Abstract
BACKGROUND/OBJECTIVES: Development of novel therapies for acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) is hindered by limited biomarkers to identify appropriate target populations and poor translatability of preclinical drug candidates. In this study, we sought to validate Fragile histidine triad diadenosine triphosphate (FHIT) deficiency as a predictive biomarker for decapping scavenger enzyme (DCPS)-targeted therapy in AML and MDS, and to identify clinically accessible proxy biomarkers. METHODS: We analyzed primary AML samples treated with the DCPS small-molecule inhibitor RG3039 to identify additional biomarkers beyond FHIT that predict treatment response. We analyzed a large cohort of clinical and multi-omics AML and MDS datasets to define FHIT-low prevalence and stratify patient populations eligible for DCPS-targeted therapy. RESULTS: Low FHIT expression predicted effective therapeutic response to DCPS inhibition; however, hydroxyurea pretreatment confounded the predictive value of FHIT. FHIT-low expression occurred in 5–24% of AML patients, with highest prevalence in pediatric populations (24.4%). In MDS, FHIT promoter methylation was present in 35.8% of patients and remained stable during treatment with the hypomethylating agent azacitidine, supporting MDS as a potential therapeutic indication. IDH2 mutation status served as a proxy biomarker for FHIT expression and DCPS sensitivity in AML. CONCLUSIONS: FHIT expression levels predict sensitivity to DCPS inhibition and a large proportion of adult and pediatric AML as well as MDS patients are eligible for DCPS-targeted therapy. IDH2 mutation status is a clinically accessible proxy biomarker for predicting FHIT expression levels and sensitivity to DCPS inhibition in AML. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s12672-026-04880-x.