Pre-Transplant Risk-Assessment for Patients with Acute Myeloid Leukemia (AML) Receiving Allogeneic Hematopoietic Stem Cell Transplants.

The presence of comorbidities was shown to have a substantial impact of the outcome of patients undergoing cancer treatment. In chronic myeloid leukemia (CML) the EBMT risk score proofed to be highly valuable in predicting the outcome of patients following allogeneic stem cell transplantation (alloSCT). We therefore investigated, whether a slightly modified EBMT risk score may be used to predict the outcome of patients with AML following alloSCT.

We retrospectively analyzed 233 patients with AML (median age: 47 years, range: 17 – 68 years) who underwent alloSCT at our institution between 1994 and 2007. 180 patients (77%) had de novo AML and 53 patients (23%) had secondary or therapy-related AML. A favorable karyotype was present in 11 patients (5%) whereas 101 (43%) or 82 patients (35%) had an intermediate risk or a poor risk karyotype. 131 patients (56%) received myeloablative conditioning (MAC) whereas 102 patients (44%) were conditioned using reduced intensity conditioning (RIC). The EBMT risk score was calculated analogous to the original score established by Gratwohl et al. (Lancet 352, 1998) and included the following variables: age (<20, 20-40 or >40 years), interval from diagnosis to alloSCT (<1 year or ≥1 year), disease stage (CR1, >CR1 or no CR), donor/recipient gender match (female donor/male recipient or other), and donor type (HLA-identical sibling or other). Altogether, 6 patients (3%) were younger than 20 years, 82 patients (35%) were between 20-40 years, and 145 patients (62%) were older than 40 years. The interval from diagnosis to alloSCT was <1 year in 180 patients (77%) and ≥1 year in 53 patients (23%). 121 patients (52%) were transplanted in CR1, 41 patients (18%) underwent alloSCT >CR1, and 71 patients (30%) had active (relapsed/refractory) disease at the time of alloSCT. A female donor/male recipient transplantation was performed in 59 patients (25%). Transplants were from a matched-related donor (MRD) in 103 patients (44%). A matched-unrelated or a mismatched-unrelated donor was chosen in 101 (44%) or in 28 patients (12%). Only 1 patient was transplanted from a haplo-identical donor.

After a median follow-up of 48 months (range: 6 – 170 months) for the surviving patients, 108 patients (46%) are alive, 101 (43%) of which are in continuous CR. Causes of death (total 125 patients (54%)) were relapse in 70 patients (30%), infections/graft versus host-disease in 54 patients (23%), or other (1 patient (0.5%)). At 10 years after alloSCT, projected overall survival (OS) or disease-free survival (DFS) were 41% or 39%. Non-relapse mortality (NRM) or incidence of relapse were 32% or 43%. Of the 233 patients, 30 (13%) had an EBMT risk score of 0-1, 48 (21%) had a score of 2, 50 (21%) had a score of 3, 40 (17%) had a score of 4, 51 (22%) had a score of 5, and 14 (6%) had a score of 6-7. OS in the different score groups were 67% (score 0-1), 50% (score 2), 48% (score 3), 33% (score 4), 23% (score 5), or 21% (score 6-7) and differed significantly between the groups (p=0.0005). NRM in patients with an EBMT risk score >4 as an abritary cut-off was significantly higher as compared to patients with a score ≤4 (53% versus 25%; p=0.009). Likewise, the incidence of relapse was significantly lower in patients with an EBMT risk score ≤3 as an abritary cut-off when compared those with a score >3 (31% versus 57%; p<0.0001). In univariate analysis, disease stage had a negative prognostic impact on OS (CR1: 54%, >CR1: 32%, no CR: 25%; p<0.0001). Likewise, OS in patients with a MRD was significantly higher as compared to patients with other donor types (MRD: 50%, other: 34%; p= 0.003). In contrast, age, interval from transplantation to alloSCT, and donor/recipient gender match did not influence OS in our analysis.

These data indicate that a modified EBMT risk score may be used to predict the outcome of patients with AML following alloSCT.

No relevant conflicts of interest to declare.