cancer cachexia syndrome in head and neck cancer patients: part ii. pathophysiology

BASIC SCIENCE REVIEW

Robert L. Ferris, MD, PhD, Section Editor

CANCER CACHEXIA SYNDROME IN HEAD AND NECK CANCERPATIENTS: PART II. PATHOPHYSIOLOGY

Jonathan George, BA,1 Trinitia Cannon, MD,2 Victor Lai, MD,1 Luther Richey, BA,1

Adam Zanation, MD,2 D. Neil Hayes, MD,3,4 Carol Shores, MD, PhD,2,4 Denis Guttridge, PhD,5

Marion Couch, MD, PhD2,4

1 Doris Duke Clinical Research Fellowship, The Verne S. Caviness General Clinical Research Center,University of North Carolina School of Medicine, Chapel Hill, North Carolina 27599-70702 Department of Otolaryngology/Head and Neck Surgery, University of North Carolina School of Medicine,G0412 Neurosciences Hospital, Chapel Hill, North Carolina 27599-7070.E-mail:3 Division of Medical Oncology, Department of Internal Medicine, University of North Carolina School ofMedicine, Chapel Hill, North Carolina 275994 Lineberger Comprehensive Cancer Center, University of North Carolina School of Medicine, Chapel Hill,North Carolina, NC 275995 Division of Human Genetics, Department of Molecular Virology, Immunology and Medical Genetics,The Arthur G. James Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio 43210

Accepted 22 December 2006Published online 27 March 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/hed.20630

Abstract: Cancer cachexia is a morbid wasting syndrome

common among patients with head and neck cancer. While its

clinical manifestations have been well characterized, its patho-

physiology remains complex. A comprehensive literature search

on cancer cachexia was performed using the National Library of

Medicine’s PubMed. The Cochrane Library and Google search

engine were also used. Recent evidence and new concepts on

the pathophysiology of cancer cachexia are summarized. Tar-

geted therapies are presented, and new concepts are high-

lighted. Cancer cachexia is characterized by complex, multilevel

pathogenesis. It involves up-regulated tissue catabolism and

impaired anabolism, release of tumor-derived catabolic factors

and inflammatory cytokines, and neuroendocrine dysfunction.

These culminate to create an energy-inefficient state character-

ized by wasting, chronic inflammation, neuroendocrine dysfunc-

tion, and anorexia. VVC 2007 Wiley Periodicals, Inc. Head Neck

29: 497–507, 2007

Keywords: cancer cachexia and anorexia; muscle wasting;

weight loss

INTRODUCTION

Despite recent advances in treatment, clinicaldilemmas persist for the surgeon treating patientswith head and neck cancer with marked weightloss. At initial presentation, it is often difficult toaccurately determine the cause of lost weight. Thepatient may be experiencing weight loss because

Correspondence to: M. Couch

Contract grant sponsor: University of North Carolina’s General ClinicalResearch Center (GCRC); contract grant number: RR00046; Contractgrant sponsor: Doris Duke Clinical Research Program.

VVC 2007 Wiley Periodicals, Inc.

Cancer Cachexia Pathophysiology HEAD & NECK—DOI 10.1002/hed May 2007 497

of cancer-induced odynophagia, aging, immobility,anorexia, chemoradiation therapy, cancer ca-chexia, starvation due to mechanical obstruction,or malnutrition due to alcoholism. Of these, can-cer cachexia is the least understood cause ofweight loss, and yet it is responsible for the great-est morbidity among patients with head and neckcancer.

The clinical manifestations of this syndromehave been characterized previously.1 They includeskeletal muscle and adipose tissue wasting, exag-gerated systemic inflammation, and anorexia.Recent findings have shed light on the pathologicbasis for these manifestations. Skeletal muscleand adipose tissue wasting appear to be mediatedby dysfunction in several overlapping pathways.First, dysregulation of muscle-specific cellularcomponents such as the ubiquitin–proteasomesystem (UPS) and the dystrophin glycoproteincomplex (DGC) facilitates muscle catabolism.2,3

Tumor-derived catabolic factors and circulatingproinflammatory cytokines also augment muscleand adipose tissue catabolism. Finally, disruptionof neuroendocrine pathways leads to anorexia.4

The aim of this review is to clarify the mecha-nisms involved in the pathogenesis of cancercachexia and to describe potential targets for anti-cachexia treatment.

SKELETAL MUSCLE

The most prominent clinical feature in cancercachexia is progressive loss of skeletal musclemass. This may approach 75% reduction in skele-tal muscle protein mass, even with a total weightloss of only 30%.1 Loss of muscle protein plays amajor role in the shortened survival time of ca-chectic cancer patients, who experience a finalcommon pathway of progressive physical disabil-ity and impairment of respiratory function. Cur-rent data suggest that loss of muscle protein ismediated primarily by accelerated protein catabo-lism, but that impaired protein anabolism is alsoimportant. These appear to be 2 dynamic andinterrelated processes that ultimately account forthe progressive muscle wasting seen in cancercachexia.

ACCELERATED MUSCLE CATABOLISM

Three important and recently identified factorsplay central roles in skeletal muscle wasting inpatients with cancer cachexia. These are the ubiq-uitin–proteasome proteolytic system, the DGC,

and proteolysis-inducing factor (PIF). The UPS isan important pathway for muscle protein degra-dation that appears to be up-regulated in animalmodels of cancer cachexia. The DGC is a cytoarch-itectural framework in muscle cells whose dys-function may help initiate muscle wasting.Finally, PIF is a tumor-derived catabolic factorthat may initiate and augment muscle wasting inhumans.

Current research on the pathogenesis of can-cer cachexia posits that exuberant release ofproinflammatory cytokines, such as interleukin(IL-) 1b, IL-6, tumor necrosis factor alpha (TNF-a), and interferon gamma (IFN-g), causes skeletalmuscle proteolysis through suppression of musclegenes and activation of ubiquitin–proteasome-mediated proteolysis.4–6 These circulating proin-flammatory cytokines inhibit the expression ofmyosin heavy chain genes, leading to dissociationof myosin from its contractile apparatus in musclecells. Free myosin is degraded by the UPS, whosecomponents are activated by these same cyto-kines. Ubiquitin–proteasome-mediated muscleprotein catabolism is also inducible by PIF, a pur-ported tumor-derived catabolic factor.7 Currentresearch also suggests that muscle breakdown incachexia involves cellular dysregulation of theDGC, amembrane structure responsible formain-taining the functional integrity of muscle cells.3

Here we will describe 3 important components ofskeletal muscle proteolysis in cancer cachexia: theUPS, PIF, and the DGC.

Up-regulation of the Ubiquitin–Proteasome System.

The UPS is a 750-kDa tube-like structure foundwithin cells. This structure is hypothesized to bean important pathway underlying catabolic dis-ease states such as starvation, sepsis, denervationatrophy, severe trauma, and cancer cachexia.8

It is the major catabolic pathway in cancercachexia,9 and its overactivation has been repro-duced by nearly all rodentmodels of cancer-associ-ated muscle wasting tested thus far.10 Further-more, research on human subjects with cancerhas also demonstrated UPS overactivation.11

Most intracellular proteins in skeletal muscleare degraded through the UPS proteolytic system.Muscle protein degradation involves enzymaticmarking of proteins with multiple ubiquitin mole-cules in a multistep process that culminates inprotein degradation within the proteasome. In thefirst step of this marking process, multiple ubiqui-tin molecules are covalently conjugated to con-tractile proteins by ubiquitin-activating (E1),

498 Cancer Cachexia Pathophysiology HEAD & NECK—DOI 10.1002/hed May 2007

ubiquitin-conjugating (E2), and ubiquitin-ligating(E3) enzymes. In the second step, tagged proteinsare recognized and degraded within the 26S pro-teasome complex, whose catalytic core is linedwith proteolytic enzymes. The proteasome thenreleases oligopeptides that are rapidly degradedinto amino acids by cytosolic peptidases. Theseamino acids are then transported to the liver,where they are converted to acute-phase proteinsin an energy-wasting cycle that exacerbatesinflammation and disrupts physiologic proteinbalance (Figure 1).

Proinflammatory cytokines such as IL-6, TNF-a, and IFN-g have been shown to up-regulatemuscle-specific expression of important compo-nents of the UPS.8 Recent data suggest that cyto-kine-mediated induction of 2 muscle-specific ubiq-uitin ligase genes, muscle ring-finger 1 (MuRF1)and muscle atrophy F box (MAFbx), may be majorsteps the induction of in skeletal muscle atrophyin cachexia.12 Importantly, TNF-a- and IL-6induce muscle-specific expression of IKK (in-hibitor of NF-jB). This activates nuclear factorkappa B (NF-jB) and causes MuRF1 up-regula-tion, which results in severe muscle wasting inmice.13 Thus, the IKK/NF-jB/MuRF1 pathway isa cytokine-inducible signaling pathway thatappears to mediate skeletal muscle wasting incancer cachexia14 (Figure 2).

There appears to be a similar pathway forinduction of the ubiquitin ligase gene MAFbx andsubsequent activation of UPS proteolysis. In cul-tured myotubes undergoing atrophy, MAFbxinduction is associated with inhibition of thePI3K/Akt pathway, a pathway normally associ-ated with muscle hypertrophy. Activation of thispathway with insulin-like growth factor 1 (IGF-1)

suppresses the induction of MAFbx and MuRF1,preventing muscle wasting.15 This indicates apotential avenue for future therapy of cachexia.

Accelerated muscle catabolism in cancer cach-exia thus appears to result from cytokine-medi-ated induction of the muscle-specific ubiquitinligase genes MuRF1 and MAFbx and consequentactivation of UPS proteolysis (Figure 3). Thedeveloping understanding of this system createsnew targets for the treatment of cachexia. Geneknockout, MuRF1 and MAFbx suppression byIGF-1, and direct proteasomal inhibition mightbecome targets for anticachexia treatment in thefuture.

Induction of Muscle Breakdown by Proteolysis-

Inducing Factor. PIF is a 24-kDa glycoproteinproduced by tumor cells in both mice and humansthat has been hypothesized to be responsible forcancer cachexia.16 PIF has been found in manyhuman tumor types, including breast, ovarian,pancreatic, and colorectal tumors,1 and has beenisolated from the urine of patients with cachecticcancer with weight loss.17 Furthermore, PIFappears to be present in patients with tumor-related cachexia, but absent in cancer patients notlosing weight, or in weight-losing patients withbenign disease.18 Longitudinal studies haveshown that cancer patients expressing PIF in

FIGURE 1. The ubiquitin–proteasome system. Proteins like

myosin are marked for degradation in the proteasome by con-

jugation with polyubiquitin chains. Proteolytic enzymes in the

proteasome complex then degrade these ubiquitinated (Ub-)

proteins into peptides. Ub, ubiquitin.

FIGURE 2. Induction of the ubiquitin ligase MuRF1 and muscle

wasting in mice. Graph and Western blot showing the relation-

ship of MuRF1 induction with decrease in fiber diameter in mice

with colon-26 adenocarcinoma. Glyceraldehyde-3-phosphate de-

hydrogenase is a housekeeping gene used as a control. MuRF1,

muscle ring-finger 1, a gene for the ubiquitin ligase enzyme;

GADPH, glyceraldehydes-3-phosphate dehydrogenase.


their urine lose weight over time, while patientsnot expressing this factor gain weight.17 PIFadministration to mice and to cultured myocytesrapidly induces muscle catabolism, while PIF-induced weight loss in mice is reversed with anti-PIFmonoclonal antibodies.18

PIF induces muscle wasting by several mecha-nisms. First, it induces the muscle-specific releaseof 15-hydroxyeicosatetraenoic acid, activatingUPS-mediated proteolysis.18 PIF also stimulatesthe activation of NF-jB in tissue culture, whichfacilitates the release of proinflammatory cyto-kines and induces UPS-mediated proteolysis.8

Thus, by stimulating the release of proinflamma-tory cytokines and activating the UPS, PIFexpression generates a series of events culminat-ing in skeletal muscle wasting.

Dysregulation of the Dystrophin Glycoprotein

Complex. New research shows that cancercachexia may also involve the loss of an essentialmuscle protein complex. The DGC is a muscle-spe-cific protein manifold that anchors sarcomeremembranes in place and prevents them from beingtorn by shear forces produced during muscle con-traction. Recent findings suggest that dysregula-tion of the DGC mediates the development of can-

cer-induced muscle wasting.3 Muscles from colon-26 adenocarcinoma-bearing mice exhibited mem-brane abnormalities associated with reduced lev-els of dystrophin and increased glycosylation ofproteins within the DGC. The dysregulation of theDGC correlates positively with weight loss inpatients with gastroesophageal adenocarcinoma.3

Also, mutant mice lacking the protein dystrophinshow enhanced tumor-induced wasting, whiletransgenic animals expressing dystrophin—spe-cifically in skeletal muscle—are spared from dis-ease.3 These findings suggest that the loss of func-tion of the DGC may cause muscle wasting in can-cer cachexia.

Whether dysfunction of the DGC initiates,maintains, or promotes muscle breakdown in can-cer cachexia is currently unclear. Overexpressionof dystrophin appears to block the induction ofMuRF1 and MAFbx but does not affect NF-jBactivation in these muscles.3 This suggests thatdysfunction within the DGC may mediate UPS-dependent muscle breakdown in a manner inde-pendent of NF-jB. However, the exact mechanismof this dysfunction remains to be elucidated(Figure 4).

IMPAIRED MUSCLE ANABOLISM

Although accelerated skeletal muscle wasting ap-pears to be a primary mediator of cancer cachexia,another important feature is impaired skeletalmuscle anabolism. One study found that the totalbody protein synthesis in healthy individuals was53%, whereas in cachectic individuals it was only8%.19 Several pathways appear to be involved inreduced skeletal muscle anabolism in cachexia.These include imbalance in the physiologic aminoacid pool, reduced myosin expression, and up-reg-ulation of the gene regulator myostatin.

FIGURE 3. TNF-a-mediated activation of ubiquitinating

enzymes. TNF-a up-regulates the expression of the ubiquitin

ligase enzyme MuRF1 in an NF-jB–dependent manner. This

may be 1 mechanism by which proinflammatory cytokines

mediate muscle wasting in cachexia. This process appears to

involve several signaling factors that are part of the NF-jB path-

way, including the Ikk complex (with subunits a, b, and g), p65,and p50. TNF-a, tumor necrosis factor alpha; NF-jB, nuclear

factor kappa B; MuRF1, muscle ring-finger 1; TNFR, tumor ne-

crosis factor receptor; IjB, inhibitor of nuclear factor kappa B;

Ijj, inhibitor of nuclear factor kappa B kinase.

FIGURE 4. The dystrophin glycoprotein complex. Simplistic view

of proteins involved in the dystrophin–glycoprotein complex. This

shows how disruption in the muscle membrane complex can dis-

rupt the intramuscular cytoskeleton, leading to muscle breakdown.


Imbalance in the Amino Acid Ratio. Protein syn-thesis in patients with cancer cachexia isimpaired by an imbalance in the physiologicamino acid ratio. As mentioned previously, freeamino acids are released by the UPS following thebreakdown of skeletal proteins. These are thentaken up by the liver, where they are converted toacute-phase proteins in order to meet increasedenergy needs of chronic inflammation in cachexia.The large amount of amino acids consumed inskeletal muscle breakdown depletes the physio-logic reserve of amino acids available for skeletalmuscle synthesis. Without the proper ratio of freeamino acids, general protein synthesis is inhib-ited. Thus, increased skeletal muscle breakdownappears to alter the free amino acid ratio, therebyinhibiting protein synthesis in skeletal muscle.

Reduced Expression of Myosin. Reduced myosinexpression has also been implicated in impairedprotein anabolism in cachexia. Cachectic tumor-bearing rats exhibit decreases in myosin expres-sion thatmay be induced by circulating proinflam-matory cytokines.20 Recent studies of culturedmyocytes demonstrated that protein catabolisminduced by TNF-a and IFN-g suppresses the pro-duction of MyoD in an NF-jB–dependent manner(Figure 5).21 MyoD is a muscle-specific nucleartranscription factor that transcribes myosinheavy chain gene. Reduced MyoD expression is

known to deplete the pool of myosin heavy chain,resulting in diminished muscle protein synthesisand cachexia. Thus, cytokine-mediated inhibitionofMyoD production ultimately suppresses myosinheavy chain expression, preventingmuscle forma-tion and leading to atrophy.

Myostatin Up-regulation. Impaired skeletal mus-cle anabolism is also caused by overexpression of amuscle gene regulator known as myostatin. Myo-statin is a muscle-specific negative regulator ofskeletal muscle growth that can suppress musclecell proliferation and differentiation in an NF-jB–independent manner.22 Muscles of tumor-bearingmice exhibit significantly higher levels of myosta-tin than muscles in non–tumor-bearing mice,22

and transgenic mice with overexpression of themyostatin gene develop a cachexia-like syndromecharacterized by severe wasting.22 Although itsrole remains to be more accurately characterized,myostatin up-regulation appears to be an impor-tant factor in impaired muscle regeneration incancer cachexia.

To summarize, skeletal muscle atrophy inpatients with cancer cachexia is characterized byaccelerated catabolism and impaired anabolism.Accelerated muscle catabolism is mediated by up-regulation of the UPS proteolytic pathway, tumorrelease of the muscle catabolic factor PIF, and dys-regulation of the DGC. Impaired muscle anabo-lism is mediated by an imbalance in the aminoacid pool, reduced myosin expression, and up-reg-ulation of themuscle gene regulator myostatin.

ADIPOSE TISSUE

Cachexia also involves abnormalities in lipid me-tabolism, resulting in marked adipose tissue lossin the cachectic patient. Indeed, body compositionanalysis of lung cancer patients with loss of 30%or more of their premorbid weight showed an 85%fall in total body fat.23 Increased adipose tissue ca-tabolism, rather than impaired anabolism, appearsto be central to the etiology of fat loss in cachecticpatients. This is characterized by increased lipoly-sis, hypertriglyceridemia, increased hepatic secre-tion of very low density lipoprotein, increased denovo fatty acid synthesis, and a futile cycle of fattyacids between the liver and adipose tissue. Adiposetissue catabolism appears to be stimulated by a tu-mor-derived factor called lipid-mobilizing factorand by circulating proinflammatory cytokines inthe tumor-bearing host.

FIGURE 5. Down-regulation of MyoD by TNF-a and IFN-g.TNF-a and IFN-g activate intracellular signaling pathways to in-

hibit MyoD expression in an NF-jB–dependent manner. MyoD,

myogenic differentiation gene; IFN-g, interferon gamma; IFNR,

interferon receptor; TNF-a, tumor necrosis factor alpha; TNFR,

tumor necrosis factor receptor; IjB, inhibitor of nuclear factor

kappa B; Ijj, inhibitor of nuclear factor kappa B kinase; NF-jB,nuclear factor kappa B; p65, p65 transcription factor, a subunit

of nuclear factor kappa B; p55, p55 transcription factor, a subu-

nit of nuclear factor kappa B. [Color figure can be viewed in the

online issue, which is available at www.interscience.wiley.com.]


Lipid Mobilizing Factor. Lipid-mobilizing factor(LMF) is a 43-kDa lipolytic factor derived fromboth tumor and brown adipose tissue. It is homolo-gous with human protein Zn-a2-glycoprotein, andhas been isolated from the urine of both cachecticcancer patients and cachectic mice.24,25 It isknown to cause a selective reduction in body fat,24

and is thought to be responsible for atrophy of adi-pose tissue in cachectic patients. One study foundthat cancer patients with weight loss had detecta-ble concentrations of LMF in their urine, whilecancer patients without weight loss did not.26

LMF directly stimulates lipolysis by down-reg-ulating lipoprotein lipase (LPL) and up-regulat-ing hormone sensitive lipase.26 This results in ele-vated levels of glycerol and free fatty acids (FFAs).Although glycerol is cycled back to the liver toserve as a substrate for gluconeogenesis, FFAs aretaken up by cells and are used as an alternate fuelsource for oxidative phosphorylation and ATP pro-duction. Circulating FFAs are oxidized in adiposetissue by mitochondrial uncoupling proteins,which are up-regulated by LMF. Therefore, LMFinduces accelerated FFA oxidation in brown adi-pose tissue through up-regulation of uncouplingproteins, inducing fat catabolism.27 Importantly,this up-regulation of uncoupling proteins mayrepresent the beginning of an important energy-wasting cycle. Uncoupling proteins normallydecrease the coupling of respiration with the phos-phorylation of ADP. Their action therefore gener-ates heat instead of ATP and acts as an \energysink," since no ATP is produced when uncouplingproteins induce protons to cross the inner mito-chondrial membrane. In a murine model ofcachexia, murine adenocarcinoma 16 (MAC16)tumors caused overexpression of uncoupling pro-tein-1 in brown adipose tissue, which resulted inincreased thermogenesis, increased energy ex-penditure, and weight loss.27

LMF therefore appears to be responsible foradipose tissue catabolism in cachexia, which itinduces by directly stimulating lipolysis and byup-regulating expression of uncoupling proteinsin brown adipose tissue, thereby increasing fattyacid oxidation.

Tumor Necrosis Factor-a. TNF-a promotes fatcatabolism by inhibiting fat differentiation andincreasing adipocyte apoptosis. The primarymechanism of TNF-a-induced fat loss in patientswith cachectic cancer involves inhibition of LPLand stimulation of LMF.4 TNF-a inhibits LPL ac-

tivity in human adipose tissue by down-regulatingLPL protein expression.28 Increased LPL activityproduces hyperlipidemia and prevents the storageof fat, while increased LMF release stimulates therelease of FFAs from adipocytes and induces theiroxidation by uncoupling proteins. Also, TNF-aand IL-1 have both been shown to inhibit glucoseand FFA transport into adipose tissue.29 This ulti-mately decreases lipogenesis in adipose tissue.Additionally, TNF-a has been implicated in down-regulating several enzymes involved in lipogene-sis, including acetyl-CoA carboxylase, fatty acidsynthase, and acyl-CoA synthase.29

To summarize, the total body fat loss in pa-tients with cachectic cancer is mediated primarilyby increased lipolysis rather than by decreased fatsynthesis. Increased lipolysis is a result of theactions of both LMF and TNF-a. The mechanismof LMF-mediated lipolysis involves increasedexpression of oxidative uncoupling proteins inbrown adipose tissue. TNF-a inhibits LPL andstimulates the release of lipid-mobilizing factor.

INFLAMMATION

In addition to skeletal muscle and adipose tissuecatabolism, cancer cachexia is characterized by aprofound chronic inflammatory state. Inflamma-tion in cachexia has been established in a numberof different animal models of cachexia.18 It is char-acterized by increased release of proinflammatorycytokines such as IL-1b, IL-6, TNF-a, and IFN-g.4

These cytokines are thought to be the principalcatabolic factors in skeletal muscle and adiposetissue wasting in cachexia.4 These cytokines pro-duce many of biochemical and metabolic dysfunc-tions seen cachexia, including hypermetabolism,anorexia, decreased muscle protein synthesis,and increased UPS-mediated muscle proteoly-sis.18 IL-6, TNF-a, and IFN-g have been shownto activate NF-jB, triggering UPS-mediated mus-cle breakdown and inhibiting muscle protein syn-thesis through reduction in MyoD expression.4

TNF-a, on the other hand, can directly inducelipolysis.4

Proinflammatory cytokines appear to potenti-ate each other’s actions. This is seen in the activa-tion of proteolysis (TNF-aþ IFN-gþ IL-1b) and inthe up-regulation of cytokine receptors (TNF-a þIFN-g).18,30 Also, cytokines such as IFN-g, IL-1b,and IL-6 are thought to be responsible for theinduction of acute-phase protein production. To-gether, increased cytokine levels have been shownto reduce survival time in cachectic patients.4


Following is a summary of the major proinflam-matory cytokines involved in the pathogenesis ofcancer cachexia.

Interleukin-1b. Derived from macrophages andlymphocytes, IL-1b concentrations increase in thecachectic state. IL-1b is thought to be partly re-sponsible for muscle wasting in cachexia, and isknown to cause effects similar to those seen byTNF-a, including stimulation of muscle catabo-lism.31,32 Also, IL-1b appears to be linked to thedevelopment of cachexia through the induction ofthe inflammatory response. In a murine model ofhead and neck cancer, mice with a double muta-tion in the Toll-like receptor-4 (TLR4) lack theability to mount an appropriate inflammatoryresponse. In this model, wild-type cogenic micereceiving injections with equal numbers of a squa-mous cell carcinoma cell line, SCCF-VII, werefound to be more cachectic and exhibited higherlevels of IL-1b than mutant mice (personal com-munication, Marion Couch, MD, PhD). IL-1b isalso thought to be responsible for the induction ofanorexia in cachectic patients. It can induce ano-rexia when administered to animals, and it mayincrease the levels of corticotrophin releasing hor-mone, an anorexigenic neurotransmitter.31 IL-1bmay also increase levels of tryptophan in the cere-brospinal fluid, increasing serotonergic neuro-transmission production in the hypothalamus andinducing anorexia.

Interleukin-6. IL-6 is a glycoprotein predomi-nantly secreted by activated immune cells. It isinvolved in the amplification of inflammatory cas-cades in the immune response, and its concentra-tions are increased in patients with cancercachexia and in patients with lung cancer inwhom IL-6 acts to enhance the acute phaseresponse.33 Elevated IL-6 concentrations are cor-related with poor nutritional status, impaired per-formance and shorter survival, indicating that theinflammatory response induced in part by IL-6may cause a substantial amount of the morbidityseen in cachexia.34 IL-6 has also been implicatedas an important mediator of cachexia in murinemodels. In BALB-C nude mice bearing colon-26adenocarcinoma tumors, IL-6 is markedly over-secreted.35 Robust production of IL-6 by tumorcells has been shown to induce cachexia in murineadenocarcinoma models, and serum IL-6 levelshave been shown to be 35% higher in cachecticmice than in noncachectic mice.4

Tissue Necrosis Factor-a. TNF-a is a 17.4-kDaprotein produced by macrophages and naturalkiller cells that plays a complex and multifacetedrole in cancer cachexia. It induces proteolysis,activates lipolysis, and suppresses expression ofenzymes involved in lipogenesis.4 TNF-a has beenimplicated in muscle wasting in several animalmodels, including the Yoshida AH-130 hepatomaand Lewis lung carcinoma.36,37 Increased concen-trations of TNF-a are seen in cancer cachexia inhumans4 and have been shown to correlate withdecreased food intake and body weight, increasedbody temperature, decreased glycogen, lipid, andprotein synthesis, and increased gluconeogenesis,lipolysis, and proteolysis.28 As mentioned pre-viously, TNF-a appears to suppress skeletal mus-cle differentiation by suppressing MyoD expres-sion, and it can down-regulate myosin heavychain in combination with IFN-g.38 Both of theseoccur in an NF-jB–dependent manner. TNF-aalso appears to be involved in adipocyte apoptosis,which it induces by activating cellular proteasesknown as caspases.39 Experimental treatment ofcachexia with the synthetic anti–TNF-a antibodyinfliximab demonstrated weight stabilization in1 of 4 patients with metastatic small cell lungcancer.14

Interferon g. IFN-g is a pleiotropic cytokine in-volved in the regulation of nearly all phases ofimmune and inflammatory responses. Producedby T lymphocytes and natural killer cells, IFN-gcauses imbalance between orexigenic and ano-rexigenic signals in the body. As mentioned above,IFN-g, when administered together with TNF-a,induces UPS-mediated proteolysis in mice. It hasalso been shown to produce progressive weightloss in mice inoculated with Lewis lung tumors.Finally, treatment with anti–IFN-g antibodiescounteracts this effect, indicating a potentialfuture treatment for cachexia.40

To summarize, the proinflammatory cytokinepathways involved in cancer cachexia are quitecomplex. At this time, relatively little is knownabout how these cytokines induce and maintaincachexia in humans. But as more is learned aboutthe ability of each human tumor system to inducecachexia, the relative contribution of these cyto-kines to the pathogenesis of cachexia will beunderstood. A final common pathway may existfor all proinflammatory cytokines involved incachexia. This might ultimately become an ave-nue for future treatment.


ANABOLIC HORMONE DYSREGULATION

In normal adults, bone growth and tissue mainte-nance rely on an intact growth hormone (GH)/in-sulin-like growth factor-1 (IGF-1) axis. There ismounting evidence that patients with cachexiamay have alterations in this axis.18 Acquired GHresistance has been reported, but the role of theGH/IGF-1 axis in catabolic states such as cancercachexia has not been adequately characterized.41

Administering GH as part of an acute therapy forICU patients has been shown to cause increasedmortality.42 Therefore, more research is neededbefore GH or IGF-1 can be considered as a possibletherapy for wasting.

NEUROENDOCRINE DYSFUNCTION

Patients with cancer frequently suffer from ano-rexia. The prevalence of anorexia in cachecticpatients may range between 15% and 40% at pre-sentation.43 This may be an effect of the tumoritself or the consequence of treatments such aschemotherapy and radiation. It appears that ano-rexia is a consequence of weight loss, rather thanits cause, since reduced food intake in cachecticpatients is preceded by tissue wasting.44 Themechanisms of appetite dysregulation in cancercachexia appear to involve disrupted communica-tion between peripheral organs and homeostaticcontrol centers in the hypothalamus. Neuroendo-crine afferent signals originating from peripheralorgans inform the brain about nutritional require-ments and energy status. In the healthy patient,peripheral nutritional signals are integrated andprocessed within homeostatic control centers inthe hypothalamus and an appropriate response iscommunicated via efferent pathways back to pe-ripheral organs. In the cachectic patient, however,disruption within this signaling system results inanorexia and reduced food intake. Three neuro-peptide mediators of this signaling pathway, lep-tin, ghrelin, and neuropeptide Y (NPY), arebelieved to be responsible for the disruption of thisfeedback loop in cachectic patients.

Leptin. A product of the ob gene, leptin is a neu-roendocrine hormone secreted by adipose tissue.It has anorectic and lipolytic properties and isknown to regulate weight by inhibiting feeding. Innormal patients, weight loss lowers leptin levels,triggering the hypothalamus to stimulate feeding.In most experimental models of cachexia, how-ever, leptin levels are elevated. This results in in-hibition of orexigenic signals. Indeed, some of the

proinflammatory cytokines implicated in cachexiacan produce chronic leptin up-regulation, result-ing in increased anorexia.45 Although elevatedleptin has been found in patients with cachecticcancer, low levels of leptin have been found in ca-chectic mice. In MAC16-tumor-bearing mice, forexample, circulating leptin levels were signifi-cantly reduced.46 This indicates that leptin maynot consistently mediate cachexia, and that tumorproducts such as LMF and PIF are able to overridethe effects of low levels of leptin and independ-ently cause appetite suppression and cachexia.46

From the above evidence, we conclude that thetrue role of leptin has not been accurately charac-terized. It is unclear at this point whether leptinplays a leading role in the development of ano-rexia in cancer cachexia, or if it is overridden bytumor-derived factors (Figure 6).

Ghrelin. Ghrelin is a novel GH-releasing pep-tide that was first isolated from the stomach.Secreted from the stomach, it stimulates foodintake and decreases energy expenditure, therebyincreasing body weight. Ghrelin circulates in thebloodstream under fasting conditions and trans-mits a hunger signal from the periphery to theCNS, where it acts directly to increase feedingand decrease sympathetic nerve activity. Ghrelinappears to act as a counterpart to leptin, whichdecreases feeding and increases sympatheticnerve activities. Indeed, cancer cachexia involvesinordinately increased levels of active ghrelin.47

This may be a compensatory response to markedweight loss in the cachectic patient.

Neuropeptide Y. NPY is an orexigenic neuro-transmitter that is involved in regulation of circa-dian rhythms, sexual functioning, anxiety, pe-ripheral vascular resistance and cardiac contrac-tility. Although widely distributed throughout thebrain, it is found abundantly in the arcuate nu-cleus of the hypothalamus. A potent feeding-stim-ulatory peptide, NPY has been shown to reverse

FIGURE 6. Leptin dysregulation in adipose tissue. Loss of fat

mass in normal weight loss leads to a drop in leptin levels,

removing the inhibitory effect of leptin on appetite. In cachexia,

elevated leptin levels inappropriately inhibit food intake.


anorexia induced by the proinflammatory cyto-kines IL-1b and ciliary neurotrophic factor.48

However, it has also been found to be dysfunc-tional in anorectic tumor-bearing rodents, makingit a possible mediator of cancer cachexia in hu-mans (Figure 7).

NEW CONCEPTS

Cancer Cachexia and Aging. As natural conse-quences of aging, patients with head and neckcancer often experience sarcopenia, physical inac-tivity, and reduced protein regeneration. Theirsarcopenia from agingmay be due to down-regula-tion of the growth hormone axis and decline in tes-tosterone concentrations, while their physicalinactivity may be multifactorial and their im-paired protein synthesis a result of resistance tomuscle-specific anabolic signals. Patients with ca-chectic head and neck cancer suffer from systemicinflammation and reduced food intake in additionto these aging-related processes. Therefore, treat-ment of patients with cachectic cancer requiresa multipronged, multidisciplinary approach.41

Treatment plans should incorporate physicaltherapy to improve mobilization, high proteinintake to maximize muscle anabolism, and neu-

traceutical or pharmacologic intervention to bluntinflammation and improve food intake.

Cancer Cachexia: An Autoimmune Disease? Finally,not all tumors are the same, and all hosts arephysiologically different. Certain tumors maysecrete more catabolic factors or induce a moreexaggerated inflammatory response in their hoststhan other tumors. However, given the same tu-mor type and burden, why do some patients withcancer develop cachexia while others do not? Wepropose that there may be single nucleotide poly-morphisms in a variety of important immunereceptors that may predispose certain patients todevelop cachexia. The presence of such polymor-phisms might explain the interindividual differ-ences seen in the clinical manifestations ofcachexia.

One set of candidates for this genetic variabili-ty is the family of Toll-like receptors (TLRs). Thesecell-surface receptors are involved in immune reg-ulation and mediate both sterile and infectiousinflammatory responses and the complexresponses involved in autoimmunity. Their stimu-lation produces a robust cytokine response, whichincludes elaboration of IL-1b, IL-6, TNF-a, andIFN-g. Ten TLR paralogs have been identified inhumans, which together recognize exogenous mol-

FIGURE 7. Summary of the pathophysiology of cancer cachexia. UPS, ubiquitin–proteasome system; DGC, dystrophin glycoprotein

complex; A.A., amino acid; MSTN, myostatin; LMF, lipid mobilizing factor; NPY, Neuropeptide Y.


ecules from a diverse array of organisms, includingbacteria, fungi, and viruses. TLRs also recognizethe important and widely distributed lipopolysac-charide (LPS) molecules on bacteria. Interestingly,LPS injection into mice produces wasting similarto cancer cachexia.49 Mice with double mutationsin the TLR gene are actually resistant to the devel-opment of such wasting, while their wild-typecounterparts are not.49 Furthermore, such mutantmice exhibit lower levels of IL-1b and are signifi-cantly less cachectic after SCCF VII tumor chal-lenge than their wild-type counterparts, as meas-ured by weight and body composition.41

Further investigation into the distribution ofsingle nucleotide polymorphisms in TLR genesmay explain interindividual differences in theclinical manifestations of inflammation in cancercachexia. This may ultimately enable us to pre-vent the development of muscle wasting, adiposetissue loss, and anorexia in patients with cancer.

CONCLUSION

Cachexia represents a complex metabolic statecharacterized by progressive weight loss, muscleand fat atrophy, and neuroendocrine dysfunctionmediated mainly by tumor- and host-derived fac-tors. Disruption of specific physiologic processesmediates the clinical manifestations of this dis-ease. For example, cytokine-mediated up-regula-tion of the UPS, tumor secretion of PIF, and dysre-gulation of the DGC mediate accelerated musclecatabolism. Imbalance in the amino acid pool,reduced myosin expression, and myostatin up-regulation result in impaired skeletal muscledevelopment, while LMF and TNF-a appear tocontrol fat wasting. A host of proinflammatorycytokines, including IL-1b, IL-6, TNF-a, and IFN-g, mediate systemic inflammation, although theexact mechanisms for their actions have notbecome clear. Additionally, neuroendocrine medi-ators like leptin, ghrelin, and NPY contribute todisruption of hypothalamic neuroendocrine path-ways and thereby induce anorexia.

Areas for treatment of cachexia are emerging.Proteasome suppression, anticytokine treatment,and inhibition of NF-jBmay develop as means foranticachexia therapy. Future therapies may alsofocus on correcting neuroendocrine deficits or pro-motion of muscle anabolism by targeting the pro-teolytic effects of inflammatory and catabolicfactors. Adequate clinical studies remain to beperformed to determine the most effective meansof anticachexia therapy.

Finally, new concepts are evolving in this field.These include the role of aging in cancer cachexiaand the role of TLR gene polymorphisms in alter-ing responses to the cachectic state. We hope fur-ther research into these areas will answer remain-ing questions about the underlying mechanismsof cachexia.

Acknowledgments. We thank Dr. Anne Voss,Senior Research Scientist, Abbott Laboratories,Inc and Corey Cannon for artistic contributions.

REFERENCES

1. Tisdale MJ. Cancer cachexia. Langenbecks Arch Surg2004;389:299–305.

2. Tisdale MJ. Cancer cachexia: metabolic alterations andclinical manifestations. Nutrition 1997;13:1–7.

3. Acharyya S, Butchbach ME, Sahenk Z, et al. Dystrophinglycoprotein complex dysfunction: a regulatory linkbetween muscular dystrophy and cancer cachexia. Can-cer Cell 2005;8:421–432.

4. Argiles JM, Busquets S, Lopez-Soriano FJ. Cytokines inthe pathogenesis of cancer cachexia. Curr Opin ClinNutr Metab Care 2003;6:401–406.

5. Deans C, Wigmore SJ. Systemic inflammation, cachexiaand prognosis in patients with cancer. Curr Opin ClinNutr Metab Care 2005;8:265–269.

6. Li YP, Lecker SH, Chen Y, Waddell ID, Goldberg AL,Reid MB. TNF-a increases ubiquitin-conjugating activityin skeletal muscle by up-regulating UbcH2/E220k.FASEB J 2003;17:1048–1057.

7. Lorite MJ, Smith HJ, Arnold JA, Morris A, ThompsonMG, Tisdale MJ. Activation of ATP-ubiquitin-dependentproteolysis in skeletal muscle in vivo and murine myo-blasts in vitro by a proteolysis-inducing factor (PIF). BrJ Cancer 2001;85:297–302.

8. Camps C, Iranzo V, Bremnes RM, Sirera R. Anorexia–cachexia syndrome in cancer: implications of the ubiqui-tin–proteasome pathway. Support Care Cancer 2006;14:1173–1183.

9. Lecker SH, Solomon V, Mitch WE, Goldberg AL. Muscleprotein breakdown and the critical role of the ubiquitin-proteasome pathway in normal and disease states.J Nutr 1999;129(Suppl):227S–237S.

10. Jagoe RT, Goldberg AL. What do we really know aboutthe ubiquitin-proteasome pathway in muscle atrophy?Curr Opin Clin Nutr Metab Care 2001;4:183–190.

11. Attaix D, Aurousseau E, Combaret L, et al. Ubiquitin-proteasome-dependent proteolysis in skeletal muscle.Reprod Nutr Dev 1998;38:153–165.

12. Bodine SC, Latres E, Baumhueter S, et al. Identificationof ubiquitin ligases required for skeletal muscle atrophy.Science 2001;294:1704–1708.

13. Cai D, Frantz JD, Tawa NE Jr, et al. IKKb/NF-jB acti-vation causes severe muscle wasting in mice. Cell2004;119:285–298.

14. Boddaert MS, Gerritsen WR, Pinedo HM. On our way totargeted therapy for cachexia in cancer? Curr OpinOncol 2006;18:335–340.

15. Stitt TN, Drujan D, Clarke BA, et al. The IGF-1/PI3K/Akt pathway prevents expression of muscle atrophy-induced ubiquitin ligases by inhibiting FOXO transcrip-tion factors. Mol Cell 2004;14:395–403.

16. Todorov P, Cariuk P, McDevitt T, Coles B, Fearon K, Tis-dale M. Characterization of a cancer cachectic factor. Na-ture 1996;379:739–742.


17. Williams ML, Torres-Duarte A, Brant LJ, Bhargava P,Marshall J, Wainer IW. The relationship between a uri-nary cachectic factor and weight loss in advanced cancerpatients. Cancer Invest 2004;22:866–870.

18. Baracos VE. Cancer-associated cachexia and underlyingbiological mechanisms. Annu Rev Nutr 2006;26:435–461.

19. Tisdale MJ. Biology of cachexia. J Natl Cancer Inst1997;89:1763–1773.

20. Ladner KJ, Caligiuri MA, Guttridge DC. Tumor necrosisfactor-regulated biphasic activation of NF-jB is requiredfor cytokine-induced loss of skeletal muscle gene prod-ucts. J Biol Chem 2003;278:2294–2303.

21. Guttridge DC, Mayo MW, Madrid LV, Wang CY, BaldwinAS Jr. NF-jB-induced loss of MyoD messenger RNA:possible role in muscle decay and cachexia. Science 2000;289:2363–2366.

22. Jones SW, Hill RJ, Krasney PA, O’Conner B, Peirce N,Greenhaff PL. Disuse atrophy and exercise rehabilita-tion in humans profoundly affects the expression ofgenes associated with the regulation of skeletal musclemass. FASEB J 2004;18:1025–1027.

23. Fouladiun M, Korner U, Bosaeus I, Daneryd P,Hyltander A, Lundholm KG. Body composition and timecourse changes in regional distribution of fat and leantissue in unselected cancer patients on palliative care—correlations with food intake, metabolism, exercisecapacity, and hormones. Cancer 2005;103:2189–2198.

24. Bing C, Bao Y, Jenkins J, et al. Zinc-a2-glycoprotein, alipid mobilizing factor, is expressed in adipocytes and isup-regulated in mice with cancer cachexia. Proc NatlAcad Sci U S A 2004;101:2500–2505.

25. Hirai K, Hussey HJ, Barber MD, Price SA, Tisdale MJ.Biological evaluation of a lipid-mobilizing factor isolatedfrom the urine of cancer patients. Cancer Res 1998;58:2359–2365.

26. Todorov PT, McDevitt TM, Meyer DJ, Ueyama H,Ohkubo I, Tisdale MJ. Purification and characterizationof a tumor lipid-mobilizing factor. Cancer Res 1998;58:2353–2358.

27. Bing C, Brown M, King P, Collins P, Tisdale MJ, Wil-liams G. Increased gene expression of brown fat uncou-pling protein (UCP)1 and skeletal muscle UCP2 andUCP3 in MAC16-induced cancer cachexia. Cancer Res2000;60:2405–2410.

28. Figueras M, Busquets S, Carbo N, Almendro V, ArgilesJM, Lopez-Soriano FJ. Cancer cachexia results in anincrease in TNF-a receptor gene expression in both skel-etal muscle and adipose tissue. Int J Oncol 2005;27:855–860.

29. Espat NJ, Moldawer LL, Copeland EM III. Cytokine-mediated alterations in host metabolism prevent nutri-tional repletion in cachectic cancer patients. J SurgOncol 1995;58:77–82.

30. Zhang Y, Pilon G, Marette A, Baracos VE. Cytokinesand endotoxin induce cytokine receptors in skeletalmuscle. Am J Physiol Endocrinol Metab 2000;279:E196–E205.

31. Turrin NP, Ilyin SE, Gayle DA, et al. Interleukin-1b sys-tem in anorectic catabolic tumor-bearing rats. Curr OpinClin Nutr Metab Care 2004;7:419–426.

32. Baracos V, Rodemann HP, Dinarello CA, Goldberg AL.Stimulation of muscle protein degradation and prosta-

glandin E2 release by leukocytic pyrogen (interleukin-1). A mechanism for the increased degradation of mus-cle proteins during fever. N Engl J Med 1983;308:553–558.

33. Seifart C, Plagens A, Dempfle A, et al. TNF-a, TNF-b,IL-6, and IL-10 polymorphisms in patients with lungcancer. Dis Markers 2005;21:157–165.

34. Falconer JS, Fearon KC, Plester CE, Ross JA, CarterDC. Cytokines, the acute-phase response, and restingenergy expenditure in cachectic patients with pancreaticcancer. Ann Surg 1994;219:325–331.

35. Strassmann G, Fong M, Kenney JS, Jacob CO. Evidencefor the involvement of interleukin 6 in experimental can-cer cachexia. J Clin Invest 1992;89:1681–1684.

36. Costelli P, Llovera M, Carbo N, Garcia-Martinez C,Lopez-Sorianoq FJ, Argiles JM. Interleukin-1 receptorantagonist (IL-1ra) is unable to reverse cachexia in ratsbearing an ascites hepatoma (Yoshida AH-130). CancerLett 1995;95:33–38.

37. Llovera M, Garcia-Martinez C, Lopez-Soriano J, et al.Role of TNF receptor 1 in protein turnover during cancercachexia using gene knockout mice. Mol Cell Endocrinol1998;142:183–189.

38. Costelli P, Muscaritoli M, Bossola M, et al. Skeletal mus-cle wasting in tumor-bearing rats is associated withMyoD down-regulation. Int J Oncol 2005;26:1663–1668.

39. Inadera H, Nagai S, Dong HY, Matsushima K. Molecularanalysis of lipid-depleting factor in a colon-26-inoculatedcancer cachexia model. Int J Cancer 2002;101:37–45.

40. Matthys P, Heremans H, Opdenakker G, Billiau A. Anti-interferon-g antibody treatment, growth of Lewis lungtumours in mice and tumour-associated cachexia. Eur JCancer 1991;27:182–187.

41. Clark RG, Robinson IC. Up and down the growth hor-mone cascade. Cytokine Growth Factor Rev 1996;7:65–80.

42. Takala J, Ruokonen E, Webster NR, et al. Increasedmortality associated with growth hormone treatment incritically ill adults. N Engl J Med 1999;341:785–792.

43. De Wys WD. Anorexia as a general effect of cancer. Can-cer Cell 1972;45:2013–2019.

44. DeWys WD. Anorexia as a general effect of cancer. Can-cer 1979;43(5, Suppl):2013–2019.

45. Inui A. Cancer anorexia-cachexia syndrome: currentissues in research and management. CA Cancer J Clin2002;52:72–91.

46. Bing C, Taylor S, Tisdale MJ, Williams G. Cachexia inMAC16 adenocarcinoma: suppression of hunger despitenormal regulation of leptin, insulin and hypothalamicneuropeptide Y. J Neurochem 2001;79:1004–1012.

47. Garcia JM, Garcia-Touza M, Hijazi RA, et al. Activeghrelin levels and active to total ghrelin ratio in cancer-induced cachexia. J Clin Endocrinol Metab 2005;90:2920–2926.

48. Sonti G, Ilyin SE, Plata-Salaman CR. Neuropeptide Yblocks and reverses interleukin-1 b-induced anorexia inrats. Peptides 1996;17:517–520.

49. Frost RA, Nystrom GJ, Lang CH. Lipopolysaccharidestimulates nitric oxide synthase-2 expression in murineskeletal muscle and C2C12 myoblasts via Toll-like recep-tor-4 and c-Jun NH2-terminal kinase pathways. Am JPhysiol Cell Physiol 2004;287:C1605–C1615.


cancer cachexia syndrome in head and neck cancer patients: part ii. pathophysiology

Documents