OUP user menu

Cancer cachexia

J.N. Gordon, S.R. Green, P.M. Goggin
DOI: http://dx.doi.org/10.1093/qjmed/hci127 779-788 First published online: 7 October 2005


Cancer cachexia is a severe debilitating disorder for which there are currently few therapeutic options. It is driven by the release of pro-inflammatory cytokines and cachectic factors by both host and tumour. Over the past few years, basic science advances have begun to reveal the breadth and complexity of the immunological mechanisms involved, and in the process have uncovered some novel potential therapeutic targets. The effectiveness of thalidomide and eicosapentaenoic acid at attenuating weight loss in clinical trials also provides a further rationale for modulating the immune response. We are now entering an exciting period in cachexia research, and it is likely that the next few years will see effective new biological therapies reach clinical practice.


Cancer cachexia is a complex disorder characterized by progressive loss of weight, in association with anorexia, asthenia (lack of energy and strength) anaemia and alterations in immune function. It is a significant cause of morbidity and mortality, occurring in up to 80% of patients with advanced cancer, and responsible for death in up to 20% of cases.1 Different tumours display varying propensities to induce cachexia, with it most commonly seen in subjects with gastrointestinal, lung and prostate cancers, in contrast to haematological and breast malignancies where it is rare. Weight loss is associated with both reduced quality of life, and shortened life expectancy, with death occurring when subjects have lost around 30% of their pre-morbid weight.2 Additionally, cachectic patients have a lower chance of responding to chemotherapy and are more prone to toxic side-effects.3

It is now apparent that, despite the common association of anorexia with cachexia, early concepts that the cachectic process simply resulted from an imbalance between energy intake and expenditure were naive. This is witnessed by the fundamental differences that can be observed between subjects with cachexia compared to those suffering from starvation (Table 1). In cachexia, weight is lost from both the fat and skeletal muscle compartments, in contrast to starvation, where it is lost preferentially from the fat compartment. Furthermore, the protein loss that does occur in starvation is split equally between skeletal muscle and visceral protein, whereas in cachexia, visceral protein is relatively preserved. Thus there is loss of liver mass in starvation, but an increase in mass in cachectic patients due to metabolic recycling activity and the acute-phase response. Finally, and most tellingly, weight loss in starvation is easily reversed by feeding, whereas it has now been clearly demonstrated through trials of oral and parenteral feeding that nutritional supplementation alone cannot reverse cachexia. In short, the altered metabolism in cancer cachexia more closely resembles that of patients suffering from severe sepsis or multiple trauma rather than starvation.4

View this table:
Table 1

Metabolic changes seen in starvation and cancer cachexia

Resting energy expenditureDecreasedIncreased
Acute phase responseNoYes
Skeletal muscleMaintainedDecreased
Adipose tissueDecreasedDecreased
Liver sizeDecreasedIncreased
Glucose intoleranceNoYes
Insulin levelsDecreasedIncreased

Although the mechanism of cancer cachexia remains incompletely understood, it is now clear that the persistent inflammatory response of the host, in conjunction with the production of specific cytokines and catabolic factors by the tumour itself, is central to disease pathogenesis.5–7 In tandem, clinical trials of agents that can modulate the immune response have shown promise in attenuating the disease process and reversing the block to protein accretion. Thus there is now real hope that over the next few years these advances will result in effective new therapies for this debilitating disorder.



Anorexia is common in cancer patients, being present in >50% of cases, and results from derangement of the complex central and peripheral signalling pathways that control food intake.8 In the hypothalamus, neuropeptide Y (NPY) stimulates food intake and the pro-opiomelanocortin/cocaine and amphetamine regulated transcript (POMC/CART) pathway inhibits food intake. The main peripheral regulators of these pathways are hormones such as gherlin, which is orexigenic (stimulates appetite), and leptin, which is anorexigenic, along with neurotransmitters such as serotonin. In anorexic rats, there is a decrease in NPY-immunoreactive neurons in the hypothalamus, compared with non-anorexic controls,9 whereas blocking the POMC/CART pathway can restore food intake in tumour-bearing animals.10,,11 Cytokines such as interleukin-1β (IL-1β), interleukin-6 (IL-6) and tumour necrosis factor α (TNFα) play an important role in the dysregulation of these systems, either through excessive negative feedback signalling from leptin, persistent stimulation of anorexigenic peptides such as corticotrophin-releasing factor, or by inhibition of the neuropeptide Y pathway.12–14 However, data from human studies are currently sparse and inconsistent, and it remains to be resolved how mechanisms that control anorexia in rodent studies relate to human neurophysiology.

Altered metabolism

In clinical studies, resting energy expenditure (REE), which in a normal population accounts for 70% of total energy expenditure, varies widely between different groups of cancer patients. In some studies REE is increased,15,,16 while in other studies it is normal or reduced.17,,18 Some of these metabolic differences relate to tumour type, with cancers such as lung and pancreas generally increasing REE, while others, such as colorectal cancer, do not.19 However, in longitudinal studies of cancer-bearing rats, there is a change in energy expenditure over time, with an initial hypermetabolic phase, followed by a period of normal energy expenditure, before a pre-terminal hypometabolic phase.20 If this is the case in humans, then some of the wide variation in REE seen in clinical studies may relate to disease phase.

It is unclear exactly how tumours influence REE, though recent studies suggest that it may be through the upregulation of uncoupling proteins, a family of mitochondrial membrane proteins, which can increase thermogenesis and energy expenditure. Uncoupling protein-1 (UCP-1), which generates heat by uncoupling the oxidation of fatty acids from the generation of ATP, is only found in brown adipose tissue (BAT), and is known to be critical to thermoregulation in rodents. Although BAT is uncommon in adults, it is present in 80% of patients with cachexia, compared with 13% of controls.21 UCP-2 and -3, which are present in skeletal muscle, may also play a significant role in energy balance and lipid metabolism, as they are upregulated in rodent models of cachexia. However, the significance of this is unclear: in one model, increased expression was also seen in pair-fed litter mates, suggesting anorexia was driving production, while in a separate model, expression was not linked to food intake.22,,23 In humans, skeletal muscle UCP3 mRNA levels have been found to be 5-fold higher in cachectic patients than in controls.24

Adipose tissue

Cachexia results in increased lipolysis, decreased lipogenesis, hyperlipidaemia, raised circulating levels of free fatty acids and glycerol, and ultimately the loss of large amounts (up to 85%) of adipose tissue.25 Some of these changes may be mediated by pro-inflammatory cytokines, such as TNFα, interferon γ (INFγ) and IL-1β, all of which can inhibit lipoprotein lipase, preventing adipocytes from storing fatty acids. However, more recently, patients with cachexia have been found to excrete a lipid-mobilizing factor (LMF) in their urine that is also produced by a mouse adenocarcinoma model.26 This is identical to the plasma protein zinc α2-glycoprotein (ZAG), which is expressed in human adipocytes and increased in mice with cancer cachexia.27,,28 ZAG stimulates adipose tissue breakdown through a cAMP-dependent pathway, and upregulates uncoupling proteins found in adipose tissue, potentially contributing to changes in REE.2,,29

Skeletal muscle

During starvation, energy utilization by the brain is switched from glucose to fat-derived ketone bodies, allowing decreased gluconeogenesis from amino acids in the liver and so, conservation of muscle mass. In contrast, patients with cachexia experience progressive severe loss of skeletal muscle with relative preservation of visceral protein reserves, and an increase in liver mass secondary to the acute phase response. The loss of skeletal muscle mass is due to a combination of reduced protein synthesis and increased protein degradation. Skeletal protein synthesis is slowed by the lack of available amino acids to act as a substrate, most having been diverted due to the increased synthesis of acute phase proteins and gluconeogenesis. Furthermore, branched-chain amino acids (BCAAs) can regulate protein synthesis directly by modulating mRNA translation.30 However, though reduced protein synthesis plays a role, protein degradation is the major cause of loss of skeletal muscle mass in cachexia. Lysosomal protease cathespins B probably play a role in early protein breakdown, as they are elevated in skeletal muscle biopsies from patients with lung cancer and minimal weight loss.31 In more established cachexia, the ubiquitin-proteasome dependent proteolytic pathway is upregulated and is the predominant pathway for protein degradation. In this pathway, myofibrillar protein is flagged for proteolysis by the attachment of a polyubiquitin chain, and can then be recognized and destroyed by the 26S proteasome, a large cylindrical structure which contains the proteolytic enzymes.1 The rate limiting step is polyubiquitination, which is controlled by specific ligases. Muscle RING Finger 1 (MuRF1) and Muscle Atrophy F-box (MAFbx) are two newly described ligases that appear to play a crucial role in the development of cachexia. These make attractive potential therapeutic targets, as they are markedly upregulated in a variety of different animal models of atrophy and cachexia, and mice deficient for either ligase remain resistant to wasting.32–34 MAFbx also mediates the degradation of MyoD, a protein that controls muscle cell growth and differentiation, via the ubiquitin-proteasome pathway.35 Thus upregulation of MAFbx can both inhibit new muscle formation, and increase skeletal muscle degradation. Finally, other ligases, such as E3alpha-II, which is inducible by TNFα and INFγ and is upregulated in cachexia, are also likely to play an important role in muscle catabolism, and may represent alternative therapeutic targets.36

Mediators of cachexia

The metabolic abnormalities that arise in cancer cachexia are driven by the production of a combination of cytokines and other cachectic factors by both the host and the tumour (Figure 1). The presence of tumour results in a persistent host inflammatory response, characterized by the production of T helper 1 (Th1) cytokines, such as TNFα, IL-1β, IL-6 and IFNγ, and the induction of an acute phase response. In the acute setting, such as infection, this response is beneficial to the host, as it upregulates the immune system and aids recovery. However, in cancer patients, the continuing presence of tumour results in a chronic inflammatory state that leads to the development of cachexia. TNFα, IL-6, IL-1β, and IFNγ, though not capable of reproducing the whole syndrome of cachexia, can all induce some features of anorexia and wasting when administered to animal models.37–41 Furthermore, the administration of anti-TNF, anti-INFγ, and anti-IL-6 antibodies can attenuate or reverse some of the symptoms of cachexia.40,42,,43 Interestingly, it has been shown in vitro that TNFα and IFNγ can act synergistically to induce muscle breakdown in skeletal myocytes, despite being unable to induce muscle breakdown when given individually.44 This is further evidence of the critical importance of the cytokine milieu in the development of cachexia.

Figure 1.

Pathogenesis of cancer cachexia. The pro-inflammatory response of the host in tandem with the production of specific cachectic factors by the tumour drives multiple metabolic changes resulting in the cancer cachexia syndrome.

Tumours also produce specific cachectic factors such as proteolysis inducing factor (PIF) and lipid mobilizing factor (LMF), which directly promote protein and fat breakdown. PIF is a glycoprotein, initially isolated from MAC16 tumour bearing mice and the urine of cachectic cancer patients, that induces protein breakdown by upregulation of the ubiquitin-proteasome proteolytic pathway.45 The presence of both these factors correlates strongly with weight loss in both animals and humans.2,,46

There is increasing evidence that both cytokines and PIF cause wasting by activation of nuclear factor kappa B (NFκB). (Figure 2) NFκB is a transcription factor that controls the expression of a number of pro-inflammatory cytokines. Following activation, it translocates to the nucleus, and binds to specific promoter regions. Both PIF and TNFα upregulate components of the ubiquitin-proteasome pathway in an NFκB-dependent manner.47,,48 Activation of NFκB by TNFα in mouse muscle cells also suppresses production of the transcription factor MyoD, preventing the repair of damaged muscle.44 It was subsequently shown that TNFα and INFγ can selectively deplete myosin in skeletal muscle by two separate pathways, one involving the inhibition of gene transcription through by MyoD and NF-κB, and the other, upregulation of the ubiquitin-proteasome pathway.49 Activation of NF-κB through transgenic expression of IκB kinase β has also been shown to result in severe muscle wasting in mice.50 In addition, inhibitors of NFκB can attenuate proteasome expression, weight loss, and protein degradation in mouse skeletal muscle.51 Finally, PIF can also activate NFκB and STAT3 in hepatocytes, resulting in increased production of IL-6, IL-8, and CRP.52 Corroboratory evidence for the importance of cytokines in cachexia comes from genetic studies. Polymorphisms of both IL-1β and IFNγ genes influence survival in subjects with advanced pancreatic cancer.53,,54

Figure 2.

Activation of NF-κB in skeletal muscle. Mediators such as TNFα and PIF bind to cell surface receptors, activating the IκB kinase complex, resulting in phosphorylation, ubiquitination and subsequent proteosomal degradation of IκBα. Freed NF-κB dimers (predominantly p50 and p65) are then able to translocate to the nucleus and activate gene transcription. This results in upregulation of the ubquitin-proteasome pathway, enhancing muscle degradation, and down-regulation of the transcription factor MyoD, inhibiting new muscle synthesis.

Other potential mediators of cachexia include myostatin and angiotensin II. Myostatin is a protein that belongs to the TGF-β super family and is exclusively expressed in smooth muscle, where it acts as a negative regulator of muscle growth. Targeted deletion of the gene in mice results in skeletal muscle hypertrophy,55 while systemic over-expression results in profound muscle and fat loss similar to that seen in human cachexia.56 Administration of a neutralizing monoclonal anti-myostatin antibody to adult mice leads to a specific increase in skeletal muscle size and strength without obvious side-effects.57 It therefore represents an intriguing potential therapeutic target, though whether it plays a specific role in the pathogenesis of cancer cachexia is currently unknown.

Angiotensin II infusion in rats results in marked weight loss accompanied by depression of circulating and skeletal muscle insulin-like growth factor 1 (IGF-1).58 Inhibition of the IGF-1 signalling pathway leads to caspase-3 activation, actin cleavage, upregulation of ubiquitin ligases, and increased apoptosis, resulting in increased protein degradation.59 Muscle-specific over-expression of IGF-1 in transgenic mice blocks the cachectic effects of angiotensin II.59 Thus, targeting the IGF-1 pathway may be beneficial in the cachexia associated with conditions such as congestive cardiac failure, where there is activation of the renin-angiotensin system, but its role in cancer-associated cachexia remains uncertain.

Treatment (Table 2)

Current therapeutic interventions in cancer cachexia are of limited benefit. Despite the fact that that nutritional intake is often significantly reduced in patients with cachexia, hypercaloric feeding does not promote weight gain. Trials of both enteral and parenteral feeding have consistently failed to show any benefit in terms of weight gain, nutritional status, quality of life, or survival, and should not be used routinely.60–62 They may have a more limited role in specific circumstances, such as helping to facilitate completion of treatment in patients with oesophageal cancer.63 However, more recently, a large randomized controlled trial comparing fish-oil-supplemented high-calorie drinks with high-calorie drinks alone did show weight stabilization in both groups, indicating some benefit from simple nutritional support.64

View this table:
Table 2

Treatments for cancer cachexia that have undergone clinical trials

DrugMode of actionEffectSide-effects
Steroids Megesterol acetate MedroxyprogesteroneAnabolic effects Appetite stimulantsTotal weight gain due to increased fat mass and fluid retention. No increase in lean body mass. Increased sense of well-beingDiabetes Osteoporosis Mood swings Thromboembolism
NSAIDsInhibits prostaglandin production. Reduces REE and acute phase responseTotal weight gain, reduced need for alternative analgesics, improved quality of life. No increase in lean body mass. Prolonged survival in one studyGI upset/haemorrhage
CannabinoidsAppetite stimulantIneffectiveNausea/vomiting
Eicosapentaenoic acid (EPA) Fish oilsInhibits NFκB Inhibits PIF Reduces pro-inflammatory cytokinesIncreased lean body mass in pilot studies. Overall ineffective at increasing weight in large RCTs—possibly due to inability of patients to achieve target doseNausea, fishy taste/odour, GI upset
PentoxifyllineInhibits TNFαIneffective
ThalidomideInhibits TNFα, Effect Th1 to Th2 shift. Inhibit NFκBWeight stabilization. Attenuated loss of lean body mass. Trend towards prolonged survivalRash, peripheral neuropathy, daytime somnolence, constipation

Appetite stimulants

Steroids and hormonal agents such as megesterol acetate are currently widely used in the treatment of anorexia and cachexia. They act through multiple pathways, such as increasing NPY levels to increase appetite, and down-regulating pro-inflammatory cytokines.65 Although they lead to short-term improvements in appetite, food intake, and wellbeing, they do not influence lean body mass or survival, and any weight increase appears to result from a combination of fat deposition and fluid retention.66–69 They are also associated with significant complications, such as thromboembolic disease, proximal myopathy, mood change, and insulin resistance. Their use should therefore be limited to the pre-terminal phase of cancer cachexia. Marijuana and derivatives have also been investigated as potential appetite stimulants. However, in a randomized trial involving 469 patients with advanced cancer, dronabinol, a synthetic form of THC (the active ingredient in marijuana), was significantly less effective than megesterol acetate in promoting weight gain.70

Non-steroidal anti-inflammatory drugs (NSAIDS)

NSAIDS are anti-inflammatory agents that inhibit cyclo-oxgenase and thus the production of prostaglandins. In a short (7-day) placebo-controlled study, ibuprofen reduced resting energy expenditure and the acute phase response, though body weight was not measured.71 In a subsequent trial, megesterol acetate and ibuprofen, given in combination, produced a greater increase in body weight than did megesterol acetate alone, although it was not clear to what extent this reflected an increase in lean body mass.72 In one study, indomethacin improved some measures of performance and prolonged survival in cachectic patients with metastatic solid tumours, but had no effect on overall body weight.73

Eicosapentaenoic acid (EPA)

EPA, the major active component of fish oil, has attracted attention as a potential nutritional supplement in cancer cachexia, due to its ability to down-regulate both pro-inflammatory cytokines and PIF.74 However, despite initial encouraging data from open-label studies, the results from two subsequent large-scale multicentre trials have been disappointing. In the European study involving a total of 200 patients, EPA-enriched oral supplements failed to show any significant benefit over oral supplements alone, although both arrested weight loss. Post hoc dose-response analysis suggested that at higher doses the n-3 fatty acid enriched supplement may be associated with lean tissue gain.64 In the North American and Canadian study involving 421 patients, EPA-enriched nutritional supplements, either alone or in combination with megestrol acetate, were no better than megesterol acetate itself at improving appetite or weight.75 Recent data from animal studies suggest that combining EPA with the leucine metabolite beta-hydroxy-betamethylbutyrate (HMB) to aid protein synthesis is more effective than EPA alone in reversing cachexia.76 It remains to be seen whether this will be effective in humans.


Pentoxifylline is a phosphodiesterase-4 inhibitor that reduces TNFα production. In a randomized placebo controlled trial, it was not effective at reversing weight loss in cachexia.77 However, pentoxifylline is not effective in the treatment of Crohn's disease either, in contrast to alternative anti-TNFα strategies. The reason for this remains unclear.


Thalidomide is a unique drug with multiple immunomodulatory properties. It is a potent inhibitor of TNFα production from lipopolysaccharide (LPS)-stimulated blood monocytes, by enhancing TNFα mRNA degradation.78,,79 It can also enhance IL-4 and IL-5 production affecting a shift from a Th1 to Th2 cytokine profile, down-regulate IL-12 production, inhibit cyclooxygenase-2 (COX-2) production, and inhibit angiogenesis.80 Recently, it has been shown to inhibit NF-κB activity through enhancing the activity of IκB kinase, which may be central to its multiple immunomodulatory effects.81,,82 We have conducted a randomized, double-blind, placebo-controlled trial of thalidomide in weight-losing patients with advanced pancreatic cancer. In this study, thalidomide significantly attenuated both total weight loss and loss of lean body mass. Weight gain was associated with an improvement in physical functioning. There was also a trend to prolonged survival, although this did not reach statistical significance.83 A further larger study involving subjects with cachexia secondary to all upper gastrointestinal tumour types is now in progress.

Future therapeutic strategies

Many other biological therapies aimed at inhibiting immunological pathways involved in cachexia are currently undergoing testing in laboratory and animal studies. These include the inhibition of pro-inflammatory cytokines, the administration of immunoregulatory cytokines, inhibition of intracellular signalling pathways, and the inhibition of protein degradation or stimulation of protein synthesis (Table 3). It is likely that several of these will proceed to clinical trials over the next few years.

View this table:
Table 3

Potential future therapeutic strategies for the treatment of cancer cachexia

TargetTherapeutic strategy
Inflammatory cytokinesChimeric or monoclonal antibodies
TNFαSmall-molecular-weight inhibitors
IL-1βSoluble receptors
Therapeutic cytokineIL-15
Other proteinsAnti-myostatin antibodies
Signalling pathwaysAntisense to NF-κB
p38 MAP Kinase inhibitors
Central appetite regulationMC4-R antagonists
Inhibition of protein degradationUbiquitin ligase inhibitors
Proteasome inhibitors—e.g. formoterol (β2 agonist)
Tumour cachectic factorsSpecific inhibitors of PIF and LMF
HormonesAngiotensin II inhibitors
  • PIF, proteolysis inducing factor; LMF, lipid mobilizing factor; NF-κB, Nuclear factor kappa B; MC4-R, melanocortin-4 receptor.


Despite major improvements in the treatment of cancer over the last 30 years, little progress has been made in the treatment of cancer cachexia. However, the past few years have seen a rapid advance in our understanding of the immunopathological basis of cachexia, with the resultant discovery of many potential new therapeutic targets. The parallel demonstration that immunomodulatory agents such as thalidomide and EPA can alter the natural history of cachexia underscores the importance of understanding this immune response. We are now at the vanguard of drug treatment of cachexia, and it is very likely we will see major developments over the coming few years. Ultimately, strategies that target specific immunological pathways and down-regulate the persistent inflammatory response should improve outcome in cancer cachexia. The challenge is to identify which pathways are important and to inhibit them in vivo. Ultimately proof of efficacy for new therapeutic rationales will only come from large-scale human studies.


View Abstract