In the present study, post-transplant outcomes for AML in MRD-negative CR1 were assessed by comparing two different MAC and RTC regimens. FluBu3 RTC showed comparable post-HSCT outcomes to BuCy MAC, with a similar OS of 84.5% and 78.3% (p=0.358) and RFS of 76.3% and 78.0% (p=0.806), respectively, at 2 years post-transplantation. The similar OS and RFS between the groups were due to the reduced NRM at the expense of relapse for the FluBu3 arm or vice versa for the BuCy arm; the incidence was 21.5% and 7.7%, for CIR (p=0.074) and 2.2% and 14.3% for NRM (p=0.0324), respectively.
Numerous attempts have been made to identify optimal conditioning for AML patients7,13,21,22; the summary of several studies is shown in Table 5 in which various conditioning regimens are presented in the order of treatment intensities based on TCI score. As an acceptable lower intensity, FluBu2 (TCI score, 1.5) showed comparable survival and relapse outcomes to BuCy in a prospective EBMT study12; however, the target population in that study was subjects with MDS and included a small portion of secondary AML. In a BMT-CTN study, where the majority of patients received FluBu4 (64%) and FluBu2 (81%) as MAC and RIC, respectively, a significant improvement in OS with MAC was observed specifically for AML patients but not for MDS patients17. For the optimal upper range of intensity, comparative studies in which FluBu4 (TCI score, 3.5) versus BuCy (TCI score, 4.0) are investigated could provide some insights. Although the use of FluBu4 has become more popular due to its advantageous ability to reduce NRM while retaining similar OS and leukemia free survival was reported in a large prospective study conducted by an Italian group23, confirming its superiority is difficult because conflicting results have been reported in some prospective studies24,25. Based on these findings, the optimal conditioning intensity for patients with AML is likely within the TCI score range of >1.5 but ≤4.0, unless patients are expected to be intolerant of the regimen.
The FluBu3 protocol we used as RTC in the present study was assessed as a valid alternative in frail patients with either myeloid or lymphoid hematologic malignancies in a prospective phase 2 trial conducted by French researchers26. In that study which included 34% AML patients, 2-year OS, disease-free survival (DFS), CIR, and NRM were 62%, 50%, 44%, and 11% in the entire cohort, respectively. In a more recent study by the Acute Myeloid Leukemia Working Party (ALWP) of the EBMT, patients with AML who underwent first allogeneic HSCT in CR1 and had intermediate or high-risk karyotype were evaluated after the use of either the FluBu3 or FluTBI800 conditioning regimens27; in the FluBu3 arm, the probabilities of OS, RFS, CIR, and NRM at 2 years were 62%, 60%, 30%, and 10%, respectively. In the present study, the respective 2-year OS, RFS, CIR, and NRM rates with FluBu3 were better at 85%, 76%, 22%, and 2%, respectively. These results may be attributed to the more favorable patient characteristics of MRD-negative CR1 patients and the inclusion of favorable-risk AML patients. Although a head-to-head comparison of results between FluBu3 and FluBu2/4 has not been released to date (ClinicalTrials.gov; identifier: NCT01985061), apparently the results from the use of intermediate dose busulfan (Bu3) correspond well with the FluBu4 arm in the Italian study in which the 2-year OS, RFS, CIR, and NRM were reportedly 62%, 59%, 32% and 8%, respectively. Compared with FluBu2, FluBu3 might be associated with better disease control; Shimoni et al. reported 2-year OS and DFS rates of 47% and 43%, respectively, when using the FluBu2 regimen28, and BMT-CTN was associated with 1.5-year RFS and CIR rates of 47% and 48%, respectively, using the RIC regimen (81%, FluBu2)17.
Although various middle dose intensity regimens are available, in a retrospective study by ALWP of the EBMT, FluTBI800 showed favorable results compared with FluBu327 despite both regimens having the same TCI score of 2.5. In the entire cohort, relapse rates at 2 years were 30% and 20% (p=0.01) for FluBu3 and FluTBI800, respectively. FluTBI800 was also associated with significantly improved RFS, OS, and GRFS, and a tendency towards reduced risk of relapse in multivariate models for patients <50 years of age. Regarding BuCy and CyTBI MAC regimens, although both can be scored as 4 based on the TCI-score system, greater toxicities and higher NRM were reported with CyTBI than with BuCy in some studies29,30. These findings may indicate the appropriateness of applying the same score system to different types of treatments with distinct properties, such as radiation and chemotherapy, is debatable31–33. Otherwise, these results may also be interpreted as showing the efficacy and/or risk are not always associated with intensity as classified, and even regimens with the same intensity score may have varying levels of efficacy and toxicity depending on the regimen-specific profiles34.
Although the focus in the present study was only on MRD-negative CR1 patients, relapse remained the primary cause of transplant failure in the FluBu3 group, with higher-risk cytogenetics and/or FLT3-ITD mutation significantly negatively affecting transplant outcomes. Notably, 75% of relapsed FluBu3 patients in this study had poor cytogenetics or FLT3-ITD mutation. Factors such as higher HCT-CI and HLA-mismatch, which primarily were associated with increased NRM and worse OS and RFS in the BuCy group, did not significantly affect post-transplant outcomes in the FluBu3 group. Overall, FluBu3 RTC does not appear to overcome inherent disease risk even with a deep response prior to HSCT. In addition, a higher risk for NRM may exist when applying conventional MAC for MMUD transplants. These findings have practical implications, such as considering patients with a higher risk feature based on their baseline cytogenetic characteristics as candidates for active post-HSCT maintenance therapy35, or for receiving a conditioning regimen of higher intensity than FluBu3 if specific maintenance therapy is unavailable. Furthermore, if the donor option is limited to MMUD, opting for less intense conditioning such as FluBu3 instead of BuCy would be a more appropriate choice.
Despite the notable results, the present study had several limitations. A major weakness was the non-interventional nature, and the baseline characteristics between the two different conditioning groups were not well balanced, potentially influencing post-HSCT outcomes. Indeed, the percentage of patients who received ATG was significantly higher in the FluBu3 arm compared with the BuCy arm, which might have contributed to a higher incidence of relapse despite the total dose of ATG used was lower (only 2.5 mg/kg) than the usual dose. In addition, because the use of FluBu3 conditioning was started later due to the changes in institutional strategy, a significant gap in follow-up periods between the two groups occurred; the FluBu3 group had a much shorter period, which may have led to a potential failure in detecting delayed hazards36. In addition, the testing technique used to define MRD negativity in the present study does not meet the current guideline recommendations19, with the exception of patients with either RUNX1-RUNX1T1, CBFB-MYH11 fusions, or NPM1 mutation. Lastly, because FluBu3 and other regimens of intermediate dose intensities were not compared in the present study, whether the FluBu3 regimen is the best-performing regimen among regimens of similar intensities could not be confirmed.
In summary, the results of the present study showed FluBu3 RTC is a viable alternative to BuCy MAC for MRD-negative CR1 AML patients undergoing transplant with comparable post-HSCT outcomes in terms of similar OS and RFS, and associated with significantly reduced NRM offset by a trend towards higher risk of relapse. Individual conditioning plans in which inherent disease risk, patient frailty, and donor availability are considered are necessary for optimal outcomes. Although the present study has some limitations, the results provide valuable insight into conditioning regimens for AML patients undergoing transplant.