Treosulfan Vs Busulfan As Part of Clofarabine Based Reduced Intensity Conditioning Regimen before Allotransplant for Myeloid Malignancies

Introduction: Clofarabine-busulfan based reduced intensity conditioning (RIC) regimens have been proved to be efficient in providing good survivals in adults allotransplanted for myeloid malignancies. Recently, treosulfan replacing busulfan as part of a fludarabine/busulfan RIC regimen was shown to provide better survival thanks to a lower non relapse mortality in older or comorbid allotransplanted patients with acute myeloid leukaemia (AML) or myelodysplastic syndrome (Beelen, Lancet Haematology 2019). Here we provide the results of allotransplant after using a clofarabine/treosulfan RIC regimen for myeloid malignancies, reporting also a comparison with a clofarabine/busulfan RIC regimen.

Methods: This was a monocentric retrospective study comparing adult patients allotransplanted for myeloid malignancies with either a Clofarabine/busulfan or a Clofarabine/treosulfan RIC regimen. The former consisted of clofarabine 30 mg/m²/day 5 days, busulfan 3.2 mg/kg/day 2 days and 2.5 mg/kg/day of rabbit anti-thymocyte globulin 1 or 2 days (CLOB2A1/A2). The latter consisted of clofarabine 30 mg/m²/day 5 days, treosulfan 10 g/m²/day 3 days and 2.5 mg/kg/day of rabbit anti-thymocyte globulin (ATG) 2 days (CLOT3A2, thereafter CloT3 group). Beside ATG, graft-versus-host disease (GVHD) prophylaxis consisted for both groups of ciclosporine and mycophenolate mofetil (MMF). All patients received peripheral blood stem cell as source of graft. Various outcomes were considered: neutrophils (>0.5 Giga/L) and platelets (>50 Giga/L) recoveries, overall survival (OS), disease-free survival (DFS), non-relapse mortality (NRM), relapse incidence (RI), acute and chronic GVHD. Stastistical analyses were performed in July 2024 using R version 4.2.2.

Results: A total of 142 adult patients (>=18 years old, yo) were considered, 108 in the CloB2A1/A2 sub-group and 34 in the CloT3A2 sub-group. Having comparable outcomes (OS, DFS, NRM, RI, acute and chronic GVHD), patients receiving CloB2A2 (n=27) and CloB2A1 (n=81) were thus combined for the purpose of this study (CloB2 group). Patients were transplanted between August 2010 and July 2023 in the same center. The use of treosulfan was part of a more recent strategy (transplants between May 2022 and July 2023), with the consequence that the median follow-up was shorter for the CloT3 sub-group (18 vs 84 months). Patient characteristics were similar between both groups in terms of median age (CloB2 65 yo vs CloT3 62 yo) and HCT-CI score (3 vs 3), disease (AML 67% vs 71%), donor type (all matched donors, median age 40 yo vs 37 yo, female donor to male recipient 12% vs 3%), ELN2017-risk for AML patients (adverse 60% vs 67%), CMV status (Recipient-/Donor- 46% vs 59%), median graft CD34+ and CD3+ cells infused. When comparing both groups, neutrophil recovery was significantly faster for the CloT3 sub-group (median 10 vs 16 days, p<0.001) while platelets recovery delay was similar (13 vs 11 days, p=0.2) Other outcomes were similar: 18-month OS (CloB2 69% vs CloT3 79%, p=0.3), DFS (63% vs 70%, p=0.4), NRM (15% vs 15%, p>0.9), RI (22% vs 15%, p=0.3), day-100 grade 2-4 (18% vs 24%, p=0.4) and grade 3-4 (13% vs 21%, p=0.2) aGVHD, chronic GVHD (41% vs 56%, p=0.06) and extensive chronic GVHD (16% vs 18%, p=0.8). The causes of death were not significantly different (p=0.4).

We then focused only on AML patients (CloB2 n=71, CloT3 n=24). The two groups had similar characteristics. 18-month OS (CloB2 72% vs CloT3 79%, p=0.6), DFS (68% vs 75%, p=0.5), NRM (9.9% vs 21%, p=0.13), day-100 grade 2-4 (18% vs 25%, p=0.5) and grade 3-4 (13% vs 21%, p=0.2), chronic GVHD (42% vs 46%, p=0.5) and moderate/severe chronic GVHD (20% vs 21%, p=0.9) were comparable. However, CloT3 was associated with significant lower 18-month RI (4.4% vs 23%, p=0.04). This was confirmed by multivariate analysis (HR: 0.14; 95%CI: 0.02-1.00, p=0.05). The main cause of deaths was GVHD in the CloT3 sub-group (n=5/5, 100%) while it was relapse in the CloB2 sub-group (n=24/35, 69%), p=0.006.

Conclusion: The use of a CloT3 RIC regimen provides similar outcomes compared to the use of a CloB2 RIC regimen in adults with myeloid malignancies receiving a PBSC matched transplant, with the exception of a faster neutrophils recovery. However, CloT3 is associated with significant lower RI but higher incidence of GVDH-related deaths in AML patients. These results have to be confirmed on larger prospective studies using maybe reinforced GVHD prophylaxis.