Transcatheter Arterial Chemoembolization in Hepatocellular Carcinoma
To the Editor:
We read with great interest the report by Bruix et al., 1 which recently appeared in HEPATOLOGY, and the related editorial by Okada. 2 The question addressed by the two articles is indeed an important one: Are transarterial locoregional procedures useful in the treatment of hepatocellular carcinoma (HCC) and, consequently, should they be used in all patients not eligible for surgery, percutaneous ethanol injection, thermal ablation, or orthotopic liver trasnplantation? Indeed, most open trials, in which the survival of treated patients has been analyzed and compared with that of historic controls, confirm the usefulness of the procedure. However, the only four published prospective randomized studies failed in providing any evidence of efficacy. 1,[3] The three studies by Pelletier et al., 3 Madden et al., and the Group d'Etude et de Traitment du Carcinome He Β΄patocellulaire 5 all presented, in several investigators' opinions 6 as well as in our own, several biases, such as the selection of patients with extremely poor prognosis and early mortality 3,4 or a quite limited number of patients included per center involved in the study. The article by Bruix et al. reports the results of a randomized, single-center, well-designed study, with a relatively large sample of patients allocated to transcathether arterial embolization (TAE) or supportive care. However, even in this well-designed study, some aspects can be identified that, in our mind, may preclude the possibility of drawing definite conclusions on the therapeutic procedure. For instance, as Dr. Okada correctly underlines, over 60% of the patients underwent a single TAE course. It has to be remarked that repeated cycles may indeed be useful, because a clear-cut prolongation of survival has been reported in patients undergoing repeated transcathether arterial chemoembolization (TACE). We have also previously published that overall survival is significantly higher in the patients who undergo more than two TACE courses (P Ο .002). In a recent reevaluation of our series of patients treated by TACE, which now includes 76 patients, again the number of procedures turned out to be significantly correlated with long-term survival (P Ο .0002 in the Cox regression multivariate analysis). In our own institution, the approach to patients' diagnosis, staging, and treatment is almost the same as in Bruix' s. The survival we obtained in patients undergoing TACE treatement is only slightly higher than that reported by Bruix et al., the figures being 83% at 1 year, 60% (vs. 49%) at 2 years, 35% at 3 years, and 24% (vs. 13%) at 4 years, whereas we still observe 13% of patients surviving at 5 years (vs. 0%). We performed TACE and not TAE, because it has been suggested that addition of a chemotherapeutic agent may increase the antineoplastic efficacy to some extent. 9 Our