Paediatric allogeneic bone marrow transplantation for homozygous beta-thalassaemia, the Dutch experience

Abstract

We reviewed the results of the Dutch paediatric bone marrow transplant (BMT) program for children receiving HLA-identical BMT for beta-thalassaemia major over an 18-year period. In all, 19 patients underwent a total of 21 transplants in our treatment centre between July 1984 and February 2002. Eight females (age 0.3-12 years; median 5 years) and 11 males (age 0.8-18 years; median 6 years) were included. Information, prospectively collected, included molecular defects, donor genotype, beta/alpha-globin expression rates, serum ferritin levels, hepato-splenomegaly, chelation history, virology screening, liver pathology together with post-transplant outcome inclusive of leucocyte chimerism. In total, 11 patients received standard busulphan/cyclophosphamide (Bu/Cy) conditioning, with or without ATG. Stable engraftment was seen in 5/11 with late rejection occurring in six patients. Of these, two children underwent a second successful SCT. For this group, overall event-free survival (EFS) and disease-free survival (DFS) were 90 (10/11) and 64% (7/11), respectively. The probability of rejection was 55%. Subsequent addition of melphalan to the conditioning regimen resulted in long-term stable engraftment in all patients with an EFS/DFS for this group of 90% (9/10). Treatment-related mortality, irrespective of conditioning, was low at 5% (1/19 patients). Veno-occlusive disease (VOD) occurred in 19% (4/21 transplants) and acute GvHD in 19% (4/21 transplants). Post-BMT beta/alpha synthetic ratio measurement monitored donor erythroid engraftment and predicted rejection with a return to transfusion dependency. Maintained full donor chimerism is indicative of stable engraftment both for leucocyte and erythroid lineages, whereas mixed donor chimerism is not. Our results emphasise the importance of the conditioning regimen and post-transplant chimerism surveillance predictive of rejection or long-term stable engraftment.