Different impact of intermediate and unfavourable cytogenetics at the time of diagnosis on outcome of de novo AML after allo-SCT: a long-term retrospective analysis from a single institution

Reduced-intensity conditioning with fludarabine and thiotepa for second allogeneic transplantation of relapsed patients with AML

A second allograft was offered to 58 relapsed AML patients after conditioning with fludarabine 90-150 mg/m(2) and thiotepa 15 mg/kg, in most cases with active disease. Median age was 53 years (range 23-69), median time to relapse after the first allo-SCT was 326 (47-2189) days and median follow-up was 6.7 years. GVHD prophylaxis consisted mainly of CsA and alemtuzumab. Response rates at 1 month were CR in 50 and persistent disease in 3/53 evaluable patients. At 3 years, the relapse incidence (95% confidence interval) was 56 (45-71)%, the TRM 31 (21-46)%, the OS rate was 18 (9-29)% and the EFS rate was 13 (5-23)%. OS improved with younger patient age, longer relapse-free interval after the first allo-SCT and the development of chronic GVHD. Patients ≥ 65 years old who relapsed >12 months after the first allograft (n=20) had a 3-year OS rate of 41 (19-62)%. Conventional cytogenetics and FLT3 mutation status did not affect outcome. Our regimen is feasible and provides at least for a subgroup of patients with AML recurrence after allo-SCT a reasonable therapeutic option in an otherwise fatal situation. Further modifications and a better understanding of the underlying biology could help lower the risk of relapse.