Background: Studies indicate enhanced diagnostic accuracy for digital image analysis (DIA) and fluorescence in situ hybridization (FISH) versus routine cytology examination (RC) when biliary strictures are evaluated. These tumor markers have not been applied to EUS-guided FNA.
Objective: Our purpose was to determine the accuracy of RC versus the composite results of DIA/FISH.
Design: Patients enrolled with known or suspected malignancy. The final diagnosis was based on strict cytopathologic and imaging criteria and 12-month follow-up.
Settings: Tertiary referral center.
Patients: A total of 39 patients were enrolled in whom each diagnostic test was performed on samples from 42 sites to evaluate lymphadenopathy (n=19), pancreatic mass (n=19), esophageal or gastric wall mass (n=3), and thyroid mass (n=1).
Interventions: EUS-guided FNA with RC, DIA, and FISH.
Main outcome measurement: Diagnostic accuracy of RC, DIA, and FISH.
Results: Malignancy was diagnosed in 30 of 42 patients, including esophageal squamous cell carcinoma, esophageal adenocarcinoma, gastric adenocarcinoma, pancreatic adenocarcinoma, pancreatic mucinous cystic neoplasia, intraductal papillary mucinous neoplasia, metastatic forearm sarcoma, small cell and non-small cell lung cancer, thyroid carcinoma, malignant GI stromal tumor, melanoma, adenocarcinoma of unknown primary, and lymphoma. The sensitivity, specificity, and accuracy of DIA/FISH versus RC for detecting malignancy were 97%, 100%, and 98% versus 87%, 100%, and 90%, respectively.
Limitations: Single-center pilot study.
Conclusions: Our findings suggest that DIA and FISH processing of EUS-guided FNA specimens provides higher diagnostic accuracy than RC does. These data suggest that these tumor markers incorporate generic targets as suggested by the high diagnostic sensitivity in this patient cohort with diverse pathologic conditions.