Introduction

Fludarabin/Busulfan (Flu/Bu) is a widely used reduced intensity conditioning (RIC) regimen for allogeneic stem cell transplantation (alloSCT) in AML. As an alternative, the sequential FLAMSA-TBI RIC regimen has been developed and has shown a substantial activity, particularly in relapsed/refractory AML patients (pts). Recently, FLAMSA-Bu with busulfan replacing TBI has also frequently been used. To evaluate their role, the acute leukemia working party (ALWP) of the EBMT conducted a retrospective study, comparing the outcomes of AML pts transplanted in CR1 or CR2 after conditioning with either Flu/Bu or one of the FLAMSA regimens.

Patients and Methods

Pts were eligible if they had received an alloSCT for AML in CR1 or CR2 between 1/2005 and 6/2016 after conditioning with either Flu/Bu (fludarabine 5x30 mg/m2, busulfan (8x0.8 mg/kg b.w. i.v.)), FLAMSA-TBI (fludarabine 4x30 mg/m2, cytarabine 4 x 2000 mg/m2, amsacrine 4x100 mg/m2)-TBI(4Gy), cyclophosphamide (2x40-60 mg/kg b.w.) or FLAMSA-Bu, in which TBI is replaced by busulfan (8x0.8 mg/kg b.w.). Data were extracted from the ALWP database. Primary endpoint was leukemia free survival (LFS) at two years. Overall survival (OS), relapse incidence (RI), non-relapse-mortality (NRM), graft-versus-host-disease (GVHD) free relapse free survival (GRFS), and rates of acute and chronic GVHD were secondary endpoints. To allow for potential confounding factors between treatment groups, propensity score matching was performed separately for the Flu/Bu group vs. FLAMSA-TBI and FLAMSA-Bu groups using the 'MatchIt' package. Factors included in the model were age per deciles, status at transplant, donor type, cytogenetics risk group, sex matching, pt and donor CMV serology and use of in vivo T cell depletion. Then, comparisons were performed using weighted Cox regression.

Results

1642 pts in the Flu/Bu group were compared to 174 and 318 in the FLAMSA-TBI and FLAMSA-Bu groups, respectively. Median follow-up of survivors was 20.1 months (0.6 - 129.0). Main differences between groups were: pts in the FLAMSA-Bu group suffered more often from secondary AML (25,7% vs.12.6% and 13.9% for Flu/Bu and FLAMSA-TBI, respectively, p<10-4). In addition, they were less often transplanted from a matched sibling donor (29.3 % vs. 46.7% vs. 38.1%, p<10-4) had more wild type NPM1 (79,1% vs. 60.5% vs. 46.2%. p=0.003) and had received more often in vivo T-cell depletion (98.9% vs. 83.8 % vs. 88.6%, p<10-4). Pts in the FLAMSA-TBI group were younger (median: 48y) than in both other groups (59.6y each p<10-4), and had more often a Karnofsky performance status >=90% (81.9%) as compared to the Flu/Bu and FLAMSA-Bu groups (74.0% and 75.8%, respectively, p=0.014).

In univariate analysis, LFS at two years was significantly higher in the FLAMSA-TBI group (60.3%) compared to FLAMSA-Bu (48.7%) and Flu/Bu (50.8%), p=0.003. Also, OS and GRFS were highest in the FLAMSA-TBI group (67% and 46% vs. 53.3% and 38.4% vs. 57.1% and 38.3%, respectively p=0.02 and p=0.02). NRM was higher in the FLAMSA-Bu group (26.6%) compared to FLAMSA-TBI (17.1%) and Flu/Bu (17.8%), p=0.02. RI was lower in both FLAMSA groups (FLAMSA-TBI 22.7%, FLAMSA-Bu 24.6%) compared to Flu/Bu (31.4%), p=10-4). Main cause of death in the FLAMSA-Bu group was infection (41.3 %). This rate was lower in the Flu/Bu and FLAMSA-TBI groups with 17.9% and 18.2%, respectively (p=10-5 and p=0.001).

Weighted Cox regression revealed that FLAMSA-Bu compared to Flu/Bu was associated with lower RI (HR 0.53 (95% CI 0.33-0.87), p=0.011), whereas NRM was higher (HR 1.88(95% CI 1.17-3.02), p=0.008). LFS, OS, GRFS, aGVHD II-IV, aGVHD III-IV, cGVHD and ext cGVHD were not different. Comparing FLAMSA-TBI with Flu/Bu there was a trend towards reduced relapse incidence for FLAMSA-TBI (HR=0.72 (95% CI 0.51-1.02), p=0.063), better LFS (HR=0.78 (95% CI 0.59-1.04), p=0.085) and GRFS (HR=0.8 (95% CI 0.62-1.03, p=0.081). All other outcome variables were not significantly different.

Conclusions

This retrospective analysis indicates that FLAMSA-Bu compared to Flu/Bu leads to substantial reduction in RI in AML pts transplanted in CR1 or CR2. Higher NRM, mainly due to infections counterbalance this positive effect. Therefore, LFS and OS are not different. Similarly, FLAMSA-TBI was associated with a trend towards reduced RI as compared to Flu/Bu. These data need confirmation in a prospective trial.

Disclosures

Schmid: MoilMed: Membership on an entity's Board of Directors or advisory committees; Incyte: Research Funding, Speakers Bureau; Novartis: Research Funding, Speakers Bureau; Celgene: Research Funding, Speakers Bureau; Jazz: Membership on an entity's Board of Directors or advisory committees. Scheid: Amgen: Consultancy, Honoraria; Takeda: Consultancy, Honoraria; BMS: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; Sanofi: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Celgene: Consultancy, Honoraria. Socié: Alexion Pharmaceuticals, Inc.: Consultancy. Savani: Jazz Pharmaceuticals: Speakers Bureau.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution