Introduction to cancer and nutrition
Cancer is the leading cause of death worldwide. In 2018 alone, over 1.7 million people were diagnosed with cancer and over 600,000 deaths have resulted from this disease.
1 From 1999 to 2015, the overall incidence of cancer decreased by 2.2% in men and remained stable in women, yet the mortality rate for both has only decreased slightly to 1.8% and 1.4% respectively.
2 These modest improvements in outcomes could be attributed to improvements in cancer therapies and detection; however, it is tempting to speculate that greater improvements could be achieved if the therapeutic effects of diet on cancer are considered.
The last decade seems to have witnessed a surge of interest in the role of diet and its effects on cancer metabolism. This interest is at least partly ignited by the rapid increase in the prevalence of obesity in the United States. Between 2015 and 2016, the National Health and Nutrition Examination Survey demonstrated 39.8% of adults and 18.5% of youth were obese.
3 A prospective study of over 900,000 adults in the United States showed a significant proportional increase between obesity and mortality risk from multiple cancers, including of the esophagus, colon and rectum, liver, gallbladder, pancreas, kidney, prostate, breast, uterus, cervix, and ovary.
4 With the high prevalence of obesity in the United States, many investigators have sought to investigate the effects of excessive nutritional intake on the outcomes of cancer patients.
In general, cancer therapies target the hypermetabolic state of cancer cells to primarily destroying rapidly dividing tumor cells. However, rapidly dividing normal cells are also affected by anti-neoplastic therapies. As our understanding of tumor proliferation and apoptosis within individual cancers expands, investigators have been making significant strides to create targeted immunotherapies that will eliminate the bystander effect seen with traditional chemotherapy such as neutropenia, mucositis, renal failure, and cardiac dysfunction. The first example of immunotherapy was developed by Milstein and Kohler in 1970. Then, in 1997, rituximab, a monoclonal CD20 antibody, became available as a targeted treatment for non-Hodgkin’s lymphoma and has since become a widely utilized therapeutic option due to its short-lived side effects that can be managed medically for 3–6 months.
5 While targeted therapy remains the goal of oncologists and patients alike, investigators are interested in how lifestyle modifications, such as dietary pattern, can impact tumor physiology and clinical outcomes. The advent of personalized medicine, precision cancer medicine, or “theranostics” further supports a departure from broad algorithmic interventions and instead focuses on the matrix of a specific patient’s oncogenomic, pharmacogenomic, epigenetic, historical, and lifestyle characteristics. Nutritional status and diet are increasingly seen in this conception as significant matrix variables in predicting risk for disease, and response to intervention.
The role of diet in cancer metabolism is certainly an area of popular interest. The American Institute for Cancer Research and the World Cancer Research Fund estimates that 30%–40% of cancers can be prevented by healthy dietary regimens, improved physical activity, and maintenance of appropriate body weight.
6 While “prevention” is likely an overstatement, reduction of risk does appear to be supported by the evidence. It has been shown in epidemiological studies of breast cancer, prostate cancer, and colon cancer that migration to different countries influences overall risk of the development of these cancers, leading to hypotheses that changes in dietary habits may alter cancer risk. Kolonel et al.
7 demonstrated in 1980 that first-generation Japanese women had a threefold increase in breast cancer compared to Japanese woman living in Japan. It has been postulated that this difference may be partially explained by the switch from a primarily plant-based diet to a high-fat, high-sugar diet. Similarly, the mortality rate from stomach cancer in European migrants to Australia decreased in parallel with the length of time migrants stayed in Australia, but risk of colorectal cancer increased proportionally to the length of stay.
8 Prostate cancer studies have shown similar results where certain lifestyle factors are associated with the progression of malignancy.
9 These epidemiologic studies suggest that changes in lifestyle and dietary influences play a role in determining the risk of various cancers.
At the time of cancer diagnosis, many patients inquire about how lifestyle modifications can slow their tumor progression. Calorie restriction (CR) is a well-established dietary intervention for preventing cancer and increasing lifespan in experimental animal models.
10 In a prostate cancer xenograft mouse model, CR alone was shown to decrease final tumor weight, plasma insulin, and insulin-like growth factor (IGF)-1 levels and increase apoptosis, overall suggesting that decreasing caloric intake may reduce tumor proliferation.
11 In human studies, a 15% caloric reduction over 4 years demonstrated a sustained reduction in plasma growth factors and hormones, which have been associated with increased risk of cancer.
10,12 The CALERIE study revealed that it was feasible for patients to comply with a 25% CR intervention over a 2-year period and later showed a reduction in markers of oxidative stress, which may inhibit cancer proliferation.
13,14 Another diet of interest is the vegan diet, which has been shown to decrease tumor markers and inhibit tumor cell growth in prostate cancer studies.
15 Optimization of dietary regimens has thus gained attention as a lifestyle modification that individuals may undertake to alter their risk of the development of malignancies.
Certain malignancies have well-documented associations with lifestyle habits, such as lung cancer with smoking and mesothelioma with asbestos exposure.
16,17 Although primary prevention of tumorigenesis is a shared goal by patients and oncologists alike, the majority of cancers do not have clear risk factors, thus posing challenges in educating patients about the proper exposures to avoid. With regard to lifestyle modifications and cancer risk, the general dietary guidelines published by the American Cancer Society advocate for the avoidance of excess weight gain, consumption of a primarily vegetarian-based diet, and limited intake of alcohol, red meat, and processed foods.
18,19 Here, we present a review of the expansive literature on the interactions between nutrition, diet, and the course of malignant neoplasms.
Cancer metabolism
Cachexia is an integral component of involuntary weight loss in the setting of disease-associated wasting. It is further and still broadly defined as an inflammatory-associated wasting of protein, particularly skeletal muscle, and loss of energy stores. Cytokines and hormones, such as leptin, insulin, several interleukins, and growth factors, elicited in disease and as principal mediators of inflammation may in effect tilt physiology toward catabolic breakdown of tissue presumably in service of mobilizing critical nutrients for central nervous system and aspects of immune function.
20
Cachexia has been shown to be associated with cancer-related mortality; however, no consistently effective therapies have been developed to prevent or hamper its progression.
20 Even for patients who are able to eat—appetite suppression or anorexia is a common cachexia symptom—and efforts to improved nutrition often offer little respite.
21 It is interesting to speculate that underreporting of cachexia as a “contributing cause of death” has perennially affected the primacy placed on research and funding of this enormously pervasive process.
As a special case, cancer-induced cachexia is described as a multifactorial metabolic disorder seen in 50% of cancer patients.
22 The exact mechanisms remain unclear, but it is thought to be typified by an increase in energy expenditure, hepatic gluconeogenesis, fat lipolysis, and skeletal muscle proteolysis leading to progressive weight loss throughout therapy.
23,24 Further complicating this phenomenon, as noted previously, is that this auto-catabolic process is not reversed by total parental nutrition or supplements.
25 It is important to note that the relationship between tumor phenotype to host genotype likely results in heterogeneity in the degree of cachexia between individuals with the same malignancy.
Even prior to diagnosis, most cancer patients will experience a significant degree of catabolism leading to both muscle and adipose depletion, in contrast to the predominant wasting of adipose tissue in anorexic patients.
26 This difference may be due to the change in resting energy expenditure (REE), which is defined as the amount of energy expended by a person at rest. A recent meta-analysis of 27 studies comparing 1453 cancer patients to 1145 control patients shows that, on average, the REE of cancer patients was 9.66% higher than that of controls.
27 It is important to note that different malignancies are known to have varying degree of metabolism, and one study found marked REE variation among different cancer groups (p < 0.001).
27 For example, Burkitt’s lymphoma has a doubling time of 24 h,
28 while Hodgkin’s lymphoma has a doubling time of 29 days.
29 Therefore, the specific type of cancer may have large effects on the overall metabolism within the patient.
It was first described in 1927 by Otto Warburg et al.
30 that tumors cells demonstrate an aberrant metabolic pathway that contributes to excessive catabolism. In normal tissues via oxidative phosphorylation, one molecule of glucose is able to produce 36 adenosine triphosphate (ATP) molecules, which are the intracellular energy used to drive cellular processing. Cells also have an alternative pathway known as glycolysis, which utilizes one molecule of glucose to produce just two ATP molecules. In a phenomenon known as the Warburg effect, tumor cells selectively upregulate key regulators of glycolysis, such as hypoxia-induced factor-1α (HIF-1α).
31 HIF-1α drives three main processes: (1) increased expression of glucose transporter-1 (GLUT-1) leading to increased glucose availability for glycolysis;
32 (2) increased pyruvate dehydrogenase kinase production leading to increased conversion of lactate to pyruvate and decreased production of acetyl CoA necessary for the Krebs cycle (produces one ATP and three nicotinamide adenine dinucleotide (NADH));
33 and (3) increased lactate dehydrogenase production leading to increased conversion of lactate to pyruvate, which serves as the primary substrate for gluconeogenesis in the liver.
31 Increased lactate acid production via glycolysis is also released systemically, where it undergoes gluconeogenesis in an ATP-consumptive process known as the futile lactate–glucose shunt, or Cori cycle.
34 Overall, these processes promote an overall energy deficient state, leading to lipolysis and proteolysis and eventually cachexia. A recent in vivo study with human pancreatic cancer cells in athymic mice reveals that the Warburg effect was present in the complex system of a live animal, thus further supporting these proposed mechanisms of cancer-induced cachexia.
24 After 90 years since the Warburg effect was proposed, it remains unclear if the Warburg effect acts independently or if physiologic interactions in response to the Warburg effect drive cancer-induced cachexia.
Systemic inflammation that occurs during tumor proliferation is thought to contribute to cachexia.
35 Elevation in pro-inflammatory markers, such as C-reactive protein and fibrinogen, has been positively correlated with the degree of muscle wasting in cancer and chronic disease.
36,37 Cytokines, including tumor necrosis factor-alpha (TNF-α), interleukin-1(IL-1), and interleukin-6 (IL-6), contribute to persistent inflammation and increase rates of gluconeogenesis, lipolysis, and proteolysis.
38 TNF-α also stimulates the expression of uncoupling proteins 2 and 3 (UCP2/3) in cachectic skeletal muscle, due to UCP’s mediated proton leakage, and decreases the coupling of respiration to ADP phosphorylation, thereby generating heat instead of ATP from brown fat.
37,39 In fact, IL-6 levels have been found to be elevated in cachexic patients when compared with patients who maintained their weight during therapy.
37 These mechanisms have been replicated in animal studies, where IL-6 blockade in mice leads to attenuation of protein degradation
40 and TNF-α blockade decreases catabolism in rat models.
41,42
The role of neuropeptides in inducing anorexia during a catabolic state is emerging as a major contributing factor to cancer-induced cachexia. There appears to be an imbalance between appetite-stimulating molecules, such as neuropeptide P (NYP), melanin-concentrating hormone, orexins, endogenous opioids, and cannabinoids compared with appetite suppressant molecules such as serotonin, peptide YY, cholecystokinin, leptin, and insulin.
43 The role of neuropeptide Y is of special consideration as it is produced primarily in the hypothalamus and has a robust feeding stimulatory effect. Overstimulation with NYP leads to obesity in rats, in addition to decreased energy expenditure and reduced brown fat thermogenesis.
43 However, NYP injection into rats with sarcomas showed a reduction in feeding behavior, and a separate study revealed that rats with malignancies demonstrated a reduced release of NYP, suggesting that tumors may induce neurohormonal changes that decrease feeding activity and compound weight loss in cancer patients.
44–46
Another hormone of interest is leptin, a protein primarily produced in white and brown adipocytes that is responsible for peripheral signaling to reduce appetite and increase energy expenditure.
47 While decreased food intake normally suppresses leptin production, anorexia has been associated with increased leptin levels in adipose tissue and plasma.
48,49 Multiple authors have reported low or undetectable circulating leptin levels in cancer patients, suggesting a degree of leptin dysregulation.
50 Elucidating the mechanisms behind neurohormonal dysregulation of leptin may provide insight into further therapeutic options in the future.
Anti-metabolic drugs and their impact on nutritional deficiencies
Cancer chemotherapy regimens target rapidly dividing cells by either inhibiting DNA or protein synthesis or restricting essential micronutrients, thus leading to cellular death. Further, studies show that between 30% and 90% of patients have inadequate dietary regimens.
51,52 Chemotherapy, along with poor nutritional intake during treatment, is a great concern for patients and physicians. Chemotherapies that affect the nutritional status of patients include anti-metabolites, such folate, purine, and pyrimidine analogs, in addition to platinum-based drugs. Due to the broad mechanism of action of these drugs, many organ systems are adversely affected.
Methotrexate is an anti-metabolite folate analog that is used in standard therapy protocols for leukemia, lymphoma, and osteosarcoma.
53–55 Methotrexate leads to cellular death by acting as a dihydrofolate inhibitor, which is responsible for the conversion of dihydrofolate to tetrahydrofolate. Inhibition of tetrahydrofolate formation reduces the availability of one-carbon fragments necessary for the production of purines and the conversion of deoxyuridylate to thymidylate for DNA synthesis and cell reproduction. Although rapidly dividing malignant cells are most affected by inhibition of DNA synthesis, other rapidly proliferating cells are affected as well, such as hematopoietic and epithelial cells, thus leading to myelosuppression and mucositis.
56 Due to these side effects, a leucovorin “rescue” is performed 24 h after methotrexate is administered. Leucovorin is a derivative of tetrahydrofolic acid, which does not require dihydrofolate reductase activity and thus “rescues” the healthy cells by allowing for de novo DNA synthesis to resume.
57 Although methotrexate is associated with favorable mortality outcomes in combination with other chemotherapies, it is difficult to predict one’s toxicity after treatment. Therefore, dose reductions are often necessary due to severe side effects from inhibition of DNA synthesis and cellular proliferation.
5-Fluorouracil (5-FU) is a competitive inhibitor of thymidylate synthase, which blocks thymidine synthesis and inhibits DNA and RNA replication.
58 5-FU is widely used in colorectal, breast, and head and neck cancers.
58 Interestingly, increased intracellular folate levels increase thymidylate synthase levels, thereby enhancing the inhibitory effects of 5-FU. This interaction between metabolites suggests that optimal nutrition can improve chemotherapy effects in specific cancer groups.
59,60 Gemcitabine has a similar mechanism of action with 5-FU; however, it is a prodrug that, when phosphorylated, interrupts DNA synthesis and inhibits DNA repair, leading to cellular death.
56 In pancreatic cancer, it has been shown that intravenous omega-3 fatty acids, along with gemcitabine, improve quality of life scores with regard to chemotherapy-induced side effects.
61 The relationship between targeted supplementation and chemotherapy optimization suggests that optimized nutrition and reduction of chemotherapy effects could be dependent on specific treatment regimens.
Platinum-based chemotherapy drugs such as cisplatin, carboplatin, and oxaliplatin are used in the treatment of cancers such as testicular, ovarian, and lung cancer, in addition to osteosarcoma and neuroblastoma.
62 After the platinum-based drug is incorporated into the cell using copper transporters, it binds at the N7 position of guanine, causing cross-linking between adjacent guanine. This leads to failed DNA repair mechanisms and eventually cellular apoptosis.
63 Two studies in patients with solid tumors showed that selenium supplementation, when given during cisplatin therapy, reduced myelosuppression and nephrotoxicity, suggesting that optimal levels of selenium could aid in the toxicity profile related to platinum-based therapies. It was not reported whether mortality or relapse rates were affected by this type of supplementation, and thus, more studies are warranted to elucidate these data (
Table 1).
64,65