PEP Topic 

Pentoxifylline is a methylxanthine derivative approved by the U.S. Food and Drug Administration for the treatment of intermittent claudication. It can inhibit tumor necrosis factor alpha production, reducing plasma levels of this cytokine, which is thought to be a mediator in cancer-associated anorexia and cachexia (Goldberg et al., 1995). Pentoxifylline was evaluated in anorexia.

Effectiveness Not Established

Guideline/Expert Opinion

Desport, J.C., Gory-Delabaere, G., Blanc-Vincent, M.P., Bachmann, P., Beal, J., Benamouzig, R., . . . Senesse, P. (2003). Standards, options and recommendations for the use of appetite stimulants in oncology (2000). British Journal of Cancer, 89(Suppl. 1), S98–S100.

doi: 10.1038/sj.bjc.6601090

Purpose & Patient Population:

To review the literature on the use of appetite stimulants in oncology by a multidisciplinary group established by the French National Federation of Cancer Centers

Type of Resource/Evidence-Based Process:

The group conducted a literature search of four databases (MEDLINE, CancerLit, Embase, and the Cochrane Library) using search phrases: appetite stimulants, anorexia/drug therapy, cachexia/drug, or appetite associated with neoplasms. The search yielded 55 reports of randomized controlled trials (RCTs) published between 1990–1999 that evaluated the appetite-stimulating effect of corticosteroids, synthetic progestogens, or other drugs. 

The group defined standards, options, and recommendations for the use of appetite stimulants.

  • Standards: Clinical situations for which there exist strong indications or contraindications for a particular intervention.
  • Options: Situations for which there are several alternatives without clear superiority of one choice over another.
  • Recommendations: Additional information to enable the available options to be ranked using explicit criteria with an indication of the level of evidence.

They also defined the level of evidence. 

  • A: High-standard, meta-analysis, or several high-standard RCTs with consistent results.
  • B: Good-quality evidence from randomized trials or prospective or retrospective studies with consistent results.
  • C: Weak methodology of studies or inconsistent results.
  • D: Lack of scientific data or case study reports only.
  • Expert Agreement: Data do not exist for the method concerned, but the experts are unanimous in their judgment.

The primary outcome used in analysis of study results was anorexia. Secondary outcomes were improved quality of life, increase in body weight, increased food consumption, decrease in nausea and/or vomiting, and improvement in anthropometric and biologic parameters.

Guidelines & Recommendations:


Corticosteroids were found to be effective appetite stimulants. Their level of evidence was B1 (good-quality evidence from randomized trials), but optimal dose and scheduling information was lacking.

Synthetic Progestogens

Megesterol acetate: Effective appetite stimulant (level B1) and beneficial effect on body weight (level B1). Minimum effective dose is 160 mg/day (level B1). Optimal dose is 480 mg/day (level C). No greater efficacy was seen with doses higher than 480 mg/day (level B1).*

Medroxyprogesterone acetate: Effective appetite stimulant (level B1). Effect on weight was not confirmed (level C). The group recommended more clinical trials to establish optimal dose and duration of therapy.

Cyproheptadine: May be an appetite stimulant, but adverse effects were reported (level C).

Dronabinol, metoclopramide, nandrolone, pentoxifylline: No appetite-stimulating effects were shown (level C). These should be used only in the setting of RCTs.

Hydrazine sulfate: Not an appetite stimulant (level A).

Nursing Implications:

Corticosteroids and progestogens can be used in the treatment of anorexia in patients with cancer, especially in patients with advanced disease and with any type of cancer. Hydrazine sulfate should not be used.

* Data from trials completed after 1999 establish the safety and efficacy of higher doses of megesterol acetate.

Research Evidence Summaries

Goldberg, R.M., Loprinzi, C.L., Mailliard, J.A., O'Fallon, J.R., Krook, J.E., Ghosh, C., . . . Shanahan, T.G. (1995). Pentoxifylline for treatment of cancer anorexia and cachexia? A randomized, double-blind, placebo-controlled trial. Journal of Clinical Oncology, 13, 2856–2859.


Intervention Characteristics/Basic Study Process:

Patients were randomized, in double-blind fashion, to pentoxifylline 400 mg by mouth three times daily or placebo three times daily. If patients lost 5% of their on-study weight, the bind was broken and patients could cross over to the pentoxifylline arm.

Sample Characteristics:

  • N = 70 patients (randomized, 35 in each arm)
  • ELIGIBILITY CRITERIA: Older than 18 years of age, Eastern Cooperative Oncology Group (ECOG) performance status 0–2, life expectancy greater than three months, weight loss greater than five pounds over two months, estimated calorie intake of less than 20 kcal/kg/d
  • EXCLUSION CRITERIA: Parenteral or tube feeding; edema or ascites; use of corticosteroids, androgens, progestational agents, or hydrazine sulfate within one month; mechanical gastrointestinal (GI) obstruction; more than five episodes of emesis per week; primary or metastatic brain tumors; hemoptysis, GI hemorrhage, or Sr. Cr more than two times normal
  • Patients were stratified by
    • Primary tumor site
    • Amount of weight loss
    • Use of chemotherapy (CDDP versus non-CDDP)
    • Sex, ECOG performance status, and MD estimate of survival.

Study Design:

Patients were randomized in a double-blind fashion.

Measurement Instruments/Methods:

  • Objective evidence of weight change (excluded if patient had edema or ascites)
  • Questionnaires to evaluate appetite and perceived food intake


No significant difference was seen between cohorts in weight gain (P = .43). According to the questionnaires, there was no evidence that pentoxifylline is more effective than placebo in enhancing appetite. Toxicity data were similar for both groups for nausea and vomiting, fluid retention, and abdominal pain. The study was closed after the planned interim analysis, and it was concluded that pentoxifylline is not an effective treatment for cancer anorexia and cachexia.


The 95% confidence interval for the difference in the percentage of weight gain for pentoxifylline minus placebo ranged from a loss of 3.3% to a gain of 1.9%; this confidence interval suggests that the percentage of weight gain for pentoxifylline-treated patients is about the same for patients receiving placebo.


  • Heterogeneous group of patients with cancers arising from a variety of primary sites

Systematic Review/Meta-Analysis

Yavuzsen, T., Davis, M.P., Walsh, D., LeGrand, S., & Lagman, R. (2005). Systematic review of the treatment of cancer-associated anorexia and weight loss. Journal of Clinical Oncology, 23, 8500–8511.

doi: 10.1200/JCO.2005.01.8010

Search Strategy:

Studies were included in the review if they reported on

  • Adult patients older than 18 years of age
  • Patients with nonhematologic malignancies
  • Patients with anorexia or symptoms of anorexia, such as lack of appetite, weight loss, poor performance status, and decreased quality of life.

Literature Evaluated:

The review involved only prospective, randomized controlled trials (RCTs; double- and single-blind or unblended and phase III trials). The quality of studies was assessed using the validated scale published by Jadad et al. (1996).

Sample Characteristics:

There were 55 studies reviewed that met the eligibility criteria.


Androgenic steroids
Androgenic steroids were studied in two studies involving 512 patients; no significant benefit was demonstrated.
Cannabinoids were studied in one RCT involving 469 patients; they did not confer an additional benefit.
Six studies investigated the use of corticosteroids in 647 patients. Some improvements in appetite were found; however, dosage and type of steroid varied such that optimal dose and duration of therapy could not be determined.
Two studies investigated the use of cyproheptadine in 344 patients; these investigations had conflicting outcomes.
Eicosapentaenoic acid (EPA)
Three studies investigated the use of EPA in 689 patients; these reported conflicting results.
Erythropoietin (EPO)
EPO was investigated in two studies involving 417 patients. In one investigation, EPO was administered in combination with a COX-2 inhibitor with and without a specialized nutritional program. The intent-to-treat analysis was negative. No differences in food intake were noted.
Ghrelin was investigated in one RCT involving seven patients; differences between groups were noted, but long-term safety data on the agent are not available.
Hydrazine sulfate
Five studies investigated the use of hydrazine sulfate in 796 patients. Multicenter RCTs in patients with lung and colon cancers did not demonstrate any benefit when compared to a placebo.
Interferon was investigated in one study involving 57 patients; no differences were found.
Melatonin was investigated in two studies involving 186 patients; they did not demonstrate any improvement in appetite or intake.
Nonsteroidal anti-inflammatory drugs (NSAIDs)
NSAIDs were investigated in two trials involving 417 patients; these investigations failed to demonstrate a benefit in the NSAID arm.
One study investigated the use of pentoxyfilline in 70 patients and found no benefit. Melatonin was investigated in two studies involving 186 patients; these did not demonstrate any improvement in appetite or intake.
Twenty-nine studies reviewed the safety and efficacy of progestins: 23 examined megesterol acetate (MA), and six investigated medroxyprogesterone acetate (MPA). Results favored progestins over placebo and found that side effects were tolerable. Ten studies assessed the influence of MA on quality of life; these demonstrated that the effect of MA on quality of life was minimal.
Two studies investigated prokinetics for anorexia in a total of 55 patients. No improvement in caloric intake or appetite was noted.
Thalidomide has not been investigated in prospective RCTs.


Multiple RCTs have been conducted to investigate the safety and efficacy of pharmacologic agents to stimulate appetite. Only two therapeutic interventions for cancer-related anorexia demonstrated enough evidence to support their use in patients with cancer: corticosteroids and progestins. Other studies had mixed outcomes, positive results in only a single randomized trial, or were not placebo-controlled.

There is strong evidence supporting the use of progestins in patients with cancer, of which the most commonly reported drugs were MA and MPA. There was increased weight with both progestins; there was also evidence of a dose-response, but higher doses did not confer any additional benefit with regard to appetite. Metaclopromide is effective for nausea and early satiety but has not been shown to directly stimulate appetite.

The RCTs did not show sufficient evidence to justify the use of dronabinol, EPA, EPO, ghrelin, interferon, melatonin, nandrolone, NSAIDs, or pentoxyfilline in cancer-related anorexia. Cyproheptadine is a weak appetite stimulant, but side effects are limiting.

Nursing Implications:

The optimal dose, start time, and duration of treatment for many appetite stimulants are still unknown. A more systematic approach to research methodology is needed. In addition, uniform outcome measures to better assess the value of various appetite stimulants are needed. These should include subjective ratings of appetite and associated symptoms (e.g.,  early satiety) and objective measures (e.g., food consumed, weight gain, weight loss).