Multivariable time-dependent analysis of the impact of azacitidine in patients with lower-risk myelodysplastic syndrome and unfavorable specific lower-risk score

Scoring systems for lower-risk myelodysplastic syndrome (LR-MDS) recognize patients with a poorer than expected outcome. This study retrospectively analyzes the role of azacitidine in LR-MDS with adverse risk score and compared to an historical cohort treated with best supportive care or erythropoiesis-stimulating agents. Overall response to AZA was 40%. One and 2-year probabilities of survival were 62% and 45% for AZA vs. 25% and 11% (P=10(-4)). In a multivariable time-dependent analysis, response to AZA (CR/PR/HI) was associated with an improved survival (HR=0.234, 95% CI, 0.063-0.0863; P=0.029). Thrombocytopenia (<50 × 10(9)L(-1)) is confirmed as an adverse parameter in LR-MDS (HR=1.649, 95% CI, 1.012-2.687; P=0.045).