Risk Factors for Outcome after Haploidentical Hematopoietic Stem Cell Transplant in Children with Very High Risk Acute Lymphoblastic Leukemia: Impact of Centre Experience. A Survey on Behalf of the ALWP and PDWP of the EBMT.

In the absence of an HLA identical donor, T-cell depleted haploidentical hematopoietic stem cell transplantation (Haplo-HSCT) is an alternative option to treat children with very high risk acute lymphoblastic leukemia (ALL). However, little data is available in children. We have analyzed 118 children (≤16 years old) with ALL transplanted with a myeloablative Haplo-HSCT from 1995 to 2004 in Europe. Only transplants with 2 or more HLA disparities out of 6 (A, B and DRB1) were included. The median age was 8.5 years and median follow-up 56 months (8–116). At transplant, 21 (18%) were in CR1, 72 (61%) in CR2 or CR3 and 25 (21%) in more advanced phase. The 3-year leukemia free survival was 32±10%, 28±5% and 0%, respectively. Therefore we restricted the analysis to patients in remission (n=93) transplanted in 30 EBMT centers. Thirty-four (37%) patients were treated in centres performing more than 10 Haplo-HSCT in the study period (3 centres). The median age of recipient and donor was 8.7 and 37 years, respectively. Fifty four (65%) received a full Haplo-HSCT and 95% of the donors were parental. Twenty seven (29%) had t(9;21) or t(4;11). The majority of patients did not receive drugs for GVHD prophylaxis, ATG/ALG was used in conditioning in 74%, and 73% received TBI. The Clinimacs® device was used in 74% of selections. The median number of CD34+ cells infused was 12.8 10⁶/Kg. Cumulative incidence with competing risk and KM estimates were used to calculate outcome probabilities. The median days of neutrophil recovery was 15 days (8–55) and 90% of patients had signs of engraftment. Acute GVHD II–IV was observed in 24% of the patients. In univariate analysis for LFS there was a trend towards better results for patients receiving higher CD34 cell dose (p=0.08) and a significant difference for patients transplanted in centres performing more Haplo-HSCT (49% versus 17%, p=0.002). Relapse incidence (RI) and non relapse mortality NRM) tended to be different between the experienced and less experienced centres. In less experienced centres NRM was 41% vs 27% (p=0.13) and RI 41% vs. 24% (p=0.10). There were patient, donor and transplant related differences between less and more experienced centres, specifically related to donor sex, Philadelphia positivity, year of transplantation, use of TBI, ATG and DLI and previous autograft. In conclusion, Haplo-HSCT is an alternative option to treat children with very high risk ALL in the absence of HLA identical donor. Centres with more experience have better LFS.