In 1999, we decided to start a phase I/II study of haploidentical transplantation for high-risk patients. The aim of the work was to implement a strategy to accelerate and strengthen the immune reconstitution by using nonspecific manipulation post-transplant and by developing specific strategies directed against viral antigens. The goal was to increase the graft-versus-leukemia effect without inducing or aggravating the deleterious graft-versus-host disease. The conditioning regimen, adapted to our group of patients, remained the same throughout. Importantly, the first recruited patients were in refractory disease, over time we were referred less advanced patients (complete remission 2 or more). There were 45 patients, all at high-risk, among which 27 were in refractory relapse. We questioned the importance of post-transplant growth factors policy and the influence of donor lymphocyte infusion. Because of the conditioning, transplant-related mortality was low at 3 months, but thereafter changed unfavourably when using granulocyte macrophage-colony stimulating factors in an increased incidence of acute graft-versus-host disease. As a whole the long-term survival of the patients was poor (18%) but improved a lot when transplanted patients were in complete remission (leukaemia-free survival of 39% at five years). Regarding the use of growth factors and donor lymphocyte infusion, granulocyte macrophage-colony stimulating factors with donor lymphocyte infusion induced a very high transplant-related mortality due to a high rate of severe graft-versus-host disease, while the combination of granulocyte colony-stimulating factors and a moderate dose of donor lymphocyte infusion was much safer but didn’t overcome the high relapse rate in refractory patients. The combination of granulocyte colony-stimulating factors and donor lymphocyte infusion might nonetheless be sufficient to decrease the infection rate in patients transplanted in complete remission. The use of granulocyte macrophage-colony stimulating factors leads to an unacceptable lethal graft-versus-host disease rate. The 39% at five years leukaemia-free survival in patients in complete remission compares favourably with what can be achieved with matched unrelated donors in complete remission 2 or more.

(BELG J HEMATOL 2013;4(4): 151–160)