Methods of tissue preparation after EUS‐guided tissue acquisition without rapid on‐site assessment: Results of a randomized study

• Published in: Journal of gastroenterology and hepatology 2023-05, Vol.38 (5), p.733-739
• Main Authors: Phan, Vinh‐An, Ruszkiewicz, Andrew, Safaeian, Romina, Zobel, Joshua, Nguyen, Nam Q.
• Format: Article
• Language: English
• Subjects: Bile Duct Neoplasms - pathology; Bile Ducts, Intrahepatic - pathology; Cholangiocarcinoma; Endoscopic ultrasound; Endoscopic Ultrasound-Guided Fine Needle Aspiration - methods; Fibroids; Fine needle aspiration; Histology; Humans; Inflammation; Lymphoma; Metastases; Neuroectodermal Tumors, Primitive - pathology; Neuroendocrine tumors; Pancreas; Pancreas - diagnostic imaging; Pancreas - pathology; Pancreatic cancer; Pancreatic Neoplasms - diagnostic imaging; Pancreatic Neoplasms - pathology; Pancreatitis; Retrospective Studies; Tissue preparation
• ISSN: 0815-9319; 1440-1746
• DOI: 10.1111/jgh.16137

In the absence of rapid on‐site evaluation (ROSE), it is not clear which method of tissue preparation is best to process tissue obtained from EUS guidance. Cytological smearing (CS), cell block (CB), and direct histology (DH) are the available techniques. Aim To compare the diagnostic yield of three techniques of tissue preparation for EUS‐guided tissue acquisition without ROSE. Methods Patients who were referred for EUS‐FNA of peri‐gastrointestinal masses were recruited. Without ROSE, each lesion was biopsied with three needle passes, and the order in which tissue is prepared was randomized to either (i) CS + CB, (ii) CB only, or (iii) DH only. The prepared specimens were reviewed. Results A total of 243 specimens were taken from 81 patients. Tissue diagnosis was achieved in 78/81 (96.3%) of patients, including 63 neoplasms (PDAC [n = 45], pancreatic neuroendocrine tumors [PNET; n = 4], cholangiocarcinoma [n = 5], metastatic disease [n = 4], lymphoma [n = 1], linitis plastica [n = 2], leiomyoma [n = 2]) and 15 benign pathologies (chronic pancreatitis [n = 8], reactive nodes [n = 5], inflammatory biliary stricture [n = 1], and pancreatic rest [n = 1]). The three non‐diagnostic cases were found to be PDAC (n = 2) and PNET (n = 1). Sensitivity and diagnostic accuracy was highest with DH (94 and 95%), which was significantly better than that by CS + CB (43 and 54%; P = 0.0001) and CB‐only preparations (32 and 48.6%; P  0.22). Conclusion Without ROSE, our findings suggest that with just a single pass, DH should be the tissue preparation method of choice given its significantly higher diagnostic accuracy compared with CS and/or CB techniques.