Yield of endoscopic ultrasound guided fine needle aspiration and clinical factors associated with accuracy in patients with pancreatic solid tumor

Purpose: Endoscopic ultrasound guided fine needle aspiration (EUS-FNA) is a potential method to diagnose pancreatic mass, but factors associated with the accurate cytological result of EUS-FNA in pancreatic solid tumor are not well known. The aim of this study was to evaluate the yield of EUS-FNA and clinical factors which might have influence on the accuracy of the result of EUS-FNA, such as physicians, location of mass, the number of needle passage attempts and size of mass.

Methods: From September 2005 to April 2010, patients who had pancreatic mass on other radiologic exam like abdominal ultrasound (US), computerized tomography (CT) or magnetic resonance imaging (MRI) in a single tertiary referral center were included. Among them, patients with pancreatic cystic or mixed tumor were excluded. Total 79 patients who met the above criteria underwent EUS-FNA for diagnosis. Their medical records were retrospectively reviewed.

Results: Total 85 EUS-FNA were performed in 64 and 15 patients who had malignant and benign mass, respectively. Rates of male were 67.2% (43/64) and 93.3% (14/15, p=0.06), and mean age were 62.8 ± 9.7 and 60.4 ± 15.0 years old in both group (p=0.57). The locations of mass were uncinate process (12, 15.2%), head (24, 30.4%), neck (7, 8.9%), body (15, 19.0%) and tail (21, 26.6%) and mean size was 3.4 ± 1.1cm. The sensitivity, specificity and accuracy of EUS-FNA in the diagnosis of malignancy using strict cytology criteria which considered only positive cytologies as true malignancy was 27.9%, 100.0% and 37.6%, respectively. Meanwhile, the sensitivity, specificity and accuracy of EUS-FNA was 72.1%, 88.2% and 72.9%, respectively, using loose cytology criteria which considered positive malignant, suspicious and atypical cytologies as true malignancy. According to both criteria, there were no evidence of any associations between EUS cytological adequacy and several clinical factors such as physicians (p=0.90), location of mass (p=0.89), numbers of needle passage attempts (<=3 versus >3, p=0.25) and size of mass (<=3cm versus >3cm, p=0.37). Complications in this study were 5 cases of minor bleeding and 3 of abdominal pain, all of which were conservatively manageable.

Conclusion: The yield of EUS-FNA is significantly influenced by different criteria. But other factors such as physicians, location of mass, the number of needle passage attempts and size of mass do not show any influence on accuracy. EUS-FNA is a feasible and safe method for diagnosis of pancreatic solid tumor but it shows a limited result without an onsite cytopathologist.