This schedule has shown an interesting activity with nearly 40% of the patients achieving CR. Moreover 4 patients experienced an inversion rate (CR with ABMT when they never achieved this status before). In terms of toxicity, this schedule seems feasible with 2/28 toxic deaths, which is in the lower part of the range of major solid tumors ABMT programs. But even if the rationale is proper, a better patients' selection is required. The CCR (continuous Complete Remission) rate is overimposable to other main studies previously published, but all our CCR were obtained in responding patients (Sensitive Relapses or unresectable PR). We may suggest that earlier transplantation is advisable when less tumor bulky is present and less clonal eterogeneity. The exact maximum tolerated dose of Carboplatin/VP 16/Ifosfamide programs has not yet clearly pointed out. The lack of major life-threatening episodes and neuro/nephrotoxicity may allow us to explore higher Carboplatin doses. Anyway the ultimate answer to the utility of ABMT trials must come from a randomized study in responding patients.