Assessment of On-Site Evaluation (OSE) and Diagnostic Yield of EUS-Fine Needle Aspiration (FNA) Versus EUS-Guided Needle Core Biopsy Using EchoTip ProCore (ETP) Device
Sharvari S Dalal, Lisa Pitelka-Zengou, Ajit Singh Paintal, Piotr Kulesza, Aparna Mahajan, Kumar Krishnan, Srinadh Komanduri, Xiaoqi Lin, Raj Keswani, Ritu Nayar. Northwestern Memorial Hospital, Chicago, IL
Background: EUS is an established procedure for diagnosing and staging gastrointestinal (GI) lesions. Usually FNA is used to obtain lesional material, but can have low yield and sensitivity for certain tumors or anatomic locations limiting material available for diagnosis. Our pilot study used a recently introduced ETP device (Cook Medical, USA) and showed it to be technically feasible and have a high diagnostic yield. Here we assessed OSE and diagnostic yield of ETP compared to FNA.
Design: 46 cases that had tandem FNAs and ETPs were evaluated. Between 3-8 FNA and 1-5 ETP passes were performed based on OSE. For all cases, FNA smears, FNA cell blocks (CB), touch preps (TP) of ETPs, and ETP slides were reviewed. Diagnostic rates of both techniques were evaluated. IHC was performed when appropriate.
Results: 38 cases with definitive diagnoses were evaluated. 6 "unsatisfactory" and 2 "atypical" diagnoses were excluded. Final diagnoses were possible by FNA/CB in 26 cases. Diagnoses could be rendered by ETP/TP in 33 cases.
There were 14 adenocarcinomas (AC) and 24 other lesions. Amongst our AC, a diagnosis was reached by FNA/CB in 12 cases and ETP/TP in all 14 cases. Subjectively, we found both techniques to be equivalent in terms of diagnostic material provided.
Amongst the 24 non-AC (1 squamous cell carcinoma, 2 lymphomas, 12 spindle cell lesions, 4 endocrine tumors, 4 inflammatory lesions, 1 benign), definitive diagnoses could be reached by FNA/CB in 14 cases, and by ETP/TP in 19 cases. Amongst our non-ACs, the most common tumors were GISTs. FNA/CB and ETP/TP were each diagnostic in 5 of 7 GISTs.
As far as OSE, specimens that were ultimately diagnostic on ETP had adequate TP OSE 60% of the time, whereas diagnostic FNA/CB specimens had adequate OSE 61% of the time. Adequacy assessments were particularly unreliable in GISTs (adequate in diagnostic cases less than half the time by either technique).
Conclusions: TP/ETP provided diagnostic material more frequently than FNA/CB in both ACs and non-ACs. The predictive rate of OSE could possibly be improved by performing crush preps rather than TPs on portions of the ETP, particularly in cases of suspected spindle cell lesions. In cases that were ultimately diagnostic, ETP provided adequate material within 4 passes 85% of the time, suggesting its utility as a stand-alone technique in settings where OSE is not available.
Tuesday, March 20, 2012 9:30 AM
Poster Session III # 63, Tuesday Morning