical epithelial cells in pancreatic mucinous cysts are a more accurate predictor of malignancy than “positive” cytology
2010
BACKGROUND: The Sendai guidelines for risk assessment of malignancy in patients with mucinous cysts lists "positive" cytology as a high-risk feature. In the current study, the authors hypothesized that a cytological threshold of high-grade atypical epithelial cells (AEC) is a more accurate predictor of malignancy. METHODS: The clinical, radiological, and cytological data of 112 patients with histologically confirmed mucinous cysts of the pancreas were reviewed. Cytology slides were blindly reviewed and cells were classified as benign, AEC, or malignant. On histology, neoplasms were grouped as benign (low-grade and moderate dysplasia) and malignant (high-grade dysplasia/carcinoma in situ and invasive carcinoma). RESULTS: There were 92 patients with an intraductal papillary mucinous neoplasm (IPMN) and 20 with a mucinous cystic neoplasm; 39 were malignant and 73 were benign (42 with low-grade dysplasia and 31 with moderate dysplasia). Only 28% (11 of 39) of the malignant cysts were cytologically malignant with a sensitivity of 29%, a specificity of 100%, and an accuracy of 75%. AEC detected 17 additional cancers (44% of all malignant cysts; 16% more than detected on the basis of "positive" cytology). By using AEC as a surgical triage threshold, the sensitivity was 72%, the specificity was 85%, and the accuracy was 80%, with similar values for small (= 3 cm) branch duct IPMN. Nine of 73 (12%) benign cysts were identified with AEC, 4 of which had moderate dysplasia. AEC had a positive predictive value of 87% for the detection of a mucinous cyst with moderate dysplasia or worse. CONCLUSIONS: AEC are a more accurate predictor of malignancy than "positive" cytology in aspirates of pancreatic mucinous cysts, including small branch duct IPMN. AEC warrant a "suspicious" interpretation for appropriate surgical triage.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
18
References
122
Citations
NaN
KQI