Background: Several studies have investigated the diagnostic accuracy of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) for pancreatic lesions, but they have included only limited patient populations. This study aimed to clarify the diagnostic accuracy of EUS-FNA in a large number of pancreatic lesions, and to describe the factors that influence it.
Methods: From March 1997 to May 2010, 944 consecutive patients who had undergone EUS-FNA for pancreatic solid lesions were evaluated retrospectively. Factors affecting EUS-FNA accuracy were then analyzed.
Results: A total of 996 solid pancreatic lesions were sampled by EUS-FNA. The overall sampling adequacy and diagnostic accuracy of these lesions were 99.3 % (989/996) and 91.8 % (918/996), respectively. The sensitivity and specificity for differentiating malignant from benign lesions were 91.5 % (793/867) and 97.7 % (126/129), respectively. The diagnostic performance was significantly higher when both cytological and cell-block examinations were carried out than with only cytological examination. In multivariate analysis, final diagnosis, location of lesion, lesion size, availability of on-site cytopathological evaluation, and experience of EUS-FNA procedure were independent factors affecting the accuracy of EUS-FNA. On-site cytopathological evaluation and lesion size were found to be the most weighted factors affecting diagnostic accuracy.
Conclusions: EUS-FNA for pancreatic solid lesions yielded a high accuracy and low complication rate. Both cytological and cell-block preparations and on-site cytopathological evaluation contributed to improve the accuracy. The diagnostic ability of EUS-FNA was less for smaller lesions, and repeated procedures may be needed if malignancy is suspected.