Relevance of postoperative peak transaminase after elective hepatectomy.

ABSTRACT

OBJECTIVES: Determine whether inflow occlusion is correlated with peak-postoperative serum-transaminases (PSTs) and whether PST is predictive of outcome after liver resections. BACKGROUND: PST is used as the surrogate of ischemia reperfusion and as the main endpoint in prospective trials of inflow occlusion. This assumption has, however, not been validated. Furthermore, the impact of PST on the postoperative course is unknown. METHODS: This prospectively designed registered study included consecutive adult patients undergoing elective hepatectomy in 9 HPB centers. Primary outcome was PST of aspartate-amino-transferase (AST) and alanine-amino-transferase (ALT). Secondary outcome was 90-day morbidity (Dindo-Clavien grades) and length of stay. Explanatory variables were preoperative (including age, sex, body mass index, comorbidities, cirrhosis, and chemotherapy), and intraoperative variables (including procedure performed, inflow occlusion and its duration, length of surgery, vasoactive drugs used, blood loss, and transfusion) were collected prospectively on a dedicated Web site. Multivariable regression models were used to identify independent predictors of PST and of morbidity. RESULTS: Between January 2013 and September 2013, 651 hepatectomies were included. Inflow occlusion was performed in 58% (intermittent in 32%, continuous in 24%) and was not performed in 42%. PST-AST (336 IU/L; interquartile range 204-573) and PST-ALT (336 IU/L; interquartile range 205-557) occurred on postoperative day 1. PST was not correlated with the duration of inflow occlusion (ρ-AST=0.20, P<0.01; ρ-ALT=0.18, P<0.01). PST was not independently associated with morbidity. Receiver operating characteristic curve identified a cutoff of 450 IU/L but this prediction's accuracy was low area under the receiver operating characteristic curve for PST-AST 0.61, confidence interval 0.56-0.66, P<0.01, and area under the receiver operating characteristic curve for PST-ALT 0.57, confidence interval 0.52-0.62, P=0.01. CONCLUSIONS: PST is not correlated with ischemia time and should not be used as a surrogate of postoperative outcome.