P3-091Impact of cytogenetic abnormalities on response to azacitidine in patients with myelodysplastic syndrome

Abstract Background The AZA-001 phase III randomized study showed that azacitidine (AZA) significantly improved overall survival (OS) compared with conventional care in high- risk myelodysplastic syndrome (MDS), but prolonged treatment of 4-6 cycles is required before a failure to achieve response can be declared. Although selection of patients who are unlikely to derive benefit from AZA has become a top clinical priority, prognostic factors for response to and survival with AZA remain unclear. Here we examined prognostic factors for AZA in MDS patients in our hospital. Patients and Methods We retrospectively analyzed MDS patients treated with AZA between November 2011 and June 2018 in our hospital. Predictors of response were baseline cytogenetic characteristics (including IPSS-R cytogenetic classification) and the number of cytogenetic abnormalities. Univariate analyses were performed with log-rank tests and OS was measured from the onset of AZA. Results The male: female ratio was 4:1. Median age was 72 years, and median number of cycles of AZA was 11. Univariate analysis identified high score for IPSS-R cytogenetic classification and large number of cytogenetic abnormalities as factors associated with poor OS. In particular, patients with 17p abnormalities, possibly related to the presence of TP53 mutation, showed significantly shorter OS. Most rapidly developed leukemia and treatment became difficult to continue (median number of cycles of AZA, 2). Conclusion In the group with a high score for IPSS-R cytogenetic classification and large number of cytogenetic abnormalities, few patients showed survival advantage or delayed transformation to a leukemic state despite receiving AZA. An alternative to AZA monotherapy is considered necessary for patients expected to show poor response to AZA.