Myeloablative Chemo-Conditioning for First Hematopoietic STEM CELL Transplantation in Children with ACUTE Lymphoblastic Leukemia in First or Second Remission

Christina Peters, Petr Sedlacek, Jean Hugues Dalle, Stelios Graphakos, Antonio Campos, Akif Yesilipek, Jacek Wachowiak, Arjan Lankester, Andrea Pession, Amir Ali Hamidieh, Marianne Ifversen, Jochen Büchner, Gergely Krivan, Franca Fagioli, Arnaud Dalissier; Myriam Labopin; Peter Bader on behalf of the EBMT Pediatric Diseases Working Party

Most children with acute lymphoblastic leukemia (ALL) with indication for allogeneic hematopoietic stem cell transplantation (HSCT) receive myeloablative conditioning with a total body irradiation (TBI)-containing regimen. To investigate the outcomes of patients (pts) who did not undergo TBI, we performed a retrospective registry based study on children below 18 years who received a myeloablative chemo-conditioning for a first allogeneic HSCT from different donors between 2000 and 2012. In this analysis, only chemotherapeutic regimens with more than 30 applications were included.

In total, 732 pts were included: 313 pts who received bone marrow (BM) or peripheral blood stem cells (PBSC) in 1^st CR, 247 pts with BM/PBSC transplantation in CR2, 85 pts and 52 pts who received umbilical cord blood (CB) in 1^st or 2^nd CR, respectively. The most commonly applied myeloablative chemo-combinations were: Busulfan (Bu)/Cyclophosphamide (Cy) (n=202), Bu/Cy/Etoposide (VP) (n=189), Bu/Cy/Melphalan (Mel) (n=93), Bu/AraC/Mel (n=80), Bu/Fludarabine (Flu)/Thiotepa (Thio) (n=62), Bu/Cy/Thio (n=53, Bu/Cy/Thio (n=53), and Bu/Flu (n=53).

313 pts received either BM or PBSC in CR1 with a median follow up of 26 months (1-156) and we compared Bu/Cy/VP vs the other chemo-conditioning regimens. The Bu/Cy/VP cohort had a longer follow up (med 37 vs. 20 months, p=0.002), pts were younger (med 3,6 vs. 6,5 years, p=0.003) and the median year of transplant was earlier (med 2009 vs. 2010, p=0.03). Donor type, CMV match, gender match, stem cell were comparable. In univariate analysis, conditioning with Bu/Cy/VP was better than all other combinations: relapse incidence (RI) 21% vs 32% (p=0.05), leukemia-free survival (LFS) 72 vs 54% (p=0.004), overall survival (OS) 79 vs 68% (p=0.03) and chronic GVHD (cGVHD) 9% vs 19% (p=0.014). Engraftment and incidence and severity of acute GVHD were similar and non- relapse mortality (NRM) was 7% vs 13% (p=0.10). Other significant influencing factors were: interval between diagnosis and transplantation below or beyond 208 days (NRM 6% vs 16%, p=0.015), donor sibling vs other (RI 35% vs 23%, p=0.01, NRM 5% vs 16%, p=0.001) and in vivo T cell depletion (TCD) vs no TCD (RI 35% vs. 19%, p=0.003; NRM 20% vs 4%, p=0.0001). In the cox model, conditioning type (Bu/CY/VP vs other), age, year of transplantation, interval from diagnosis to transplant, donor type, stem cell source and in vivo TCD were evaluated. For LFS only BU/CY/VP was associated with better outcome (p=0.004, HR .52), RI was lower after Bu/Cy/VP (HR .54, p=0.02), NRM was higher in pts older than 4,6 years (p=0.02, HR 2,48) and after TCD HSCT (p=0.01, HR 9,13) and OS was best after Bu/Cy/VP (p=0.03, HR 0.57).

We conclude that omission of TBI is feasible for children who undergo first allogeneic HSCT in first or second complete remission. The combination of busulfan, cyclophosphamide and etoposide resulted in better LFS and OS with less NRM and RI for children who received bone marrow or peripheral blood stem cells in CR1. These observations should be the basis for prospective trials in homogenous patient groups.