Targeting the Bone Marrow Microenvironment to Overcome Hypomethylating Agent Resistance in Myelodysplastic Neoplasms

Objectives Myelodysplastic neoplasm (MDS) is a malignancy of haematopoietic stem and progenitor cells (HSPC), often resulting in bone marrow (BM) failure or progression to acute myeloid leukemia (AML). Nearly 50% of patients fail to respond to standard hypomethylating agents (HMAs), 5-azacytidine (AZA) or decitabine (DAC), necessitating better therapies. Background HSPCs in HMA non-responders exhibit higher quiescence and elevated focal adhesion kinase (FAK) activation, promoting stromal adhesion and reducing AZA uptake (Unnikrishnan et al., Cell Reports 2017). This study explores targeting HSPC-stromal interactions by combining FAK inhibitors with HMAs to enhance HSPC cycling and HMA uptake. Methods I evaluated colony-forming potential (CFU-Cs), cell cycle progression, HMA incorporation, and DNA demethylation in MDS/AML cell lines and primary HSPCs treated with HMA/FAK inhibitor combinations in stromal co-culture. Additionally, I analyzed marrow samples from the CELESTIAL-MDS phase 1 trial (NCT05636514), where patients received oral DAC (ASTX727) with or without the dual FAK inhibitor defactinib (DEF). Results In vitro studies on cell lines, combining DEF with HMAs reduced DNA demethylation caused by HMAs, while GSK2256098 increased it. HMA treatment decreased CFU-Cs in KG1 and MDSL cells co-cultured with stromal cells, with no significant impact from FAK inhibitors. Cell cycle effects varied: DAC with GSK2256098 promoted cycling, while HMAs with DEF delayed it. Ex vivo co-cultures of stromal cells with AZA-resistant MDS HSPCs showed DEF with DAC reduced LINE-1 demethylation, while GSK2256098 with AZA enhanced it. GSK2256098 improved CFU-C capacity in resistant cells, whereas DEF with DAC had minimal effect. Cell cycle responses were patient-dependent but trended toward quiescence with DEF. In CELESTIAL trial samples, DAC alone induced LINE-1 demethylation in the first cycle. DEF with DAC also caused demethylation, albeit to a lesser extent. Ex vivo studies using biopsy HSPCs mirrored these findings. Conclusion Preliminary data suggest treatment responses differ with HMA (AZA or DAC) and FAK inhibitor (DEF or GSK2256098) combinations, highlighting variability across patient samples. These findings support the need for personalized therapies in MDS. Ongoing trial data will provide further insights.