Highlight
– Metformin induces reactive oxygen species (ROS) and ferroptotic cell death in primary AML samples ex vivo, with greatest effect in cases having disrupted lipid metabolism (notably IDH2- and FLT3-mutant disease).
– Lipidomic remodeling with increased triglycerides and polyunsaturated fatty acids (PUFAs), upregulation of lipid-droplet genes including DGAT1, and CD36-mediated uptake are key determinants of metformin sensitivity.
– DGAT1 inhibition synergizes with metformin; iron chelators antagonize its effect—supporting ferroptosis as the operative cell-death mechanism and pointing to rational combination strategies.
Background
Metabolic reprogramming is a hallmark of cancer and a recognized therapeutic vulnerability in acute myeloid leukemia (AML). Many AML cells and leukemic stem cells rely on oxidative phosphorylation (OXPHOS) to sustain survival and chemoresistance, and targeting mitochondrial metabolism has emerged as a promising strategy in preclinical studies.[3] However, many potent OXPHOS inhibitors have dose-limiting toxicities that hinder clinical translation. Metformin, a widely used and generally safe antidiabetic drug, inhibits mitochondrial complex I and reduces OXPHOS at pharmacologic and higher concentrations. Interest in repurposing metformin for oncology rests on its safety profile and pleiotropic metabolic effects, but mechanisms of anti-leukemic activity and predictive biomarkers for response are incompletely defined.
Study design
The study by Sternadt and colleagues (Blood Advances, 2025)[1] interrogated metformin sensitivity across a genetically diverse panel of primary AML samples. Investigators isolated CD34+/CD117+ blasts from clinical specimens and applied label‑free quantitative proteomics. They performed single-sample gene set enrichment analysis (ssGSEA) focusing on metabolic pathways and correlated enrichment scores with ex vivo metformin sensitivity. Functional assays measured reactive oxygen species (ROS), cell death modality, and the effects of lipid manipulations. Lipidomic profiling assessed fatty acid and triglyceride changes after treatment. Genetic perturbations (CD36 knockdown) and pharmacologic modulators (palmitate, DGAT1 inhibitor, iron chelators) were used to dissect mechanism and synergy.
Key findings
The report combines proteomics, functional assays, and lipidomics to generate a coherent mechanistic story. The principal findings are:
1. Metformin triggers ROS generation and ferroptosis in primary AML ex vivo
Ex vivo metformin treatment increased cellular ROS and induced a form of non-apoptotic, iron-dependent cell death consistent with ferroptosis. Ferroptosis is characterized by iron-dependent lipid peroxidation and is distinct from apoptosis and necroptosis.[2] The cell death was mitigated by iron chelation, corroborating iron dependence, and by interventions that limit lipid peroxidation—supporting ferroptosis as the dominant mechanism.
2. Sensitivity correlates with disturbed lipid metabolism and specific genotypes
Proteomic signatures and ssGSEA showed that AML samples with enriched lipid metabolic pathways were more susceptible to metformin. IDH2- and FLT3-mutant samples were particularly sensitive. IDH mutations are known to alter cellular metabolism and redox balance through production of the oncometabolite 2-hydroxyglutarate, which can influence NADPH and lipid handling; FLT3 mutations are also linked to metabolic reprogramming—providing biological plausibility for genotype-associated vulnerability.[4]
3. Lipid uptake and remodeling modify metformin response
Supplementation with palmitate (a saturated fatty acid) enhanced metformin sensitivity in IDH2-mutant cells, while CD36 knockdown reduced sensitivity, indicating that fatty acid uptake via CD36 contributes to metformin-induced ferroptosis. Lipidomic analysis revealed increases in triglycerides and polyunsaturated fatty acids (PUFAs) after metformin exposure, suggesting active lipid remodeling that increases pools susceptible to peroxidation.
4. Lipid droplet machinery is engaged and DGAT1 is a potential combinatorial target
Metformin upregulated genes involved in lipid droplet formation, including DGAT1—an enzyme catalyzing final triglyceride synthesis and lipid droplet biogenesis. Pharmacologic inhibition of DGAT1 produced strong synergy with metformin, amplifying cell death, consistent with the concept that blocking protective lipid storage or altering lipid partitioning can sensitize cells to ferroptosis.
5. Iron chelation antagonizes metformin’s effect
Iron chelators reduced metformin-induced cell death, reinforcing the requirement for iron in the observed ferroptotic mechanism and highlighting potential therapeutic interactions—e.g., patients on iron chelation might derive less benefit if ferroptosis induction is a desired mechanism.
Mechanistic interpretation
Metformin’s primary mitochondrial effect—complex I inhibition—reduces NAD+/NADH turnover and impairs electron transport, leading to increased mitochondrial ROS in susceptible cells. Excess ROS can initiate lipid peroxidation when cellular pools of PUFAs and accessible iron are present, culminating in ferroptotic death.[2] The study indicates that AML cells with active lipid uptake and remodeling pathways accumulate PUFA-containing lipids and triglycerides, creating a substrate pool for peroxidation. DGAT1-mediated triglyceride formation and lipid droplet dynamics may represent protective or buffering mechanisms; inhibiting DGAT1 prevents sequestration of harmful lipid species, thus amplifying ferroptosis. CD36-mediated uptake of exogenous fatty acids, especially saturated FAs such as palmitate, modulates membrane composition and could influence the ratio of saturated to unsaturated lipids, thereby affecting susceptibility to peroxidation. This integrated lipid-centric view links mitochondrial inhibition to ferroptotic death via substrate availability and storage dynamics.
Clinical and translational implications
The work suggests several clinically relevant points:
- Biomarker-driven repurposing: Proteomic or transcriptomic signatures of lipid metabolism, or specific genotypes (IDH2, FLT3), may identify AML patients more likely to respond to metformin-based strategies.
- Combination strategies: Co-targeting lipid handling (DGAT1 inhibitors), fatty acid uptake (CD36 antagonists), or exploiting iron metabolism could potentiate metformin efficacy. Conversely, iron chelation could blunt benefit and should be considered when designing trials.
- Dose and pharmacokinetics caution: Many in vitro studies use metformin concentrations higher than achievable plasma levels with standard dosing. Translational development will require careful dose-finding, potentially localized delivery, or combinations that lower the required metformin exposure.
- Patient selection and safety: Given metformin’s safety record in diabetes, rapid translation is attractive, but AML patients have comorbidities (renal dysfunction, lactic acidosis risk) that must be accounted for—especially if higher-than-usual doses or combinations are used.
Expert commentary and limitations
This study elegantly connects mitochondrial inhibition to ferroptosis through lipidomic remodeling, strengthening the conceptual framework for targeting metabolic vulnerabilities in AML. Prior work has established OXPHOS dependence in certain AML subsets and highlighted the promise of mitochondrial targeting.[3] The demonstration that metformin—an approved, well-tolerated agent—can engage ferroptosis in primary AML is compelling, particularly because it identifies tractable biomarkers (lipid signatures, DGAT1, CD36) and actionable combinations.
However, limitations must be stressed. The bulk of the data are ex vivo and rely on cellular and biochemical assays; in vivo validation in animal models and patient-derived xenografts is necessary to assess pharmacokinetics, tumor microenvironment effects, and tolerability at effective doses. Metformin’s in vitro active concentrations often exceed therapeutic plasma levels, so combinatorial partners that lower the effective dose will be critical for clinical translation. Heterogeneity of patient samples and complexity of lipid metabolism across patients may complicate biomarker development. Finally, ferroptosis is a context-dependent process and interactions with host iron metabolism, inflammation, and concurrent medications (e.g., iron chelators) require systematic evaluation.
Next steps and research priorities
Priority actions to translate these findings include:
- Preclinical in vivo studies testing metformin alone and in rational combinations (DGAT1 inhibitors, CD36 modulators) using orthotopic AML models and patient-derived xenografts, with attention to dose, exposure, and toxicity.
- Development and validation of predictive biomarkers (proteomic lipid signatures, DGAT1/CD36 expression, and mutational status like IDH2/FLT3) that can stratify patients for early-phase trials.
- Phase I/II biomarker-driven clinical trials assessing safety, pharmacodynamics (ROS, lipid peroxidation markers), and early efficacy signals, while monitoring interactions with iron metabolism and standard AML therapies.
- Mechanistic studies to map how lipid droplet dynamics and specific lipid species dictate ferroptosis susceptibility in different AML subclones, including leukemic stem cells.
Conclusion
The study by Sternadt et al. positions metformin as a potential metabolic inducer of ferroptosis in AML, with sensitivity shaped by lipid uptake, remodeling, and storage pathways. The data support a precision-medicine approach—using metabolic and genetic biomarkers to select patients and combining metformin with agents targeting lipid handling to amplify ferroptotic death. While promising, the translational path requires rigorous in vivo validation, careful dosing strategies, and early-phase trials to establish safety and proof-of-concept in patients.
Funding and clinicaltrials.gov
The original Blood Advances manuscript reports the study details and author list (Sternadt et al.); specific funding sources and trial registrations should be referenced in the primary publication. As of this writing, clinical trials repurposing metformin for AML driven by ferroptosis have not been widely reported; prospective clinical translation would benefit from registration and biomarker-guided trial design on ClinicalTrials.gov.
References
[1] Sternadt D, Pereira-Martins DA, Chatzikyriakou P, Albuquerque-Simões L, Yang M, Silveira DR, Wierenga AT, Weinhäuser I, Hogeling SM, Oudejans L, Casares-Alaez P, Sarry JE, Frezza C, Huls GA, Quek L, Schuringa JJ. Metformin induces ferroptosis associated with lipidomic remodeling in AML. Blood Adv. 2025 Oct 31:bloodadvances.2025016155. doi: 10.1182/bloodadvances.2025016155. Epub ahead of print. PMID: 41172231.
[2] Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell. 2012;149(5):1060–72.
[3] Lagadinou ED, Sach A, Callahan K, Rossi RM, Neering SJ, Minhajuddin M, et al. BCL-2 inhibition targets oxidative phosphorylation and eradicates quiescent leukemia stem cells. Cell Stem Cell. 2013;12(3):329–341.
[4] Figueroa ME, Abdel-Wahab O, Lu C, Ward PS, Patel J, Shih A, et al. Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and block hematopoietic differentiation. Cancer Cell. 2010;18(6):553–567.
AI-friendly thumbnail prompt
A high-detail scientific illustration: cultured acute myeloid leukemia cells under a fluorescence microscope showing bright lipid droplets (red) and ROS/lipid peroxidation markers (green), a stylized metformin molecule semi-transparently overlaid in the corner, an iron atom icon and small DGAT1/CD36 labels; cool laboratory background, high contrast, modern editorial style suitable for a medical journal cover.