In acute myeloid leukemia (AML), karyotype at diagnosis is one of the most powerful independent prognostic factors after allogeneic stem cell transplantation (Allo-SCT) but this was based on the data of patients mostly in remission. There has been little consensus about the impact of karyotypic abnormalities on All-SCT in non-remission, especially monosomal karyotype (MK) and complex karyotype (CK).


We retrospectively analyzed the outcome of 515 consecutive AML patients not in remission who underwent Allo-SCT for the first time at Toranomon Hospital between January 2008 and 2018. Patients with therapy-related AML (n=32) and transformed AML from ET/PMF (n=15), with active infection at transplantation, in poor condition as ECOG PS of 3 or more (n=38), and lacked karyotype information at diagnosis (n=59) were excluded from this study. Patients in remission at transplantation (n=77) were also excluded. Cytogenetic abnormalities at diagnosis were categorized based on 2017 ELN risk stratification, but genetic abnormalities such as FLT3-ITD, NPM1 were not considered in this study, because of lack of genetic information in most of the patients.


Two hundred and ninety-four patients (n=197 male; n=97 female) were included in this study. The median age at transplantation was 60 years (range, 17-77), with a median HCT-CI score of 2 (0-7). Underlying diseases were AML-NOS in 91, AML with recurrent genetic abnormalities in 31, and AML-MRC in 172. Transplanted cell origins were cord blood in 252 (86%) patients, related bone marrow (BM) or peripheral blood (PB) in 22 (7%), and unrelated BM in 20 (7%). Patients were categorized into three groups according to the cytogenetic abnormalities at diagnosis: favorable (n= 17 (6%)), intermediate (n=192 (65%)), and adverse (n=85 (29%)).

With a median follow-up of 35 (range, 1-122) months, the 3-year probabilities of overall survival (OS), progression free survival (PFS), relapse rate (RR) and non-relapse mortality (NRM) for entire population were 40.3%, 36.3%, 31.7%, and 32.0%, respectively. Patients in adverse group showed a higher RR and a lower PFS compared with those in intermediate/favorable group (42.0% vs 27.7% in RR (P = 0.02), 29.1% vs 39.1% in PFS (P = 0.04), both of which were also confirmed in multivariate analysis.

Among adverse group (n=85), patients with both MK and CK (n=42) showed a higher RR, and a lower PFS compared with those without MK or CK (HR 2.29 (1.13-4.65), p=0.02 in RR, HR 1.91 (1.11-3.27), p=0.02 in PFS), in multivariate analysis. Among adverse group except for the patients with both MK and CK (n=43), 3-year PFS and RR were 44.5%, and 30.4%, which were comparable to intermediate group (35.7%, and 29.4%) (Fig. 1), indicating Allo-SCT overcomes negative impact of adverse cytogenetic abnormalities other than MK or CK in PFS and RR.

Among patients with both MK and CK, no factors were identified to have an impact on PFS. Of particular interest is that CBT was the only factor associated with decreased RR compared with other source (HR 0.41 (0.17-0.99), p=0.048) in multivariate analysis. Among those with both MK and CK receiving CBT in our cohort, the 3-year PFS, RR, and NRM were 16.1%, 45.9%, and 38.0%, respectively, which were almost comparable to previously reported outcome of those with both MK and CK in remission.


This retrospective study demonstrated that cytogenetic classification based on 2017 ELN risk stratification well predicted PFS, and RR for those in non-remission. The presence of both MK and CK is a poor prognostic factor on Allo-SCT in AML adverse group patients, while, for those in adverse group without MK or CK, Allo-SCT reduced relapse rate to the level of intermediate group.


Yamamoto:Bristol-Myers Squibb: Honoraria.

Author notes


Asterisk with author names denotes non-ASH members.