Seven hundred forty-seven hepatectomies in the 1990s: An update to evaluate the actual risk of liver resection

Background: Recent reports highlighting reduced mortality rates to less tha n 1% after hepatic resections have evaluated the management of selected pat ients. The current risk of liver resection in unselected patients needs to be more clearly defined to appreciate the actual risk of new indications. Study Design: From 1990 to 1997, 747 consecutive patients, including 16 ope rated in emergency, underwent hepatic resection. Resection was indicated fo r malignancy in 473 patients (63%). Major resections were performed in 333 patients (45%). An underlying liver disease, including cirrhosis (n = 239) and obstructive jaundice (n = 4), was present in 253 patients (35%). Multiv ariate analysis of the risk factors for postoperative mortality, morbidity, and transfusion after stratifying patients for the circumstance of the ope ration and the pathological changes of the remnant liver was performed. Results: There was no intraoperative death and the overall mortality rate w as 4.4%. This rate was 25% after emergency liver resection and 3.9% after e lective liver resection (p < 0.001). After elective resection, mortality wa s significantly higher in patients with cirrhosis (8.7%) or obstructive jau ndice (21%) than in patients with a normal liver (1%; p < 0.001). Analysis of this subgroup of 478 patients with normal liver showed that the mortalit y rate was 0% in 220 patients operated for a benign disease and in 263 pati ents who underwent minor resections. All five deaths occurred in patients w ith a malignancy and resulted from extrahepatic complications. In patients with a malignancy, the only independent predictor of death was an associate d extrahepatic procedure. The incidence of postoperative complications was 22% and was influenced by the American Society of Anaesthesiology (ASA) sto re, extent of resection, presence of a steatosis, and an associated extrahe patic procedure. The incidence of major complications was 8% and of reopera tion 3%. Perioperative blood transfusion was required in 112 of 478 (23%) a nd was not associated with increased mortality Conclusions: The 1% basic risk of elective liver resection on normal liver suggests that indications of resection for malignancy could be extended, un less an associated extrahepatic procedure is needed. Because of this low ba sic risk, future studies evaluating resection on normal liver should not co nsider in-hospital mortality as the only end point. (J Am Cell Surg 2000;19 1: 38-46. (C) 2000 by the American College of Surgeons).