Glucose Availability as a Promoter of Cancer Growth
Taken together, increased glucose flux and metabolism promotes several hallmarks of cancer such as excessive proliferation, anti-apoptotic signalling, cell cycle progression and angiogenesis. It does so, however, at the expense of substrate inflexibility compared to normal cells. It is clear that the high proliferative phenotype can only be sustained as long as a steady supply of substrates for ATP production is available. Thus, with progressive tumorigenesis, cancer cells become more and more 'addicted' to aerobic glycolysis and vulnerable to glucose deprivation. Indeed, several studies have shown that malignant cells in vitro quickly lose ATP and commit apoptosis when starved of glucose.[78–80] Masur et al. showed that diabetogenic glucose concentrations (11 mM) compared to physiological ones (5.5 mM) lead to altered expression of genes that promote cell proliferation, migration and adhesion in tumor cell lines from several organs including breast, colon, prostate and bladder. Adding insulin to the high-glucose medium further enhanced proliferation rates by 20–40% and promoted activation of the PI3K pathway. The question is whether altered blood glucose levels have similar effects on tumor growth in vivo. Theoretically, low blood glucose might cut some of the most hypoxic tumor cells from their diffusion-limited fuel supply. Gatenby and Gillies originally proposed this mechanism as an explanation for necrotic areas often found within tumor tissue, but they later revised this hypothesis based on a mathematical model that predicted only a modest decline of glucose concentrations with distance from the closest blood vessel. There are, however, several lines of evidence pointing towards a strong correlation between blood glucose levels and tumor growth in vivo that might indicate other important effects mediated by glucose. For example, the reduction of plasma glucose levels in tumor-bearing animals induced through calorie restriction may be responsible, directly or indirectly, for the significantly prolonged survival compared to normal-fed controls.[83,84] In 1962, Koroljow reported the successful treatment of two patients with metastatic tumors by an insulin-induced hypoglycemic coma. Hyperglycemia, on the other hand, is a predictor of poor survival in patients with various cancers[22,26,86–88] and has been positively correlated to an increased risk for developing cancer at several sites including the pancreas, esophagus, liver, colon, rectum, stomach and prostate in large cohort studies.[25,89,90]
Indirect Effects of Glucose Availability
Besides delivering more glucose to the tumor tissue, hyperglycemia has two other important negative effects for the host: First, as pointed out by Ely and Krone, even modest blood glucose elevations as they typically occur after a Western diet meal competitively impair the transport of ascorbic acid into immune cells.[88,91] Ascorbic acid is needed for effective phagocytosis and mitosis, so that the immune response to malignant cells is diminished. Second, it has been shown in vitro and in vivo that hyperglycemia activates monocytes and macrophages to produce inflammatory cytokines that play an important role also for the progression of cancer[92–94] (see below). Third, high plasma glucose concentrations elevate the levels of circulating insulin and free IGF1, two potent anti-apoptotic and growth factors for most cancer cells. Free IGF1 is elevated due to a decreased transcription of IGF binding protein (IGFBP)- 1 in the liver mediated by insulin. Due to expression of GLUT2, the β-cells of the pancreas are very sensitive to blood glucose concentration and steeply increase their insulin secretion when the latter exceeds the normal level of ~5 mM. In the typical Western diet consisting of three meals a day (plus the occasional CHO-rich snacks and drinks), this implies that insulin levels are elevated above the fasting baseline over most of the day. Both insulin and IGF1 activate the PI3K/Akt/mTOR/HIF-1α pathway by binding to the IGF1 receptor (IGF1R) and insulin receptor (IR), respectively (Figure 2). In addition, insulin stimulates the release of the pro-inflammatory cytokine interleukin (IL)-6 from human adipocytes. Thus, it could be hypothesized that a diet which repeatedly elevates blood glucose levels due to a high GL provides additional growth stimuli for neoplastic cells. In this respect, Venkateswaran et al. have shown in a xenograft model of human prostate cancer that a diet high in CHO stimulated the expression of IRs and phosphorylation of Akt in tumor tissue compared to a low CHO diet. In colorectal, prostate and early stage breast cancer patients[23,98] high insulin and low IGFBP-1 levels have been associated with poor prognosis. These findings again underline the importance of controlling blood sugar and hence insulin levels in cancer patients. Dietary restriction and/or a reduced CHO intake are straightforward strategies to achieve this goal.
Altered Nutritional Needs of Cancer Patients
Cancer patients and those with metabolic syndrome share common pathological abnormalities. Since 1885, when Ernst Freund described signs of hyperglycemia in 70 out of 70 cancer patients, it has been repeatedly reported that glucose tolerance and insulin sensitivity are diminished in cancer patients even before signs of cachexia (weight loss) become evident.[100–102] Both diabetes and cancer are characterized by a common pathophysiological state of chronic inflammatory signalling and associated insulin resistance. In cancer patients, insulin resistance is thought to be mediated by an acute phase response that is triggered by pro-inflammatory cytokines such as tumor necrosis factor (TNF)-?. and IL-6 In animal and human studies, removal of the tumor resulted in improved glucose clearance, suggesting that these cytokines are secreted, at least in part, from the tumor tissue itself.[104,105] The impact on the metabolism of the host is illustrated in Figure 3. In the liver, the inflammatory process leads to increased gluconeogenesis that is fuelled by lactate secreted from the tumor as well as glycerol from fatty acid breakdown and the amino acid alanine from muscle proteolysis. Gluconeogenesis is an energy-consuming process and might contribute to cancer cachexia by increasing total energy expenditure. Despite increased lipolysis, hepatic production of ketone bodies is usually not enhanced in cancer patients.[107,108] This is in contrast to starvation, where the ketone bodies acetoacetate and β-hydroxybutyrate counteract proteolysis by providing energy for the brain and muscles. In muscle, glucose uptake and glycogen synthesis are inhibited already at early stages of tumor progression, while fatty acid oxidation remains at normal levels or is increased.[110,111] In the latter case, more fat has to be provided from lipolysis in the adipose tissue. In addition, muscles progressively lose protein to provide amino acids for hepatic synthesis of acute-phase proteins and as precursors for gluconeogenesis. Thus, insulin resistance contributes to fat loss and muscle wasting, the two hallmarks of cancer cachexia. At the same time, it makes more glucose in the blood available for tumor cells.
Development of the cachectic state via sustained inflammatory signaling. Glucose metabolism in peripheral tissues is impaired already at early stages, while hepatic gluconeogenesis increases during tumor progression at later stages.
Fat and Ketone Bodies: Anti-cachectic Effects
It therefore seems reasonable to assume that dietary carbohydrates mainly fuel malignant cells which express the insulin-independent glucose transporters GLUT1 and GLUT3, while muscle cells are more likely to benefit from an increased fat and protein intake. This was summarized as early as in 1977 by C. Young, who stated that lipid sources predominate the fuel utilization of peripheral tissue of patients with neoplastic disease compared to healthy subjects. In addition, most malignant cells lack key mitochondrial enzymes necessary for conversion of ketone bodies and fatty acids to ATP,[40,113,114] while myocytes retain this ability even in the cachectic state. This led some authors to propose a high-fat, ketogenic diet (KD) as a strategy to selectively improve body composition of the host at the expense of the tumor.[113,115,116] The traditional KDs, which recommended protein and CHO to account, in combination, for roughly 20 E% (in the incorrect assumption that they were equivalent due to gluconeogenesis) and fat for the remaining 80 E%, have been widely used to treat childhood epilepsia since the 1920s. KDs are also used to treat adiposity and currently adult epilepsy. In the 1980s, Tisdale and colleagues investigated the effects of a ketogenic diet consisting mainly of medium chain triglycerides (MCTs) on two aggressive animal tumor models that were known to lack the ability to utilize ketone bodies. While the diet had no effect on rats bearing the Walker 256 sarcoma, it decreased the cachectic weight loss in proportion to its fat content in mice bearing the mousespecific colon carcinoma MAC16. For the latter, they further proved an anti-cachectic effect of a ketogenic diet in which the MCTs were replaced with long chain triglycerides (LCTs), although to a somewhat lesser extent. Contrary to LCTs, MCTs do not require transport in chylomicrones, but readily reach the liver where they are metabolized to yield high amounts of ketone bodies. Interestingly, administration of insulin was able to reduce the weight loss similar to the ketogenic MCT diet, but at the expense of a 50% increase in tumor size, which could be counteracted by addition of β-hydroxybutyrate in the drinking water. The supporting effect of insulin on tumor growth has been known since 1924, when Händel and Tadenuma described the nourishing effect of insulin on tumor tissue in an animal model, showing evidence that reducing insulin might reduce tumor growth.
Clinical Studies on Fat and Cachexia
Clinical studies investigating the anti-cachectic effects of high-fat diets are, however, rare. Fearon et al. administered a 70% MCT diet supplemented with β-hydroxybutyrate parenterally to five late-stage cachectic patients. After seven days on the diet, mean body weight had increased by 2 kg and their physical performance status had improved. Nebeling et al. investigated the effects of a MCT-based ketogenic diet taken ad libitum (60% MCT oil, 20% protein, 10% CHO, 10% other fats) on body weight and glucose metabolism in two pediatric patients with advanced-stage astrocytoma. Within 7 days on the diet, blood glucose levels had decreased to normal, while glucose uptake by the tumor estimated from FDG-PET scans had decreased by an average value of 21.8%. Notably, body weight remained stable throughout the study period of 8 weeks. In a randomized controlled study, Breitkreuz et al. showed that by supplementing the normal diet of 11 under-nourished, non-diabetic patients suffering from metastatic gastro-intestinal cancers with a fat-enriched liquid supplementation for 8 weeks, it was possible to reverse the loss of body weight and lean tissue mass and to improve several quality-oflife parameters in the treatment group, while the control group continued to lose body and lean tissue weight. The supplement contained 66% energy from fat, of which 45% were monounsaturated, 27% saturated (both LCT and MCT) and 28% polyunsaturated; mean energy intake ranged between 1000 and 2000 kcal/day and tended to be higher in patients receiving the additional fat drink.
Nutr Metab. 2011;8(75) © 2011 BioMed Central, Ltd.