Can we predict responsiveness to hypomethylating agents in AML?

DNA-hypomethylating agents (HMAs) were developed as nonintensive treatment alternatives to standard chemotherapy in older, unfit patients with acute myeloid leukemia and myelodysplastic syndrome. Given their distinct effects on the methylome and transcriptome of malignant cells compared to cytarabine (Ara-C) and other cytotoxic drugs not inhibiting DNA methyltransferases, it is of great interest to define their specific clinical ``signature.'' Here, we present and discuss clinical, genetic, and epigenetic predictors of responsiveness to HMAs. Indeed, mounting evidence supports the notion that HMAs are not "just another kind of low-dose Ara-C." Not only patient factors (age, performance status, comorbidities, etc.), blast counts, and early platelet response, but also adverse genetics (monosomal karyotype and/or a TP53 mutation) have predictive potential. Given the surprising-and initially counterintuitive-responses observed in patients with the latter features, these are subject to mechanistic studies to elucidate their as yet unresolved interaction with HMAs. Finally, other potential biomarkers for HMA response such as elevated fetal hemoglobin might also contribute to overcome the present challenges in predicting responsiveness to HMAs.