Experience of endoscopic ultrasound-guided fine needle aspiration in a regional teaching hospital.

Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) was first reported in 1991 and 1992 for gastric submucosal tumor and pancreas cancer. Subsequently, the indications have expanded to mediastinal lesions, liver, spleen, adrenal gland, ascites, pleural effusion, intra-abdominal lymph nodes. We report our experience of EUS-FNA and the number of cases required during the learning curve. From May 2005 to December 2007, we did a retrospective analysis of 45 cases (median age: 68 [range] [37-87] years; 34 men and 11 women) of EUS-FNA punctures on solid masses. EUS-FNA was performed with linear echoendoscope (Olympus GF-UCT2000, EUC2000 unit) using 22 gauge needles (Olympus power-shot needle and Wilson Cook echo-tip needle). The procedure was done by a single endoscopist. Patients were divided into 2 groups, the first 30 patients (group A) and the last 15 patients (group B). The median size of tumor was 2.6 cm (range: 1-14 cm) in group A and 2 cm (range: 0.7-3.5 cm) in group B (p=0.023). The median number of punctures was the same in both groups, i.e. 2 (range: 1-4). The sensitivity, specificity, positive predictive value, negative predictive value, accuracy in groups A and B were 55.1% vs. 85.7%, 100% vs. 100% 100% vs. 100%, 7.1% vs. 33.3%, 56.6% vs. 86.6% respectively. Group B patients had smaller tumor size, but with higher sensitivity and accuracy rates. Only one patient in group A needed 4 units of blood transfusion. EUS-FNA is a safe intervention; the sensitivity and accuracy rate can be improved after a learning curve of 30 cases.