Staging laparoscopy for pancreatic cancer should be used to select the best means of palliation and not only to maximize the resectability rate.

Staging laparoscopy, based on the assumption that endobiliary stenting is the best palliation, allegedly saves an "unnecessary" laparotomy for incurable pancreatic cancer. Our aim was to determine survival of patients with clinically resectable pancreatic cancer that is found to be unresectable intraoperatively and thereby infer appropriate utilization of staging laparoscopy. A retrospective analysis was undertaken of 148 patients with ductal adenocarcinoma (1985 to 1992) with a clinically resectable lesion based on current imaging techniques. All were considered candidates for resection but were deemed unresectable at operation because of metastases to the liver (group I; 29 patients), the peritoneum (group II; 22 patients), or distant lymph nodes (group III; 44 patients) or because of vascular invasion (group IV; 53 patients). Overall median survival was 9 months (range 1 to 53 months), but by group was as follows: group I, 6 months; group II, 7 months; group III, 11 months; and group IV, 11 months. Individual comparisons showed shorter survival for patients with distant nodal, liver, or peritoneal metastases than with nodal or vascular involvement (P<0.03). Staging laparoscopy should be performed to identify patients with liver or peritoneal metastases who have an expected survival of approximately 6 months, in whom short-term endoscopic palliation is satisfactory. Extended laparoscopy to identify lymph node or vascular involvement is contingent upon which palliation (operative vs. endoscopic) is considered most appropriate. Because we believe operative bypass provides better, more durable palliation in this latter group, we have not adopted extended laparoscopy.