Fludarabine + melphalan vs fludarabine + busulfan: Effects on transplantation outcomes in older patients with MDS

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is currently the only curative treatment available for patients with myelodysplastic syndromes (MDS), and the efficacy of allo-HSCT is largely dependent on the intensity (and type) of pretransplantation conditioning regimen. Data from recent clinical trials are divided regarding conditioning regimens; however, in one study, better outcomes were seen in association with myeloablative conditioning (MAC), while in another, better outcomes were seen with reduced intensity conditioning (RIC). Similarly, conflicting results have been reported in the relapsed setting.¹

While MAC regimens may be the best option for certain patients with MDS, it is important to keep in mind that many MDS patients are older and, due to concerns regarding comorbidities and unfitness, most will be treated with RIC regimens. In a recent study published in Transplantation and Cellular Therapy, Betül Oran et al. compared the two most commonly used RIC regimens—fludarabine + intermediate-dose melphalan (FluMel) and fludarabine + intermediate-dose busulfan (FluBu)—to determine which regimen is most effective in older patients with MDS.¹

Study design

This was a retrospective cohort analysis of data reported to the Center for International Blood and Marrow Transplant Research (CIBMTR). Results reported to the CIBMTR between 2007 and 2016 were included for patients with MDS age ≥60 years who underwent human leukocyte antigen (HLA)-matched related or unrelated donor HSCT following a RIC regimen (defined as fludarabine with either a total melphalan dose ≤150 mg/m² or intravenous busulfan ≤7.2 mg/kg).

Baseline characteristics

A total of 1,045 patients were included in the analysis (FluMel, n = 448; FluBu; n = 597) and were categorized according to disease characteristics using the revised International Prognostic Scoring System (IPSS-R) and CIBMTR risk scoring. The median patient age was 67 years in each treatment group (Table 1).

Table 1. Selected patient characteristics by RIC regimen*

CIBMTR, Center for International Blood and Marrow Transplant Research; FluBu, fludarabine + busulfan; FluMel, fludarabine + melphalan; HLA, human leukocyte antigen; HSCT, hematopoietic stem cell transplantation; IPSS-R, revised International Prognostic Scoring System; KPS, Karnofsky performance status; MDS, myelodysplastic syndromes. *Adapted from Oran et al.1
Variable, % (unless otherwise stated)	FluBu (n = 597)	FluMel (n = 448)	p value
Age, years			0.005
60‒69	74	81
≥70	26	19
Median age (range), years	67 (60‒83)	67 (60‒77)	0.23
KPS			0.13
90‒100	15	13
<90	10	11
IPSS-R before HSCT			0.71
Very low	12	10
Low	25	24
Intermediate	31	30
High	14	17
Very high	6	6
CIBMTR MDS risk score			0.18
Intermediate	47	42
High	31	31
Very high	4	5
Donor type			0.08
HLA-identical sibling	26	31
Well-matched unrelated (8/8)	74	69
Graft type			0.002
Bone marrow	3	7
Peripheral blood	97	93
Median follow-up of survivors (range), months	38 (4‒101)	35 (6‒120)

No differences were observed between the groups regarding:

• Sex
• Therapy-related MDS
• IPSS-R and CIBMTR risk categories
• Hematopoietic cell transplantation-specific comorbidity index (HCT-CI)
• Cytomegalovirus (CMV) serostatus

While treatment with graft-versus-host disease (GvHD) prophylaxis was uniform between the groups, the regimen more commonly included antithymocyte globulin (ATG) and alemtuzumab in the FluBu group vs the FluMel group (39% vs 31%; p<0.001).

Results

OS at 3 years posttransplantation

On univariate analysis, estimates of overall survival (OS) at 1 year (FluBu, 61%; FluMel, 63%) and 2 years (FluBu, 49%; FluMel, 51%) posttransplantation were similar regardless of the conditioning regimen; a difference was observed at 3 years (FluBu, 39%; FluMel, 45%).

On multivariate analysis, FluMel was associated with a decrease in OS, but only in the first 3 months following transplantation; after this point, it was associated with improved OS (p<0.001). After adjusting for MDS risk group by CIBMTR and HCT-CI, FluMel was associated with improved OS compared with FluBu (46% vs 39%; p = 0.03) as shown in Table 2.

Table 2. Adjusted probabilities of transplantation outcomes by conditioning regimen*

aGvHD; acute GvHD; cGvHD, chronic GvHD; FluBu, fludarabine + busulfan; FluMel, fludarabine + melphalan; GRFS, GvHD relapse-free survival; GvHD, graft-versus-host disease. *Adapted from Oran et al.1
Outcome	FluBu regimen % (95% CI)	FluMel regimen % (95% CI)	p value
Overall survival, years
1	61 (57‒65)	63 (59‒68)	0.4
2	48 (44‒52)	52 (47‒57)	0.24
3	39 (34‒43)	46 (41‒51)	0.03
Disease-free survival, years
1	40 (36‒44)	48 (43‒52)	0.02
2	32 (29‒36)	40 (36‒45)	0.01
3	27 (23‒31)	35 (30‒40)	0.01
Relapse incidence, years
1	44 (40‒47)	26 (22‒30)	<0.0001
2	47 (43‒51)	28 (24‒33)	<0.0001
3	50 (46‒54)	32 (27‒37)	<0.0001
Treatment-related mortality, years
1	16 (13‒19)	26 (22‒30)	<0.0001
2	22 (19‒26)	33 (28‒38)	0.0005
3	28 (24‒33)	36 (31‒41)	0.03
aGvHD grade
II‒IV at 60 days	24 (21‒28)	35 (30‒39)	0.0003
II‒IV at 6 months	40 (36‒44)	44 (39‒48)	0.19
III‒IV at 60 days	9 (7‒11)	16 (13‒20)	0.0004
III‒IV at 6 months	17 (14‒20)	19 (15‒22)	0.48
cGvHD, years
1	42 (38‒46)	35 (30‒39)	0.02
2	49 (45‒53)	45 (40‒50)	0.18
3	49 (45‒53)	48 (43‒53)	0.75
GRFS
60 days	84 (81‒87)	76 (71‒80)	0.0005
6 months	40 (36‒44)	43 (38‒48)	0.37
1 year	14 (11‒17)	22 (18‒26)	0.0009
2 years	7 (5‒10)	14 (11‒18)	0.002
3 years	6 (4‒9)	10 (7‒13)	0.1

Regarding cause of death:

• The most common causes of death among the FluMel recipients who died (n = 238) were:
  • Relapse in 67 patients (28%)
  • GvHD in 50 patients (21%)
  • Infection in 46 patients (19%)
  • Organ failure in 32 patients (13%)
• The most common causes of death among the FluBu recipients who died (n = 368) were:
  • Relapse in 184 patients (50%)
  • GvHD in 57 patients (15%)
  • Infection in 45 patients (12%)
  • Organ failure in 27 patients (7%)

FluMel: Improved DFS in older patients with MDS

Disease-free survival (DFS) was lower in the FluBu cohort compared with the FluMel cohort, even after adjusting for CIBMTR risk category; multivariate analysis showed that the favorable impact of FluMel was only apparent after 3 months posttransplantation (p<0.001; Table 2).

Relapse and TRM

• FluMel was associated with a reduced risk of relapse after adjusting for CIBMTR MDS risk category, even in the high- and very-high risk groups (Table 2).
  • The favorable impact on risk of relapse persisted throughout the posttransplantation period.
• After adjusting for HCT-CI, donor type, and CIBMTR MDS risk group, treatment-related mortality (TRM) was higher in the FluMel cohort (1 year adjusted incidence of 26%) compared with the FluBu cohort (16%).
  •  This was limited to the first 3 months posttransplantation (p<0.0001); beyond 3 months posttransplantation, FluMel was not associated with an increased risk of TRM (p = 0.84).

GvHD-related outcomes

• After adjusting for donor type and HCT-CI, FluMel was associated with increased incidences of Grade II‒IV and Grade III‒IV acute GvHD (aGvHD) in the first 60 days posttransplantation (p<0.0001 and p<0.001, respectively) but the increased risk did not persist beyond 60 days.
• Similarly, GvHD relapse-free survival (GRFS) estimates were worse in the FluMel cohort in the first 60 days posttransplantation (p<0.0001) due to an increased incidence of aGvHD.
  • After 60 days, FluMel was associated with better GRFS compared with FluBu (p<0.0001).
• There were no differences in the incidence of chronic GvHD between the FluMel and FluBu groups after adjusting for other significant variables (p = 0.17).

Conclusion

In this comparison of two RIC regimens in older (≥60 years) patients with MDS, FluMel was associated with a significant improvement in DFS, and a decreased risk of relapse compared with FluBu. An improvement was also seen in OS, despite a higher incidence of TRM in the first 3 months posttransplantation. These improvements were observed across disease risk categories.

Despite comparison groups that were similar in most patient, disease, and transplantation characteristics (excepting the use of ATG or alemtuzumab for GvHD prophylaxis), this study had several limitations, including the inability to account for certain variables (reasons for selecting one regimen over the other, dosing details, and molecular information, in particular). Recent retrospective data have shown that pharmacokinetic-guided busulfan dosing may result in more standardized systemic exposure and better leukemia control, but patients receiving busulfan in this way were unable to be identified. Based on the improved outcomes seen in this study, however, the authors concluded that FluMel should be considered the RIC regimen of choice for older patients with MDS, barring any contraindication.