The aim of the study was to determine whether simple routine parameters evaluating the first session of transarterial chemoembolization (variation in alfa-fetoprotein concentration, tumor lipiodol uptake, and post-embolization syndrome) can predict survival of patients treated for hepatocellular carcinoma.

Seventy-two patients treated with transarterial chemoembolization and evaluated one month after the first sessions with CT scan were included. Transarterial chemoembolization session included hepatic arteriography, lipiodol and doxorubicin (50 mg) emulsion injection, followed by gelatin sponge embolization. The following variables were studied in univariate and multivariate analysis: 6 recorded at the first session (age, cirrhosis etiology, Child-Pugh class, tumor number, largest lesion size, and alpha-fetoprotein concentration), and 5 recorded after the first session (variation in alfa-fetoprotein concentration, tumor lipiodol uptake, post-embolization syndrome, mean interval between each session, and associated treatment).

Mean follow-up was 22.7 months (4-106). Mean survival was 30.4 months (95% CI: 23. 3-37.5). Actuarial survival at 1, 2, 3 and 5 years was respectively 65.5%, 44%, 29.5%, and 18%. The only independent prognostic factors in multivariate analysis were the Child Pugh class and the mean interval between sessions (P < 0.001 and < 0.01 respectively). None of our criteria evaluating the first TACE session significantly influenced survival.

The 3 parameters (variation in alfa-fetoprotein concentration, tumor lipiodol uptake and post-embolization syndrome) after the first transarterial chemoembolization did not predict survival. They could not be used to determine which patient could benefit from repeated transarterial chemoembolization sessions.