Non-myeloablative conditioning prior to allogeneic stem cell transplantation in patients with acute myeloid leukemia

In most patients with acute myeloid leukemia (AML), the only curative option is allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, patients with AML are frequently elderly with comorbidities which limits their eligibility for myeloablative conditioning (MAC) followed by allo-HSCT. The use of non-myeloablative (NMA) conditioning has enabled patients, who might otherwise have been deemed ineligible, to access allo-HSCT. NMA conditioning relies more on the graft-versus-leukemia response than on the anti-leukemic effect of the conditioning regimen. The two main NMA options are low dose chemotherapy (fludarabine/busulfan; FB2) or low dose radiotherapy (fludarabine/total-body-irradiation; FluTBI2Gy).

Thomas Heinicke of the Department of Hematology and Oncology, Otto-von-Guericke University, Magdeburg, DE, and colleagues recently published the results of their retrospective analysis of outcomes following the two NMA conditioning regimens prior to allo-HSCT for AML in Bone Marrow Transplantation. They performed a registry analysis using the European Society for Blood and Marrow Transplantation (EBMT) central database to compare outcomes in patients who had undergone allo-HSCT (from matched sibling donors [MSD] or matched unrelated donors [MUD]) for AML treatment during their first complete remission (CR1) using NMA conditioning with either FB2 or FluTBI2Gy.

Study design and patient characteristics:

• FB2 conditioning: fludarabine (30mg/m²) for five consecutive days, combined with eight doses of intravenous busulfan (0.8mg/kg body weight), with (79.8% of FB2 patients) or without (20.8%) in vivo T-cell depletion (TCD). 64% of patients who received MSD transplantation received TCD, and in patients who received MUD transplantation 94% received TCD
• FluTBI2Gy conditioning: fludarabine (30 mg/m²) for three consecutive days, combined with total body irradiation (TBI) at a single dose of 200 cGy. The use of TCD was not allowed
• Endpoints: the primary outcome was leukemia-free survival (LFS), though the study also captured relapse incidence (RI), non-relapse mortality (NRM), overall survival (OS), refined graft-versus-host disease-free relapse-free survival (GRFS), neutrophil engraftment, and graft-versus-host-disease (GvHD; both acute [aGvHD] and chronic [cGvHD])
• Patient characteristics: There were some relevant differences in patient characteristics between the two treatment groups, such as median follow up (35.7 vs 26.7 months in FluTBI2Gy and FB2 respectively) and in the number of patients having a Karnofsky performance score of ≥ 90% (67.09% vs 59.44% in FB2 and FluTBI2Gy respectively) (Table 1)

Table 1: Patient and disease characteristics

IQR; interquartile range, secAML; secondary AML, MSD; matched sibling donor, FB2 fludarabine/busulfan; FluTBI2Gy fludarabine total-body-irradiation (2Gy), MUD; matched unrelated donor, GvHD; graft-versus-host-disease, CSA; cyclosporine, MTX; methotrexate, MMF; mycophenolate mofetil, TACRO; tacrolimus, TCD in vivo T-cell depletion. * p values from comparisons of missing/incomplete data

 

 

FB2

(n= 553)

FluTBI2Gy

(n= 535)

p value

Follow up (reverse Kaplan-Meier; months)

Median (IQR)

26.7 (12.3–51.1)

47.7 (20.2–74.5)

< 0.001

Age at HSCT (years)

Median (IQR)

64.8 (62.2–67.2)

65.3 (62.7–68.2)

0.0023

Time between diagnosis and HSCT (months)

Median (IQR)

5.4 (4.4–6.9)

4.6 (3.6–5.8)

< 0.0001

Year of treatment

Median (IQR)

2014 (2005–2017)

2012 (2003–2017)

< 0.0001

Donor age (years)

 

Median (IQR)

Missing

47.8 (30.7–59.3)

103 (18.62%)

49.4 (29.7–61.7)

206 (38.5%)

0.44

< 0.0001*

Diagnosis

De novo

secAML

408 (73.78%)

145 (26.22%)

417 (77.94%)

118 (22.06%)

0.11

Cytogenetics

Good

Intermediate

Poor

NA/failed

12 (2.17%)

291 (52.62%)

119 (21.52%)

131 (23.69%)

7 (1.31%)

223 (41.68%)

78 (14.58%)

227 (42.43%)

0.59

 

 

< 0.0001*

Karnofsky performance score

< 90

≥ 90

missing

130 (23.51%)

371 (67.09%)

52 (9.4%)

161 (30.09%)

318 (59.44%)

56 (10.47%)

0.01

 

0.56*

Donor type

MSD

MUD 10/10

270 (48.82%)

283 (51.18%)

240 (44.86%)

295 (55.14%)

0.19

GvHD prevention

CSA

CSA + MTX

CSA + MMF +/- MTX

TACRO +/- other

Other

161 (29.11%)

207 (37.43%)

155 (28.03%)

17 (3.07%)

13 (2.35%)

7 (1.31%)

6 (1.12%)

405 (75.7%)

112 (20.93%)

5 (0.93%)

< 0.0001

TCD

No

Yes

115 (20.8%)

438 (79.2%)

535 (100%)

0

< 0.0001

Key findings:

• Analysis of the role of TCD in the group of patients who received MSD transplantation found no differences in outcome during univariate analysis except for a lower incidence of cGvHD and extensive cGvHD and a higher probability of GRFS in the FB2 group who had received TCD (Table 2)
• For patients who received MUD transplantation, there was no comparison of patients with and without TCD as there were very few in the FB2 arm and none in the FluTBI2Gy arm. Differences between the two treatment arms were only observed in cGvHD, extensive cGvHD and GRFS (Table 3) with FB2 conditioning demonstrating a reduced incidence of cGvHD and extensive cGvHD, and an increased probability of GRFS

Table 2: Univariate analysis of incidence of GvHD and probability of GRFS in MSD transplanted patients

95% CI; 95% confidence interval, MSD; matched sibling donor, cGvHD; chronic graft-versus-host-disease, GRFS; refined graft-versus-host disease-free relapse-free survival, TCD in vivo T-cell depletion, FB2; fludarabine/busulfan, FluTBI2Gy; fludarabine total-body-irradiation (2Gy)

 

FluTBI2Gy, No TCD

FB2, TCD

FB2, No TCD

p value

cGvHD, incidence

43.6% (95% CI: 36.8−50.3)

30.5% (95% CI: 22.9−38.4)

43.4% (95% CI: 30.8−55.3)

0.01

Extensive cGvHD, incidence

25.1% (95% CI: 19.3−31.3)

13.7% (95% CI: 8.4−20.4)

24.6% (95% CI: 14.9−35.5)

0.007

GRFS

31.4% (95% CI: 25.2−37.6)

35.2% (95% CI: 27−43.4)

25.8% (95% CI: 15.6−36)

0.07

Table 3: Univariate analysis of incidence of GvHD and probability of GRFS in MUD transplanted patients

95% CI; 95% confidence interval, MSD; matched sibling donor, cGvHD; chronic graft-versus-host-disease, GRFS; refined graft-versus-host disease-free relapse-free survival, TCD in vivo T-cell depletion, FB2; fludarabine/busulfan, FluTBI2Gy; fludarabine total-body-irradiation (2Gy)

 

FluTBI2Gy

FB2

p value

cGvHD, incidence

56.6% (95% CI: 49.8–62.9)

32.7% (95% CI: 26.6–38.9)

< 0.0001

Extensive cGvHD, incidence

34.2% (95% CI: 27.9−40.6)

11.9% (95% CI: 7.9–16.7)

< 0.0001

GRFS

26% (95% CI: 20.4−31.6)

42.4% (95% CI: 35.9−48.9)

< 0.001

• Multivariate analysis of the MSD group found that NRM was higher (HR 2.14; 95% CI, 1.04–4.38; p = 0.04 for patients treated with FB2 without TCD when compared with FluTBI2Gy without TCD
• For patients treated with MUD transplantation, there were significant lower rates of cGvHD, extensive cGvHD and GRFS based on univariate analysis (Table 3). These findings were supported by multivariate analysis; HR of FluTBI2Gy vs FB2 was 2.44 (95% CI, 1.60−3.72; p < 0.0001) for cGvHD, 4.59 (95% CI, 2.35−8.96; p < 0.00001) for extensive cGvHD, and 1.35 (95% CI, 1.02–1.79; p = 0.03) for GRFS

Thomas Heinicke and colleagues discussed how results from this study differed from the phase II trial by Didier Blaise² in terms of aGvHD and cGvHD incidence, relapse rates and NRM, but felt that it was due to differences in patient selection, busulfan formulation, post grafting immunosuppression, and donors. They argued that the lower incidence of cGvHD and extensive cGvHD seen in the FB2 treatment group may be effected by the difference in TCD between the two treatment groups (79% of the FB2 group received TCD, but none in the FluTBI2Gy had TCD), and that the use of TCD in the MSD group treated with FB2 decreased rates of cGvHD, which was in agreement with previous studies.² The authors went on to discuss the protective effect that TCD seemed to offer against cGvHD, which had also been seen in other studies^4-7 with comparable rates of aGvHD and cGvHD.⁸

The study, being retrospective, has limitations due to the unknown reason behind patients being given each treatment regimens, unknown type and dose of ATG (for TCD), and missing data on residual disease. Despite the limitations, Heinicke and colleagues feel the results were important to the allo-HSCT field due to the large homogenous study cohort. They concluded by highlighting that in this group of patients, those receiving transplants from MUDs with FB2 conditioning and TCD had a lower incidence of cGvHD, extensive cGvHD, and improved GRFS when compared to FluTBI2Gy without TCD.