The Sonobrush: A Novel EUS-Guided Biopsy Device for Pancreatic Masses

Abstract

Background: Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) has become a powerful technique in the diagnostic evaluation of pancreatic masses. Accuracy of current EUS-FNA techniques are limited due to a low cellular yield, particularly in pancreatic carcinoma. The aim of this study was to determine the feasibility, safety and efficacy of a novel EUS-guided brush biopsy in the acquisition of tissue from solid pancreatic masses. Methods: Patients referred for EUS-FNA evaluation of pancreatic masses were candidates for the study. Masses were first aspirated in a standardized fashion using two passes of a 22 ga FNA needle (Wilson-Cook). Following FNA, a 19 ga needle containing a brush device was placed into the mass (Mediglobe). The brush was gently agitated within the mass to obtain cellular tissue. FNA samples were analyzed by direct smear. Brush biopsy samples were rinsed in saline and processed by either ThinPrep (Cytyc Corp.) and/or a cell block. Ease of use was ranked on a 5 point subjective scale (1=very easy →5= very difficult). Also recorded were: length of time required, cellularity of the sample (graded from 0=acellular → 3=high cellularity) and ability of the device to provide an accurate diagnosis. Results: Eleven patients were enrolled with 6 women and an average age of 65.5 years. The sonobrush deployed successfully in all cases and there were no complications. Average time for brush biopsy was 137 seconds as compared to 85 seconds for regular FNA. Ease of use of FNA was rated with an average of 1.6 while brush biopsy ranked an average of 2.1. Average cellularity of cytologic sampling by FNA was 2.7, 1.4 for sonobrush thin preps and 2.0 for sonobrush cell block preparations. FNA was diagnostic in 11/11 cases: 8 adenocarcinomas and 2 endocrine tumors (PETs). Sonobrush was diagnostic in 7/11. In three cases, however, Sonobrush produced small architecturally important tissue fragments that enhanced cytological diagnosis (one adenocarcinoma was found to be invasive; tissue in cell blocks of the 2 PETs allowed confirmatory ancillary studies.) Conclusion: EUS-guided brush biopsy is a feasible and safe means of obtaining tissue for the diagnosis of pancreatic masses and may complement FNA. In addition, larger groups of cells obtained by this technique may allow for more specific diagnosis in some cases. Further studies are required to delineate the full potential of this new device.