Usefulness of EUS in the differenzial diagnosis of malignant versus benign of pancreatic endocrine tumors

Background: Pancreatic endocrine tumors (PETs) develop in relatively few patients. It is difficult to diagnose whether it is malignant or benign PETs by EUS findings.

Objective: The aim of this study was to investigate the usefulness of EUS in the differentiation between malignant and benign PETs.

Patients and methods: Forty one patients with histologically or cytologically proven PET who underwent EUS-guided fine needle aspiration (EUS-FNA) between February 1998 and March 2010 were analyzed. The tumor's diameter, margin, echogenic pattern, ultrasonographic texture, and the presence of lateral shadowing, anechoic area (cystic lesion) and obstruction of the main pancreatic duct were documented as EUS findings. EUS findings, Ki-67 index and clinical variables were evaluated for their relationship to malignancy by univariate and multivariate analyses.

Results: EUS-FNA revealed PETs in 38 patients (93%). In the other 3 patients (7%), EUS-FNA were negative for PET, and examinations of the resection specimen confirmed the diagnosis of PET. Based on the 2004 WHO classification, well-differentiated endocrine tumors benign behavior (subgroup 1–1), well-differentiated endocrine tumors uncertain behavior (subgroup 1–2), well differentiated endocrine carcinoma (subgroup 2) and poorly-differentiated endocrine carcinoma (subgroup 3) were observed in 10 (24%), 13 (32%), 16 (39%) and 2 (5%) patients, respectively. Subgroup 1–1 and 1–2 were placed in benign group. Subgroup 2 and 3 were placed in malignant group. Median diameter of tumor was 2.0cm (range 0.6–6.7), and median age was 58 years old (range 23–81). EUS revealed hypoechoic tumor in 37 (90%), tumor with irregular margin in 12 (29%), heterogenic tumor in 24 (59%), tumor with posterior reinforcement in 16 (39%), tumor with anechoic area in 8 (20%) patients. Univariate analysis to identify predictors of malignancy showed tumor's diameter, tumor's ultrasonographic texture, tumor's margin, presence of symptom, and Ki-67 index were significant. Multivariable logistic regression analysis showed that tumor's diameter was the only independent factor.

Conclusion: EUS-FNA is a valuable method in the recognition of PETs. However, only EUS findings does not appear to have sufficient activity in differentiation between benign and malignant PETs.