Influence of cytogenetic abnormalities on outcome after allogeneic bone marrow transplantation for acute myeloid leukemia in first complete remission.

Cytogenetic abnormalities detected at diagnosis are recognized as important in predicting response to chemotherapy in acute myeloid leukemia (AML). However, there is controversy concerning the prognostic significance of karyotype for outcome after allogeneic bone marrow transplantation (allo-BMT) performed in first complete remission (CR1). This single-institution report describes allo-BMT for AML in CR1 and the effect of diagnostic cytogenetic findings on the results of that treatment. Between August 1981 and December 1999, 93 patients underwent related donor (n = 82) or unrelated donor (n = 11) BMT. Conditioning and GVHD prophylaxis were achieved predominantly with busulfan and cyclophosphamide and with cyclosporine and methotrexate, respectively. Seventy-nine (85%) of 93 patients had successful marrow karyotyping at diagnosis, and the patients were categorized into 3 prognostic groups based on the British Medical Research Council AML 10 trial classification: 15 patients(19%) were classified as having favorable risk [inv(16), t(8;2 1), t(15;17)]; 55 (70%) as having intermediate risk [no abnormality, +8, +21, +22, del(7q), del(9q), 11q23 rearrangement, and other numerical or structural abnormalities]; and 9 (11%) as having adverse risk [-5, del(5q), -7, 3q rearrangements, > or = 5 abnormalities, t(6;9), t(9;22)]. The median follow-up was 93 months (range, 16-241 months). The overall survival (OS) rate, event-free survival (EFS) rate, relapse rate, and treatment-related mortality (TRM) were not statistically different between the groups. The 5-year actuarial EFS rates for favorable, intermediate, and adverse risk groups were 58% (95% confidence interval [CI], 29%-79%), 58% (95% CI, 43%-70%), and 67% (95% CI 28%-88%), respectively. Reclassification of patients into cytogenetic prognostic subgroups according to Southwest Oncology Group criteria did not change these results. In univariate analysis, the only variable found to have a prognostic influence on OS (P = .04) and TRM (P = .03) was the type of donor (unrelated donor was linked to a worse prognosis), which was confirmed in multivariate analysis. Our study suggests that presentation karyotype has less prognostic significance for outcome following allo-BMT than for outcome following conventional chemotherapy. In particular, AML patients with poor prognostic cytogenetic changes in CR1 who are unlikely to be cured with chemotherapy alone may benefit from allo-BMT.