Significant Reduction of Treatment Related Mortality after Stem Cell Transplantation in 647 Children and Adolescents with ALL within BFM Trials during the Last 12 Years: A Long Term Analysis from 1996 to 2007

Although allogeneic stem cell transplantation (alloSCT) effectively prevents relapse of acute lymphoblastic leukemia (ALL), transplant related mortality (TRM) associated with the procedure may counterbalance that beneficial effect. To asses the risk for TRM after alloSCT we analyzed the different transplantation outcome data within the Berlin-Frankfurt-Münster (BFM)-trials. Between January 1996 and December 2003 403 patients were transplanted within ALL-BFM and ALL-Rez BFM trials (185 in CR1, 218 in CR2 or after subsequent relapse). These chemotherapy protocols included recommendations for alloSCT, however donor selection, conditioning regimen and GvHD prophylaxis frequently were modified by transplant centers. In 2003 the BFM Study Group initiated a prospective international multicenter trial (ALL-SCT-BFM 2003) aiming to standardize and harmonize the SCT procedure in order to minimize risk of TRM. Between January 2004 and December 2007 244 patients (109 in CR1, 135 in CR2 or after subsequent relapse) were recruited by 27 participating SCT centers in Austria, Germany and Switzerland. Indications for SCT were poor response to induction treatment, either detected by morphology on day 33 (NRd33) or by minimal residual disease load measured after 12 weeks of treatment, cytogenetic aberrations [t(9;22), t(4;11)], early bone marrow relapses of ALL, or any subsequent ALL relapse. More than 80% of all patients who were enrolled in ALL-SCT-BFM 2003 received the recommended conditioning regimens and GVHD prophylaxis (TBI+VP16 for MSD, TBI+VP16+ATG for MD, and TBI+Flu+VP16+ATG for MMD; in children younger than 2 years, TBI was substituted by BU+Cy; CSA in MSD-SCT, CSA plus short MTX in MD-SCT). Patients of both time periods (1996 to 2003 vs 2004 to 2007) were comparable regarding gender, immunophenotype, cytogenetic aberrations, NRd33 and type/time of relapse (if SCT in CR2). However, in the second time period there was a trend to older age at diagnosis (54.5% vs. 43.9% patients older than 10 years). Moreover, unrelated donors were used more frequently over time (61.9% vs 50.1%). Nevertheless TRM was significantly lower within ALL-SCT-BFM 2003 (11.5% vs 19.9%; p(Chi-Square)=0.006). In a logistic regression analysis including age above 10 and donor other than matched related as covariables the p-value was 0.001. TRM reduction over time was observed in MSD transplants (n=229) from 13.6% to 6.7% (p(Chi-Square)=0.12), in unrelated donor SCT (n=353) and mismatched related donor SCT (n=65) TRM decreased from 23.7% to 13.6% (p(Chi-Square)=0.01). Despite a trend to higher age at diagnosis and a higher percentage of unrelated donor transplantations, TRM reduced significantly since start of ALL-BFM SCT 2003. The experience of participating centers as well as high compliance to the recommended regimen, improved donor selection by HLA-high resolution typing, better diagnostic measures and (preemtive) treatment of infections, and close and prospective monitoring of all patients may contribute to the reduced TRM after alloSCT in children with ALL.