Standard versus alternative myeloablative conditioning regimens in allogeneic hematopoietic stem cell transplantation for high-risk acute leukemia.

BACKGROUND AND OBJECTIVES: To analyze the results of standard versus alternative myeloablative conditioning regimens in allogeneic hematopoietic stem cell transplantation for high-risk acute leukemia. DESIGN AND METHODS: From October 1986 to February 2000, 104 consecutive patients (male: n = 63; median age: 21, range 1.3-44.2 years) with high-risk acute leukemia underwent a non-T-cell depleted graft from an HLA-identical sibling following a standard or alternative myeloablative conditioning regimen. Sixty patients were affected by acute lymphoblastic leukemia (ALL) and 44 by acute myeloid leukemia (AML); the phase at transplant was >= 2nd complete remission (CR) in 76, untreated 1st relapse with < 20% blasts in 11, refractory leukemia or overt resistant relapse in 17. Pre-transplant regimens consisting of either 12 Gy fractionated total body irradiation (TBI) or 16 mg/kg busulphan (BU) combined with cyclophosphamide (CY) were defined standard (n = 38), whereas all other myeloablative regimens (TBI plus 60 mg/kg etoposide and three-drug combinations) were considered alternative (n = 66). RESULTS: No significant differences in terms of baseline characteristics, incidence and severity of either acute or chronic graft-versus-host disease (GVHD) were observed between the two groups, but a significantly higher proportion of patients prepared with an alternative regimen were not evaluable for chronic GVHD (36% vs 16%) (p = 0.026). Sixty-six patients died, 38 of relapse, 26 of transplant-related mortality (TRM) and 2 of other causes. Thirty-eight patients are still alive with a follow-up ranging from 0.7 to 13.8 years (median, 7.1 years); only 1 of 39 patients who relapsed after transplant is alive in CR at 5.7 years from relapse. At the median follow-up, the actuarial probabilities of overall survival, relapse and TRM for patients conditioned with standard and alternative regimens are respectively 52% vs 25% (95% CI, 36-68% vs 13-37%; p = 0.0163), 34% vs 58% (95% CI, 18-51% vs 43-73%; p = 0.0377) and 25% vs 32% (95% CI, 9-40% vs 19-44%; p = ns). After adjustment for diagnosis, age, period, leukemia phase, duration of 1st CR, GVHD prophylaxis and donor-recipient sex combination, the multivariate analysis showed that alternative regimens are associated with a significantly worse survival (hazard ratio 2.31; p = 0.0071) and relapse rate (hazard ratio 2.75; p = 0.0187). INTERPRETATION AND CONCLUSIONS: From this retrospective analysis we can conclude that the alternative myeloablative conditioning regimens we used did not improve the outcome of patients transplanted for high-risk acute leukemia.