It is unclear how best to treat children with acute lymphoblastic leukemia who are in a second remission after a bone marrow relapse. For those with HLA-identical siblings, the question of whether to perform a bone marrow transplantation or to continue chemotherapy has not been answered.

We compared the results of treatment with marrow transplants from HLA-identical siblings in 376 children, as reported to the International Bone Marrow Transplant Registry, with the results of chemotherapy in 540 children treated by the Pediatric Oncology Group. A preliminary analysis identified variables associated with treatment failure in both groups. We selected cohorts by matching these variables. A possible bias associated with differences in the interval between remission and treatment was controlled for by choosing matched pairs in which the duration of the second remission in the chemotherapy recipient was at least as long as the time between the second remission and transplantation in the transplant recipient. A total of 255 matched pairs were studied.

The mean (±SE) probability of a relapse at five years was significantly lower among the transplant recipients than among the chemotherapy recipients (45 ±4 percent vs. 80 ±3 percent, P<0.001). At five years the probability of leukemia-free survival was higher after transplantation than after chemotherapy (40 ±3 percent vs. 17 ±3 percent, P<0.001). The relative benefit of transplantation as compared with chemotherapy was similar in children with prognostic factors indicating a high or low risk of relapse (the duration of the first remission, age, leukocyte count at the time of the diagnosis, and phenotype of the leukemic cells).

For children with acute lymphoblastic leukemia in a second remission, bone marrow transplants from HLA-identical siblings result in fewer relapses and longer leukemia-free survival than does chemotherapy.