Predictors of peri-opertative morbidity and liver dysfunction after hepatic resection in patients with chronic liver disease

Background: Hepatic resection in patients with chronic liver disease (CLD) is associated with a risk of post-operative liver failure and higher morbidity than patients without liver disease. There is no universal risk stratification scheme for CLD patients undergoing resection. Objectives: The aim of the present study was to evaluate the association between routine pre-operative laboratory investigations, model for end-stage liver disease (MELD), indocyanine green retention at 15 min (ICG15) and post-operative outcomes in CLD patients undergoing liver resection. Methods: A retrospective review of patients undergoing resection for hepatocellular carcinoma (HCC) at the University Health Network was preformed. ICG15 results, pre- and post-operative laboratory results were obtained from clinical records. Adjusted odds ratios (AOR) were calculated for associations between pre-operative factors and post-operative outcomes using multivariate logistic regression adjusting for patient age and number of segments resected. Results: Between 2001 and 2005, 129 CLD patients underwent surgical resection for HCC. Procedures included 51 (40%) resections of <= 2 segments, 52 (40%) hemihepatectomies and 25 (19%) extended hepatic resections. Thirty-and 90-day post-operative mortality was 1.6% and 4.1%, respectively. Prolonged (>10 days) hospital length of stay (LOS) was independently associated with an ICG15 > 15% {AOR [95% confidence interval (CI)] = 8.5 (1.4-51)} and an international normalized ratio (INR) > 1.2 [AOR (95% CI) = 5.0 (1.4-18.6)]. An ICG15 > 15% and MELD score were independent predictors of prolonged LOS. An ICG15 > 15% was also independently associated with MELD > 20 on post-operative day 3 [AOR (95% CI) = 24.3 (1.8-319)]. Conclusions: Elevated ICG retention was independently associated with post-operative liver dysfunction and morbidity. The utility of ICG in combination with other biochemical measures to predict outcomes after hepatic resection in CLD patients requires further prospective study.