Inflammation and hyperinsulinemia

To understand the link between obesity and cancer, we must start with the changes that take place in adipose tissue with weight gain. Adipose tissue is composed of adipocytes and their progenitor cells along with endothelial cells, pericytes, monocytes, and macrophages.15 Excess energy intake leads adipose tissue to expand through either hypertrophy (increased adipocyte size), hyperplasia (increased adipose number), or both.16 As opposed to animal models in which de novo adipogenesis and hyperplasia are noted after excess energies lead to adipocytes reaching their upper volume limit, in adult humans adipocyte number and turnover stability indicate that hypertrophy may be the main mechanism to accommodate increased energy storage.16-18 Hypertrophic adipose tissue due to a relative deficiency of vasculature can be hypoxic and associated with necrotic-like abnormalities, thus impairing function, inducing inflammation, and increasing cell death.19, 20

Hypertrophic adipocytes are associated with increased expression and secretion of proinflammatory adipokines and cytokines, leading to recruitment of immune cells, such as macrophages.21 In obese participants, macrophages and other inflammatory cells may comprise up to 50% of adipose tissue cellular content.22 Increase in macrophages and other inflammatory cells further contributes to both local and systemic inflammation by leading to further increase in production of cytokines such as interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α).23 Cytokines such as TNF-α interfere with insulin signaling, leading to adipocyte insulin resistance and increased release of free fatty acids (FFA).24 Excess FFAs are taken up by nonadipose tissue, such as liver, skeletal muscle, and β-cells of the pancreas, leading to insulin resistance, steatosis, lipotoxicity, and metabolic dysfunction.24

In a chronic state, systemic insulin resistance is associated with higher insulin levels and eventual hyperglycemia. Studies performed in animal models of hyperglycemia in a low insulin state through apoptosis of β-cells, have noted that epigenetic modification at the level of DNA and chromatin induced by hyperglycemia can, themselves, stimulate tumor growth.10 Additionally, insulin resistance and hyperinsulinemia also leads to increased circulating levels of insulin-like growth factor-1 (IGF-1) through reduction in IGF-binding proteins and increased transcription and secretion of IGF-1.15 Activation of the IGF-1 insulin pathway can stimulate intracellular signaling through mitogenic-activated protein kinases and is correlated with development of a wide range of cancers including breast, colon, prostate, thyroid, liver, and pancreas.25 Downstream substrates involved in mediating the action of the insulin receptor including insulin receptor substrate-1 (IRS-1) and IRS-2 have also been linked with tumor cell proliferation and progression.25 Evaluation of a prospective cohort from the National Health and Nutrition Examination Survey was conducted to assess the impact of hyperinsulinemia on cancer outcomes.26 After adjusting for multiple variables, cancer mortality was significantly higher in individuals with hyperinsulinemia (hazard ratio, 2.04; 95% CI, 1.27–3.28; P = .004).

Warburg effect: Glucose dependence of cancer cells

In 1920's, Otto Warburg et al observed and reported that cancer cells metabolize glucose in a different manner from normal tissue.27 In the presence of oxygen, most differentiated cells tend to metabolize glucose through oxidative phosphorylation by initial conversion to pyruvate and subsequent entry into mitochondrial tricarboxylic acid cycle to generate adenosine triphosphate (ATP) and carbon dioxide. Under anaerobic conditions, pyruvate is converted to lactate-generating 2 ATP compared with 38 ATP via oxidative phosphorylation per glucose molecule.28 Regardless of the availability of oxygen, it seems that cancer cells are converting pyruvate to lactate, a term that has been coined aerobic glycolysis or Warburg Effect. This has been confirmed by positron emission tomography (PET) with fluorine-18 fluorodeoxyglucose showing increased update by most cancer cells.29 Further study has shown that net glucose uptake and lactate release by cancer tissue can exceed that of nonmalignant tissue by 30- and 43-fold, respectively.30

Much speculation persists as to why tumor cells would use alternative methods of energy generation aside from oxidative phosphorylation. Mitochondrial DNA mutations and alterations in the expression of nuclear-encoded mitochondrial proteins have been demonstrated in many types of cancer including head and neck, prostate, ovary, and liver cancers.30 The mitochondrial pathology seen in many types of cancer results in abnormalities in the number of mitochondria, structural abnormalities in mitochondrial cristae, and alterations in mitochondrial lipids, enzymes of the electron transport chain, and mitochondrial-associated membranes.31 These changes lead to dysregulated mitochondria that impedes oxidative phosphorylation, favoring other mechanisms of energy generation such as cytosolic glycolysis and substrate level phosphorylation reactions that involve transfer of a phosphate from a metabolic substrate to adenosine diphosphate in order to form ATP.31, 32 Additionally, production of lactate from glucose also occurs 10–100 times faster than oxidation and over a given time period; ATP generation can be comparable through both manner of glucose metabolism.33 In situations of limited energy resources, cells with a higher rate of ATP production may gain a selective advantage.34 The increased glucose consumption can also serve to support cell proliferation by using excess carbon for de novo generation of nucleotides, lipid, and protein.34 However, some aggressive tumors express the glycolytic pyruvate kinase M2 (PKM2) isoform, which produces less ATP than the PKM1 isoform.35 In these tumor cells, mitochondrial substrate level phosphorylation can compensate with the production of high-energy phosphates through the sequential metabolism of glutamine to glutamate, alpha-ketoglutarate, succinyl coenzyme A, and finally succinate.31, 36 In these cells, glutamine becomes one of the primary fuel sources for ATP synthesis through the glutaminolysis pathway, with glucose carbons used primarily for cell proliferation.

Cancer cells and other rapidly proliferating cells are also in need of nicotinamide adenine dinucleotide (NAD) phosphate (NADPH) and NADH to act as reducing agents in biosynthetic pathways such as de novo lipid synthesis.34 Increased glucose uptake by these cells allows for generation of these reducing agents through the pentose phosphate pathway, which oxidizes glucose to produce two molecules of NADPH and ribose-5-phosphate. NADPH can act as cofactor for thiol systems such as glutathione/glutathione peroxidase system, which are responsible for detoxifying H2O2 and other organic peroxides to prevent oxidative damage. Studies have confirmed the importance of pentose phosphate pathway by demonstrating increased activity in cancer cells, and noting increased oxidative stress in glucose deprived states.37

Through these mechanisms, aerobic glycolysis can continue to support biosynthesis and maintain a high flux of substrates through anabolic pathways. The ratio of NAD+/NADH and glucose metabolism can also influence histone acetylation and have wide-ranging effects impacting DNA repair and regulation of growth genes.38 The generation of lactate and its subsequent secretion also tends to impact the tumor microenvironment through a decrease in pH.39 The resulting acidic environment can be toxic to normal cells, leads to a degradation of extracellular matrix by proteinases, and can also increase angiogenesis, thus promoting local tumor invasion. In vitro studies evaluating tumor invasiveness have noted that in a heterogeneous acidic environment, tumor invasiveness was the highest in areas of lowest pH.39

Ketogenic diet and cancer

Ketogenic diet is traditionally a diet that consist of fat-to–carbohydrate and protein ratio of 3:1 or 4:1 by weight. More recently, adjustments to macronutrient composition including the addition of medium chain triglycerides have been made to increase tolerance and palatability. As the precise consistency varies across studies, ketogenic diet has been referred to as a diet that limits carbohydrate intake to a degree that results in the liver and, to an extent, the heart, gastrointestinal tract, and kidneys, oxidizing fatty acids to produce ketone bodies, which then become the predominant source of energy for the body. The three main ketone bodies include acetoacetate, β-hydroxybutyrate, and acetone. Typically, acetone is further metabolized to pyruvate, lactate, and acetate or breathed off giving the individual in ketosis a particular fruity breath.40 Acetoacetate and β-hydroxybutyrate are transported in the blood to other tissues such as the brain, in which they are converted to acetyl-CoA to enter the citric acid cycle. Typically, after 3 days of fasting, 30%–40% of total energy can be met by ketones. Some tissue such as the brain can obtain 60%–70% of their energy from ketones under fasting conditions.40

The benefit of ketosis in seizure disorder through prolonged fasting have been known since the time of Hippocrates.30 The term ketogenic diet was coined by Dr Wilder after he demonstrated that the benefits of fasting could be achieved through dietary modification and not just fasting.41 However, with the introduction of antiseizure medications, the use of ketogenic diet decreased until Dr Atkins generated greater interest in the 1970s using low-carbohydrate diets for weight loss. Given the association between obesity and diet to malignancy, there has been an increased interest in the use of ketogenic diet in cancer as adjuvant therapy. Preclinical and some clinical trials have demonstrated ketogenic diet can be inexpensive and relatively well tolerated with the potential to improve metabolic abnormalities, lower inflammation, limit tumor growth, and protects healthy cells from damage from chemotherapy and radiation.40

Many of the metabolic abnormalities that have been associated with obesity and may contribute to development of cancer are rapidly reversed under ketogenic diet (Figure 1). In a 12-week study in 40 participants with atherogenic dyslipidemia, carbohydrate-restricted diet (% carbohydrate/fat/protein of 12:59:28) compared with an energy-equivalent low-fat diet (56:24:20) led to a 10% weight loss, 12% reduction in glucose, and 50% reduction in insulin from baseline.42 More importantly, despite the high fat content, triglycerides were reduced by 51% and High-density lipoprotein (HDL)ok increased by 13%. Other long-term trials have confirmed similar findings. A 12-month trial in overweight adults with type 2 diabetes noted that a low-carbohydrate ketogenic diet (20–50 g/day of carbohydrates) compared with an energy-matched diet with 45%–50% of energies from carbohydrate resulted in more significant reduction in metabolic markers including glycated hemoglobin, more weight loss, and a larger reduction in diabetes medications.43

Summary of benefits of ketogenic diet in malignancy. ADP, adenosine diphosphate; ATP, adenosine triphosphate; GH, growth hormone; HCA2, hydroxy-carboxylic acid receptor 2; IGF-1, insulin growth factor-1; NAD, nicotinamide adenine dinucleotide; NADH, nicotinamide adenine dinucleotide hydrogen; NADP, nicotinamide adenine dinucleotide phophate; NADPH, nicotinamide adenine dinucleotide phosphate hydrogen; NMDA, N-methyl-D-aspartate; 5-P, 5-phosphate; PI3K, phosphatidylinositol-3 kinase; ROS, reactive oxygen species. Used with permission of Mayo Foundation for Medical Education and Research

In addition to these metabolic parameters, another target for ketogenic diet has been insulin-activated enzyme phosphatidylinositol-3 kinase (PI3K) (Figure 1). Many tumors have mutations in gene encoding this enzyme and PI3K inhibitors have been developed. Unfortunately, use of PI3K inhibitors results in hyperglycemia, which in turn raises insulin levels and reactivates this pathway, thus compromising the effectiveness of treatment.44 In mouse models of cancer, ketogenic diet along with metformin and sodium-glucose cotransporter 2 (SGLT2) were evaluated to assess if they could mitigate the hyperglycemia associated with PI3K inhibitors.44 In contrast to metformin, which had minimal effect, both SGLT-2 inhibitors and ketogenic diet decreased hyperglycemia and decreased release of insulin in response to PI3K inhibitors.

With a diet that limits intake of carbohydrate, glucose availability for glycolysis is limited, preventing the formation of pyruvate and subsequent conversion to lactate to generate ATP (Figure 1). Lipid and ketone bodies become the predominant source of energy and require the cell to use mitochondria, which may be dysfunctional in cancer cells as noted above. This combined with limited availability of glucose for the pentose phosphate pathway and generation of NADPH, leads to oxidative stress in cancer cells compared with normal cells.30 Evidence of this oxidative stress was demonstrated in mouse models of neuroblastoma through measurement of activation of energy sensor adenosine monophosphate (AMP)-activated protein kinase (AMPK).45 In response to ketogenic diet, they found higher activation of AMPK in tumor tissue but not in normal tissue. It is important to note that a few mitochondrial enzymes are key in metabolism of ketone bodies for energy, and tumor cells tend to express these enzymes in varying amounts. Zhang et al demonstrated that expression of genes encoding the two ketolytic enzymes 3-hydroxybutyrate dehydrogenase 1 (BDH1) and succinyl-CoA:3-oxoacid CoA transferase 1 (OXCT1) in tumor cells correlated with response to ketogenic diet.46

In addition to these metabolic benefits, ketogenic diets perhaps through action of ketone bodies can have a number of other benefits including modulation of signaling molecules, gene expression, as well as reduction in inflammation (Figure 1).40 Beta-hydroxybutyrate and acetone have been noted to modulate the signaling of N-methyl-D-aspartate (NMDA), which is physiologically relevant as NMDA receptor expression has been observed in various types of cancers.47 Similarly, hydroxy-carboxylic acid receptor 2 (HCA2) is activated by beta-hydroxybutyrate. HCA2 is described as a tumor suppressor and activates specific macrophages that have neuroprotective effects.48 Ketogenic diet and ketone bodies have also been associated with anti-inflammatory effects through reduction in cytokines such as TNF-α, IL-1, and IL-6.40 A recent study in a mouse model of colon cancer noted that a ketogenic diet was associated with lower tumor weight as well as plasma IL-6 levels.49 Additionally, blood ketone body concentrations were also negatively correlated with tumor weight. Ketogenic diet also reduces inflammation through suppression of NLRP3 inflammasome, which is a multiprotein complex that controls the activation of capase-1 and subsequent release of proinflammatory cytokines.50 Beta-hydroxybutyrate was shown to inhibit assembly of NLRP3 inflammasome.50

Despite these theoretical benefits, clinical trials with ketogenic diets are quite limited.51 A recent review of clinical trials regarding use of ketogenic diets in cancer noted that of the approximately 30 trials, the vast majority were case reports or pilot/feasibility studies, with most focusing on tolerability of ketogenic diet in this population as well as impact on body weight, glucose, and other metabolic paramaters.40 As an example, Champ et al retrospectively evaluated 53 patients undergoing treatment for high-grade glioma who had adequate glucose values recorded.52 Six of the 53 patients were on ketogenic diet during treatment with micronutrient composition of 77% fat, 8% carbohydrate, and 15% protein. They noted that the mean blood glucose for patients on a standard diet was 122 mg/dl, whereas that of patients on a ketogenic diet was 84 mg/dl and decreased from 142.5 mg/dl prior to initiation of ketogenic diet.

Cohen et al randomized 73 women with ovarian or endometrial cancer to either a ketogenic diet (energy from fat, protein, and carbohydrate of 70:25:5) or a diet recommended by American Cancer Society.53, 54 A total of 45 women completed their assigned 12-week diet intervention with 16 dropping out because of scheduling conflicts. After 12 weeks, ketogenic diet resulted in lower overall (35.3 kg vs 38.0 kg, P < .05) and central (3.0 vs 3.3 kg, P < .05) fat mass.53 There was also a 21.2% reduction in visceral fat mass with ketogenic diet compared with 4.6% with American Cancer Society diet along with a more significant reduction in insulin, C-peptide levels, glucose, and IGF-1 levels. Ketogenic diet also showed more significant improvement in physical function scores, reports of fatigue, as well as cravings for starchy foods and fast-food fat.

Klement et al also conducted a randomized control trial utilizing ketogenic diet in cancer patients.55 Their original study protocol was designed to enroll three cohorts including patients with breast, rectal, and head and neck cancer and included two interventions: ketogenic breakfast with a ketogenic drink and whole-food ketogenic diet. However, ketogenic breakfast drink was not tolerated by many, thus limiting the data to individuals with nonmetastatic breast cancer comparing either whole-food ketogenic diet (n = 29) or standard diet (n = 30).55 They noted that the ketogenic diet was well tolerated and resulted in body weight loss of 0.4 kg per week as well as fat mass. Although fat-free mass and skeletal muscle mass declined initially, it was subsequently preserved. Insulin and IGF-1 levels also decreased more in ketogenic diet group.

Ketogenic diet adverse effects

There are potential risks to ketogenic diet. Initially, some gastrointestinal distress, especially due to the high fat content of the diet, may be noted, although many studies have noted that these diets are subsequently well tolerated.30, 40 The composition of these diets may also raise concern for development of micronutrient and vitamin deficiencies and the need for supplementation. Hayashi et al noted that 6 months of ketogenic diet resulted in decline in trace minerals including selenium, zinc, and copper.56 Kidney stones have also been reported especially in children starting ketogenic diet for refractory epilepsy. McNally et al noted that only 2% of children treated with potassium citrate developed kidney stones while on ketogenic diet compared with 10.5% of those who did not receive potassium citrate.57 Additionally, ketogenic diet have also been noted to result in reduction in bone mineral density in children but may be due to developing skeleton.58, 59