Sugar fuels cancer


  • Richard T. Penson MD

    Corresponding author
    1. Division of Hematology Oncology, Massachusetts General Hospital, Boston, Massachusetts
    • Division of Hematology Oncology, Massachusetts General Hospital, Yawkey 9064, Massachusetts General Hospital, 32 Fruit Street, Boston MA 02114===

    Search for more papers by this author
    • Fax: (617) 724-6898

  • See referenced original articles on pages 1021-7, this issue.


Pre-op plasma glucose is a prognostic factor in ovarian cancer, and may reflect important metabolic and oncogenic pathways that link obesity and cancer.

Lamkin et al.1 report on the prognostic implication of pre-op glucose in ovarian cancer, identifying plasma glucose as an important prognostic factor. This nicely constructed study used an independent validation set, and it unequivocally confirms the association between higher plasma glucose and poorer survival, as has been reported in other malignancies. It is not clear whether the observation is predictive of a truly exploitable vulnerability, but it does appear to be prognostically important. The investigators did not include weight in the multivariate analysis to demonstrate that glucose is independent of this potentially significantly confounding prognostic factor. However, this observation very much fits our increasing appreciation of the molecularly driven phenotype of cancer.

Ovarian carcinoma cells express excess transporters that promote growth as part of the phenotype of an oncogene-addicted mutant,2 such as glucose transporter proteins and the folate receptor.3, 4 Minor degrees of aberration in multiple domains appear to characterize the more common epithelial cancers, rather than what is seen in some rarer tumors (eg, epidermal growth factor receptor [EGFR] mutant nonsmall cell lung cancer), sarcomas, and hematologic malignancies, that may depend on a single, necessary, and sufficient oncogene.

Curiously, there appears to be a “sweet new role for EGFR in cancer,” as the editorial for a very interesting new concept eloquently headlined in Cancer Cell recently. EGFR, a receptor tyrosine kinase associated with cell proliferation and survival, is overactive in many tumors of epithelial origin, and 30% of ovarian cancers. EGFR may, in part, function to prevent autophagic cell death by maintaining intracellular glucose level through interaction and stabilization of the sodium glucose cotransporter 1 (SGLT1).5 Although mutated EGFR does not appear to play a significant role in ovarian cancer progression, or be a valid target, perhaps a low-level, over-expression of this key epithelial receptor may play into multiple pathways promoting cell survival.6

Glucose undergoes oxidative phosphorylation in the matrix of the mitochondria, producing adenosine triphosphate (ATP) and pyruvate. Under aerobic conditions, pyruvate is broken down to carbon dioxide (CO2) and water (H2O), releasing energy in a cycle named for those who described it, Albert Szent-Györgyi and Hans Adolf Krebs, and for which the latter was awarded a Nobel Prize in 1953. Higher glucose levels stimulate insulin secretion, which drives glucose into cells, to produce the energy needed for macromolecular synthesis or high rates of proliferation.

Otto Warburg, the 1931 Nobel laureate in medicine described the seminal observation of metabolism in tumors, which utilize more anaerobic glycolysis (50% cf, 10%) and less oxidative phosphorylation.7 The cellular control of this balance, involves oncogenes and tumor suppressor genes that are also fundamentally involved in the initiation and progression of cancer. While the “Warburg effect” was initially thought to be caused by defects in oxidative phosphorylation as part of the malignant phenotype, a better understanding of the malignant genotype suggests that it is a “positive” adaption that fuels macromolecule synthesis and the growth and survival of cancer cells.8

In the western world, we eat a staggering amount of carbohydrate, with each of us consuming, on average, more than 100 pounds of sugar a year, and we all put on 10% more fat every decade. In the last 30 years, the prevalence of obesity in the United States has doubled to 30%.9 Obesity is now recognized as an increasingly important etiologic factor in cancer.10 Obesity causes insulin resistance, androgen excess, anovulation, and chronic progesterone deficiency. Obesity is all too often complicated by diabetes and the other problems of a sedentary lifestyle. This scourge of the aging western population, may have contributed to the risk of cancer, in part, by its effect on glucose, or what happens physiologically to control the glucose load.

Encouragingly, a number of studies are now reporting data to support the advice that losing weight counts.11 We should eat minimal animal fat, less red meat, and more antioxidants in fresh fruit and vegetables. Diet and exercise are simple but hard (15% to 50% attrition within a year) for those who commit to reshaping their lifestyle. Exercise may also be a way to impact cancer mortality: The Nurses' Health Study demonstrated that breast cancer patients who engaged in at least 3 hours of moderate physical activity a week after diagnosis had a 40% to 50% lower risk of cancer-related death.12

So, does sugar cause cancer? During the early 1970s, laboratory studies linked saccharin with the development of bladder cancer, but subsequent studies cast doubt on the findings, and it was removed from the US National Toxicology Program's Report on Carcinogens in 2000. Likewise, the concerns about a link between aspartame and lymphoma or brain tumors have not been substantiated. There is no clear evidence that sugar or artificial sweeteners directly cause cancer, and they present no intrinsic risk in our diet.13 However, a sweetened appetite may significantly exacerbate weight gain and contribute to adverse outcomes.

Diabetes is predictive of a poorer prognosis with many conditions, classically myocardial infarction, and 30-day mortality is twice as high (12.5% vs 6.2 %) in insulin-dependant diabetics.14 The association among sugar consumption, serum glucose, and hyperinsulinemia, or insulin resistance, remains controversial; however, there is, at least, some evidence that they are risk factors for cancer, particularly from a recent meta-analysis of case-controlled and cohort studies and the Health Professionals Follow-Up Study. The meta-analysis of 15 studies included more than 2,500,000 participants and estimated that the risk of colorectal cancer was 30% higher in diabetics (relative risk [RR], 1.30; 95% confidence interval [CI], 1.20-1.40).15 According to both the Nurses' Health Study and the Health Professionals Follow-Up Study, there was a similar increase in risk among men with high dietary glycemic load (RR, 1.32; 95% CI, .98-1.79).16

The relative contribution to the risk of cancer from glucose, insulin, or other linked growth factors is unclear and, to a degree, inseparable; but, at least, some excess risk comes from endogenous insulin, insulin-like growth factors (high insulinlike growth factor 1 [IGF1] and low IGF binding protein-3), and insulin therapy, with one case control reporting a doubling in the incidence of colorectal cancer in subjects who had used more than a 1 year of insulin (RR, 2.1; 95% CI, 1.2-3.4).17

Fasting serum glucose, obesity, and insulin resistance are also risk factors for cancer mortality. In a Korean study, 14,578 men, with a first cancer diagnosis, derived from a cohort of 901,979 male government employees and teachers, participated in a national health examination program in 1996. Subjects with a fasting serum glucose level above 126 mg/dL had a higher mortality rate for stomach (hazard ratio [HR], 1.52; 95% CI, 1.25 to 1.84) and lung (HR, 1.48; 95% CI, 1.18 to 1.87) cancer.18 Fasting insulin was associated with excess mortality in a cohort of 512 women, with early-stage breast cancer without known diabetes, who were in the highest (>51.9 pmol/L) versus the lowest (<27.0 pmol/L) insulin quartile (RR, 3.1; 95% CI, 1.7 to 5.7).19

One possible clinical scenario that may illuminate the overlap of oncogene, glucose, and cancer is the increased risk of cancer from excess growth hormone in patients with acromegaly. Although this is variable, it is likely larger than the risk of cancer associated with diabetes (RR, 2-14).20 The studies are hampered by their small size but hint at a synergy between the risks from growth hormone and insulin. Perhaps reassuringly, to date studies in small numbers of children with growth hormone for idiopathic short stature have not shown an excess cancer incidence. The risk of acromegaly is thought to be mediated by the tissue mediator of circulating growth hormone, IGF1.

Disentangling the overlapping risks of growth hormone, insulin, and glucose will more likely come from basic science than epidemiology. Certain molecular signaling pathways that appear to contain useful anti-cancer targets have provided clinical evidence of how closely the metabolic pathways may be linked. New agents that target phosphatidylinositol 3-kinases or its downstream targets v-akt murine thymoma viral oncogene homolog (AKT) and mammalian target of rapamycin (mTOR), inhibit tumor cell proliferation and survival, and, as a side effect, cause hyperglycemia and hyperlipemia. This pathway helps control cellular growth and metabolism. Craig Thompson, Chairman of the Department of Cancer Biology and Medicine at the University of Pennsylvania, who coined the term “metabolomics,” has investigated the hypothesis that animal biology has been driven by the scarcity of nutrients, and the requirement to control replicative capacity by apoptosis. Interestingly, the Thompson lab has shown that the antidiabetic drug metformin selectively impairs p53-deficient tumor cell growth. In vitro metformin treatment appears to force the metabolic conversion that p53-deficient cells, dependent on the Warburg effect, are unable to perform. A number of clinical trials are now examining this agent, alone and in combination, as a cancer therapeutic.

In the near future, we will very likely be evaluating some of the new agents presently in phase I clinical trial with 18F-2-fluoro-2-deoxy-D-glucose positron emission tomography imaging as one of the most exciting non-invasive ways to image and evaluate tumor biology, and the “clinical sugar cycle” is complete.

As is so often the case, we finish uncertain in the end. But, we are far from the end of the story, and the rapid expansion of overlapping fields is opening up our understanding of the fundamental processes of life and growth, and it is also showing us how much we don't know. Could it be that cancer is more similar to normal tissue than we ever thought, and the same pressures on growth, more profoundly important than we ever feared?