Haplo-identical transplant: long-term follow-up of a three-steps phase I/II study of the impact of post-transplant growth factors and GSF-primed DLI in advanced acute leukaemia patients

P. Lewalle (1), R. Rouas (1), A. Delforge (1), P. Crombez (1), N. Meuleman (1), J. Maertens (2), D. Bron (1), P. Martiat (1)

Author address: 

(1)Institut Bordet (ULB) (Brussels, BE); (2)KULeuven (Leuven, BE)


Introduction: we initiated this study in 2000, at a time where most referred patients (without a matched donor) were in very advanced stage (most of them refractory or progressive), in order to evaluate the impact of G-CSF and GM-CSF post-transplant, and the role of G-CSF primed DLI. It is worth noting that, with time, the proportion of patients in CR (mainly CR2 and CR3) increased. Patients and methods: a total of 45 patients entered this study of whom 17 were in refractory or progressive relapse, 9 in PR post relapse, 16 in CR 2 or 3 and 2 in CR1. The three steps consisted of (1) G-CSF + DLI (10 4 per kg) monthly from day 30 if no GVHD, (2) GM-CSF from day 5 + one single DLI at day 30, GM-CSF (day 5 to 9) without DLI for patients who could not benefi t from NK alloreactivity (ALL and some AML). There were 9 patients in step I, 12 in step II and 22 in step III. The median reinfused cell dose was 4.4x10 6 /kg and 4.8x10 4 /kg for CD34+ and CD3+ cells respectively. Results: for step I (2 CR), 1/9 grade III GVHD, TRM = 0 at day 100 and 2 at 3 years, a low incidence (25%) of CMV and aspergillosis, but 6/9 eventually relapsed. For step II, the TRM was 8/12 due to a high incidence of GVHD, which offset the low relapse rate (2/12). For step III, grade III GVHD incidence was 16/22, TRM 13/22 (8/9 if no GM-CSF) and LFS was 5/22 (all CR patients). The relapse rate was 2/22. In both step II and III patients, the incidence of CMV reactivation and aspergillosis was above 50%, probably due to a higher GVHD rate and its treatment. Patients with CMV reactivation could be treated in a pre-emptive way by the infusion of ex vivo generated specifi c T cells, but this was effective only in patients with a mild GVHD treatment. Conclusions: none of these three approaches could result in long-term LFS in patients refractory or progressive. The use of GM-CSF could maintain some of them leukemia free at the cost of ultimately lethal GVHD. The use of G-CSF and G-CSF primed DLI resulted in a low incidence of GVHD and infection, but was effective in CR patients only. Nevertheless, this could be the approach of choice for patients in any CR, but we do not have enough evidence for this,given the low number of CR patients in this series. The overall LFS for the patients in any CR was 39% at 9 years, which compares favorably with MUD transplant and is thus a reasonable alternative for patients in CR at need of a transplant.

abstract No: 


Full conference title: 

Annual Meeting of the EBMT, 37th
    • EBMT 37th (2011)