HMA-induced oncogene reactivation as a driver of disease progression in MDS

Hypomethylating agents (HMAs), including azacitidine and decitabine, are widely used for high-risk myelodysplastic syndromes (MDS) and selected AML, yet most patients eventually relapse, and survival after HMA failure remains dismal. HMAs were long assumed to function primarily by reactivating tumor suppressor genes through DNA demethylation. However, our recent work demonstrated that HMAs paradoxically upregulate the oncofetal gene SALL4, and that this induction strongly correlates with inferior overall survival, challenging existing paradigms. Emerging data suggest that additional cancer-germline antigen genes (CGAGs) may be similarly reactivated, generating oncogenic programs that facilitate resistance and leukemic progression. To investigate this, we profiled paired pre- and post-HMA patient marrow samples using RNA-seq and methylation analysis. We observed consistent post-treatment upregulation and promoter hypomethylation of SALL4 and multiple CGAGs, including PIWIL2, HORMAD1, and DDX43. Induction was more pronounced in patients with progression or treatment failure. Functional studies in MDS cell models revealed that forced expression of SALL4 or selected CGAGs enhanced proliferation, impaired myeloid differentiation, and reduced sensitivity to HMAs. Conversely, shRNA-mediated depletion mitigated these phenotypes. CRISPR-DiR-driven locus-specific demethylation of SALL4 recapitulated HMA-associated activation and conferred a proliferative advantage, confirming a causal role for promoter demethylation in SALL4 reactivation. We next evaluated therapeutic vulnerabilities associated with this pathway. Treatment with a newly developed small-molecule SALL4 degrader (SH6) reduced viability in SALL4-high models and reversed HMA-induced resistance phenotypes. Combination treatment with HMA + SH6 demonstrated enhanced cytotoxicity compared to either agent alone. Ongoing studies are expanding this approach to additional CGAG targets and mapping resistant subpopulations using single-cell transcriptomic and epigenetic profiling. Together, these findings reveal that HMA therapy can trigger unintended activation of oncogenic CGAG programs that drive MDS progression. They support a new mechanistic framework in which treatment-induced epigenetic reawakening promotes malignant fitness, and they identify SALL4 and CGAGs as high-value biomarkers and therapeutic targets. This work motivates development of companion diagnostic strategies and combination regimens pairing HMAs with targeted degraders to prevent therapy-induced disease acceleration and improve patient outcomes.