Is pancreatoduodenectomy justified for periampullary cancers with regional lymph node involvement?

Justification of pancreatoduodenectomy for highly malignant periampullary cancers with regional lymph node involvement is questioned. Attempting to clarify the therapeutic dilemma, we compared the prognoses of resectable periampullary cancers with and without lymph node involvement, as well as unresectable cancers with lymph node involvement. The medical records of surgical patients with periampullary adenocarcinoma were reviewed. We compared the survival times of resectable cancers with (resectable TanyN1M0) and without (resectable TanyN0M0) regional lymph node involvement, and the survival times of resectable cancers with lymph node involvement (resectable TanyN1M0) and unresectable cancers with lymph node involvement (unresectable TanyN1M0). There were 138 resectable and 115 unresectable periampullary cancers including 117 cancers of the pancreatic head, 117 cancers of ampulla of Vater, 7 cancers of duodenum, and 12 cancers of distal common bile duct. The prognosis was very poor in cancer of the pancreatic head. Neither resectability nor status of lymph node involvement could influence the survival time of the cancer of pancreatic head. In resectable non-pancreatic periampullary cancers, the prognosis was significantly better in the group without lymph node involvement. However, once the lymph node was involved, the outcomes were the same in the resectable and unresectable groups. Although pancreatoduodenectomy does not seem to be justified for periampullary cancers with regional lymph node involvement, this procedure should be considered for periampullary cancers without nodal involvement.