Predicting survival in patients with liver cancer considered for transarterial chemoembolization

Introduction. Transarterial chemoembolization (TACE) has been used extensively to treat tumours confined to the liver in patients unsuitable for surgical resection. This study attempts to identify patients with liver cancer most likely to benefit from this type of treatment. Patients and methods. All patients undergoing TACE for liver cancer between 1989 and 2001 were included in the study. Results. In a group of 137 consecutive patients undergoing TACE, univariate analysis identified a number of pre-treatment factors that were associated with poor prognosis. Multivariate analysis of these factors subsequently identified three pre-treatment factors; age greater than 60, serum alkaline phosphatase concentration >120U/l and albumin less than 35 g/l; that were independently and significantly associated with reduced survival duration. A scoring system was devised with one point allocated for each adverse factor which produced median survivals related to points scored as follows, 0 points—20 months, 1 point—12 months, 2 points—7 months and 3 points—4 months. To validate this scoring system the next 40 consecutive patients undergoing TACE were studied prospectively. These patients had median survival durations related to points scored as follows 0 points not calculable, 1 point—10 months, 2 points—7 months, 3 points—4 months. Conclusion. This simple scoring system can be used to predict prognosis in patients with liver cancer and may assist in clinical decision making in the selection of patients likely to benefit from TACE.