LAUSR

BACKGROUND Major liver resection for perihilar cholangiocarcinoma (PHC) is associated with a 22% to 33% postoperative liver failure incidence. The aim of this study was analyze the predictive value of future liver remnant (FLR) volume for postoperative liver failure after resection for PHC and to develop a risk score to improve patient selection for portal vein embolization. STUDY DESIGN A consecutive series of 217 patients underwent major liver resection for PHC between 1997 and 2014 at 2 Western centers; FLR volumes were calculated with CT volumetry; other variables included jaundice at presentation, immediate preoperative bilirubin, and preoperative cholangitis. The FLR volume was categorized as <30%, 30% to 45%, or >45%. A risk score for postoperative liver failure (grade B/C according to the International Study Group of Liver Surgery criteria) was developed using multivariable logistic regression with 5 predefined variables. RESULTS Postoperative liver failure incidence was 24% and liver failure-related mortality was 12%. Risk factors for liver failure were FLR volume <30% (odds ratio 4.2; 95% CI 1.77 to 10.3) and FLR volume 30% to 45% (odds ratio 1.4; 95% CI 10.6 to 3.4). In addition, jaundice at presentation (odds ratio 3.1; 95% CI 1.1 to 9.0), immediate preoperative bilirubin >50 μmol/L (>2.9 mg/dL) (odds ratio 4.3; 95% CI 1.7 to 10.7), and preoperative cholangitis (odds ratio 3.4; 95% CI 1.6 to 7.4) were risk factors for liver failure. These variables were included in a risk score that showed good discrimination (area under the curve 0.79; 95% CI 0.72 to 0.86) and ranking patients in 3 risk sub-groups with predicted liver failure incidence of 4%, 14%, and 44%. CONCLUSIONS The selection of patients for portal vein embolization using only liver volume is insufficient, considering the other predictors of liver failure in PHC patients. The proposed risk score can be used for selection of patients for portal vein embolization, for adequate patient counseling, and identification of other modifiable risk factors besides liver volume.