Predictive factors for outcome of allogeneic hematopoietic cell transplantation for adult acute lymphoblastic leukemia.

Between January 1990 and December 1997, 182 adults with acute lymphoblastic leukemia (ALL) received allogeneic hematopoietic cell transplants according to Fred Hutchinson Cancer Research Center protocols. Patients eligible for transplantation included those in first remission, especially those at high risk of relapse (n = 41), and any patient in second or later remissions (n = 46) or in relapse (n = 95). The median patient age was 29.4 years (range, 18.0-57.6 years), and the median duration of disease was 13.3 months (range, 2.4-221.9 months). Fifty-six patients had Philadelphia chromosome-positive ALL. Most patients (n = 169) received a conditioning regimen of cyclophosphamide 120 mg/kg plus 12.0 to 15.75 Gy of total body irradiation and a combination of cyclosporine and methotrexate as graft-versus-host disease (GVHD) prophylaxis. One hundred twenty-one patients received stem cells from HLA-identical donors (88 related donors and 33 unrelated donors), and 61 received stem cells from HLA-mismatched donors (26 related donors and 35 unrelated donors). Actuarial disease-free survival at 5 years was 21% for all patients, 43% for patients in first remission, 24% for patients in second or later remissions, and 9% for patients in relapse. Univariate and multivariate Cox regression analyses were performed to identify factors associated with survival, relapse, nonrelapse mortality, and disease-free survival. Factors significantly associated (P <.01) with improved survival and disease-free survival included younger age and being in first remission. Lower disease-free survival was associated with receiving cyclosporine alone as GVHD prophylaxis (P <.01). Risk of relapse correlated only with disease status at transplantation: patients who underwent transplantation in relapse had a 9-fold increased risk compared with patients who underwent transplantation in first remission. Acute or chronic GVHD had no significant effect on relapse. Increased nonrelapse mortality was associated with HLA-mismatched donors, a positive cytomegalovirus serology before transplantation, and GVHD prophylaxis with only cyclosporine. Patients with Philadelphia chromosome-positive ALL had survival and relapse rates similar to patients with normal cytogenetics.