Allogeneic hematopoietic stem cell transplantation (HSCT) is a potentially curative therapy for patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS). Despite an improvement in results of allogeneic HSCT during the past decade mainly due to a lowered transplantation-related mortality, morbidity is still high, which makes allogeneic HSCT with myeloablative conditioning appropriate only for younger patients with good performance status.1 Therefore, allogeneic HSCT was not actively performed because a large number of AML or MDS patients are elderly who are at a high-risk based on hematopoietic cell transplantation-specific comorbidity index (HCT-CI).

The advent of reduced-intensity conditioning (RIC) regimens has allowed safe application of HSCT for many patients who are older or having significant comorbidities. Retrospective comparisons of traditional myeloablative regimens to RIC regimens have shown that RIC regimens can decrease transplantation-related mortality. However, relapse is increased, resulting in similar overall survival (OS) to myeloablative conditioning.234 More recent series have suggested that dose intensity of myeloablative conditioning can be partially decreased, thereby lowering associated toxicity without sacrificing efficacy.56 However, these retrospective analyses have an inherent selection bias due to differences in characteristics of patients eligible for each modality. Currently, many institutions often use a combination of fludarabine and busulfan as RIC regimen. In addition, it is unclear which busulfan dose is appropriate in terms of efficacy vs. safety in patients who can't receive standard conditioning. The selection of appropriate conditioning is a major concern for the attending physician, which must be carefully judged considering the various conditions associated with the patient.

Treosulfan (L-treitol-1,4-methanesulfonate) as a prodrug of a bifunctional alkylating cytotoxic agent has been applied to pediatric and adult patients with hematological malignancies with low risks of organ toxicity and treatment-related mortality.789101112 As this drug does not require enzymatic activation or hepatic metabolism, drug level monitoring and adjustments are not needed owing to its low inter- and intra-patient variabilities. Treosulfan exhibits strong immunosuppressive characteristics with low proinflammatory cytokine release.13 This facilitates stem cell engraftment. It is associated with a lower risk of graft versus host disease (GVHD).14 These properties make treosulfan an attractive candidate for use in conditioning regimens before allogeneic HSCT. The combination of treosulfan and fludarabine (Treo/Flu) has been introduced over the last few years.9101112 Recent study showed that treosulfan is not inferior to busulfan when used in combination with fludarabine as a conditioning regimen for allogeneic HSCT for older or comorbid patients with AML or MDS.15 However, further studies are needed before treosulfan based regimen can be used as a standard conditioning regimen in this population. And the equivalent dose intensity of treosulfan to busulfan has not been defined yet. Furthermore, Korean data on treosulfan-based conditioning regimen are scarce. There is a need for active research activities on this.

Therefore, we conducted a retrospective study to investigate the efficacy of Treo/Flu conditioning compared to a popular regimen of RIC using busulfan and fludarabine (Bu/Flu) for patients with AML or MDS who were not eligible for a standard myeloablative conditioning.

We retrospectively identified and collected clinical data of 52 patients with AML or MDS after HSCT from matched related donor or unrelated donor, haploidentical donor between May 2011 and December 2019 using a combination of intravenous infusion of Bu/Flu or Treo/Flu. All patients were ineligible for standard conditioning regimen due to high risks of transplantation-related toxicity and mortality. Their high-risk features were age ≥ 60 years, severe liver or lung toxicities after prior chemotherapies, or persistent severe organ dysfunctions. We compared clinical outcomes of HSCT using Treo/Flu conditioning (n = 13) and those of HSCT using Bu/Flu conditioning (n = 39), a widely used regimen.

Serologic human leukocyte antigen (HLA) typing was sufficient for choosing matched sibling donors. High-resolution DNA typing was used to identify unrelated matched donors or haploidentical donors. Treosulfan-based conditioning consisted of treosulfan (42 g/m², given in daily single doses on day −7 to −5 in a dose of 14 g/m²) and fludarabine (150 mg/m², intravenously given divided from day −7 to −3 in a dose of 30 mg/m²). RIC with Bu/Flu consisted of 3.2 mg/kg/day of intravenous busulfan for 2 or 3 days (a total of 6.4–9.6 mg/kg) and 30–40 mg/m²/day of fludarabine for 4 or 6 days (a total of 160–180 mg/m²). In vivo T cell depletion with anti-thymocyte globulin (ATG) was administered according to donor source and the risk of GVHD, with total doses of ATG (thymoglobulin) ranging from 4 mg/kg to 9 mg/kg. Grafts comprised peripheral blood stem cells (PBSCs) mobilized by granulocyte colony-stimulating factor. Tacrolimus with or without low dose methotrexate or mycophenolate mofetil only was used to prevent GVHD. Methotrexate was used at a dose of 5 mg/m² on days 1, 3, 6, and 11.

Bone marrow recovery was assessed using the day of neutrophil engraftment (defined as the first of 3 consecutive days of an absolute neutrophil count of ≥ 500/uL) and platelet engraftment (defined as the first of 3 consecutive days when the platelet count exceeded 20,000/uL without transfusion requirement). Disease relapse was defined according to standard hematologic criteria. Non-relapse mortality (NRM) was defined as death from any cause in the absence of prior disease recurrence. OS was calculated from the day of HSCT until death from any cause or last follow-up. Relapse-free survival (RFS) was defined as survival without relapse. Patients with no event were censored at the last contact. The cause of death was categorized according to standard criteria. The cause of death of patients who experienced relapsed disease at any time before death was considered as relapse related. Acute and chronic GVHD were defined according to published consensus criteria.1617

The effect of age in post-transplant outcomes is a matter of debate with conflicting results.181920 Various models including HCT-CI,21 comorbidity-age index,19 and Charlson comorbidity index22 are widely used for pre-transplant risk stratification prior to allogeneic HSCT based on patient’s comorbidities. However, there is a limit to customized evaluation for each patient. These prediction systems did not include age, disease status, or source of graft known to be potential prognostic factors. If a patient is over 65 years of age or older and in the high-risk group based on the HCT-CI score, the predicted NRM is more than 40%, which is a typical example of trouble. It remains entirely up to the attending physician to decide whether to proceed with transplantation in complex and difficult patients.

We compared a treosulfan-based preparative regimen with busulfan-based regimen, a widely accepted RIC regimen for older or comorbid patients with AML or MDS. In this study, all patients in the Treo/Flu group were reluctant to use conventional RIC regimen. Most patients were 65 years old or older who were in the high-risk group based on HCT-CI. Thus, NRM of these cases was expected to be greater than 30%. About 70% of this group underwent haplo-identical stem cell transplantation. Although treating AML and MDS with haploidentical HSCT can achieve outcomes comparable to those of transplantation with other stem cell sources,232425 relapse and infections remain causes of death that restrict further improvement in clinical outcomes. Thus, the effect of donor type might have contributed to the result of increased infection complications and high NRM in the Treo/Flu group. In addition, as patients in the Treo/Flu group had other concomitant poor prognostic factors, their survival was expected to be poor.

As described in the results, the Bu/Flu group tended to show better results. These differences were judged to be due to difference in NRM at 1 year after transplantation. In the subgroup analysis for the HCT-CI high-risk group, there were no differences in survival outcomes between the two groups. The difference in NRM at 1 year was also significantly reduced. Therefore, if a patient considering an allogeneic HSCT based on a general RIC regimen is concerned about high NRM, the Treo/Flu conditioning regimen may be an effective alternative. Table 2 compares the side effects that occurred during the period from pre-transplantation treatment process to bone marrow function recovery and shows the advantage of Treo/Flu over Bu/Flu. In the Treo/Flu group, the number of cases of conditioning-related mucositis requiring treatment was less than one fourth. In addition, the frequency of nausea or vomiting requiring drug treatment was decreased. Moreover, the duration of symptom was also shorter. Therefore, in the Treo/Flu group, which contains many high-risk HCT-CI patients or elderly, this regimen is beneficial for maintaining good general condition as it can reduce drug-related side effects during or after the conditioning process.

Main reasons for the increase in NRM in the Treo/Flu group were an increase in the incidence of acute GVHD and more than two-fold increase in infectious complications during the follow-up period. Such difference in NRM was thought to be due to the increased frequency of acute GVHD, which increased the active use of immunosuppressive drugs, leading to increased infectious complications and ultimately increased mortality. This problem was thought to be related to the high proportion of haplo-identical transplantation. Although there was no significant difference in the overall frequency of infectious complications, it is also unclear what could explain the significant increase in the frequency of EBV reactivation in the Treo/Flu group. Potential causes for this problem may be differences in the frequency of ATG use,26 haplo-identical transplantation,27 and acute GVHD,2829 which have been reported as risk factors for EBV reactivation. Beelen et al.15 have reported a phase 3 clinical study, showing that treosulfan-based regimens are non-inferior to busulfan when they are combined with fludarabine. In addition, NRM is low, proving that treosulfan-based regimens are safe regimens. It is difficult to compare our study directly with a large-scale randomized-controlled study conducted with 460 subjects. Compared to the study of Beelen et al.,15 the median age of patients was much higher in the Treo/Flu group of our study. In addition, more than half of our patients were older than 65 years of age. Besides, in the Treo/Flu group of our study, the proportion of patients with HCT-CI score of 3 or higher was more than 90% and the transplantation using haplo-identical donor accounted for about 70%. These might have adversely affected their transplant outcomes. Compared to the study of Beelen et al.,15 in the Treo/Flu group of our study, the percentage of complete donor chimerism at 100 days after transplantation was lower (69.2% vs. 86.4%), although the incidence of acute GVHD above grade 2 was similar (46.2% vs. 52%). However, the frequency of chronic GVHD was lower in our study (30.8% vs. 51%). During a two-year follow-up, the relapse rate was similar, although there was a difference in mortality. Due to higher NRM (30.8% vs. 10%), 24-month RFS and OS were lower (38.5% vs. 64% and 46.2% vs. 71.3%, respectively) in our study. These differences might have been influenced by baseline characteristics of subjects. When comparing survival outcomes with Bu/Flu group as a control, outcomes were similar in both studies (56.4% vs. 50.4% for RFS, 60.0% vs. 56.4% for OS).

Although advantages of treosulfan-based conditioning regimen have been confirmed in a lot of clinical studies,78910111215 there are a few related studies on Korean. This combination as an effective RIC regimen has not been of interest to local hematologists or transplant-related medical physicians. We were able to find very few relevant data, which were included in some systemic analyses of HSCT in patients with RIC or nonmalignant diseases.3031 In the future, it is necessary to conduct a well-designed clinical study with a lot of patients using the Treo/Flu conditioning regimen so that conclusive results can be obtained in Korea.

As a limitation of this study, the overall sample size was small. The number of patients belonging to the Treo/Flu group was also small. Therefore, statistical analysis power was limited. In addition, disease types evaluated as prognostic factors affecting survival outcomes tended to differ between the two groups. The survival outcome of AML patients was evaluated to be lower than that of MDS. The difference was particularly significant for secondary leukemia. Therefore, differences in baseline characteristics before transplantation might have hindered objective comparative evaluation. As an additional problem, although we have a good understanding of the clinical importance of risk stratification based on cytogenetic and molecular studies, we were unable to proceed with the analysis due to insufficient conditions. Molecular tests including next generation sequencing, were conducted at my institution from 2017. About 60% of the subjects were able to obtain cytogenetic and molecular data that could be applied to the recently introduced risk stratification, so the analysis was not conducted.

In conclusion, this study suggests that Treo/Flu conditioning can be an alternative to Bu/Flu regimen in patients with high-risk based on HCT-CI because there was no statistically significant difference in clinical outcomes according to the type of conditioning. This regimen is expected to increase in clinical application as a RIC regimen for patients with a high-risk of NRM, especially in elderly over 65 years of age, which are recently increasing. A randomized prospective study is needed to better define the role of treosulfan-based conditioning in allogeneic HSCT in Korea.