Fine‐needle biopsy of the pancreas: Results of 204 routinely performed biopsies in 190 patients

Abstract

Abstract Two‐hundred and four fine‐needle aspiration biopsies of the pancreas have been performed in 190 patients during a 12‐year period. Sixty‐one of these were performed percutaneously guided by endoscopic retrograde cholangio‐pancreatography, percutaneous transhepatic cholangiography, angiography, or ultrasonography; and 143 were taken intraoperatively. In 77 (67%) out of 115 patients with pancreatic cancer, a correct cytological diagnosis was obtained. Two biopsies were reported as malignant in 1 patient who ultimately was found to have chronic pancreatitis (false positives). The frequency of “not representative” biopsies varied from 20.8% in patients with suspected cancer biopsied intraoperatively to 48.4% in patients biopsied preoperatively. A correct cytological diagnosis of malignancy was obtained preoperatively in 54.6% of patients with cancer, in 60.0% of patients evaluated without later operation, and in 71.1% of patients biopsied during laparotomy for suspected pancreatic cancer. The overall false‐negative rate was 9.8%. The predictive value of a positive test was almost 100%, whereas the predictive value of a negative test was only 69.6% (total material). Analyses may indicate that a more aggressive approach with multiple punctures may lower the not representative biopsy rate and increase the diagnostic accuracy in patients with pancreatic cancer.