Hematopoietic Cell Transplantation-Specific Comorbidity Index (HCT-CI) Combined with Leukemic Risk Factors Predict Outcome of AML Patients after Allogeneic Stem Cell Transplantation Following Reduced Intensity Conditioning (RIC).

Purpose: The hematopoietic cell transplantation-specific comorbidity index (HCT-CI) is used to assess the impact of comorbidities on the outcome of patients following alloSCT. Nonetheless, in AML patients leukemia-associated risk factors, i.e. cytogenetics, the presence of secondary (sAML) or therapy-related AML (tAML), or disease status prior to alloSCT, are highly relevant with regards to the long-term outcome. We therefore investigated whether the HCT-CI combined with the analysis of a defined set of AML-specific high risk factors may be used to predict the overall outcome of AML patients after alloSCT following RIC.

Patients and Methods: 90 patients with high-risk AML (median age 51 (17 – 68) years) who underwent alloHSCT at our institution between 1999 and 2007 were retrospectively analyzed (median follow-up of the surviving patients: 16 (range 2 – 113) months). 50/90 patients (56%) were in CR1, 12/90 (13%) were in CR2, 2/90 (2%) were >CR2, and 26/90 (29%) had refractory or relapsed disease at the time of transplantation. 57/90 patients (63%) had de novo AML and 33/90 patients (37%) had either secondary AML (sAML) or therapy-related AML (tAML). Whereas 4/90 patients (4%) had a favorable risk karyotype, 51/90 patients (57%) or 29/90 patients (32%) had an intermediate risk or a poor risk karyotype as defined by the SWOG/ECOG criteria (Slovak et al., Blood 2000). Notably, 18/90 (20%) patients had an intermediate risk HCT-CI (1–2 points) and 72/90 patients (80%) had an unfavorable score (>2 points). As a preparative regimen all patients received RIC, which consisted of fludarabine 180 mg/m2 + oral busulfane 8 mg/kg + anti-thymocyte globulin 30 mg/kg. As stem cell source peripheral blood stem cells (PBSC) were used in 85/90 patients (94%), whereas 5/90 patients (6%) received bone marrow (BM). 38/90 patients (42%) were transplanted from a matched related donor. A matched unrelated or a mismatched unrelated donor was available in 36/90 patients (40%) or in 16/90 patients (18%). Graft-versus-host disease (GvHD) prophylaxis consisted of cyclosporin (CSA) + mycophenolate mofetil (MMF).

Results: Projected overall survival (OS) or disease-free survival (DFS) of the whole cohort at 1, 3, and 5 years was 60%, 44%, and 44% or 56%, 46%, and 46%. 49/90 patients (54%) are in CCR. Causes of death were relapse (23/90 patients (26%)) or TRM (18/90 patients (20%)). Notably, the HCT-CI alone was not sufficient to predict the overall outcome our cohort of patients after RIC-alloSCT. Therefore, depending on the presence or absence of at least one additional leukemia-specific high-risk factor, i.e. disease status prior to alloSCT >CR1, tAML, or poor risk karyotype, patients were grouped into four subgroups: group I (HCT-CI <4, no high risk), group II (HCT-CI >4, no high risk), group III (HCT-CI <4, high risk), and group IV (HCT-CI >4, high risk). Projected OS in at 1, 2, and 3 years after alloSCT was 73%, 59%, and 59% (group I), 86%, 57%, and 57% (group II), 56%, 46%, and 37% (group III), or 33%, 16%, and 16% (group IV), which differed statistically significant between the 4 subgroups (p = 0.04). In turn, there was no statistically significant difference in TRM between groups I – IV (p = 0.61).

Conclusions: These results indicate that the combined analysis of TRM-related factors, as assessed by the HCT-CI, and a limited set of leukemia-specific risk factors, i.e. cytogenetic risk, disease status before alloSCT, and the presence of tAML, may be useful in predicting the overall outcome of patients with AML after alloSCT following RIC.