Background. Acute myeloid leukemia (AML) is an aggressive disease that shows dysregulated cell metabolism and a high relapse rate. To date, less invasive tools are urgently needed to explore disease dynamics in AML. In this regard, Extracellular Vesicles (EV) from plasma represent a promising tool for liquid biopsy. Also, the material extracted from blood (including circulating leukemic stem cells, LSC) can be used for clinical purposes revealing clinically relevant (metabolic) vulnerabilities. Along with EV profiling in AML patients, we explored the clinically relevant (metabolic) signature carried by CD34+ cells in circulation.

Methods. Peripheral blood (PB) was collected from AML patients at diagnosis (n=70) and healthy donors (HD, n=20). EVs were purified from platelet-poor plasma by size exclusion chromatography and ultrafiltration. Quantitative lipidomic profiling was performed by Liquid Chromatography coupled with High-Resolution Mass Spectrometry (LC/HRMS). Single Cell ENergetic metabolism by Profiling Translation inhibition (SCENITH) has been performed in fresh whole blood.

Results. We profiled the surface protein of isolated circulating EVs by AML patients at diagnosis. Compared to EV from HD, leukemic EV from newly diagnosed AML patients were mainly enriched in cancer stem cell marker, CD44. Intriguingly, the area under the curve (AUC) confirmed a reliable performance for employing this marker on EVs to discriminate EVs from AML patients. Among other 36 markers explored, CD4 (immune marker), CD105 (mesenchymal stromal cell marker), CD14 (monocyte marker), and CD133-1 (stem cell marker) were highly expressed only in leukemic EVs. In particular, CD11c as the well-known dendritic marker was solely expressed in EVs from some AML patients and totally absent in healthy EVs. By contrast, in circulating leukemic EVs, we observed depletion in markers such as CD62P, CD41b, and CD41a suggesting an impaired function in leukemic megakaryocytes and platelets.

From the clinic point of view, according to European LeukemiaNet (ELN) 2017 risk classification, we observed that CD9, CD29, CD31, and CD42a were enriched in the favorable AML group in comparison with adverse/intermediate-risk AML patients. Notably, non-survivor AML patients showed a significant reduction in the expression of EV markers including CD11c, CD69, and ROR-1.

Then, to further highlight crucial characteristics in EV composition, we performed an untargeted lipidomic analysis on circulating EVs. Fatty acids (FA) and Diacylglycerols (DG) were significantly overexpressed in leukemic EVs compared to healthy EVs. Leukemic EVs were depleted in Cholesteryl ester (CE) suggesting an increase in membrane stiffness. Interestingly, the lipid CE18:2 was found high in adverse-risk AML patients encouraging further studies on EV lipidomic profiles for better AML stratification.

An in-depth investigation of the metabolic dependencies of leukemia (stem) cells from blood might be crucial for detecting novel metabolic vulnerabilities in AML patients. Thus, we applied the innovative SCENITH approach in fresh blood from AML patients at diagnosis. We found that circulating AML CD34+ cells were highly dependent on glucose oxidation and less prone to use fatty acids and amino acids as sources for ATP production when glucose oxidation was inhibited. Also, AML CD34+ cells in circulation reported a higher glycolytic capacity compared to mitochondrial dependence. Towards a clinically relevant implication, SCENITH analysis showed that mitochondrial dependence was solely increased in adverse and intermediate-risk AML patients compared to favorable AML patients. Interestingly, the percentage of mitochondrial dependence in AML CD34+ was positively correlated to the expression of CD24 (cancer stem marker) observed on leukemic EVs suggesting a novel potential link between EV profiling and energy metabolism for leukemia.

Conclusion. Overall, these results demonstrate that an abnormal signature characterizes the phenotype of circulating EVs in AML patients. Indeed, the depletion of selected EV markers could be associated with high-risk AML disease and supports their prognostic relevance. This work might reveal novel metabolic vulnerabilities in the AML scenario by exploiting both EV-based liquid biopsy and single-cell metabolic phenotyping.

Cavo:AbbVie, Amgen, Bristol Myers Squibb/Celgene, Pfizer, GlaxoSmithKline, Sanofi, Roche, Takeda: Consultancy, Honoraria; Janssen: Honoraria, Speakers Bureau. Curti:Jazz Pharma: Membership on an entity's Board of Directors or advisory committees; Abbvie: Membership on an entity's Board of Directors or advisory committees; Pfizer: Membership on an entity's Board of Directors or advisory committees; Novartis: Membership on an entity's Board of Directors or advisory committees.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution