Outcome and Risk Factors for Therapy-Related Myeloid Neoplasms Treated with Allogeneic Stem Cell Transplantation

Background

Therapy-related myeloid neoplasms (t-MN) have a poor prognosis with conventional chemotherapy and a median survival of ≤1 year. Allogeneic stem cell transplantation (HCT) is considered the only effective treatment. EBMT and CIBMTR analyzed their registry data on HCT against t-MN and reported its outcome and developed scoring systems using identified risk factors in HCT. This study aimed to investigate outcome and risk factors in adult patients with t-MN who underwent HCT in Japan.

Methods

Data source

For this retrospective observational study, recipients' clinical data were provided by the Transplant Registry Unified Management Program of Japan society for Hematopoietic Cell Transplantation.

Definition

Therapy-related myeloid neoplasms include therapy-related AML (t-AML), t-MDS, and t-MDS/MPN occurring as late complications of cytotoxic chemotherapy and/or radiation therapy administered for a prior neoplastic or non-neoplastic disorder. Cases with past history of MDS or MPN were excluded because it was not possible to clearly distinguish their neoplasms from an original myeloid neoplasm or therapy-related neoplasm.

The primary outcome of the analysis was overall survival (OS), whereas the secondary endpoints included incidence of relapse and transplantation-related mortality (TRM).

Cytogenetic risk groups in AML were classified according to the MRC (2010). In MDS, three risk groups were defined using the revised International Scoring System: favorable (very good and good), intermediate (intermediate), poor (poor and very poor). HLA mismatch was defined as incompatibility between the recipient and donor when at least a single allele mismatch was detected at HLA-A, -B, and -DR in related and unrelated donors. In cord blood, HLA mismatch was defined as at least two antigen mismatches detected at HLA-A, -B, and -DR.

Statistical analysis

OS was estimated using the Kaplan-Meier method and was compared using the log-rank test. Cox's proportional-hazards regression model was used for multivariate analysis of prognostic factors. Cumulative incidence curves were used in a competing-risk setting to calculate the probability of relapse and TRM. Relapse and therapy-related deaths were considered a competing risk event for each other and were compared using Gray's test.

Results

Between 1992 and 2012, 641 adult patients who had undergone HCTs for confirmed t-MN (not de novo myeloid neoplasms) were identified. Median age at HCT was 53 (range; 16-80) years, with a female proportion of 49%. In total, 414 (64.6%) patients had AML, 215 (33.5%) had MDS, and 12 (1.9%) had MPN. Approximately 50% of patients had a prior history of lymphoma, 14.6% had acute leukemia, and 13.3% had breast cancer. Karyotype was available in 310 cases. Favorable karyotypes were detected in 11 (3.5%) cases, whereas intermediate and poor karyotypes were detected in 122 (39.4%) and in 177 (57.1%) cases, respectively. HLA mismatched HCT was 25.7%. Bone marrow or peripheral HCT from related donors was used in 189 (28.7%) cases, whereas bone marrow HCT from unrelated donor was used in 228 (35.7%) cases, and cord blood HCT was used in 229 (35.6%) cases. Approximately 30% of patients were in remission at HCT.

Overall survival was 45.3% [95% confident interval (CI) 45-48], 33% (CI 29-37), and 28% (CI 24-32) at 1, 3, and 5 years, respectively. In multivariate analysis, the significant factors for poor survival were ECOG Performance Status (PS) 2-4 (HR 1.99; CI 1.45-2.72), disease status of non-remission at the time of HCT (HR 1.82; CI 1.33-2.50), patient age of >55 years (HR 1.72; CI 1.32-2.24), and poor karyotype (HR 1.54; CI 1.15-2.05).

The cumulative incidence of non-relapse mortality and relapse at 3 years was 37% and 36%, respectively. In a multivariate analysis, the significant factor for non-relapse mortality was patient age of >55 years (HR 1.48; CI 1.09-2.02). For relapse, poor karyotype (HR 2.29; CI 1.65-3.18) was a significant factor.

Conclusions

PS 2-4, non-remission, higher patient age, and poor cytogenetics were predictive of poor survival. The outcome of HCT against therapy-related myeloid neoplasms in Japan was comparable to those of EBMT and CIBMTR.