Acute myeloid leukemia (AML) is a heterogeneous group of diseases characterized by uncontrolled proliferation of hematopoietic stem cells in the bone marrow. Malignant cell growth is characterized by disruption of normal intracellular signaling, caused by mutations or aberrant external signaling. The phosphoinositide 3-kinase (PI3K)-Akt-mammalian target of rapamycin (mTOR) pathway (PI3K-Akt-mTOR pathway) is among one of the intracellular pathways aberrantly upregulated in cancers including AML. Activation of this pathway seems important in leukemogenesis, and given the central role of this pathway in metabolism, the bioenergetics of AML cells may depend on downstream signaling within this pathway. Furthermore, observations suggest that constitutive activation of the PI3K-Akt-mTOR pathway differs between patients, and that increased activity within this pathway is an adverse prognostic parameter in AML. Pharmacological targeting of the PI3K-Akt-mTOR pathway with specific inhibitors results in suppression of leukemic cell growth. However, AML patients seem to differ regarding their susceptibility to various small-molecule inhibitors, reflecting biological heterogeneity in the intracellular signaling status. These findings should be further investigated in both preclinical and clinical settings, along with the potential use of this pathway as a prognostic biomarker, both in patients receiving intensive curative AML treatment and in elderly/unfit receiving AML-stabilizing treatment.

Clonal heterogeneity detected by karyotyping is a biomarker associated with adverse prognosis in acute myeloid leukemia (AML). Constitutive activation of the phosphatidylinositol-3-kinase-Akt-mechanistic target of rapamycin (PI3K-Akt-mTOR) pathway is present in AML cells, and this pathway integrates signaling from several upstream receptors/mediators. We suggest that this pathway reflects biologically important clonal heterogeneity. We investigated constitutive PI3K-Akt-mTOR pathway activation in primary human AML cells derived from 114 patients, together with 18 pathway mediators. The cohort included patients with normal karyotype or single karyotype abnormalities and with an expected heterogeneity of molecular genetic abnormalities. Clonal heterogeneity reflected as pathway mediator heterogeneity was detected for 49 patients. Global gene expression profiles of AML cell populations with and without clonal heterogeneity differed with regard to expression of ectopic olfactory receptors (a subset of G-protein coupled receptors) and proteins involved in G-protein coupled receptor signaling. Finally, the presence of clonal heterogeneity was associated with adverse prognosis for patients receiving intensive antileukemic treatment. The clonal heterogeneity as reflected in the activation status of selected mediators in the PI3K-Akt-mTOR pathway was associated with a different gene expression profile and had an independent prognostic impact. Biological heterogeneity reflected in the intracellular signaling status should be further investigated as a prognostic biomarker in human AML.

Acute myeloid leukemia (AML) is an aggressive blood cancer resulting in accumulation of immature, dysfunctional blood cells in the bone marrow. Changes in cell metabolism are features of many cancers, including AML and this may be exploited as a therapeutic target. In this study we investigated the in vitro antileukemic effects of seven metabolic inhibitors that target different metabolic pathways. The metabolic inhibitors were tested on AML cells derived from 81 patients using proliferation and viability assays; we also compared global gene expression and proteomic profiles for various patient subsets. Metformin, 2DG, 6AN, BPTES and ST1326 had strong antiproliferative and proapoptotic effects for most patients, whereas lonidamine and AZD3965 had an effect only for a minority. Antiproliferative effects on AML cells were additive when combined with the chemotherapeutic agent AraC. Using unsupervised hierarchical clustering, we identified a strong antiproliferative effect on AML cells after treatment with metabolic inhibitors for a subset of 29 patients. Gene expression and proteomic studies suggested that this subset was characterized by altered metabolic and transcriptional regulation. In addition, the Bcl-2 inhibitor venetoclax, in combination with 2DG or 6AN, increased the antiproliferative effects of these metabolic inhibitors on AML cells. Therapeutic targeting of cellular metabolism may have potential in AML, but the optimal strategy will likely differ between patients.

Acute myeloid leukemia (AML) is an aggressive malignancy, and many elderly/unfit patients cannot receive intensive and potentially curative therapy. These patients receive low-toxicity disease-stabilizing treatment. The combination of all-trans retinoic acid (ATRA) and the histone deacetylase inhibitor valproic acid can stabilize the disease for a subset of such patients. We performed untargeted serum metabolomic profiling for 44 AML patients receiving treatment based on ATRA and valproic acid combined with low-dose cytotoxic drugs (cytarabine, hydroxyurea, 6-mercaptopurin) which identified 886 metabolites. When comparing pretreatment samples from responders and non-responders, metabolites mainly belonging to amino acid and lipid (i.e., fatty acid) pathways were altered. Furthermore, patients with rapidly progressive disease showed an extensively altered lipid metabolism. Both ATRA and valproic acid monotherapy also altered the amino acid and lipid metabolite profiles; however, these changes were only highly significant for valproic acid treatment. Twenty-three metabolites were significantly altered by seven-day valproic acid treatment (p < 0.05, q < 0.05), where the majority of altered metabolites belonged to lipid (especially fatty acid metabolism) and amino acid pathways, including several carnitines. These metabolomic effects, and especially the effects on lipid metabolism, may be important for the antileukemic and epigenetic effects of this treatment.

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is used in the treatment of high-risk acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS); however, the treatment has high risk of severe transplantation-related mortality (TRM). In this study, we examined pretransplantation serum samples derived from 92 consecutive allotransplant recipients with AML or MDS. Using nontargeted metabolomics, we identified 1274 metabolites including 968 of known identity (named biochemicals). We further investigated metabolites that differed significantly when comparing patients with and without early extensive fluid retention, pretransplantation inflammation (both being associated with increased risk of acute graft-versus-host disease [GVHD]/nonrelapse mortality) and development of systemic steroid-requiring acute GVHD (aGVHD). All three factors are associated with TRM and were also associated with significantly altered amino acid metabolism, although there was only a minor overlap between these three factors with regard to significantly altered individual metabolites. Furthermore, steroid-requiring aGVHD was especially associated with altered taurine/hypotaurine, tryptophan, biotin, and phenylacetate metabolism together with altered malate-aspartate shuttle and urea cycle regulation. In contrast, pretransplantation inflammation was associated with a weaker modulation of many different metabolic pathways, whereas extensive fluid retention was associated with a weaker modulation of taurine/hypotaurine metabolism. An unsupervised hierarchical cluster analysis based on the 13 most significantly identified metabolites associated with aGVHD identified a patient subset with high metabolite levels and increased frequencies of MDS/MDS-AML, steroid-requiring aGVHD and early TRM. On the other hand, a clustering analysis based on metabolites that were significantly altered for aGVHD, inflammation, and fluid retention comparison groups identified a patient subset with a highly significant association with TRM. Our study suggests that the systemic pretransplantation metabolic profiles can be used to identify patient subsets with an increased frequency of TRM.