A single-centre comparison of thiotepa-treosulfan versus thiotepa-busulfan based conditioning regimens in adults with haematological malignancies undergoing allogeneic haematopoietic stem cell transplant.

Abstract Introduction Choice of conditioning chemotherapy plays a crucial part in haematopoietic stem cell transplantation (HSCT) outcomes. Treosulfan- and busulfan-based reduced intensity conditioning has been associated with lower graft versus host disease (GVHD), lower treatment related toxicity and improved non-relapse mortality (NRM). Expounding on that, the addition of the alkylator thiotepa to fludarabine and busulfan has been shown to be effective and safe. Yet there is limited data on safety and utility in double alkylator thiotepa and treosulfan conditioning in adults with haematological malignancies. Aims In this single centre retrospective analysis, we compared the transplant outcomes and adverse events between thiotepa and treosulfan-based versus thiotepa and busulfan-based conditioning. The primary endpoint was 1-year overall survival (OS). Secondary endpoints were 100-day NRM, incidence of GVHD, 1-year cumulative incidence of relapse (CIR) and graft vs host disease-free, relapse-free survival (GRFS). Adverse events were also evaluated. MethodsAdult patients who underwent allogeneic HSCT for haematological malignancies at a tertiary medical centre in Singapore between January 2020 to April 2025 and received thiotepa with treosulfan or thiotepa with busulfan based regimens were included. Treosulfan-based group: thiotepa-treosulfan-fludarabine (TTF) or thiotepa-etoposide-cyclophosphamide fludarabine-treosulfan (TEC-FT) regimens. Busulfan-based group: thiotepa-busulfan-fludarabine (TBF) or thiotepa-etoposide-cyclophosphamide busulfan-fludarabine (TEC-BF) regimens. Clinical characteristics and outcome measures were collected and analysed using Kaplan-Meier for survival analysis and Fine-Gray model for cumulative incidence of relapse. ResultsA total of 49 patients were analyzed. Fourteen patients were given thiotepa and treosulfan-based conditioning while 35 patients were given thiotepa and busulfan-based conditioning. Patients in the treosulfan group were older, with median ages of 60 years old in the treosulfan group vs 54 years old in the busulfan group (p-value = 0.023). Patients in the treosulfan group had a lower proportion in complete remission (CR) going into transplant compared to that of the busulfan group (CR of 21% vs 37%, p-value = 0.035). There was no significant difference in haematopoietic cell transplantation-specific comorbidity index (HCT-CI) scores, Eastern Cooperative Oncology Group (ECOG) status, and donor source between the two both groups. The treosulfan group had a significantly higher proportion of sequential conditioning regimens compared to the busulfan arm (32% vs 2.9%, p-value = 0.005). No difference in OS was demonstrated at 1-year between treosulfan-based vs busulfan-based conditioning (65.6 vs 63.6%). When evaluating secondary endpoints, the 100-day NRM and incidence of acute GVHD ≥3 were similar. In both groups, 100-day NRM was 14%. Incidence of acute GVHD grade 3 and above was 29% in the treosulfan arm and 11% in the busulfan arm respectively. There was no difference in 1-year CIR (27.2 vs 21.9%) or GRFS (53.0 vs 63.5%) for thiotepa and treosulfan-based vs thiotepa and busulfan-based conditioning. Data from 6 patients had yet to mature to 365 days at the time of analysis hence were excluded from the 1-year OS, CIR and GRFS data analysis. Lastly, there was no significant difference in adverse events including the incidence of veno-occlusive disease (VOD), liver derangements, infections, mucositis and bleeding. Discussion We observed similar effectiveness and safety between thiotepa combined with treosulfan-based compared to thiotepa combined with busulfan-based conditioning. Notably, these outcomes were observed despite the treosulfan arm comprising of an older patient population, a higher proportion of individuals not in CR prior to transplant and undergoing sequential conditioning regimens, as well as more patients considered at an elevated risk for VOD due to a greater prevalence of pre-transplant liver dysfunction, iron overload, or portal hypertension. The rates of 100-day NRM and acute GVHD incidence were not significantly different. There were similar 1-year OS, CIR, GRFS rates and adverse events. Limitations of this retrospective analysis include small sample size predisposing to potential sampling bias. Randomized prospective studies are needed to assess the potential of double alkylator combination of thiotepa and treosulfan in adults with haematological malignancies.