An Alternative Donor Is a Valid Option Compared to a Matched-Unrelated in Allogeneic Stem Cell Transplantation for Acute Lymphoblastic Leukemia in CR2: A Report of 841 Patients from the EBMT Acute Leukemia Working Party

Background: In acute lymphoblastic leukemia (ALL) in CR1, the indications of allogeneic stem cell transplantation (allo-SCT) remain under debate, given the recent availability of new therapeutic tools including novel chemotherapy agents, monoclonal antibodies and T cell cellular therapies as well as sensitive molecular techniques to assess minimal residual disease (MRD). In contrast, allo-SCT for ALL in CR2 is a well-established procedure. At the same time, unmanipulated grafts are increasingly used in the haplo-setting, and innovative regimens for GvHD prophylaxis have been adopted with encouraging results.

Aims :

The current study aimed to compare the outcomes of ALL patients (pts) who do not have a matched sibling and received allo-SCT in CR2 from a matched-unrelated donor 10/10 (MUD) versus alternative donors: unrelated donor 9/10 (UD), cord blood unit (CBU) and haplo-identical T replete donor (HD).

Methods:

Patients with ALL in CR2 who underwent a first allo-SCT with a MUD or an UD, a CBU or HD, reported between 2005 and 2015 to the registry of the EBMT ALWP, were included.

The major endpoints were to assess overall survival (OS), leukemia-free survival (LFS), relapse incidence (RI), non-relapse mortality (NRM), and graft-versus-host disease-free, relapse-free survival (GRFS)

Results:

Four hundred twenty eight pts received a MUD 10/10, 171 pts an UD 9/10, 148 pts a CBU and 94 a T replete HD. Median age was lower in the CBU group compared to the other groups (26.1 yrs (18-76) vs 31.6 (18-74) , p<10^-4). There was a higher proportion of patients who underwent allo-SCT within 18 months following diagnosis in the HD group :42.5% vs 32.5%, p=0.05 (36.9% in MUD 10/10, 28% in UD 9/10, and 24% in CBU groups, respectively). The median time to relapse to transplant was similar in the 4 groups (p=0.68). The number of Philadelphia chromosome positive ALL pts and of positive MRD at transplant were comparable in the 4 groups (p=0.60 and p=0.74, respectively).

At 2 years, for the overall population, LFS was 36.9% (95% CI: 33.2-40.6), OS 42.6% (95% CI: 38.8-46.5), RI 32.4% (95% CI: 28.9-35.9), NRM 30.7% (95%CI: 27.4 -34.1), and GRFS 30% (95%CI: 26-33). As to the outcomes, in univariate analysis, no differences were found between the 4 groups. In multivariate analysis, 3 predictive factors were associated with an improved LFS: the time from diagnosis to transplant >18months (DxTx>18 m)(HR=0.66, 95%CI, 0.54-0.81, p<10^-4), a Karnofsky score at transplant ≥90% (KS) (HR=0.77, 95%CI, 0.62-0.97, p=0.02),and the year of transplant (HR=0.95, 95%CI: 0.91-0.99, p=0.02), whereas age was associated with a lower LFS (HR=1.103 per 10 years, 95%CI:1.023-1.19, p=0.01).For OS, HD compared to MUD 10/10, patient age, and CMV positivity were associated with lower OS (HR=1.42, 95%CI 1.02-1.99, p=0.04; HR=1.10, 95%CI 1.02-1.19, p=0.018; and HR=1.27, 95%CI 1.02-1.58, p=0.03; respectively), whereas 3 predictive factors were associated with better OS: DxTx>18m, KS ≥90% and the year of transplant. In multivariate analysis for RI, only DxTx>18m was a protective factor (HR=0.49, 95%CI, 0.38-0.65, p<10^-4). In multivariate analysis for GVHD II-IV, a HD was associated with a higher risk compared to MUD 10/10 (HR=0.028, 95%CI: 1.05-2.30, p=0.03). At 2 years, in univariate analysis, the cumulative incidence of chronic GVHD was not statistically different between the 4 groups and no risk factors were identified in multivariate analysis. In multivariate analysis for NRM, patient age and CMV positivity were factors associated with a higher NRM (HR=1.14, 95%CI, 1.03-1.27, p=0.015; HR=1.50, 95%CI, 1.09-2.05, p=0.011), while year of transplant was the only factor associated with lower NRM (HR=0.94, 95%CI, 0.88-0.99, p=0.027). Finally, considering GRFS at 2 years, DxTx>18m and the year of transplant were associated with better GRFS (HR=0.69, 95%CI, 0.57-0.84, p<10^-3 and HR=0.96, 95%CI, 0.92-0.99, p=0.04).

Conclusion

Allo-SCT may rescue more than one third of pts with ALL in CR2. Importantly, once a matched sibling donor is not available the donor type being unrelated or alternative did not have any impact on patients' LFS, RI, NRM and GRFS. When compared to a MUD 10/10, an T replete HD was associated with lower OS and higher aGVHD II-IV. Interestingly, DxTx>18m was a major prognostic factor for both LFS and RI. Therefore, as compared to a MUD 10/10, a UD 9/10, a CBU, but also a T replete HD could be considered as valid options in adult pts with ALL in CR2.