Autologous Versus Related Versus Unrelated Hematopoietic Stem Cell Transplantation for Adult Patients with NPM1 Mutated/FLT3-ITD Negative Acute Myeloid Leukemia in First Remission: A Study from the Acute Leukemia Working Party (ALWP) of the European Society of Blood and Marrow Transplantation (EBMT)

Introduction: Acute myeloid leukemia (AML) in first remission (CR1) with intermediate-risk cytogenetics and isolated NPM1 mutation carries a good prognosis. There is not a clear benefit to perform allogeneic stem cell transplantation (allo-SCT) in CR1, and this procedure is not usually recommended in these patients. High dose chemotherapy followed by autologous stem cell transplantation (auto-SCT) is a good therapeutic alternative in good-risk AML in CR1 associated with lower toxicity. Nevertheless, relapse remains as high as 40% in this subgroup. To evaluate the best consolidation strategy in AML with isolated NPM1 mutation, we retrospectively compared auto-SCT, related (MRD) and fully matched (10/10) unrelated (MUD) allo-SCT within the EBMT registry.

Methods: We selected de novo adult AML with intermediate-risk cytogenetics harbouring an isolated NPM1 mutation without FLT3-ITD and transplanted in CR1 between 2005 and 2015.

Results: Two hundred and fifty-six patients have been allocated. One hundred and twenty-five patients received an auto-SCT, 72 a MRD and 59 a MUD. Median age at SCT was 52.5-year-old (range, 18.9-77) and median follow-up was 30 months (range, 1.6-109.7). A normal karyotype was found in 86% of auto-SCT and 74% of allo-SCT (p=0.02). Molecular status was available in 55% of patients, with a higher proportion of complete molecular remission (CMR) before auto-SCT than before allo-SCT (84% versus 63%, respectively, p=0.005). There was no difference in the number of induction courses to reach CR1 between groups. The majority of patients have a good performance status at the time of SCT. Peripheral blood was more frequently used as stem cell source in auto-SCT compared to allo-SCT (97% versus 77%, respectively, p<0.01). Total body irradiation was administered in 23% of allo-SCT and in 3% of auto-SCT (p<0.01). Among allo-SCT, a myeloablative conditioning regimen was administered in 57% of the patients. In vivo T-cell depletion was more frequently in MUD than in MRD (70 versus 40%, respectively, p=0.001).

The 2-year leukemia-free survival (LFS) was 62% in auto-SCT, 69% in MUD and 81% in MRD (p=0.02 for MRD versus auto-SCT or MUD). The 2-year overall survival (OS) was not significantly different among auto-SCT, MUD and MRD, reaching 82%, 81% and 85%, respectively (p=0.88). The 2-year non-relapse mortality (NRM) was 2.5% in auto-SCT and 7.5% in allo-SCT (p=0.04), without any difference between MRD and MUD. The 2-year cumulative incidence of relapse (RI) was significantly higher after auto-SCT (30%) than after MUD (22%) and MRD (12%, p=0.01). Among patients transplanted in CMR, RI was higher after auto-SCT and LFS was superior after allo-SCT (RI: 21 vs. 11 vs 4%; LFS: 61 vs 74 vs. 91 after auto-SCT, MUD, and MRD, respectively, p=0.04), although this difference did not translate into a different OS (82 vs. 77 vs. 90 after auto-SCT, MUD, and MRD, respectively). In the 28 patients not transplanted in CMR who received an allo-SCT, 2-year RI, LFS and OS were 33% (15-53), 63 (43-83), and 81 (65 - 96%), whereas only 10 patients with persistant MRD received an auto-SCT precluding firm conclusions.

In multivariate analysis, MRD versus auto-SCT but not MUD versus auto-SCT was significantly associated with a lower RI (p<0.01 and p=0.13, respectively) and a better LFS (p=0.01 and p=0.31, respectively). Age correlated significantly with higher NRM (p<0.01). Donor type, and age were not significantly associated with OS.

Conclusion: Allo-SCT using a sibling donor appears as a good consolidation therapy to prevent relapse for young patients with AML and isolated NPM1 mutation in CR1. Auto-SCT was followed by higher relapse risk and inferior LFS, but similar OS to both allo-SCT modalities. In patients with molecular persistence, allo-SCT was followed by a relatively favourable outcome, but the low number of patients with molecular data included in this series precludes more firm conclusions. The strategy for patients not transplanted in CMR remains to be further evaluated.