In 1981, Doll and Peto published a widely quoted estimate that 35% of all cancer deaths may be avoided by changes
in diet (7). Willett updated this estimate, narrowing the confidence interval but still concluding that about 32% of all cancer in the United States may be avoided by dietary modifications (range, 20% to 42%) (8). Given insights concerning the relationship between obesity and cancer risk, an updated estimate likely would be higher (9). Especially strong associations are observed between diet and cancer risk for colorectal cancer, breast cancer, prostate cancer, pancreas cancer, endometrial cancer, and gall bladder cancer (8). These data are of even greater significance in light of improved trends in cancer survival. More patients are surviving cancer and are therefore at risk for developing second primary cancers. These cancer survivors are an accessible and receptive group for education about dietary cancer prevention and secondary interventions.

Anorexia and weight loss are frequent findings in patients with cancer. As many as 40% of patients with cancer present with weight loss, and the prevalence of the cancer cachexia syndrome (CCS; see elsewhere in this volume) is as high as 80% in those with advanced malignancies (10, 11). The extent of weight loss at the time of diagnosis is of prognostic significance. For any given tumor type, survival is shorter in patients with significant pretreatment weight loss (more than 10% of usual body weight) (Table 88.1) (10, 11). A weight loss of greater than 2.75% per month has been linked to decreased survival (12). Additionally, weight loss is a significant contributor to symptom distress in patients with cancer. Changes in body image and associated fatigue can contribute to depression and social withdrawal. Observation of these changes in a loved one also may have profound effects on family and friends (13, 14, 15). Early recognition of these consequences of weight loss may afford the best opportunity to prevent the debilitating consequences. These issues may be especially problematic in children and the elderly (16). In the therapeutic armamentarium, surgery is the modality most frequently used to actually cure cancer. Numerous studies dating back at least 75 years demonstrate the increased morbidity and mortality of major operations in malnourished patients (17, 18, 19, 20). A study estimated a fivefold increase in mortality in underweight patients (body mass index [BMI] <18.5 kg/m²) undergoing major intra-abdominal cancer surgery (21). In fact, some patients may not be candidates for potentially curative cancer surgery because of the overwhelming risk of life-threatening complications as a result of malnutrition.

Overnutrition also is being seen more frequently in patients with cancer as a result of the rising incidence of overweight and obesity in the United States and the link between obesity and cancer risk. Its impact on patients with cancer is only just being elucidated. Cancer is usually thought of as a wasting disease. Weight loss and cachexia are hallmarks of cancer, in much the same sense that tuberculosis was referred to as “consumption” because of the gradual but relentless undernutrition that developed during the course of the disease and marked its progression toward lethality. Although this remains true for some cancers on presentation or as advanced, incurable disease develops, overnutrition also is becoming a significant problem in patients with cancer. Obesity is associated with higher risk of death from a variety of cancers, including cancers traditionally associated with wasting such as liver, pancreas, gastric, and esophageal cancer. Especially strong associations are seen in women for uterine cancer (6.25 relative risk [RR]), cervix cancer (3.20 RR), and breast cancer (2.12 RR). It is estimated that excess weight contributes to 14% of all cancer deaths in men and 20% in women (22). The causes of this are unclear, but many hypotheses seem plausible (23). Strong animal evidence and some human data exist that caloric restriction increases longevity and prevents cancer (24). Obesity may interfere with cancer detection, because physical exam findings can be masked. It is difficult to properly dose chemotherapy and plan radiation therapy in overweight and obese patients (25, 26). Surgical morbidity is higher in obese patients, and obesity may make it technically more difficult to perform precise surgery to assure adequate margins and lymph node harvest (21). Weight gain and physical activity are associated with decreased survival in a number of site-specific cancers, including breast and colorectal cancer, possibly as a result of the effect of adipose tissue on hormonal mediators such as estrogens, insulin, insulinlike growth factor-I (IGF-1), and adipokines, and the overall host inflammatory milieu (27).

Multiple metabolic and cytokine-induced changes, and clinical factors contribute to the development of malnutrition in patients with cancer (Table 88.2). They are detailed in the discussion of CCS. However, the interplay of these factors varies from patient to patient. Anorexia is a prominent contributor to weight loss in many patients with cancer. Its causes are complex and likely relate to an altered metabolic milieu resulting from cytokine and metabolic derangements (28, 29). Anorexia is not usually the primary cause of weight loss; it is a secondary effect that contributes to the downward cycle often observed in patients with cancer who lose weight. A number of lines of evidence support this contention. Parenteral nutrition (PN) support may be used to adequately deliver required energy and nutrients to undernourished patients with cancer, but this usually does not reverse the weight loss or the metabolic stigmata of CCS (30). The clinical and metabolic features of starvation and cachexia in humans differ markedly, suggesting decreased intake is not the underlying cause of cancer-associated weight loss. Additionally, cancer weight loss may precede changes in appetite (31). Finally, in some situations, the perceived anorexia is actually an adaptive decrease of food intake in response to weight loss (32). Therefore, merely telling patients to eat more and better is unlikely to reverse the presence of CCS.

Other factors that contribute to weight loss in patients with cancer include mechanical factors, the side effects of cancer therapy, and psychosocial factors. The psychological factors associated with cancer that may alter food intake include pain, anxiety, depression, and social isolation. Mechanical causes may be a direct effect of tumor, or may relate to complications of therapy. Tumors may cause obstruction of the gastrointestinal (GI) tract; they may involve or compress hollow viscus, altering gastric and small bowel compliance. Cancer and cancer surgery may be complicated by the development of GI fistulas, with resultant effects on nutrition status, nutrient absorption, and fluid and electrolyte balance. Symptoms that relate to these mechanical issues include alterations in taste, early satiety, pain, cramps, vomiting, diarrhea, and constipation, all of which may exacerbate anorexia. Cancer treatments may induce anorexia and weight loss. The postoperative state is invariably accompanied by a temporary catabolic state and decreased nutrient intake, which can be prolonged if surgical complications occur. Chemotherapy often induces transient nausea and vomiting or injury to GI mucosa with resultant stomatitis, mucositis, diarrhea, and/or typhlitis. These may be particularly severe in profoundly neutropenic patients, such as those receiving chemotherapy for leukemias and lymphomas and those undergoing high-dose chemotherapy with either autologous or allogeneic bone marrow reconstitution. Radiation therapy can cause acute GI injury. It also may cause chronic radiation enteritis with malabsorption and stricture formation. These side effects of treatment may cause many of the symptoms noted in relation to mechanical factors.

It has long been recognized that caloric restriction with a carefully balanced diet avoids specific nutrient deficiencies, increases maximum life span, and has cancerpreventive effects in mammals and primates; data suggest that this may also be true in humans. Purported mechanisms invoke adaptive changes to the stress of caloric restriction, including decreased oxidative stress, decreased plasma levels of inflammatory cytokines, reduced production of anabolic factors and hormones, and protection against age-related deterioration of immune surveillance. In addition, caloric restriction affects a variety of processes related to carcinogenesis, including DNA repair, and removal of damaged cells through apoptosis and auto phagy (36). The IGFs may be important mediators of these effects (37). Conversely, caloric excess may lead to opposite effects, facilitating carcinogenesis. These effects may be an important element in the link between cancer
and obesity (37). Caloric excess may also affect existent cancers. Numerous animal models demonstrate stimulation of tumor growth in animals receiving parenteral and enteral nutrition (EN). This also may be the case in humans. It appears that overall caloric load is a significant contributor to this effect (38).

Obesity effects endocrine regulation, in part through effects of adipose tissue metabolism on sex hormone levels and breakdown. This may account for the especially strong relationship between obesity and the risk of breast and endometrial cancer in women (22). Furthermore, the interrelationship of obesity and leptin gradually is being elucidated. Data suggest possible direct causal links between leptin and cancer (39).

Dysregulated cell proliferation is a hallmark of cancer. Cancer cell proliferation is sensitive to the presence of required nutrients. This has raised concern that feeding patients with cancer may have the unwanted effect of stimulating tumor growth. There have been only a few relevant studies in humans. Administration of PN in patients with cancer has been shown to effect tumor ploidy and cellular proliferation in (40, 41). This effect may depend on the composition of the nutrition support (42, 43). Other studies do not demonstrate these effects (38). It has been suggested that the potential proliferation stimulating effects of nutrition support actually may be used therapeutically to increase chemotherapy effectiveness. In this regard, one study of patients undergoing high-dose chemotherapy and stem cell transplant demonstrated a long-term survival benefit in patients receiving PN versus those who received a standard oral diet (44).

Apoptosis and autophagy are important cellular processes that help prevent malignant transformation and may be exploited to therapeutic benefit. Apoptosis is the process of programmed cell death by which damaged cells undergo “suicide,” in part to prevent damaged DNA from propagating malignant cell lineages. Autophagy involves the bulk degradation of cytoplasmic organelles and proteins. By orchestrating “cellular recycling,” autophagy plays key roles in the quality control of cellular components and supplies nutrients and materials for newly constructed structures in cells under metabolic stresses. The physiologic relevance of autophagy in tumor formation and progression is still controversial. The cytoprotective function of autophagy in cells subjected to starvation might enhance the survival of tumor cells that are often subject to metabolic stresses in vivo. A tumor-suppressive function of autophagy also has been suggested. The loss of autophagy may induce genomic instability, suggesting that intact autophagy pathways play a role in suppression of tumor formation and growth (45). In fact, the role of dysregulation of apoptosis and autophagy may rival that of proliferation in the malignant transformation of normal cells. Both processes are influenced by nutrient milieu. Carotenoids (e.g., lycopene), flavonoids (genistein), stilbenes (resveratrol), polyphenols (curcumin), and isothiocyanates all have been demonstrated to induce apoptosis in cancer cells preferentially over normal cells (46). Metabolic stress induced by nutrient starvation is the major trigger of autophagy identified to date. Cancer cells may be particularly sensitive to this trigger (47).

The essential features of cancer cells (dysregulated proliferation, apoptosis, and autophagy) are entwined with intrinsic, characteristic, metabolic differences in comparison with normal cells. The Warburg effect on carbohydrate metabolism has been alluded to previously. Similar protean changes in cellular metabolism are seen in lipid, protein, and nucleotide synthetic and degradative pathways (48). These unique characteristics of cancer cells potentially could be exploited to therapeutic benefit (49, 50). Particular attention has been focused on glutamine (GLN) in this regard (51). To date, however, no dietary changes have been demonstrated to be of benefit in patients with an established cancer.

Both organic and inorganic micronutrients may affect tumor growth and metastasis. For example, vitamins A and D both have effects on angiogenesis; optimization of the intake and tissue levels of these micronutrients potentially may be exploited to suppress tumor growth (52). Folate helps to regulate DNA expression through methylation. This may affect aging and the development of age-related cancers. Whether this may be exploited to treat established cancers is uncertain (53). Similarly, inorganic micronutrients such as selenium have been demonstrated to modulate cancer growth in vitro and in vivo (54). To date, however, no demonstrated effectiveness exists in humans with an established cancer.

Many nutrients have been postulated to influence cancer risk and progression through their effects on immune function. These effects are numerous and varied (55). So-called immune-enhancing EN formulas containing mixtures of GLN, arginine, nucleic acids, and omega-3 (n-3) fatty acids appear to improve surgical outcomes in patients with cancer but do not play an obvious direct role in cancer therapy. GLN is the most extensively studied single nutrient. Analysis of GLN metabolism in the tumorbearing host suggests that in patients with cancer GLN may be conditionally essential. Peripheral muscle stores of GLN are reduced in patients with cancer. It is hypothesized that GLN supplementation in patients with cancer may restore immunocompetence and gut barrier function
by providing substrate to GLN-requiring tissues, such as lymphocytes involved in cancer control, that are made conditionally deficient by the presence of a tumor (56).

Nutrigenomics may be defined as the interaction between nutrition and the genome. Genetic polymorphisms can alter the response to dietary components by influencing the absorption, metabolism, or site of action. Likewise, variation in DNA methylation patterns and other epigenetic events that influence overall gene expression can modify the response to food components and vice versa. Furthermore, variation in the ability of food components to increase or depress gene expression may account for some of the observed inconsistencies in the response to food components. This nascent field is likely to play an increasingly important role in the use of diet and nutrients to treat cancer (57, 58).

The metabolic changes seen in CCS are multiple and variable. They are generally characterized by decoupling of supply and demand, resulting in excessive substrate cycling and turnover (67).

The most striking feature of the energetics of the metabolic response to cancer is its variability. In comparison with control groups, patients with cancer may have reduced, normal, or increased energy expenditure (76, 77, 78, 79, 80, 81). The variability is in part caused by the heterogeneity of “cancer,” but is also likely owing to differences in host responses to tumor and to the presence of comorbid conditions such as infection. Estimation of energy needs in patients with cancer is problematic because of this heterogeneity in energy expenditure.

Decreased skeletal muscle mass is a hallmark of CCS (29, 67). An apparent failure exists in patients with cancer of the normal mechanism of protein metabolism adaptation seen during simple starvation (82). Despite protein depletion, protein turnover remains normal or is even increased. This appears to result from a combination of decreased synthesis and increased proteolysis. Proteolysis-inducing factor (PIF), detected in the urine of patients with cancer with cachexia, is associated with decreased plasma amino acid levels and decreased protein synthesis (83). PIF activates an RNA-dependent protein kinase, which in turn activates nuclear factor-κB (NF-κB). NF-κB in turn activates the ubiquitin proteasome proteolytic pathway. This NF-κB pathway is proposed as the primary proteolytic pathway in CCS (59).

Depletion of fat stores is a characteristic feature of CCS, accounting for the wasted, “skin-and-bones” appearance of many patients with cancer. Increased turnover of glycerol and fatty acids compared with normal subjects is observed. Glucose infusion fails to suppress lipolysis in patients with cancer (84). Adipose cells from cachectic patients demonstrate increased lipolytic activity (29). TNF-α may play a role in lipolysis by inhibiting lipoprotein lipase, thereby preventing the ability of adipocytes to extract fatty acids from circulating lipoproteins (i.e., low-density lipoprotein) (59). Lipid-mobilizing factor (LMF) also has been linked with increased lipolysis, increased free fatty acid turnover, and increased serum glycerol. LMF appears to increase lipolysis via increases in hormone sensitive lipase (59). Increased lipid levels in the blood seen in patients with cancer may help the host by fueling the generalized increased substrate turnover characteristic of CCS. Unfortunately, the same lipids also may be used by the tumor to meet essential requirements for polyunsaturated fats such as linoleic and arachidonic acids (85).

Alterations in carbohydrate metabolism are also commonly seen in cancer cachexia. Weight loss in CCS is often associated with glucose intolerance and diminished insulin sensitivity (86, 87). This may be a result of insulin resistance or decreased leptin levels or both (87, 88). Gluconeogenesis may be increased as a result of up-regulated Cori cycle activity in response to tumor production of lactic acid (89, 90). Increased hepatic gluconeogenesis is observed and may result from increased peripheral release of other glucose precursors, especially alanine and glycerol (91, 92). Cytokine mediators of CCS increase glucose demand, which induces gluconeogenic enzymes in the liver, further driving glucose synthesis (89). Host energy depletion can result from increased hepatic gluconeogenesis. Recycling of precursors to produce glucose is an energy-consuming process. The magnitude of this effect may be clinically significant in some patients (82, 93).

A myriad of chemical, metabolic, and clinical factors are implicated in the pathogenesis of CCS. This complexity goes a long way in explaining the historic intractability of CCS to clinical interventions.