Indication of the extent of hepatectomy for hepatocellular carcinoma on cirrhosis by a simple algorithm based on preoperative variables.

OBJECTIVE: To produce a model indicating the extent of hepatectomy for hepatocellular carcinoma on cirrhosis based on easily available preoperative data. DESIGN: Retrospective study based on multicenter prospectively updated databases. SETTING: Two tertiary referral centers specializing in hepatobiliary surgery. PATIENTS: A total of 466 patients undergoing hepatectomy for hepatocellular carcinoma on cirrhosis between 1995 and 2006. MAIN OUTCOME MEASURES: To create a decision tree for safe liver resection based on factors affecting irreversible postoperative liver failure (IPLF). RESULTS: A total of 23 patients (4.9%) developed IPLF. The model for end-stage liver disease (MELD) score (categorized as &lt;9, 9-10, and >10; P < .05 for all comparisons) and extent of hepatectomy were independent predictors of IPLF. In patients with a MELD score of less than 9, the IPLF rate was 0.4%. In patients with a MELD score of 9 or 10, the IPLF rate was 1.2% for resections of less than 1 segment, 5.1% for segmentectomies or bisegmentectomies, and 11.1% for major hepatectomies. In this category of MELD, serum sodium levels identified a low-risk group (sodium >/=140 mEq/L; to convert to millimoles per liter, multiply by 1.0) not experiencing IPLF and a high-risk group (sodium &lt;140 mEq/L) in which resections of less than 1 segment led to an IPLF rate of 2.5% and resections of 1 segment or more led to an IPLF rate of more than 5% (P < .05). In patients with a MELD score of more than 10, the IPLF rate was more than 15% in all types of hepatectomies. CONCLUSION: A simple algorithm based on the MELD score and serum sodium level can indicate the maximum tolerable extent of hepatectomy for hepatocellular carcinoma on cirrhosis.