As noted above, the weight loss associated with malignant disease involves a reduction of both body fat and body protein. In the normal adult mammal, proteins are continuously synthesized and degraded. This dynamic process, termed protein turnover, is regulated by a variety of hor- monal as well as nutritional factors. In the healthy adult, a state of nitrogen balance occurs, with protein synthesis equal to protein degradation. In the tumor-bearing host, this relationship is dis-
Figure 17.1 Postulated pathogenic mechanism involved in the development of cancer anorexia. In the tumor-bearing animal, plasma tryptophan (TRP) concentrations increase, facilitating the competition with other large neutral amino acids (LNAA) in crossing the blood-brain barrier via the specific L-transport sys- tem. BCAA, branched-chain amino acids; Phe, phenylalanine; Tyr, tyrosine; Met, methionine; CSF, cere- brospinal fluid. As a consequence of the successful competition of TRP across the blood-brain barrier, TRP concentrations in the central nervous system (CNS) increase, leading to enhanced serotonin (5-HT) synthe- sis in the hypothalamus, ultimately leading to anorexia. (Adapted from Laviano et al., 1996, with permis- sion of the authors and publisher.)
turbed, so that there is usually a net loss of nitrogen from nonmalignant tissues as protein degra- dation exceeds synthesis. The relationship of the host to the neoplasm with respect to protein and amino acid metabolism is depicted diagrammatically in Figure 17.2.
As depicted in the figure and noted in numerous studies, protein degradation in the cancer- bearing host occurs to the greatest extent in the musculature of the organism. In experimental systems, protein synthesis in muscle decreases (Paxton et al., 1987) while protein degradation in this same tissue increases dramatically, as measured by several different methods (Tessitore et al., 1987; Lazarus et al., 1999). In a mouse model, this enhanced protein degradation in muscle was associated with a rise in PGE2, one of the prostaglandins, (Smith and Tisdale, 1993). In turn, the protein degradation in muscle could be inhibited by administration of the ω-3 polyunsatu-rated fatty acid, eicosapentaenoic acid (Beck et al., 1991). These factors may be considered as a direct or indirect effect of the neoplasm on the host via known and unknown humoral influences (Figure 17.2). In the human, Emery et al. (1984) reported that protein synthesis in muscle could account for only about 8% of total body protein synthesis in cancer patients, as compared with
53% in normal control subjects. In contrast to skeletal muscle, protein synthesis in liver in- creased during cachexia (Paxton et al., 1987), although—in one experimental system—as tumor growth continued, protein degradation became predominant in that organ (Tessitore et al., 1987). Unlike these other tissues, the adrenals in experimental systems become enlarged and remain so during the entire experimental period (Tessitore et al., 1993). This is probably due to a general- ized stress response, both psychologically and physically, which in turn results in increased lev- els of gluconeogenic hormones such as glucagon and cortisol (cf. Hulsewé et al., 1997). Although overall protein synthesis in livers of cancer-bearing animals increased, there was no change in the synthesis of serum albumin, the major protein component of serum, in either hu- mans (Lundholm et al., 1980) or experimental animals (Ove et al., 1972); however, in tumor- bearing rats, albumin catabolism was increased, an effect abolished by adrenalectomy (Jewell and Hunter, 1971). But in patients with hepatocellular carcinoma, O’Keefe et al. (1990) demon- strated that amino acid incorporation into serum albumin and other serum proteins was higher than that in controls or in patients with hepatic metastasis.
As noted in Figure 17.2, the intermediary between muscle and liver and, in turn, the neo- plasm is the blood amino acid concentrations. In general, as noted earlier in this chapter, patients exhibit decreases in the concentrations of gluconeogenic amino acids, including the branched- chain neutral amino acids (Figure 17.1). This is in contrast to individuals exhibiting severe mal- nutrition, in whom the concentration of these amino acids is usually normal or increased (cf. Tisdale, 1997). As might be expected of a rapidly growing tissue, neoplastic cells under both in vitro and in vivo conditions have a high capacity for concentrating amino acids. Furthermore, tumor-bearing animals fed a protein-containing diet were actually observed to excrete less nitro- gen and consequently appeared to be in a more positive nitrogen balance than controls without neoplasms (cf. Goodlad, 1964). This is also true of human patients with advanced cancer (Bren- nan and Burt, 1981). In animal experiments, analysis of the tumors showed them to contain more nitrogen than was retained during the experimental period. Thus, part of the tumor protein must have been obtained at the expense of the tissues of the host. This ability of the tumor to draw directly on its host tissue proteins is also illustrated by the observation that tumor growth in ani- mals maintained on a protein-free diet was still about three-quarters of the rate observed in protein-fed animals. Observations such as these in both humans and animals have led investiga- tors, originally Mider et al. (1948), to describe the growing tumor as a nitrogen trap.
Although the original studies leading to the concept of the nitrogen trap were carried out with only a few neoplasms, there is substantial reason to believe that the concept is applicable to many forms of cancer. Some neoplasms appear to have special types of nitrogen traps in the form of enzymatic capacities for the selective degradation of essential amino acids (Pitot et al.,1961). When such a circumstance is present, the host becomes deficient in an essential amino acid (e.g., threonine) while the neoplasm retains its ability to synthesize protein and to grow. Another specific amino acid “trap” has been shown for the nonessential amino acid glutamine in experimental systems (cf. de Blaauw et al., 1997). While glutamine is the most abundant free amino acid in the organism, because of the increased protein synthesis and need for this amino acid as well as its concomitant breakdown by neoplasms, glutamine may become a “condition- ally” essential amino acid because the capacity for the endogenous biosynthesis of this amino acid is exceeded by tissue utilization (cf. de Blaauw et al., 1997). Stein (1978) has suggested that the progressive weight loss, abnormal gluconeogenesis, and lactate recycling seen in cachexia may be the result of an imbalance of essential amino acids within the host that results from se- lective uptake of essential amino acids. This imbalance may prevent the host from effectively decreasing gluconeogenesis as long as the tumor persists in causing an amino acid imbalance. The capacity of a neoplasm to act as a selective nitrogen trap is undoubtedly associated with the characteristic of relative autonomy of the tumor cell, but other factors may also play a role. Sidransky and Verney (1979) demonstrated that the livers of hepatoma-bearing rats do not re- spond to several nutritional stresses, such as a protein-free diet or tryptophan administration, as do the livers of non-tumor-bearing animals. Thus the neoplasm, directly or indirectly, may alter the regulatory mechanisms of normal tissues. On the other hand, it would appear that the break- down of protein in muscle is not necessarily due to an enhancement of protein degradation but rather to an inhibition of protein synthesis, leading to a greater net degradation of muscle protein (Svaninger et al., 1983).
Many neoplasms, especially those exhibiting relatively rapid growth during the process of cachexia, have a decidedly greater ability to concentrate many if not most amino acids within their cells (cf. Goodlad, 1964). Shortly after the administration of a labeled amino acid, the “pools” of amino acids within various normal tissues are labeled to a much greater extent than those in neoplastic tissue. Some years ago, Busch and associates (1961) showed that when la- beled amino acids were administered to animals in the form of proteins such as albumin, tumor proteins were found to have a much greater specific activity than proteins in nonneoplastic tis- sue. This suggests that the amino acids in the form of proteins are better substrates for protein synthesis in neoplasms and that tumors perhaps have a selective advantage in extracting whole proteins from the bloodstream. Furthermore, humans and animals with large or rapidly growing neoplasms usually exhibit hypoalbuminemia.
The amino acid requirements of tumors do not appear to differ significantly from those of normal tissues except with regard to nutritional requirements apparently associated with growth and possible specific abnormalities seen in certain neoplasms. A peculiarity of tumor nutrition has been exploited therapeutically by the use of the enzyme asparaginase to treat certain leuke- mias and lymphomas in which the tumor cells themselves require the amino acid asparagine for growth and survival (Uren and Handschumacher, 1977). The use of this enzyme to deplete the host of a specific amino acid that apparently does not significantly compromise normal tissue metabolism differs from other studies, in which diets missing certain amino acids were fed in an attempt to determine the effect of nutritional deficiencies on tumor growth. Other attempts have also been carried out with enzymes that selectively degrade other amino acids, such as phenyl- alanine and methionine, in experimental systems. However, unlike the asparaginase system, the amino acids degraded by these enzymes are essential for normal cells in general and, although inhibition of tumor growth was demonstrated, such preparations have not come into clinical use. Because of the metabolic importance of glutamine (see above), glutaminase prepared from bacterial sources has been used in both experimental and human systems with some success (Roberts et al., 1971; Spiers and Wade, 1976), but it has not found general use, possibly because of its toxicity. Furthermore, comparable experiments have been attempted with amino acid–deficient or –limited diets, again with variable results (Lorincz et al., 1969). On the other hand, the use of vitamin-deficient diets has received some attention, especially since at least one vitamin analog, methotrexate (an analog of folic acid), is used routinely in clinical chemother- apy (Jolivet et al., 1983). In this instance it has been demonstrated that, in experimental animals, the drug is more effective when folic acid–deficient diets are fed (Rosen et al., 1964). Patients receiving methotrexate for prolonged periods must also be carefully monitored to ensure that the drug therapy, while combating neoplastic growth, does not produce a folic acid deficiency.
Many patients with advanced cancer also exhibit significant anemia. Although the exact mechanism for this anemia is not clear at present, there is substantial evidence that humoral fac- tors produced by the neoplasm (Zucker et al., 1980) or by the host in response to the neoplasm (Ludwig and Fritz, 1998) result in a decreased production of the hormone erythropoietin. The decrease in the production of this hormone appears to be the general common denominator seen in cancer-related anemia (cf. Spivak, 1994) and is likely due to the effect of cytokines produced as a result of neoplastic interaction with the host. A diagram of the proposed pathophysiology of the anemia of cancer is seen in Figure 17.3. The nomenclature of the various cytokines indicated in the figure is discussed further in Chapter 19. Besides the general mechanism, there are other causes of anemia in cancer patients, such as direct blood loss, autoimmune hemolysis, cachexia, nutritional deficiencies, and the therapy of the neoplasm itself (Ludwig and Fritz, 1998).