Background: This study compares liver resection (LR) or transplantation (LTx) in an attempt to reevaluate the indications for treatment.
Study design: One hundred fifty-four LRs and 121 LTxs performed from 1985 to 1999 were considered. Survival and recurrence rate, together with age, gender, liver disease, Child-Pugh classification, alpha-fetoprotein (AFP), tumor capsule, vascular invasion, size, number of nodules, histologic grade, and pTNM were considered. Followup was completed in all cases (mean +/- SD = 3.2 +/- 2.9 years).
Results: The 5- and 10-year actuarial survival rates were 61.7% and 59.8% in LTx and 46.9% and 28.0% in LR (p = 0.08). Recurrence-free survival was 85.9% and 85.9%, respectively, in LTx and 42.8% and 30.7% in LR (p < 0.0001). In both groups, size, capsule, AFP, vascular invasion, grade, pTNM, Child-Pugh classification, and age were all significantly related to survival and cancer recurrence. pTNM, AFP, Child-Pugh classification, and age, in LR, and capsule, AFP, and viral cirrhosis, in LTx, were significant independent variables in Cox's regression model for survival. Only AFP, vascular invasion, and grade were significant in both groups for recurrence.
Conclusions: LTx offers better recurrence freedom than LR, but longterm survival is not significantly different in the two series. A strict selection should be made to optimize graft allocation. Size and multifocality should not be considered absolute contraindications for LTx. AFP, vascular invasion, and grade are more likely to reflect the risk of recurrence of the disease. LR should be considered in patients who do not fulfill transplant criteria and also in some categories of patients with certain tumor characteristics (small resectable tumors in well-compensated cirrhosis).