by Stephen Western

Glycolysis, a metabolic process which occurs in all cells, is literally the enzymatic cleavage of a glucose molecule into two molecules of pyruvate, in a series of reactions which yield two molecules of ATP, the cellular energy storing molecule. In normal cells, pyruvate is further metabolized in the cell mitochondria through the process of aerobic respiration involving the citric acid cycle (Krebs cycle) and oxidative phosphorylation. This yields about 18 molecules of ATP per molecule of pyruvate. Clearly, aerobic respiration within mitochondria is a far more efficient ATP generating process than glycolysis, and glycolysis typically serves as a prelude, providing the pyruvate for further mitochondrial metabolism.

In contrast, the metabolism of cancer cells is characterized by an increased rate of glycolysis, and a decrease in mitochondrial respiration, even in the presence of oxygen. This is termed aerobic glycolysis, or “the Warburg effect”. This increased glycolysis typically begins as an adaptation to low oxygen conditions (hypoxia), as glycolysis is not an oxygen-requiring process. In this case, the pyruvate resulting from glycolysis is further metabolised to lactate, which is then ejected from the cell, acidifying the extracellular environment. In addition to sparing oxygen, this increased glycolytic rate in cancer cells also provides carbon for macromolecular synthesis, an essential requirement for cell proliferation.


Dichloroacetate (DCA) has been used for decades to treat metabolic disorders involving lactic acidosis. In simple terms, DCA shifts the balance of pyruvate metabolism towards energy (ATP) production in the mitochondria and away from conversion into lactate outside the mitochondria. More precisely, DCA inhibits the activity of pyruvate dehydrogenase kinases (PDKs), allowing increased activity of pyruvate dehydrogenase, and the further metabolism of pyruvate within the cell mitochondria. The resulting enhancement of mitochondrial function restores the apoptotic potential of the cell (the cell's ability to die) (1, 2).

Avastin and DCA work together in vivo

A study (14) published in 2013 tested the effects of bevacizumab (Avastin), DCA, or both drugs combined in athymic nude mice implanted with subcutaneous U87 and U118 glioblastoma tumours. First, the effects of bevacizumab alone were studied. Application of the drug led to increased expression of the two hypoxia markers, hypoxia-inducible factor 1 alpha (HIF-1a) and carbonic anhydrase IX (CA9). Downstream HIF transcriptional targets involved in glycolysis, including glucose transporters GLUT1 and GLUT3 and pyruvate dehydrogenase kinases 1 and 3 (PDK1 and PDK3), were strongly increased. In contrast, genes involved in mitochondrial oxidative metabolism were repressed, including pyruvate dehydrogenases and members of all five OXPHOS complexes. These results show that in this model, bevacizumab leads to lower oxygen levels (hypoxia) and a compensation for this by increasing glucose flux and glycolysis at the expense of oxidative mitochondrial metabolism. In the mice, tumours became completely resistant to bevacizumab by about 40 days after implantation.

Next, the researchers tested DCA and the combination of DCA and bevacizumab in the same mouse models. While DCA alone led to relatively minor changes in tumour growth rate, the combination therapy strongly decreased tumour volume compared to untreated controls. In the U87 model this effect was significant before day 15, and in the U118 model, tumour growth was nearly completely stopped by the combination therapy, with only a very minor increase in tumour volume at day 105. When the tumours were analyzed for molecular markers, combination treated tumours had significantly less staining for Ki-67, a marker of proliferation, compared to untreated and bevacizumab-only treated tumours, suggesting a cytostatic mechanism for the combination therapy. As DCA is an available drug with modest side-effects, its use in combination with bevacizumab (Avastin) should be seriously considered by patients and researchers.

Phenylbutyrate and DCA combined

In one experiment (17), mice were given either saline, 250 mg/kg of phenylbutyrate, 250 mg/kg of dichloroacetate, or both drugs combined, by oral gavage. Both phenylbutyrate alone and DCA alone significantly increased the activity of pyruvate dehydrogenase complex in the brains of the mice. The largest effect was seen with the combination of both drugs, which significantly increased the activity of pyruvate dehydrogenase complex beyond the level of either drug alone. While this was not a tumour study, the increase in pyruvate dehydrogenase activity with DCA and phenylbutyrate would be expected to enhance mitochondria function in GBM cells, reduce the harmful output of lactic acid into the tumour environment, and increase apoptosis (cell death) of the cancer cells. As explained below, sodium phenylbutyrate could also disrupt tumour metabolism by reducing circulating levels of glutamine. While 27 grams of sodium phenylbutyrate per day has been considered the maximum tolerated daily dose, one recurrent anaplastic astrocytoma patient in a small clinical trial (18) was reported to have a complete response on lower doses of sodium phenylbutyrate (starting at 18 grams per day in three divided doses, later lowered to 9 grams per day, then 4.5 grams per day).

Who is most likely to benefit from DCA?

Glioblastomas are characterized by rapid growth, large areas of hypoxia (low oxygen), necrotic areas of anoxia (no oxygen) and extracellular acidification due to the conversion of pyruvate to lactate. For these reasons, glioblastomas are the most highly glycolytic gliomas, and most glioblastoma patients would likely benefit from DCA as did four of five patients in a small DCA pilot trial (2). DCA has never been tested formally in lower grade gliomas.

The presence in a tumour of HIF-1 and its glycolysis-related target genes may indicate who will benefit from DCA treatment. Several studies have measured HIF-1 levels in gliomas of different grades. In the first study (3), none of the eight grade I or six grade II gliomas were positive for HIF1a, while two of seven (29%) anaplastic astrocytomas and 15 of 42 (36%) of glioblastomas were positive. Another study (4) found that only two of ten (20%) anaplastic astrocytomas were positive for HIF-1a and glucose transporter 1 (GLUT-1), while all 11 of 11 glioblastomas were positive for these markers. A third study (5) found that 37.5% out of 24 cases of diffuse astrocytoma (grade II), 27.5% out of 40 cases of anaplastic astrocytoma, and 83.3% out of 42 cases of glioblastoma tested positive for HIF-1.

It thus appears that HIF-1 expression is very common in glioblastomas while relatively rare in lower grade astrocytomas. The relationship between HIF-1 expression and IDH1 mutations was not examined in these studies.

Other markers of glycolysis are similarly rare or absent in lower grade gliomas versus glioblastomas. As an example, one study (6) found that hexokinase 2 (HK2), an enzyme of the first stage of glycolysis, was expressed in eight of twelve (67%) glioblastomas, while five lower grade astrocytomas were negative. HK2 was found especially in the pseudopalisading cells around areas of necrosis in glioblastomas, an area associated with very low oxygen levels.

Another clue indicating a decreased rate of glycolysis in lower grade gliomas compared to glioblastomas comes from FDG-PET scans. FDG (fluorodeoxyglucose) is a glucose analog created by the insertion of a radioactive fluorine isotope onto a glucose molecule. FDG-PET scans are used in tumour imaging, essentially showing regions of increased glucose uptake. Many lower grade gliomas have an uptake rate of FDG similar to that found in normal white matter, making these tumours difficult to visualize by FDG-PET (7).

Glutamine depletion and sodium phenylbutyrate

A great number of cancer patients are aware of the role of glucose as a fuel for tumour proliferation, and many adopt dietary or pharmacological strategies to reduce or stabilize blood glucose levels. Less well known is the equally important role of glutamine, the predominant amino acid in the bloodstream, as a carbon source for proliferative cancer growth. Glutamine and glutamine-derived molecules may be used as a fuel source in mitochondria, and also may provide nitrogen for amino acid synthesis.

Sodium phenylacetate and sodium phenylbutyrate (Buphenyl) are both FDA approved drugs used clinically for treatment of hyperammonemia and urea cycle disorders. Both drugs have also been tested in cancer trials as cell differentiating agents. Phenylbutyrate is largely converted to phenylacetate after ingestion, which then reacts with glutamine in the liver and kidney to form phenylacetylglutamine, which is excreted in the urine. One of the effects is a lowering of circulating glutamine levels (15). Due to this ability, phenylacetate was proposed as an anticancer drug as far back as 1971 (16). Phenylbutyrate is considered a prodrug of phenylacetate.

For more information on these drugs see the Repurposed Drugs page.


  1. Anticancer drugs that target metabolism: is dichloroacetate the new paradigm? Papandreou et al. 2010.

  2. Metabolic modulation of glioblastoma with dichloroacetate. Michelakis et al. 2010.

  3. Expression of HIF-1alpha and iNOS in astrocytic gliomas: a clinicopathological study. Giannopoulou et al. 2006.

  4. Differential expression of HIF-1 in glioblastoma multiforme and anaplastic astrocytoma. Mayer et al. 2012.

  5. The expression of hypoxia-inducible factor-1 in primary brain tumors. Reszek et al. 2013.

  6. Hexokinase 2 is a key mediator of aerobic glycolysis and promotes tumor growth in human glioblastoma multiforme. Wolf et al. 2011.

  7. PET imaging of gliomas. Bansal et al. 2011.

  8. Transformation by the R-enantiomer of 2-hydroxyglutarate linked to EGLN activation. Koivunen et al. 2012.

  9. Increased mitochondrial activity in a novel IDH1-R132H mutant human oligodendroglioma xenograft model: in situ detection of 2-HG and alpha-KG. Navis et al. 2013.

  10. Lactate dehydrogenase A silencing in IDH mutant gliomas. Chesnelong et al. 2013.

  11. Individualizing chemotherapy using the anti-diabetic drug, metformin, as an "adjuvant": an exploratory study. Bradford et al. 2013.

  12. Metabolic alterations due to IDH1 mutation in glioma: opening for therapeutic opportunities? Mustafa et al. Letter to the editor. January 9, 2014.

  13. Systemic treatment with the antidiabetic drug metformin selectively impairs p53-deficient tumor cell growth. Buzzai et al. 2007.

  14. Dichloroacetate reverses the hypoxic adaptation to bevacizumab and enhances its antitumor effects in mouse xenografts. Kumar et al. 2013.
    READ ABSTRACT (Email me for a PDF copy)

  15. Phenylbutyrate-induced glutamine depletion in humans: effect on leucine metabolism. Darmaun et al. 1998.

  16. Phenylacetic acid as a potential therapeutic agent for the treatment of human cancer. Neish WJ, 1971.

  17. Differential inhibition of PDKs by phenylbutyrate and enhancement of pyruvate dehydrogenase complex activity by combination with dichloroacetate. Ferriero et al. 2015.
    READ ABSTRACT Email me for a PDF copy

  18. Complete response of a recurrent, multicentric malignant glioma in a patient treated with phenylbutyrate. Baker et al. 2002.
    READ ABSTRACT Email me for a PDF copy

This page was created on 12/08/2013 and last updated on 04/21/2019

Our privacy / cookie policy has changed.
Click HERE to read it!