Yield of EUS-guided FNA of pancreatic masses in the presence or the absence of chronic pancreatitis

Abstract

Background: Evaluation of a focal pancreatic mass in the setting of chronic pancreatitis (CP) is a diagnostic challenge. The objectives of the study were to compare the diagnostic yield and accuracy of EUS-guided FNA (EUS-FNA) in the evaluation of pancreatic-mass lesions in the presence or the absence of CP and to identify predictors of CP before EUS-FNA of pancreatic-mass lesions.

Methods: The study design was analysis of data collected prospectively on all patients with solid pancreatic-mass lesions who underwent EUS-FNA at a tertiary referral center. A total of 282 consecutive patients underwent 300 EUS-FNA procedures of pancreatic-mass lesions over a 3-year period. The diagnostic yield and the accuracy of EUS-FNA was compared between patients with and without CP. CP was defined by the presence of more than 4 EUS criteria.

Results: Final diagnosis was adenocarcinoma in 210 (70%), benign disease in 64 (21%), other pathology in 19 (6%), and indeterminate in 4 (2%); 3 patients (1%) were lost to follow-up. CP was noted in 75/300 (25%) patients. A lower sensitivity for EUS-FNA was observed in patients with CP than in those without CP (73.9% vs. 91.3%; p = 0.02). While patients with CP had a higher negative predictive value (88.9% vs. 45.5%; p < 0.001), no significant differences were observed for specificity (100% vs. 93.8%), positive predictive value (100% vs. 99.5%), and accuracy (91.5% vs. 91.4%) between those with and without CP. False-negative cytology was encountered in 24 cases: 6/71 (8%) with CP vs. 18/222 (8%) without CP. Patients with CP required more EUS-FNA passes to establish a diagnosis vs. those without CP (median, 5 vs. 2; p < 0.001). On multivariable analysis, age < 50 years (p < 0.001), male gender (p < 0.001), black race (p = 0.001), and the absence of jaundice (p = 0.005) were significantly associated with CP. The impact of EUS-FNA on long-term clinical management was not analyzed. The impact of individual EUS features of CP on sensitivity of EUS-FNA was not evaluated. By protocol, mass lesions that were benign required more passes to definitively exclude malignancy.

Conclusions: EUS-FNA has a low sensitivity for pancreatic-mass lesions in the setting of CP. This decreased sensitivity can be overcome by performing more numbers of passes at FNA, which improves diagnostic accuracy. Demographic features and clinical presentation are predictive of underlying CP in patients with pancreatic-mass lesions.