Per-Pass Performance Characteristics of Endoscopic Ultrasound-Guided Fine-Needle Aspiration of Malignant Solid Pancreatic Masses in a Large Multicenter Cohort.

Abstract

Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) is widely performed for the evaluation of pancreatic masses. We evaluated the performance characteristics of EUS-FNA in obtaining a diagnosis of pancreatic malignancy. We performed a multicenter study of patients who underwent EUS-FNA for a solid pancreatic mass. Endoscopic ultrasound-guided FNA was standardized using a 25-gauge needle, slow-pull stylet technique for specimen acquisition, and on-site cytopathology. For the primary analysis, only malignant cytology was regarded as positive. A secondary analysis was performed in which malignant and/or suspicious cytology was regarded as positive. A total of 138 patients underwent EUS-FNA. In the primary analysis, the sensitivity of EUS-FNA for malignancy was 56.7% on first pass, 73.3% on second pass, 83.3% on third pass, 89.2% on fourth pass, and 90.8% on fifth pass, with no increase beyond the fifth pass. In the secondary analysis, the sensitivity was 75.0% on first pass, 89.2% on second pass, 93.3% on third pass, and 95.8% on fourth pass, with no increase beyond the fourth pass. No significant relationship was seen between lesion size and diagnostic yield. Using a 25-gauge needle, the maximal diagnostic yield of EUS-FNA for a solid pancreatic mass is reached after 4 needle passes.