Characterization of Metabolic Response to AG-348, an Allosteric Activator of Red Cell Pyruvate Kinase, in Healthy Volunteers and Pyruvate Kinase Deficiency Patients

Pyruvate kinase (PK) deficiency is a glycolytic enzymopathy that causes lifelong chronic hemolytic anemia. AG-348 is an allosteric activator of the red cell isoform of pyruvate kinase (PK-R) that is in clinical development to treat PK deficiency. Phase 1 studies of AG-348 in healthy volunteers (NCT02108106, NCT02149966) have been completed, and a phase 2 study in patients with PK deficiency is in progress (DRIVE PK, NCT02476916). We have previously reported that in the healthy volunteer studies, AG-348 induced changes in levels of the metabolites adenosine triphosphate (ATP) and 2,3-diphosphoglycerate (2,3-DPG) that are consistent with PK-R activation. A preliminary analysis of data showed that nine out of 18 DRIVE PK patients achieved a maximal increase in hemoglobin (Hb) levels of >1.0 g/dL.

In this study, we characterize whole blood metabolism of healthy human subjects (from the multiple-ascending dose study NCT02149966) as well as PK-deficient patients (from DRIVE PK), before and after administration of AG-348, with a focus on the flux through the PK-R reaction. Whole blood taken from healthy subjects/DRIVE PK patients on each respective study was incubated with a stable isotope tracer, [^U-13C₆]-glucose. Glycolytic flux through PK-R was estimated by kinetic flux profiling based on the isotope labeling, as well as by the specific lactate production rate, with the two methods giving consistent results.

Blood cells from healthy subjects exhibited classic red blood cell (RBC) metabolism, with the majority of glucose catabolized through glycolysis. AG-348 was shown to significantly increase the maximal PK-R protein activity in these subjects, and also to have metabolic effects consistent with PK-R activation, marked by decreased concentrations of glycolytic intermediates such as 2,3-DPG and phosphoenolpyruvate (PEP). Increases in ATP concentrations were also observed, with the magnitude and kinetics of the increase strongly suggesting enhanced adenosine salvage or synthesis. The overall glycolytic rates, however, did not change significantly after two weeks of AG-348 dosing, revealing the homeostatic regulation of RBC glycolysis in healthy blood.

Five of the first 18 DRIVE PK patients underwent an extensive sampling protocol for metabolic analysis. This analysis revealed a number of distinct metabolic qualities in PK-deficient patients at baseline compared with healthy subjects. These included significantly reduced rates of lactate production and high concentrations of nucleotides, amino acids, and Krebs cycle intermediates. ¹³C labeling was observed in Krebs cycle intermediates, demonstrating significant respiratory metabolism in whole blood cells from PK-deficient patients, while incomplete labeling of glycolytic intermediates suggested the presence of a metabolically inactive cell subpopulation. These observations are most consistent with the hypothesis that PK-deficient whole blood is dominated metabolically by immature erythrocytes that retain residual mitochondrial activity.

With AG-348 treatment, three of the five DRIVE PK patients had increases in Hb of >1.0 g/dL. In those three patients, we observed increased incorporation of ¹³C label into glycolytic intermediates such as 2,3-DPG, suggesting an increase in metabolically active erythrocytes. First order flux estimates based on either 2,3-DPG labeling kinetics or lactate production rates showed a >0.1 mmol/L/hr (>50%) increase in glycolytic flux. Neither of the two DRIVE PK patients that did not have an Hb increase of >1.0 g/dL showed significant metabolic changes.

In conclusion, metabolic profiling and stable isotope tracing experiments in blood from healthy subjects treated with AG-348 revealed strong homeostatic regulation of glycolysis even in the presence of activated PK-R. Analysis of data from a small number (n=5) of PK-deficient patients treated with AG-348 for two weeks showed that the three patients with Hb increases >1.0 g/dL also had increased glycolytic flux. While the small number of patients makes these results preliminary, it is the first demonstration of a direct link between increased red cell glycolysis induced by the PK-R activator AG-348 and the resulting hematological response as assessed by increases in Hb levels. Updated analyses including additional patients will be presented as more data are collected in the ongoing study.