Cytidine deaminase activity predicts clinical outcome of azacitidine in real-world MDS patients

Abstract Azacitidine (Vidaza®, Aza) is a mainstay for treating acute myeloid leukemia (AML) in patients unfit for induction chemotherapy and other high risk myelodysplastic syndromes (MDS). Only half of patients usually respond and almost all patients will relapse eventually. There are no predictive factors for response to Aza. Aza is detoxified in the liver by a single enzyme, cytidine deaminase (CDA). CDA is a ubiquitous enzyme coded by a highly polymorphic gene, with subsequent great variability in resulting liver activity. Here we investigated on the link between CDA phenotypic status, toxicity and efficacy of Aza in adult patients treated for AML and MDS. A total of 104 patients were screened in this retrospective, real-world study. Of the 104 patients treated, 67.3% received more than 3 cycles. Treatment was postponed due to Aza-related toxicity in 66% (before Course-2) and 53% (before Course-3) of the patients. Median overall survival (OS) were 15 months (95% CI: 10–22) and 13 months (95% CI: 9–21) in AML and high MDS cohorts, respectively. Responding patients at 3-months showed prolonged survivals, with landmark OS of 17 and 14 months for AML and MDS, respectively. Half of the patients had deficient CDA activity (i.e., < 2 UA/mg), with a lower proportion of deficient patients in MDS patients (34%) as compared with AML patients (67%). In MDS patients, CDA deficiency was correlated with longer landmark OS (14 VS. 8 months; p = 0.03 multivariate analysis), but not in AML patients. In addition, our data suggest that delay in Aza treatment is not associated with lack of efficacy, and thus should not be considered as a signal to shift to other drugs. Overall, our data suggest that CDA is an independent covariate could thus be a marker for predicting clinical outcome in MDS patients treated with Aza.