Therapy-Related Acute Myeloid Leukemia and Myelodysplastic Syndrome in Lymphoma Patients Receiving Hematopoietic Cell Transplantation

Therapy-related acute myeloid leukemia and myelodysplastic syndrome (t-AML/MDS) represent severe late effects in patients receiving hematopoietic cell transplantation (HCT) for malignant lymphoma (ML). The option of high-dose therapy with autologous HCT or allogeneic HCT with reduced-intensity conditioning remains controversial in patients with relapsed ML. We retrospectively analyzed incidence, outcome, and risk factors for the development of t-AML/MDS in ML patients treated with autologous or allogeneic HCT. A total of 13,810 ML patients received autologous (n=9963) or allogeneic (n=3847) HCT between 1985 and 2012 were considered. At a median overall survival (OS) of 52 and 46 months in autologous and allogeneic HCT groups, respectively, ML patients receiving autologous HCT (1.38%; 95% confidence interval [CI], 1.09-1.68; at 3 year after autologous HCT) have a significant risk for developing t-AML/MDS compared to allogeneic HCT (0.37%; 95%CI, 0.06-0.67; at 3 year after allogeneic HCT). Median time from HCT to t-AML/MDS after autologous and allogeneic HCT was 957 and 414 days, respectively. Significant risk factors for the development of t-AML/MDS after autologous and allogeneic HCT are high-stage risk (P=0.04) or secondary malignancies (P<0.001) and receiving cord blood stem cell (P=0.03) or radiotherapy-conditioning (P=0.002), respectively. The additional data collection from 46 transplant centers was available for 62 of 220 patients with t-AML/MDS: 56/211 (26.5%) of t-AML/MDS patients receiving autologous HCT and 6/10 (60%) receiving allogeneic HCT. Information on cytogenetic and/or molecular genetic evaluation was available for 62 patients: 41/56 patients (73%) receiving autologous HCT and 6/6 patients (100%) receiving allogeneic HCT had cytogenetic abnormalities. Abnormalities in chromosome 5 and/or chromosome 7 were found in 28/56 (50%) after autologous HCT and 5/6 (83%) after allogeneic HCT. Balanced translocations, such as mixed-lineage leukemia rearrangement 11q23 and t(15;17), were detected in 4/56 patients (7%) after autologous HCT and 0/6 patients after allogeneic HCT. Complex karyotypes were observed in 17/56 patients (30%) after autologous HCT and 4/6 patients (67%) after allogeneic HCT. Allogeneic HCT for t-AML/MDS was available for 20/62 patients (32%): 18/56 (32%) after autologous HCT and 2/6 (33%) after allogeneic HCT. Among 56 patients in the autologous HCT group, t-AML/MDS patients who underwent allogeneic HCT had a significantly better OS from onset of t-AML/MDS compared with those who did not undergo allogeneic HCT (median OS 72 vs. 60 months, P=0.043), whereas the number of patients with high-disease risk was significantly higher in patients who did not undergo allogeneic HCT. Among 6 patients in the allogeneic HCT group succumbed to their disease: 2/2 patients who underwent allogeneic HCT died due to transplant-related toxicities and 4/4 patients who did not undergo allogeneic HCT died due to secondary malignancies in the CR state of lymphoma after the primary allogeneic HCT. When compared with the group of patients with t-AML/MDS after autologous and allogeneic HCT, the number of patients with Ann Arbor stage 4 and the percentages of patients receiving conditioning with etoposide were significantly higher in the autologous HCT group. The number of HCT, the percentages of patients receiving conditioning with total body irradiation and fludarabine, the longer duration of G-CSF, the later neutrophil engraftment and higher scores of revised international prognostic scorin system for MDS were significantly higher in the allogeneic HCT group. In conclusion, t-AML/MDS is rare but ML patients with t-AML/MDS after HCT are associated with a poor prognosis. Thus, novel treatment strategies for ML should aim at a reduction of chemotherapy and a reduced number of HCT to reduce the risk of developing t-AML/MDS. In addition, strategies to select ML patients more carefully for autologous HCT to exclude ML patients with high-stage risk may allow the identification of individual ML patients at particular high risk for t-AML/MDS. Allogeneic HCT might be an alternative to autologous HCT in ML patients at high risk for t-AML/MDS. Finally, the further optimization of allogeneic HCT might increase the cure rates among ML patients diagnosed with t-AML/MDS after autologous HCT.