Surgical palliation for pancreatic cancer. The UCLA experience

We reviewed the records of 340 patients with a tissue diagnosis of pancreatic cancer treated at UCLA Medical Center between 1973 and 1988. Sixty-one patients underwent pancreatic resection (group I), 173 had some form of surgical palliation (group II), and 106 had neither (group III). The diagnosis was made 1 to 2 months more quickly in the last 8 years of the review than in the first 8 years, but the effect of early diagnosis on curability was negligible. Biliary obstruction was best treated by cholecystojejunostomy or choledochojejunostomy, which were equally effective. Anastomoses to the jejunum were safer and more effective than were those to the duodenum for the relief of biliary obstruction. Gastrojejunostomy should be performed prophylactically as well as therapeutically. It was effective and safe in both settings. Surgical palliation for pancreatic cancer was generally effective and was associated with an operative mortality rate of less than 10%. However morbidity was high, with significant complications occurring in one third of cases.