Variable Outcome of Allogeneic Stem Cell Transplant According to the Different Levels of Pre-Transplant Minimal Residual Disease, in Adult Patients with Acute Myeloid Leukemia

Despite advances in treatment and supportive therapies, outcome of acute myeloid leukemia (AML) remains dismal with approximately 40% of younger and less than 20% of elderly patients becoming long-term survivors. Stem cell transplantation (SCT) is a well-established post-remission therapeutic option, and in both its variants, autologous (AuSCT) and allogeneic (ASCT), it is often incorporated in modern treatment programs for it may reliably improve long-term prognosis. Ten years ago we demonstrated that the levels of minimal residual disease (MRD) before autologous stem cell transplant (AuSCT), assessed at the post-consolidation time point by multiparametric flow cytometry (MFC), affected outcome (1). Moreover, we have preliminarily observed that in MRD positive (MRD^pos) patients, allogeneic stem cell transplant (ASCT) attenuates the negative prognostic impact of pre-transplant MRD positivity by conferring a significant survival advantage in terms of either overall (OS) or disease free survival (DFS) (2). At variance with this, others have shown that even in the setting of ASCT, pre-transplant MRD positivity is associated with a poor prognosis regardless of the graft-versus-leukemia (GVL) effect (3). The aim of the present analysis was to evaluate, in 81 MRD^pos patients submitted to ASCT (45) or AuSCT (36), the impact on clinical outcome of different MRD levels. As previously reported, counting 3.5x10^-4 (0.035%) residual leukemic cells (RLCs) or more in the bone marrow (BM) upon full hematological recovery after consolidation cycle, was regarded as a condition of MRD positivity. Patients with or above 3.5x10^-4 RLC were arbitrarily divided into 3 different cohorts: 1) ≥0.035%≤0.1% (13 patients, 6 ASCT and 7 AuSCT); 2) >0.1%≤1% (52 patients, 31 ASCT and 21 AuSCT); 3) >1% (16 patients, 8 ASCT and 8 AuSCT). In the category no. 2, ASCT gave a significant 5-years OS (64,9% vs 17,9%, p<0.001) and DFS (66,5% vs <10%, p<0.001) advantage. In the category no. 3, even if the low numbers suggest caution in drawing any conclusions, the patients showed a similar poor outcome both after AuSCT or ASCT. In multivariable analysis, backward, forward and stepwise regressions confirmed the role of ASCT vs. AuSCT (OS: HR 0.53, SDF 95% 0.30-0.95, p=0.039; DFS: HR 0.40, SDF 95% 0.23-0.70, p=0.001) and MRD^pos vs MRD^neg status (OS: HR 1.89, SDF 95% 1.03-3.46, p=0.032; DFS: HR 2.20, SDF 95% 1.24-3.93, p=0.007). These preliminary results indicate that in pre-transplant MRD positive patients the log-term outcome of ASCT may be different according to the amount of RLC. In low tumor burden (< 1%) MRD positive patients, conditioning regimens and GVL effect can still offer a chance of cure similarly to MRD negative patients. In high tumor burden (>1%) MRD positive patients, ASCT does not guarantee a long-term control of leukemia and further reduction of relapse risk is expected to be facilitated by novel strategies.