Diabetes is associated with increased perioperative mortality but equivalent long-term outcome after hepatic resection for colorectal cancer.

Diabetes is associated with alterations in liver metabolism and immune response that may influence postoperative recovery and long-term survival after hepatectomy for cancer. Patients with type I or type II diabetes mellitus submitted to a potentially curative hepatic resection for metastatic colorectal cancer were identified from the prospective database, and compared with patients with hepatic colorectal metastases submitted to resection during the same time interval, but without diabetes mellitus. Data on operative morbidity and mortality and long-term survival were analyzed. Between December 1990 and July 1999, a total of 727 patients underwent hepatic resection, 61 of whom (8.1%) had type I and type II diabetes mellitus. Operative mortality in the diabetic patients was significantly greater than in nondiabetic patients (8% vs. 2%, P < 0.02). Among patients with diabetes mellitus, four of the five perioperative deaths were due to liver failure after major hepatic resection (lobectomy or greater). All four of these patients had significant parenchymal abnormality (three with steatosis). Long-term survival was identical to that in nondiabetic control subjects. We conclude that the presence of diabetes is associated with a higher incidence of perioperative mortality. In patients with diabetes mellitus and parenchymal steatosis, major hepatic resection should be undertaken with caution.