Cell block technique and cytological smears for the differential diagnosis of pancreatic neoplasms after endosonography-guided fine-needle aspiration.

Cytological smear is widely employed to analyse specimens obtained from endosonography-guided fine-needle aspiration (EUS-FNA), but false-negative or inconclusive results may occur. A better diagnostic yield can be obtained from processing cell blocks. We compared the effectiveness of the cell block technique and cytological smear in the diagnosis of pancreatic neoplasms. From January 1997 to December 2006, 611 patients with pancreatic tumors were evaluated by EUS-FNA. Surgery was performed in 356 cases, and the other 255 patients were followed clinically for an average of 12.8 months. In total, 282 (46.2%) patients were evaluated with cytological smears, and 329 (53.8%) were evaluated using only cell blocks. Malignant disease was detected in 352 (57.6%) cases, in which adenocarcinoma accounted for 236 (67%) cases. A benign disease was found in the other 259 cases, including 35.1% focal chronic pancreatitis and 32.4% pseudocysts. Aspiration samples were satisfactory in 595 (97.4%) patients after an average of 2.2 (1-4) passes of the needle. Regardless of the cytopathological examination technique, EUS-FNA confirmed malignancy in 269 of 352 (76.4%) cases, and a benign disease in 257 of 259 (99.2%) cases. For patients who received surgery with histologically confirmed lesions, the sensitivity specificity, positive and negative predictive values, and accuracy of the smears versus cell blocks in diagnosing pancreatic tumors were 61% versus 85.2% (P<0.001), 100% versus 93.1%, 100% versus 98.4%, 36% versus 55.1% (P=0.046) and 68% versus 86.5% (P<0.001), respectively The cell block technique demonstrated a hig her sensitivity, negative predictive value and accuracy than cytological smears.