High-Dose Melphalan-Based Sequential Conditioning Chemotherapy Followed By Allogeneic Hematopoietic Stem Cell Transplantation (HSCT) in Adult Patients with Relapsed or Refractory Acute Myeloid Leukemia (AML)

Introduction: The majority of patients with newly diagnosed AML fail to achieve long-lasting remission, even after initial complete remission (CR). Response rates to common salvage regimens reach up to 50% in patients with late relapse and decrease to 10-15% in patients with refractory AML or early relapse. Without allogeneic HSCT the long-term overall survival (OS) for patients with refractory AML hardly reaches 5%. Survival rates of AML patients post allogeneic HSCT are significantly higher for patients transplanted in CR as compared to those transplanted with active leukemia. Common conditioning regimens followed by allogeneic HSCT of patients with active AML have shown disappointing results. Therefore, others and we developed the concept of sequential therapy as a preparative regimen for allogeneic HSCT to optimize the balance between treatment-related toxicity and antileukemic efficacy.

Methods: From April 2005 to May 2016, 292 patients with R/R-AML received an allogeneic HSCT at the BMT centers of the German university hospitals of Muenster and Essen. The median age was 56 years (range 17 -74). The remission status prior to transplant was primary refractory (144 patients), secondary refractory (97 patients) or relapsed AML (51 patients). All patients were treated with a single infusion of high dose melphalan, followed by dose-adapted conditioning consisting of fludarabine (120 mg/m²) and 8 Gy TBI or followed by fludarabine (150 mg/m²) and treosulfan (42 g/m²). Median time between melphalan and continuation of the therapy was 5 days (range 1-18 days), resulting in an overall duration of conditioning prior to HSCT of 11 days in most patients. Additional pretreatment with antithymocyte globulin (ATG) was applied to 157 of 222 patients with an unrelated donor and in 44 of 70 patients with a HLA-identical sibling donor. Most patients (except 2) received cyclosporine as acute GvHD prophylaxis in combination with methotrexate (89 patients) or mycophenolate mofetil (197 patients) or both (4 patients).

Results: Median follow-up after allogeneic HSCT of all patients alive at last follow-up was 4.8 years. OS rates at 3 years were 34% for primary refractory patients, 29% for secondary refractory patients, and 41% for patients with untreated relapsed AML. Factors associated with an inferior OS were older age, a time interval of 5-9 months between first diagnosis and allogeneic HSCT, and transplantation from an HLA-mismatched unrelated donor. Most importantly, leukemic burden was associated inversely with a 3-year OS of 51%, 31%, and 22% for patients with leukemic blast infiltration of <20%, 20-50%, and >50%, respectively (p <.001). Particularly patients transplanted from HLA-mismatched unrelated donors had a significantly higher non-relapse-related mortality (NRM) rate of 55% (95%CI 27 to 56%) at one year, compared to those transplanted from HLA-matched related (34%, 95%CI 24 to 48%) or unrelated donors (28%, 95%CI 22 to 37%). In addition, adverse risk factors for NRM were older age, pre-existing infection prior to start of conditioning therapy and a higher comorbidity index (HCT-CI >3). Multivariate analysis of risk factors revealed that high disease burden before conditioning (p <.001), longer time interval between first diagnosis and transplantation (p .039), and transplantation from an HLA-mismatched unrelated donor (p .049) were independently associated with inferior overall survival. High risk disease according to cytogenetic and molecular genetic aberrations as an adverse risk factor showed a trend in this analysis (p .052).

Conclusion: Allogeneic HSCT remains the most potent treatment option for patients with R/R-AML. High-dose melphalan based sequential conditioning chemotherapy followed by an allogeneic HSCT allows for achieving long-term remission in a substantial proportion of patients with active R/R-AML with otherwise detrimental prognosis. Since allogeneic HSCT is superior to other alternative treatment strategies evaluated in the past, direct allogeneic HSCT should be evaluated prospectively against remission induction therapies followed by an allogeneic HSCT in patients with R/R AML.