LAUSR

Simple Summary Around 30% of children with Down Syndrome (DS) will develop Transient Abnormal Myelopoiesis (TAM) and 20% of them will progress to Acute Myeloid Leukemia (AML), mostly Megakaryoblastic Leukemia (AMKL). The optimal balance between treatment intensity and treatment-related toxicity has not yet been defined; neither the prognostic factors that determine the risk of developing AML nor the outcome. The aims of our retrospective study were to analyze the demographic/biological features of this population, identify possible risk factors and the optimal treatment. We observed that early intervention in TAM is effective to prevent a dismal outcome. The strongest poor-prognostic factor of DS-AML was sporadic DS-AML (non-AMKL immunophenotype), as well as complex karyotype and young age. Classical Myeloid Leukemia associated with DS (ML-DS) good outcome is mainly related to their low relapse rate. Even though the augmented sensitivity to chemotherapy seen in DS patients must be kept in mind, our data do not support the omission of high doses of cytarabine in ML-DS. Abstract Children with Down syndrome (DS) are at an increased risk of developing clonal myeloproliferative disorders. The balance between treatment intensity and treatment-related toxicity has not yet been defined. We analyzed this population to identify risk factors and optimal treatment. This single-center retrospective study included 78 DS patients <16 years-old with Transient Abnormal Myelopoiesis (TAM, n = 25), Acute Myeloblastic Leukemia (DS-AML, n = 41) of which 35 had classical Myeloid Leukemia associated with DS (ML-DS) with megakaryoblastic immunophenotype (AMKL) and 6 sporadic DS-AML (non-AMKL). Patients with DS-AML were treated according to four BFM-based protocols. Classical ML-DS vs. non-DS-AMKL were compared and the outcome of ML-DS was analyzed according to treatment intensity. Only four patients with TAM required cytoreduction with a 5-year Event-Free Survival probability (EFSp) of 74.4 (±9.1)%. DS-AML treatment-related deaths were due to infections, with a 5-year EFSp of 60.6 (±8.2)%. Megakaryoblastic immunophenotype was the strongest good-prognostic factor in univariate and multivariate analysis (p = 0.000). When compared ML-DS with non-DS-AMKL, a better outcome was associated with a lower relapse rate (p = 0.0002). Analysis of administered treatment was done on 32/33 ML-DS patients who achieved CR according to receiving or not high-dose ARA-C block (HDARA-C), and no difference in 5-year EFSp was observed (p = 0.172). TAM rarely required treatment and when severe manifestations occurred, early intervention was effective. DS-AML good outcome was associated with AMKL with a low relapse-rate. Even if treatment-related mortality is still high, our data do not support the omission of HDARA-C in ML-DS since we observed a trend to detect a higher relapse rate in the arm without HDARA-C.