Comparison of reduced-intensity conditioning regimens in patients with acute lymphoblastic leukemia >45 years undergoing allogeneic stem cell transplantationâ€”a retrospective study by the Acute Leukemia Working Party of EBMT

The optimal reduced-intensity conditioning (RIC) for patients with acute lymphoblastic leukemia (ALL) undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT) remains unclear. We retrospectively analyzed 417 patients > 45 years with ALL in first complete remission who underwent a matched sibling or unrelated allo-HSCT and compared outcomes between fludarabine/busulfan (FLUBU, n= 127), fludarabine/melphalan (FLUMEL, n= 190), and fludarabineTBI (FLUTBI, n= 100) conditioning. At 2 years, there were no differences between the groups in terms of cumulative incidence (CI) of relapse (40% for FLUBU vs 36% for FLUMEL vs 41% for FLUTBI, p= 0.21); transplant-related mortality (TRM) (18% for FLUBU, 22% for FLUMEL, 14% for FLUTBI, p= 0.09); overall survival (55% for FLUBU, 50% for FLUMEL, 60% for FLUTBI, p= 0.62) or leukemia-free survival (43% for FLUBU, 42% for FLUMEL, 45% for FLUTBI, p= 0.99), but GVHD-relapse-free survival was significantly lower in the FLUTBI group than FLUBU and FLUMEL group (18% vs 35% vs 28%, p= 0.02). However, this difference was lost in the multivariate analysis when adjusted for the in vivo T-cell depletion. Finally, the FLUMEL regimen was shown to be an independent risk factor for a higher TRM (HR 1.97, 95% CI 1.05–3.72, p= 0.04). We conclude that the three most popular RIC regimens yield similar transplant outcomes. * Zinaida Peric zinaida.peric@mef.hr 1 University Hospital Centre Zagreb, School of Medicine, University of Zagreb, Zagreb, Croatia 2 EBMT Paris study office/CEREST-TC, Saint Antoine Hospital, INSERM UMR 938, University Pierre et Marie Curie, Paris, France 3 Erasmus MC Cancer Institute, University Medical Center Rotterdam, Rotterdam, The Netherlands 4 University College London Hospital, London, UK 5 Royal Marsden Hospital, London, UK 6 University Hospital Birmingham NHS Trust, Queen Elizabeth Medical Centre, Edgbaston, Birmingham, UK 7 Nottingham University Hospital, Nottingham, UK 8 University Hospital Maastricht, Maastricht, The Netherlands 9 University Hospital Centre Lapeyronie, Montpellier, France 10 Hospital St. Louis, AP-HP, University of Paris, Paris, France 11 Hospital Clinic, Institute of Hematology & Oncology, Barcelona, Spain 12 University Medical Centre, Utrecht, The Netherlands 13 Leeds Teaching Hospitals Trust, Leeds, UK 14 Maria Sklodowska-Curie Institute—Oncology Center, Gliwice Branch, Gliwice, Poland 15 Hematology Division, Chaim Sheba Medical Centre, TelHashomer, Israel 16 Saint Antoine Hospital, University Pierre et Marie Curie, Paris, France 12 34 56 78 90 () ;,: 12 34 56 78 90 (); ,:


Introduction
Long-term outcomes of older adults with acute lymphoblastic leukemia (ALL) remain poor, with an estimated 5-year leukemia-free survival (LFS) of~30-40% [1][2][3]. These results have been obtained with chemotherapy alone and are partly due to the inability of older adults to tolerate intensive regimens used in pediatric and young adult populations. The use of conventional myeloablative allogeneic hematopoietic stem cell transplantation (allo-HSCT) has been shown to improve survival rates in adults by 45-75% [4,5]. However, transplantrelated mortality (TRM) after myeloablative allo-HSCT is substantial, ranging between 33 and 58% [6], increases with age, and is higher for adults with impaired performance status [7,8]. In such patients, reduced-intensity conditioning (RIC) may offer the chance of a potentially curative strategy by obtaining a graft-versus-leukemia effect without the associated toxicities of myeloablative conditioning (MAC). On the other hand, the risk of relapse after RIC regimens may be greater than that after MAC regimens [8][9][10].
Although several RIC regimens have been developed over the last decades, their cytotoxic and immunosuppressive effects are different, and this may influence transplant outcome. However, to date there have been no large prospective studies comparing outcomes of different RIC regimens in patients with acute leukemias, and the optimal RIC regimen in allo-HSCT remains unclear. The most widely used RIC regimens are fludarabine with intermediate doses of busulfan (6.4 mg/kg), fludarabine with intermediate doses of melphalan (140 mg/m 2 ), and fludarabine with low-dose total body irradiation (TBI, 2 Gy). Several retrospective studies have compared these regimens, but with contradictory results [11,12]. This is probably due to small population numbers, different diseases being analyzed together and neither age limit for enrollment nor dosage of drugs in regimens being fixed. Furthermore, these studies focused mostly on acute myeloid leukemia and included only small numbers of patients with ALL.
We therefore took advantage of the European Society for Blood and Marrow Transplantation (EBMT) dataset, and retrospectively compared outcomes of these three most popular RIC conditioning regimens following allo-HSCT from a matched sibling donor or an unrelated donor in a large homogeneous population of patients with ALL aged 45 years or older undergoing transplant in first complete remission (CR1).

Study design and data collection
This is a registry-based retrospective study. Data were provided and the study design was approved by the Acute Leukemia Working Party (ALWP) of the EBMT group registry, in accordance with the EBMT guidelines for retrospective studies. The EBMT is a voluntary working group of more than 600 transplant centers which are required to report all consecutive stem cell transplantations and followups once a year. Audits are routinely performed to determine the accuracy of the data. Since 1990, patients have been able to provide informed consent to authorize the use of their transplant information for research purposes. The ALWP of the EBMT granted ethical approval for this study.

Patient selection
Patients were selected according to the following criteria: (1) aged 45 years and older at the time of transplantation, (2) a diagnosis of ALL, with available data on the immunophenotype and Ph-positivity, (3) in CR1 (4) initial allo-HSCT between 2005 and June 2016, (4) HLA-matched related or unrelated donor (fully matched or mismatched at one HLA locus), (5) received peripheral blood hematopoietic stem cells (PBSC), (6) underwent the RIC conditioning regimen. Patients who received a previous allo-HSCT or T-depleted grafts were excluded. Indication for RIC allo-SCT depended on each center's policy. The RIC regimen was defined as the use of fludarabine associated with intermediate doses of intravenous busulfan (FLUBU, busulfan at 6.4 mg/kg), intermediate doses of melphalan (FLUMEL, melphalan at 140 mg/m 2 ), or low-dose TBI (FLUTBI; TBI at 2 Gy).

Endpoints and definitions
The primary endpoint was overall survival (OS). Secondary endpoints were cumulative incidences (CI) of relapse, TRM, acute and chronic graft-versus-host disease (GVHD), LFS and GVHD free, relapse-free survival (GRFS). Acute and chronic GVHD were graded according to previously published criteria [13,14]. OS was defined as the probability of survival, TRM as death without evidence of relapse, LFS as survival with no evidence of relapse or disease progression. GRFS was defined as survival with no previous grades III-IV acute GVHD, no severe chronic GVHD and no relapse.

Statistical analysis
The main patient characteristics were compared using the Mann-Whitney test for quantitative variables and chisquare test or Fisher's exact test for categorical variables. Probabilities of OS, LFS, and GRFS were estimated using the Kaplan-Meier method, and the differences between groups were compared using the log-rank test. GVHD, relapse, and TRM were calculated using the CI method and analyzed in a time-dependent fashion. Differences between groups were compared using the Gray's test. For acute and chronic GVHD or relapse, death of the patient was considered as a competing risk of the event. For TRM, the competing event was relapse. Factors differing between the groups in terms of distribution and factors significantly associated with the outcome were included in the multivariate analysis. Multivariate analyses were performed using the Cox proportional-hazard model. All tests were two-sided and P values < 0.05 were considered as indicating a statistically significant association. Analyses were performed using the R statistical software version 3.2.3 (available online at http://www.R-project.org).

Patient characteristics
A total of 417 patients were included in this study; 127 patients in the FLUBU group, 190 patients in the FLUMEL group, and 100 patients in the FLUTBI group. Patient characteristics of each group are summarized in Table 1. The median follow-up of patients was significantly longer (p = 0.001) in the FLUTBI group (51 months, range 34-69) than in the FLUBU group (35 months, range,  and FLUMEL group (23 months, range, 20-26). Patients in the FLUBU group were significantly older (median 59 years, range 45-71) than patients in the FLUMEL (median 54 years, range 45-74) and the FLUTBI (median 57 years, range 45-72) groups, (p = 0.001). Incidence of Ph+ ALL was lower in the FLUMEL group compared with FLUBU or FLUTBI groups (52% vs 69%, p < 0.001). Most patients in the FLUBU group received ATG (88%), while most of the FLUMEL patients received Campath (71%) as GVHD prophylaxis. Only 12% of the patients received in vivo Tcell depletion in the FLUTBI group (11 ATG and 1 Campath). The rest of the demographic and transplant characteristics were comparable between the three groups.

Discussion
To our knowledge, this is the first study comparing outcomes of the most used RIC conditioning regimens in adults with ALL. We compared RIC allo-HSCT after FLUBU, FLUMEL, and FLUTBI conditioning in 417 patients with ALL in CR1 and found similar transplantation outcomes in terms of OS, LFS, and relapse. However, lack of in vivo T-cell depletion with the FLUTBI regimen yielded more cGVHD and a lower GRFS, while FLUMEL emerged as an independent predictor of TRM in the multivariate analysis.
One criticism of RIC regimens is that many of them do not include TBI, which is thought to reduce the risk of CNS relapse in ALL [26]. This finding is mostly based on MAC and RIC comparisons, where TBI is usually added to MAC regimens [16,26]. Moreover, a recent large Centre for International Blood and Marrow Transplant Research (CIBMTR) study comparing myeloablative TBI-and busulfan-based regimens confirmed a protective role of TBI for relapse in a multivariate analysis [27]. Furthermore, a multicentric study coordinated by the Fred Hutchinson Cancer Research Center evaluated an FLUTBI RIC regimen in patients older than 50 years, with comorbidities or prior transplantation and found a remarkable 3-year OS of 62% for patients in CR1 with relapse ranging from 15 to 32% depending of the Ph+ status [20]. This contrasts with our study where the addition of TBI did not provide better antileukemic control since there was no significant difference in relapse incidence between the FLUTBI group in comparison with FLUBU and FLUMEL groups (41% vs 40% vs 36%, p = 0.21). However, the low dose of TBI used in this study (2 Gy) may have been insufficient to protect against CNS relapse and also we have previously shown that there is wide variation in TBI delivery among the centers which leads to potential obstacles when analyzing TBI data [28,29].
PBSC is a common source of stem cells in RIC allo-HSCT and all patients in our study received PBSC. Previous data comparing BM and PBSC in ALL RIC patients are lacking and the only data available are from the AML setting or from analysis of AML and ALL together, with contradictory results. A large CIBMTR study in AML patients found no differences between BM and PBSC outcomes in RIC allo-HSCT [30]. On the contrary, a previous EBMT study of RIC-allo HSCT in AML and patients with ALL, found higher OS, LFS, and relapse incidence but at the expanse of more chronic GVHD after the use of PBSC compared with BM [31]. In our study, the only significant difference between RIC regimens was found in the incidence of chronic GVHD (significantly higher in the FLUTBI compared with FLUBU and FLUMEL group; (39% vs 16% vs 12%, p = 0.001). This led to a significantly lower GRFS in the FLUTBI group but the difference was lost on multivariate analysis when adjusted for the use of ATG or Campath, traditionally used in the FLUBU and FLUMEL conditioning. Most of the patients in our study who received the FLUTBI regimen (88%) did not receive ATG or Campath, and this highlights the importance of in vivo T-cell depletion in RIC regimens, particularly when PBSCs are used. It is generally accepted that old age itself is not a contraindication for RIC allo-HSCT in patients with good performance status. However, large registry studies have shown that, when stratified by age, patients older than 66 years have higher rates of TRM and decreased OS [32]. Of course, the older population also has a worse performance status and more comorbidities which makes it difficult to discern whether age or performance status contribute more to poorer outcomes. Nevertheless, in our study increasing age emerged as the main risk factor for worse outcomes; it independently predicted higher rates of TRM and relapse and lower OS, LFS, and GRFS. Therefore, our results support the finding that in older adults, age may still modify the impact of poor performance status, and transplant, even with RIC, should be undertaken with caution.
Despite comparable outcomes between RIC regimens, the outcomes reported in our study are still unsatisfactory, with comparable LFS of <50% in all groups (43% in FLUBU vs 42% in FLUMEL vs 45% in FLUTBI, p = 0.99). This highlights the importance of developing strategies for preventing relapse after allo-HSCT. Minimal residual disease (MRD) has been shown to be the strongest predictor of outcome after allo-HSCT [33][34][35][36][37]. Strategies to improve allo-HSCT outcome in MRD-positive patients include pretransplant elimination of MRD with potent new drugs such as blinatumomab [38], pre-transplant adjustment of ATG doses based on lymphocyte counts [39], as well as posttransplant pre-emptive donor lymphocyte infusion (DLI) [40]. A step further is the prevention of relapse in MRD-negative high-risk patients and includes tyrosine kinase inhibitor maintenance therapy in Ph-positive [41][42][43], or prophylactic DLI in Ph-negative patients. In relapsed patients, major improvements have been made with bispecific and drug-conjugated antibodies (blinatumomab and inotuzumab ozogamicin), while exciting new strategies include genetically engineered T-lymphocytesthe chimeric antigen receptor T cells [44][45][46].
Our analysis has some limitations, mainly due to its retrospective design and some significant differences between populations' characteristics. Furthermore, it was neither possible to provide the details of comorbidities nor further information on MRD in patients before transplant, which could have affected transplant outcomes. Nevertheless, this is the largest study of patients with ALL receiving RIC allo-HSCT reported so far, leading to some important conclusions.
In summary, the three most popular RIC preparative regimens (FLUBU, FLUMEL, and FLUTBI) yield similar transplantation outcomes in adults with ALL. However, FLUMEL conditioning seems to be associated with higher transplant-related toxicity, while more chronic GVHD in the FLUTBI group is mainly related to the low use of in vivo T-cell depletion.
Author contributions ZP assembled and analyzed data and wrote the first version of the manuscript. AN, SG, and MM designed the study, supervised research, analyzed data, and helped with writing the manuscript. EP, ML, and CP assembled the data, performed statistical analysis, and commented on the manuscript. All other co-authors collected data, recruited patients, and helped with writing the manuscript. All authors approved submission of the manuscript for publication.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.