Treosulfan Compared to Total Body Irradiation-Based Preparative Regimens Before Allogeneic Stem Cell Transplantation for Acute Myeloid Leukemia : A Retrospective Long-Term Study.

Background: The tolerability and efficacy of high-dose dihydroxy-busulfane (Treosulfan [Treo]), a water-soluble bifunctional alkylating agent with profound stem cell toxicity, as the major myelotoxic component of the preparative regimen preceding allogeneic stem cell transplantation (alloSCT) in patients (pts) with acute myeloid leukemia (AML) or myelodysplastic syndromes (MDS) has been recently demonstrated in phase II clinical trials. Since no prospective trials directly comparing Treo-based conditioning to a myeloablative total body irradiation (TBI) regimen will be available in the near future, the major clinical endpoints after Treo-based conditioning in comparison to a myeloablative TBI regimen were evaluated in adult pts with AML.

Methods: A total of 203 pts who underwent alloSCT between 1999 and 2005 were included in this bi-centre retrospective study, of whom 46 pts (23%) had received a Treo-based regimen (3 x 14 g/ m²), while 157 pts had been conditioned with TBI (4 x 2.5 Gy to 6 x 2 Gy), each combined with either fludarabine (5 x 30 mg/m²) or cyclophosphamide (2 x 60 mg/kg). Both cohorts were well comparable in terms of patient age, donor type, disease stage, cytogenetic risk category, and the hematopoietic cell transplantation comorbidity index (HCT-CI). The median follow-up time of all pts is 58 months (mo) (range 4 – 110 mo).

Results: For 125 pts transplanted in complete remission (CR)(CR1: 61 pts, CR2: 64 pts), non-relapse mortality (NRM) at 100 days, 1, and 3 years (yrs) after transplant was 6% (95% confidence limits [CL]: 0%–15%), 13% (95%-CL: 9%–24%), and 30% (95%-CL: 13%–30%) after Treo-compared to 10% (95%-CL: 4%–16%), 16% (95%-CL: 9%–24%), and 22% (95%-CL: 13%–30%) after the TBI-based regimen (ns). Similarly, no significant difference of NRM between the two regimens was detectable for the 78 pts in more advanced disease stages. Further, the post transplant risk of hematologic relapse (RR) was not influenced by the conditioning regimen: The RR for CR pts at 1 and 3 yrs was 13% (95%-CL: 1%–25%) and 16% (95%-CL: 3%–30%) after Treo-compared to 25% (95%-CL: 16%–34%) and 33% (95%-CL: 23%–43%) after TBI-based conditioning (ns). The corresponding figures for more advanced stages were 43% (95%-CL: 16%–70%) and 57% (95%-CL: 27%–87%) compared to 29% (95%-CL: 17%–40%) and 33% (95%-CL: 21%–46%), respectively (ns). Overall survival (OS) for all CR pts at 1 and 3 yrs was 77% (95%-CL: 63%–92%) and 54% (95%-CL: 36%–73%) after Treo-compared to 68% (95%-CL: 58%–67%) and 49% (95%-CL: 38%–60%) after TBI-based conditioning (ns). For pts transplanted in CR1, OS at 1 and 3 yrs reached 79% (95%-CL: 61%–97%) and 63% (95%-CL: 41%–85%) after Treo- and 70% (95%-CL: 58%–82%) and 53% (95%-CL: 39%–66%) after TBI-based conditioning (ns). In more advanced stages, OS at 1 and 3 years declined to 29% (95%-CL: 0%–7%) and 21% (95%-CL: 0%–50%) after Treo- and to 29% (95%-CL: 15%–44%) and 16% (95%-CL: 1%–32%) after TBI-based conditioning, respectively (ns). Major determinants of OS and RR were the disease stage (CR vs more advanced stages) and the cytogenetic risk category (low-intermediate vs high risk). The HCT-CI (stratified to <=2 or >2) had no apparent influence on NRM, RR, or OS in this analysis.

Conclusion: This retrospective long-term analysis supports that Treo-based conditioning leads to at least equivalent results regarding the major clinical endpoints of alloSCT when compared to a standard myeloablative TBI-based conditioning regimen in adult AML pts. An international prospective multicenter randomized trial has now been launched (EudraCT-No. 2008-002356-18) to compare Treosulfan-based conditioning with a reference reduced intensity conditioning regimen in AML and MDS pts.