Background: Long term results after liver resection for hepatocellular carcinoma (HCC) are disappointing because the disease tends to recur. In this study, the authors assessed prognostic factors affecting long term outcome, in the hope that these factors might be used in selecting HCC patients for surgery.
Methods: During the period 1977-1995, 100 consecutive patients underwent curative liver resection; 78 of 100 had HCC arising on preexisting cirrhosis (53 Child's Class A and 25 Child's Class B). Thirty-five prognostic factors were evaluated for their association with overall survival (OS) and disease free survival (DFS) in univariate and multivariate analysis (Cox proportional hazards model).
Results: There were four postoperative deaths. Seven patients died in hospital of hepatorenal failure: six had Child's Class B cirrhosis and had undergone preoperative chemoembolization. Of the remaining 89 patients, 50 developed recurrence. All surviving Child's Class B patients had recurrence. Five-year OS, postoperative deaths included, was 38% (median, 36 months). Five-year DFS, postoperative deaths excluded, was 26% (median, 21 months). Independent prognostic factors for DFS were Child's class, glutamic-oxaloacetic transaminase, gamma-glutamyltransferase, alpha-fetoprotein, number of tumor nodules, width of resection margins, preoperative chemoembolization, and experience of the team that performed the surgery. Factors with an independent effect on OS were Child's class and width of resection margins.
Conclusions: Liver resection can provide long term DFS in HCC patients with normal liver function. In patients with liver function impairment or an inadequate resection margin, recurrences are almost certain to occur. Preoperative chemoembolization significantly prolongs DFS but may increase the risk of postoperative liver failure in patients with liver function impairment.