Haploidentical hematopoietic stem cell transplantation (haplo-HSCT) using T-cell-replete grafts and post-transplantation cyclophosphamide (PTCY) provides a popular curative approach for older patients (pts) with high-risk (HR) MDS/AML. The sequential therapeutic concept herein, is used to optimize disease control and gain time, especially in patients suffering from active disease at time of haplo-HSCT. Yet, sequential conditioning regimens prior to haplo-HSCT are often associated with a considerable risk of severe adverse events especially in older comorbid patients. Previous studies in the HLA-matched setting have demonstrated feasibility and safety of treosulfan-based reduced intensity conditioning (RIC) by stable engraftment and low non-relapse mortality (NRM). However, data for treosulfan-based conditioning in the unmanipulated HLA-haplo-HSCT setting in HR AML/MDS pts are rare, especially in the context of sequential conditioning.

Here we report on a matched-pair analysis of 26 patients treated with either a treosulfan- or melphalan-based sequential conditioning for haplo-HSCT using PTCY as GvHD prophylaxis in HR MDS/AML.

We retrospectively analyzed the outcome and toxicity profile of 26 patients undergoing sequential haplo-HSCT at our center between January 2009 and June 2019. Thirteen patients with HR AML/MDS and >54 years old who underwent sequential haplo-HSCT using treosulfan (3x10g/m2) for RIC were considered for potential matching with recipients (n=30) of a sequential melphalan-based RIC regimen. Matching criteria comprised (1) disease activity (blast yes or no), (2) disease status (relapse, refractory, high-risk cytogenetics), (3) HCT-CI and (4) age (+/- 5 years). Post-grafting immunosuppression consisted of cyclophosphamide, tacrolimus and MMF.

Thirteen patients undergoing treosulfan-haplo-HSCT were successfully pair-matched with thirteen recipients of melphalan-based haplo-HSCT, respectively ((1) p=1.0; (2) p=1.0; (3) p=1.0; (4) p=0,76). Median age of the entire cohort was 63 years (54-71). Each group consisted of two MDS patients and eleven AML patients. All recipients treated with treosulfan showed neutrophil engraftment with a median of 20 days while only 69% of the melphalan treated patients engrafted (median=19 days, p= 0.9). In the melphalan group one graft rejection occurred, three patients died in early aplasia. Acute GvHD °II-IV occurred in 23% and 44% of the patients treated with treosulfan or melphalan, respectively. Severe (°III-IV) non-hematologic regimen-related toxicities were seen in 2/13 pts of the treosulfan and 7/13 pts of the melphalan group, predominately affecting the GI-tract in both. NRM at day +100 was 0% and 31% (p=0.06) for the treosulfan and melphalan group, respectively. Thereby, infections made up for most NRM events. Conversely, CI of relapse at one year was 23% vs 0 % (p=0.004) for treosulfan vs melphalan. OS at 1-year (treosulfan group: 69% vs melphalan gropup: 62%) and PFS at 1-year (treosulfan group: 69% vs melphalan group: 62%) did not differ significantly between the groups (p=0.72)

Sequential haplo-HSCT using treosulfan and PTCY in older advanced MDS/AML patients is safe, resulting in lower NRM at the expense of higher relapse incidence compared to the melphalan-based sequential conditioning approach. Treosulfan-based RIC in haplo-HSCT though might be an alternative in older pts with low leukemic burden. However for disease control its intensity should be reconsidered and all available post-grafting mantainance strategies applied.


Fraccaroli:Medac: Other: Travel Grant. Buecklein:Celgene: Research Funding; Amgen: Consultancy; Pfizer: Consultancy; Gilead: Consultancy, Research Funding; Novartis: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution