Autologous versus allogeneic unrelated donor transplantation for acute lymphoblastic leukemia: comparative toxicity and outcomes.

For patients with high-risk or relapsed acute lymphoblastic leukemia (ALL) lacking a related histocompatible donor, autologous (Auto) and unrelated donor (URD) transplantation are available options. We compared outcomes and toxicities in 712 patients with ALL (517 URD, 195 Auto) in first complete remission (CR1) or second complete remission (CR2) who underwent transplantation. All patients were <50 years old, although URD patientswere younger (median age, 14 versus 18 years, P < .002). The proportion of patients in CR1 versus CR2 was similar (36% versus 38%, P = .57), but more URD recipients than Auto recipients had high-risk karyotypes (25% versus 13%, P = .003) and white blood cell (WBC) counts > or =50 x 10(9)/L (33% versus 14%, P < .001). Engraftment was similar in URD and Auto recipients. Ex vivo purging delayed but did not prevent engraftment after Auto transplantation. Transplantation-related mortality was higher after URD transplantation (42%+/-8%) than after Auto transplantation (20%+/-12%) in CR1 (P = .004) and also in CR2. Conversely, relapse was more frequent after Auto transplantation in CR1 (Auto, 49%+/-12% versus URD, 14%+/-5%) and CR2 (64%+/-8% versus 25%+/-5%) (P < .0001). These findings showed net similar outcomes for these 2 transplantation choices. Transplantation in CR1 yielded similar 3-year survival rates for URD (51%+/-7%) and Auto (44%+/-12%), as did transplantation in CR2 (40%+/-6% versus 32%+/-9%, respectively). Multivariate regression analysis identified significantly better disease-free survival after the first 6 months in matched URD versus Auto in younger patients, in those in CR2 with CR1 >1year, WBC <50 x 10(9)/L, performance status > or =90%, and in those who have undergone transplantation since 1995. These comparative data suggest that both matched URD and Auto transplantation can yield extended survival. Although URD transplantation offers substantially better protection against leukemic relapse, improvements in allotransplantation safety and refinements in patient selection are required to better aid treatment decision making for the best overall survival.