Anorexia as a Treatment Target
In a survey of 1,000 patients with advanced cancer, Donnelly et al found anorexia to be one of the most common adverse symptoms reported.[13] Despite the fact that several clinical trials have failed to demonstrate an improvement in survival or global quality of life, an argument can be made for attempting to palliate cancer-associated anorexia because it is a prevalent, noxious, and worrisome symptom for cancer patients and their families.[14]
An important first step in the palliation of this symptom is a discussion with patients and family members about realistic expectations with respect to food intake and about the importance of assuaging guilt arising from failure to meet expectations of dietary intake (Figure 1). Sometimes such a discussion leads exclusively to an acceptance of anorexia and a validation of realistic eating goals. At other times, it also leads to the prescription of an orexigenic agent. Several potential orexigenic agents have been studied, most notably progesterones and corticosteroids.
Megestrol Acetate: Megestrol acetate (Megace) and medroxyprogesterone acetate (Depo-Provera) are progestational agents that improve appetite in patients with advanced cancer. Megestrol acetate has been studied more extensively. Although its mechanism of action is not fully understood, studies have suggested that it downregulates certain cytokines implicated in cancer-associated anorexia. Of 15 published placebo-controlled trials on megestrol, 13 demonstrated that the drug improves appetite in the cancer setting.
In a North Central Cancer Treatment Group trial, Loprinzi et al examined 133 cancer patients and found that those who received megestrol acetate at a dose of 800 mg/d manifested both an increase in appetite and an increase in nonfluid weight.[15] The optimal dosing of megestrol acetate was determined in a follow-up study from this same group.[16] This study demonstrated a direct dose-response effect of megestrol acetate as an orexigenic agent, with doses from 480 to 800 mg/d proving superior to a dose of 160 mg/d. The higher dose of 1,280 mg/d was not more effective.
More recently, megestrol acetate oral suspension has gained popularity. This formulation is associated with better compliance, less cost, and improved bioavailability. A reasonable approach to the treatment of cancer-associated anorexia is to begin with 400 mg/d of oral suspension megestrol acetate and consider an escalation to 600 to 800 mg/d over 2 to 4 weeks if patients fail to derive a favorable initial response.
Megestrol acetate is well tolerated. Apart from a slightly increased risk for thrombophlebitis (especially with concomitant chemotherapy) and usually occult suppression of the pituitary-adrenal axis, this agent has few side effects. With regard to the latter potential side effect, case reports and a few studies have suggested transient adrenal insufficiency with its use. It is recommended that megestrol acetate treated patients receive supplemental corticosteroids if they experience a serious infection or trauma or if they undergo surgery.
Corticosteroids: Corticosteroids represent another class of agents often prescribed for cancer-associated anorexia. In the first randomized, placebo-controlled trial for cancer-associated anorexia and weight loss, Moertel et al found that dexamethasone improved appetite in patients with advanced cancer.[17] These findings have been replicated in other studies. In addition, Loprinzi et al recently compared three different agents for the treatment of cancer-associated anorexia.[18] Although the anabolic agent fluoxymesterone was not effective, the other two agents—dexamethasone and megestrol acetate—were. Both were comparable in efficacy, although dexamethasone was associated with a significantly worse side-effect profile.
In view of such data, we generally recommend that dexamethasone be reserved for cancer patients with a markedly poor prognosis, because troublesome side effects such as myopathy are unlikely to occur in the short term. Additionally, dexamethasone is a reasonable choice in patients at higher risk for thromboembolic phenomena, and megestrol acetate is a reasonable choice for most other cancer patients who desire treatment for cancer-associated anorexia.
Other Orexigenic Agents: In addition to the agents noted above, several other potential orexigenic agents have been tested in randomized, controlled trials. For example, in a recent study of 100 patients with metastatic cancer, Lissoni and others found that melatonin, at an oral dose of 20 mg/d, resulted in a significantly lower rate of weight loss, compared to placebo-treated patients.[19]
The cannabinoid dronabinol (Marinol) also carries appetite-stimulatory properties, and the North Central Cancer Treatment Group recently completed a phase III trial comparing the orexigenic properties of single-agent dronabinol vs megestrol vs both agents combined. This trial of 477 patients with advanced cancer found that megestrol was superior to dronabinol and equivalent to combination therapy in its appetite-stimulatory properties (manuscript submitted).
Cyproheptadine, a histamine and serotonin antagonist, is another agent that has manifested minimal orexigenic properties in a large, placebo-controlled trial.[20] While not generally recommended for the palliation of cancer-associated anorexia, this agent does appear to improve appetite in carcinoid patients by means of inhibition of serotonin.
Finally, a randomized, placebo-controlled trial examined pentoxifylline (Trental), an inhibitor of tumor necrosis factor (TNF)-alpha, which is a cytokine implicated in anorexia. This study concluded that pentoxifylline yielded no appetite-stimulatory effects in patients with advanced cancer.[21]
Tissue Wasting as a Treatment Target
The main reason that nutritional support and orexigenic agents have benefited only subsets of cancer patients is because cancer-associated weight loss is a distinct entity, separate from starvation. Several lines of evidence draw this distinction (Table 2).
Loss of Lean Tissue: The hallmark of cancer-associated weight loss is a disproportionate and excessive loss of lean tissue that is not generally observed with simple starvation. In a study of 50 cancer patients, Cohn et al assessed body composition with tritiated water, prompt gamma neutron activation, and total body potassium measurement.[22] Results were compared to those of age- and sex-matched controls. Cancer patients who were losing weight manifested a loss of both fat and lean tissue. However, the loss of lean body tissue, most notably skeletal muscle, was dramatic. In contrast, starvation is characterized by a preferential loss of fat tissue and a relative preservation of lean tissue.[23]
Resting Energy Expenditure: In some cancer patients, an increase in resting energy expenditure contributes to the energy deficits that eventually lead to weight loss. Elevations in resting energy expenditure have been observed in patients with lung cancer—even early-stage non-small-cell lung cancer[24]—hematologic malignancies, and sarcomas.[25-27] Although other studies have suggested that malignancies, such as those arising from the gastrointestinal tract, do not predispose to this hypermetabolism,[28] the studies referenced above draw a further distinction between cancer-associated weight loss and starvation. In the latter, resting energy expenditure drops.[23]
The mechanisms behind these metabolic aberrations in cancer patients remain unexplored, but recent studies have invoked uncoupling proteins. Bing et al have found that concentrations of mRNA for uncoupling protein-1 (UCP-1) were significantly increased in brown adipose tissue in a murine tumor model when compared to levels in controls.[29] This distinction in metabolic rates between individuals with cancer and those undergoing food deprivation suggests that aggressive feeding alone cannot reverse all the alterations observed in cancer-associated weight loss.
Mediators Activated in Cancer-Associated Weight Loss: A variety of different mediators appear to become activated in cancer-associated weight loss in contrast to simple starvation. Todorov et al have discovered a 24-kd proteoglycan from tumor homogenate in the MAC16 tumor line.[30] They have termed this substance proteolysis-inducing factor. This substance appears to have direct proteolytic activity on muscle, and antibodies to this substance prevent weight loss in animal models. This proteoglycan is also specific for cancer-associated wasting, as opposed to starvation and as opposed to cancer that exists in the absence of wasting.
In addition to this purported mediator of cancer-associated wasting, other mediators that appear to play an active role in this entity include TNF-alpha, interleukin (IL)-1-beta, IL-6, and ciliary neurotrophic factor.[31-35] Recent studies have suggested that some of these mediators spawn a variety of other events, such as activation of the ubiquitin-proteasome system, and thereby lead to lean tissue wasting.[36]
Agents That May Halt the Wasting Process: Our improved understanding of the pathophysiology of cancer-associated weight loss has led to several clinical trials aimed at halting the wasting process. Agents that seem the most promising include: (1) eicosapentaenoic acid (EPA), (2) thalidomide (Thalomid), (3) adenosine triphosphate (ATP), and (4) nonsteroidal anti-inflammatory agents such as ibuprofen (Table 3.)
Eicosapentaenoic Acid—EPA is an omega-3 fatty acid derived from fish oil. It appears to attenuate cancer-associated tissue wasting and, in one study, actually resulted in an increase in weight.[37,38] Barber et al studied 20 patients who were losing weight and had unresectable pancreatic cancer. With the initiation of two daily cans of fish oil, each of which contained 1.09 g of EPA, patients manifested a median weight gain of 1 kg at 3 weeks and 2 kg at 7 weeks.
Gogos et al conducted a placebo-controlled trial in 60 patients with advanced cancer.[39] Half the patients received a fish oil-based supplement and half received a placebo. The group that received the supplement achieved a slight survival advantage compared to the group that did not. Although the potential mechanisms behind EPA’s weight-sustaining effects have not been fully elucidated, a recent animal study by Hussey and Tisdale suggests that the proteolysis-inducing factor may be suppressed with its administration.[40]
Currently, the North Central Cancer Treatment Group is undertaking a multicenter trial that examines EPA vs megestrol vs a combination of both agents in the treatment of cancer-associated weight loss and anorexia.
Thalidomide—The drug thalidomide may also have a potential role in controlling the wasting process in cancer patients. Trials in acquired immunodeficiency syndrome (AIDS) patients have led to further investigation of this agent in cancer patients.
Recent preliminary data by Boasberg and others suggest beneficial effects in cancer patients.[41] In this study of 15 cancer patients who were losing weight with metastatic disease, administration of oral thalidomide at a dose of 100 mg/d at bedtime, and subsequently escalated to 200 mg/d, resulted in weight stability. Because thalidomide appears to shorten the half-life of mRNA TNF-alpha, investigators have postulated that cytokine inhibition may be the means by which any potential inhibition of tissue wasting may occur.
Adenosine Triphosphate—Preliminary data in lung cancer patients led to a recently published trial that examined the use of ATP in patients with advanced non-small-cell lung cancer. The rationale for this trial rests on the fact that the metabolic rate in non-small-cell lung cancer patients is increased, leading to "spent" energy that drives weight loss. Since ATP is a key energy source, direct administration of this agent may compensate for the spent energy observed in non-small-cell lung cancer patients.
To test this hypothesis, Agteresch et al randomized 58 patients to either receive ATP or not.[42] Although the study was not double-blinded, a 0.2-kg increase in weight was noted in the ATP-treated group over a 4-week period, in contrast to a 1-kg drop in weight in the patients who did not receive ATP. Other parameters such as muscle strength, quality-of-life scores, functional scores, and psychological scores were statistically different between the two groups, with the ATP-treated patients demonstrating more favorable scores. A promising agent, ATP deserves further investigation in this setting.
Ibuprofen—The fact that many of the mediators of cancer-associated weight loss are in fact inflammatory markers, including TNF-alpha, IL-1-beta, and IL-6, has given rise to studies that have examined nonsteroidal anti-inflammatory agents as a means of treating cancer-associated weight loss. Ibuprofen has been the most extensively studied of these drugs. In several small studies, ibuprofen was shown to abrogate the inflammatory response and lead to subsequent modest increases in weight.[43,44] These preliminary results also appear promising and suggest a need for further investigation of this agent in this setting.
Conclusions
In summary, loss of appetite and weight are prevalent among cancer patients and predict a poor prognosis. Under a few clinical circumstances—such as perioperatively in severely malnourished patients, stem cell or bone marrow transplant patients, or head and neck cancer patients—increased caloric intake appears to improve prognosis and reduce morbidity. Orexigenic agents also play an important role in reducing the morbidity associated with anorexia. However, even the most effective of these agents, megestrol and dexamethasone, provide benefit only to subgroups of patients.
Our improved understanding of the pathophysiology of cancer-associated weight loss and the acknowledgment that it is a "derangement of the constitution" and distinct from simple starvation have given rise to several pharmacologic interventions focusing on the abrogation of tissue wasting. These interventions, which require further study, may give rise to more effective treatment options for weight-losing cancer patients in the future.
