Abstract
Of 144 patients with suspected pancreatic disease in whom a 75Se-selenomethionine scan was performed, endoscopic retrograde pancreatography (ERP) was successful in 108 (75%). The final diagnosis is known in 100 patients and has been compared with scan and ERP findings. A normal scan reliably indicated a normal pancreas, but the scan was falsely abnormal in 30%. ERP distinguished between carcinoma and chronic pancreatitis in 84% of cases but was falsely normal in five patients with pancreatic disease. In extrahepatic biliary disease both tests tended to give falsely abnormal results. A sequence of tests to provide a rapid and reliable assessment of pancreatic function should be a radio-isotope scan, followed by ERP if the results of the scan are abnormal, and a Lundh test if the scan is abnormal but the findings on ERP are normal.