Background: Although reinduction rates are good for children with relapsed acute lymphoblastic leukaemia there is no consensus on whether bone marrow transplantation (BMT) is the most effective treatment to prolong second remission.
Patients and methods: Analyses comparing the outcome of related donor allogeneic BMT (related allograft) with chemotherapy are unreliable because of selection biases. To avoid these biases, the MRC UKALL R1 trial was analysed by HLA-matched donor availability.
Results: No significant difference in outcome was found between the donor and no donor groups. The donor group had a non-significant eight-year event-free survival (EFS) advantage of 8%, (95% confidence interval -9%-24%) over the no donor group. Patients with a first remission less than two years appeared to benefit most from having a donor, although the effect was only marginally significantly different from patients with longer first remission. Analysis by treatment received gave similar results, with BMT patients having a 5% (P = 0.8) eight-year EFS advantage over patients who received chemotherapy.
Conclusions: Related allograft was not found to be significantly better than chemotherapy, but there was the possibility of a moderate EFS benefit with related allograft. especially in patients with a short first remission.