Comparison of autologous bone marrow transplantation and intensive chemotherapy as postremission therapy in adult acute myeloid leukemia. The Groupe Ouest Est Leucémies Aiguës Myéloblastiques (GOELAM).

Three intensive consolidation strategies are currently proposed to younger adults with acute myeloid leukemia (AML) in first complete remission (CR): allogeneic or autologous bone marrow transplantation (BMT) and intensive consolidation chemotherapy (ICC). Patients aged 15 to 50 years with de novo AML received an induction treatment with 7 days of cytarabine and either idarubicin or rubidazone. After achievement of a CR, patients up to the age of 40 and having an HLA-identical sibling were assigned to undergo an allogeneic BMT. All the other patients received a first course of ICC with high-dose cytarabine and the same anthracycline as for induction. They were then randomly assigned to either receive a second course of ICC with amsacrine and etoposide or a combination of busulfan and cyclosphosphamide followed by an unpurged autologous BMT. Of 517 eligible patients, 367 had a CR, but only 219 (59.5%) actually received the planned intensive postremission treatment (73 allogeneic BMT, 75 autologous BMT, and 71 ICC). With a median follow-up of 62 months, the 4-year disease-free survival (DFS) of the 367 patients in CR was 39.5%. The 4-year overall survival (OS) of the 517 eligible patients was 40.5%. In multivariate analysis, DFS and OS were influenced only by the initial white blood cell count and by the French-American-British classification. The type of postremission therapy had no significant impact on the outcome. There was no difference in the 4-year DFS and OS between 88 patients for whom an allogeneic BMT was scheduled (respectively, 44% and 53%) and 134 patients of the same age category and without an HLA-identical sibling (respectively, 38% and 53%). Similarly, there was no difference in the outcome between autologous BMT and ICC. The 4-year DFS was 44% for the 86 patients randomly assigned to autologous BMT and 40% for the 78 patients assigned to ICC (P = .41). The 4-year OS was similar in the two groups (50% v 54.5%, P = .72). The median duration of hospitalization and thrombocytopenia were longer after autologous BMT (39 v 32 days, P = .006, and 109.5 v 18.5 days, P = .0001, respectively). After a first course of ICC, a second course of chemotherapy is less myelotoxic than an unpurged autologous BMT but yields comparable DFS and OS rates.