Sensitivity of EUS-Guided Fine Needle Aspirate vs Fine Needle Biopsy with a Newly Designed Needle for Pancreas Cancer in a Large, Referral Population

Abstract

Purpose: EUS-fine needle aspiration (FNA) is the preferred method for diagnosing pancreas cancer in the U.S. Reported sensitivity varies but remains between 80-90% in most studies. Specificity approaches 100%. A new needle has been designed to generate fine needle biopsies (FNB) and enhance tissue acquisition for FNA in the hope of increasing yields. Methods: We reviewed all patients who underwent EUS for a pancreatic mass by 1 endosonographer since the introduction of the ProCore FNB needle in 2010, and compared these results to patients who underwent EUS-FNA with a standard hollow needle from the same manufacturer (Echotip, Cook Medical) by the same individual 2 years earlier. FNA techniques and cytologic preparations were similar. Final diagnosis was confirmed by histologic examination of surgical specimens or clinical follow-up. Diagnostic yields were compared. Results: Between January 2005 and September 2008, a total of 113 patients underwent EUS-FNA for pancreas masses. 72 were ultimately diagnosed with a malignancy, of which EUS-FNA was positive in 61 (84%). A median of 4 passes were made per case with a 19, 22 or 25 gauge needle. Final diagnosis included pancreatic adenocarcinoma (60), neuroendocrine tumor (5), lymphoma (3), and metastatic carcinoma (2). There was one case of bleeding. Between October 2010 and November 2011, 62 patients underwent EUS assessment for presumed pancreas cancer. 13 underwent EUS-FNA and 49 EUS-FNB with a ProCore needle.17 cases were performed with a 19 gauge needle, the remainder with a 22 gauge needle. 47 patients proved to have malignancies (9 FNA, 38 FNB). These included 42 adenocarcinoma, 4 neuroendocrine tumors and 1 metastatic breast cancer. Cytology was positive in only 4 of 9 (44%) cancers in the FNA group with a median of 4 needle passes but in 36 of 38 FNB cases (95%) (median 4 passes). 1 patient with initially negative FNB was positive on a second FNB attempt. Pancreatitis was the only complication and occurred in 1 FNA patient. No false positive results were encountered in either group. The sensitivity for FNB was significantly greater than FNA within the last year (p=0.002). When FNB from group 2 is compared to FNA from Group 1 only, the difference does not reach significance (P=0.13). If all FNA cases are combined, FNB sensitivity exceeds FNA at borderline significance (P=0.054). If the larger groups are compared by patient rather than by procedure, FNB sensitivity exceeds FNA (P=0.039). Conclusion: Both EUS-FNA and FNB effectively collect tissue for the diagnosis of pancreas cancer. FNB appears to be more sensitive than FNA for the diagnosis of pancreas cancer. Disclosure: Dr. Schembre is a patent holder for the ProCore needle and serves on the speaker bureau for Cook Medical. Dr. Brandabur has no financial conflicts to report.