RALLE pilot: response-guided therapy for marrow relapse in acute lymphoblastic leukemia in children.

Despite improved treatment results of childhood acute lymphoblastic leukemia (ALL), 20% to 30% have a relapse, and then the outcome is very poor. We studied 40 children with ALL marrow relapse piloting an ALL relapse protocol with well-known drugs and drug combinations by using a concept of response-guided design. We also measured response in logarithmic fashion. Our primary end points were achievement of M1 marrow status, minimal residual disease status below 10(-3), and second remission. The remission induction rate was 90% with 10% induction mortality. After the A blocks (dexamethasone, vincristine, idar-ubicin and pegylated L-asparaginase), 85% had M1 status, 39% had minimal residual disease <= 1 x 10(-3), and 66% had 2 to 3 log response. After B1 block (cyclo, VP-16) the figures were 92%, 58%, and 83%, respectively. Twenty-five of 40 patients received allogeneic stem cell transplantation. Three-year event-free survival of the whole cohort was 37%, and the relapse rate was 38%. Three-year event-free survival by risk group was 53% for late, 34% for early, and 21% for very early relapses. An ALL marrow relapse nonresponsive to steroids, vincristine, asparaginase, anthracyclines, and alkylating agents is uncommon, and these classic drugs can still be advocated for induction of ALL relapse. The problems lie in creating a consolidation capable of preventing particularly posttransplant relapses.