EUS-FNA with rescue fluorescence in situ hybridization for the diagnosis of pancreatic carcinoma in patients with inconclusive on-site cytopathology results

Background: Detection of chromosomal abnormalities by fluorescence in situ hybridization (FISH) analysis has not been well-studied in FNA samples of pancreatic masses. Selective use of FISH in patients with inconclusive on-site cytopathology results may improve the sensitivity of EUS for malignancy. Objective: To determine the sensitivity and specificity of FISH analysis in patients with inconclusive on-site cytopathology results. Design: Consecutive patients with suspected pancreatic malignancy, nonrandomized cohort study. Final diagnosis was based on either surgical biopsy or disease progression on extended follow-up or death. Setting: Academic center, tertiary-care referral cancer center. Patients: A total of 212 EUS examinations were performed in 206 patients for solid pancreatic lesions over a 24-month period (January 2009-December 2010). FISH analysis was done for 69 patients with inconclusive or nonavailable on-site cytology results. Intervention: EUS-guided FNA (EUS-FNA) of solid pancreatic masses with cytology and FISH analysis for polysomy of chromosomes 3, 7, and 17 and deletion of 9p21. Main Outcome Measurements: Sensitivity/specificity of cytology, FISH, and a composite of cytology and FISH. Results: Patients with positive on-site cytology (110), neuroendocrine tumors (22), insufficient follow-up (1), FISH not obtained (3), and renal cancer with pancreatic metastasis (1) were excluded. Sixty-nine patients comprised the study cohort, 54 with malignancy and 15 with benign disease. Sensitivity for malignancy of cytology, FISH analysis, and the combination were 61%, 74%, and 85%, respectively (P =.009). FISH detected an additional 13 cases of pancreatic adenocarcinoma missed by cytology. There was no false-positive FISH analysis in 15 patients with benign disease. No major complications occurred from EUS-FNA. Limitations: Single center, selected patients underwent FISH analysis, limited number of patients with benign disease. Conclusion: In patients with suspected pancreatic cancer, FISH analysis can detect additional cases missed by cytology without compromising specificity. FISH analysis to detect polysomy of chromosomes 3, 7, and 17 and deletion of 9p21 should be considered when cytology is negative for malignancy in patients with a known pancreatic mass.