Lipidomic reprogramming links VEGF-dependent CNS adaptation to ferroptosis sensitivity in acute lymphoblastic leukemia.

While central nervous system (CNS) involvement in acute lymphoblastic leukemia (ALL) remains a major cause of treatment failure, underlying molecular mechanisms are not fully understood. Based on previous work implicating vascular endothelial growth factor A (VEGF), we dissected its contribution to VEGF receptor 2 (VEGFR2)-mediated CNS entry versus niche-dependent survival and evaluated VEGF-targeted strategies in translational xenograft models. Using the NOD/SCID/huALL model enabling temporally defined intervention before and after CNS dissemination, bevacizumab consistently reduced established CNS ALL but had only minor effects on CNS entry, indicating a predominant contribution of VEGF in CNS ALL survival. Consistently, neither genetic depletion nor pharmacologic inhibition of murine Vegfr2 impaired transendothelial migration (TEM) in vitro. To further substantiate the role of VEGF in CNS ALL, manifest disease was treated in vivo with the multi-kinase inhibitor Sorafenib that was previously shown effective in refractory CNS leukemia patients. Sorafenib reduced CNS ALL with the highest efficacy in this compartment. In vitro, Sorafenib decreased TEM but also induced lipid peroxidation in ALL cells accompanied by ferroptosis induction. Importantly, lipidomic profiling of CNS-derived ALL cells revealed enhanced peroxidizability linked to increased ferroptosis susceptibility, providing a mechanistic link for lipid peroxidation-driven cell death in the CNS. Collectively, our findings identify VEGF signaling and lipidomic remodeling as complementary mechanisms of CNS adaptation in ALL. While VEGF supports CNS persistence, lipidomic remodeling generates a highly peroxidizable state associated with increased ferroptosis susceptibility. These findings suggest that inhibiting VEGF signaling and simultaneously inducing ferroptosis may represent a therapeutic strategy for CNS ALL.