High leukemia-free survival after TBI-based conditioning and mycophenolate mofetil-containing immunosuppression in patients allografted for chronic myelomonocytic leukemia: a single-center experience
This retrospective single-center analysis studied the impact of the conditioning and the graft-versus-host disease (GVHD) prophylaxis on outcome in unselected patients allografted for chronic myelomonocytic leukemia (CMML) and acute myeloid leukemia (AML) secondary to documented prior CMML. A total of 44 patients (median age 61 years) allografted between 2002 and 2019 in our institution were analyzed. Fifteen patients had secondary
AML. The conditioning regimen was fractionated 6-8 Gy total body irradiation (TBI) in combination with fludarabine in 33 (75%) patients. Eleven patients (25%) received alkylator-based conditioning therapy without TBI. For GVHD prophylaxis, a calcineurin inhibitor (CNI) backbone in combination with methotrexate (MTX) or mycophenolate mofetil (MMF) was applied in 21 and 23 patients, respectively. All patients allografted from an unrelated donor (UD) received antithymocyte globuline. In univariate analysis of the entire cohort, TBI-based conditioning and MMF-containing immunosuppression were associated with improved leukemia-free survival (LFS, HR 0.16, P < 0.001 and HR 0.41, P = 0.030, respectively). After stratification according to conditioning and GVHD prophylaxis into four groups (TBI-MMF [n = 17], TBI-MTX [n = 16], alkylator-MMF [n = 6], alkylator-MTX [n = 5]), TBI-MMF was associated with improved overall survival (OS) and LFS (P = 0.001 and P < 0.001, respectively). Patient and disease characteristics did not differ between the groups. The associations of TBI-based conditioning and MMF with prolonged LFS were observed across the CMML (n = 29), secondary AML (n = 15), and UD allograft (n = 34) subgroups. In summary, our study suggests that allografting based on intermediate-dose TBI conditioning and MMF-containing GVHD prophylaxis is associated with increased disease control in CMML. Larger (registry-based) studies are warranted to confirm our findings.
