Autologous Hematopoietic Stem Cell Transplantation (autoHSCT) in CLL: First Results of An EBMT Randomized Trial Comparing Autotransplant Versus Wait and Watch.

Between November 2001 and July 2007, 223 patients were enrolled (SFGM-TC/FCLLG n=98, MRC n=62, GCLLSG n=32, SAKK n=10, other EBMT centers n=17). There were 74% males and 26% females. Binet stages were progressive A 13%, B 67%, C 20%; 59% were in CR, and 41% in very good or nodular PR. Of note, SFGM-TC/FCLLG included only patients in CR. 82% of the patients were enrolled in 1^st, and 18% in 2^nd remission. Patients were randomized between group 1 (autoHSCT n=112) and group 2 (observation n=111) after an induction treatment which was left at the discretion of the investigators. Median PFS was 43 months in the observation group but not reached in the autoHSCT group; 5-year PFS was 48% and 65%, respectively (p=0.005). Accordingly, autoHSCT halved the relapse risk (5-year relapse incidence 25% vs. 51%; HR 0.4 [0.23-0.71], p=0.002). Cox modeling for randomization arm, Binet stage, disease status, line of treatment, contributing group (country), and the interaction between randomization arm and contributing group confirmed that autoHSCT significantly improved PFS (HR 0.41 [0.23-0.75] p=0.004). The beneficial effect of autoHSCT was stable over all contributing groups although patients accrued by SFGM-TC/FCLLG overall had a significantly better PFS than patients from other countries (HR 0.2 [0.08-0.55], p=0.001). At 5 years, the probability of OS was 92% and 91% for autoHSCT and observation, respectively. Significant differences in terms of non-relapse death were not observed. At the last follow up, among 205 evaluable patients, 186 are alive (147CR, 39 relapse), 19 died (14 from relapse and 5 from non-relapse causes) . In conclusion, in patients with CLL in first or second remission, consolidating autoHSCT reduces the risk of progression (PFS) by more than 50%, but has no effect on overall survival.

No relevant conflicts of interest to declare.