Hematopoietic stem cell transplant (HSCT) versus intensive chemotherapy in infant acute lymphoblastic leukemia (ALL)

9514 Background: Infants (<366 days) with ALL and an MLL rearrangement (MLL-R), have a dismal prognosis. Early relapse is common and event free survival (EFS) is poor. The Children's Oncology Group (COG) parallel pilot trials, 1953 and 9407, were designed to prospectively compare early HSCT to intensified chemotherapy in infants with MLL-R ALL. This is the largest prospective trial investigating the role of HSCT in such patients (pts) in first remission (CR-1). Methods: From 1997–2000, 186 pts were registered (132 MLL-R). Both trials were identical through Induction, Induction Intensification and Re-Induction (Re-I), then diverged. Pts with MLL-R ALL with matched/one antigen mismatched - related/unrelated donors were to undergo HSCT (mandated on 1953/optional on 9407) at completion of Re-I. The HSCT regimen was ara-c, cyclophosphamide, steroid and total body irradiation; graft vs. host prophylaxis was cyclosporine. Results: Compliance with protocol mandated conditioning for HSCT was poor and 28/53 HSCT pts received non-protocol preparative regimens. By life table comparisons, the 5 year HSCT EFS was 50.9% (SE=9.9%, RHR 1.13) vs. 48.7% (SE=10.1%) p=0.68 in 47 control pts (chemotherapy-treated, MLL-R), who survived 143 days (median time to HSCT = 143 days). By statistical regression analysis the RHR for HSCT was 1.454 (90% CI 0.911, 2.318), p-value = 0.19, showing a trend in favor of chemotherapy alone. Neither donor source nor preparative regimen affected HSCT outcome. Chemotherapy outcomes were similar on 1953 and 9407. Events for HSCT pts were most common in the first 6 months: 20 events (11 HSCT related deaths/9 relapses). In the control group, 7/9 events in this time period were due to relapse. Conclusion: These results show no advantage for HSCT in CR1 in infants with MLL-R ALL. These data do not support the routine use of HSCT in CR1 in MLL-R infants. No significant financial relationships to disclose.