Background: The International Study Group of Liver Surgery (ISGLS) defined posthepatectomy liver failure as pathological values for the international normalized ratio (INR) and bilirubin 5 days after liver resection. The occurrence of biliary leakage was defined as a drainage bilirubin to serum bilirubin ratio > 3 at day 3 or later after resection or interventional surgical revision due to biliary peritonitis. A confirmatory explorative analysis was carried out.
Patients and methods: The study involved an evaluation of primary liver resection from the years 2009 and 2010. Primary endpoints were the incidence of posthepatectomy liver failure and biliary leakage in accordance with the ISGLS definition. Secondary endpoints were complications and 90-day mortality. Results are displayed as median values (minimum and maximum).
Results: A total of 214 liver resections were included from the years 2009 and 2010. Patients were an average of 61.5 years old (min. 18, max. 83 years). The incidence of liver failure was 7.4 % (16 out of 214) and fatal in 7 patients. In 31 % (65 out of 214) a biliary leakage occurred, 14 (23 %) patients developed a type B, 1 patient(5 %) a type C leakage and 50 leakages were clinically inapparent. The incidence of clinically relevant biliary leakages was 7 % (15 out of 214). The sensitivity of the definition was 100 % and the specificity 75 %. The incidence of Dindo-Calvien complications > 3b was 10.2 %, of sepsis 5.6 % and the 90-day mortality was 6.5 %. Multivariate analysis did not reveal independent predictive factors for biliary leakage or liver failure.
Conclusion: The definition for posthepatectomy liver failure was found to be valid in this cohort. The incidence of postoperative biliary leakage is over-estimated with the current definition and delivers a large number of false positive results without clinical relevance.