Several randomized controlled trials, recently reviewed by Mock, indicate that exercise is beneficial in the treatment or prevention of fatigue in cancer patients and cancer survivors. Although the numbers of subjects in these studies were small,[52-135] the impact of exercise was pronounced, with reduction in fatigue typically in the 40% to 50% range. Improvements in overall quality of life were noted in three out of four studies. Numerous nonrandomized studies also suggest that exercise has beneficial effects on cancer-related fatigue. Certain caveats regarding these data should be noted. The participants in these studies were probably highly selected and motivated, resulting in good adherence to the exercise programs. The majority of exercise trials focused on patients with early-stage cancer, particularly breast cancer, although benefits were also observed in patients with prostate cancer and other malignancies. Exercise programs must be individualized; some types may present risks for patients with severe cytopenias, bone metastases, compromised pulmonary function, or cardiovascular disease. No sham-controlled studies have been performed (in fact, it is difficult to imagine how they could be designed). One cannot exclude the possibility that some of the reported benefits of exercise represent a placebo effect. However, it can be argued that the distinction between a placebo effect and a "real" improvement in a subjective symptom such as fatigue is meaningless when the intervention has negligible risks and cost. Sleep disturbances may be a prime cause of fatigue in some patients. Mood disorders, pain, cough, and other symptoms may disrupt sleep. Even when the duration of sleep is adequate or increased, sleep patterns may be aberrant, leading to disturbances in circadian rhythms. Some evidence suggests that energy conservation strategies may improve sleep patterns and daytime functioning in patients with cancer-related fatigue. The impact of sedatives in cancer patients suffering from disordered sleep has not been studied. Debilitating fatigue may contribute to the high incidence of depression in cancer patients. Conversely, in some patients the vegetative symptoms of depression may exacerbate fatigue. The vicious cycle of depressive symptoms and fatigue has not been explored in detail, but, not surprisingly, strong correlations between mood states and fatigue have been reported in numerous studies.[20,22,36- 42] In some cases, psychological distress appears to be the single most powerful predictor of fatigue. There is evidence that behavioral or psychological interventions may ameliorate fatigue. Support groups, individual counseling and education, and psychotherapy have been reported to reduce cancer-related fatigue.[ 2] However, studies of these interventions have generally been small and of limited applicability. Given the multiple etiologies of cancerrelated fatigue, and the cultural and psychological diversity of those who suffer from it, no single approach is likely to be helpful in all patients. Pharmacologic Approaches
- Traditional Stimulants-Since the pathophysiology of cancer-related fatigue remains obscure in many cases, pharmacotherapy is either empiric, or geared toward reversing possible contributing factors. Stimulant medications such as methylphenidate(Drug information on methylphenidate) and dextroamphetamine are probably the most widely prescribed medicines for cancer-related fatigue when no reversible cause is evident. Methylphenidate has been shown to ameliorate fatigue in people with AIDS. In cancer patients, small crossover studies suggest a benefit, and larger randomized controlled trials are ongoing. The adverse effects of these medicines are a concern, however. Anorexia and irritability are common, and hypertension, tachycardia, and arrhythmias may occur. There is some evidence that these medications lower the seizure threshold. Methylphenidate inhibits the metabolism of warfarin(Drug information on warfarin) and some anticonvulsants. Pemoline(Drug information on pemoline) has central nervous system (CNS) effects similar to the other stimulants but lacks sympathomimetic effects. Its interactions with other medications have not been well studied.
- Modafinil-Modafinil (Provigil) is a wakefulness-promoting agent that has been approved for use in narcolepsy. Its mechanism of action is unknown, but it is not a sympathomimetic agent and does not appear to act on dopaminergic pathways in the brain. The drug lacks the cardiovascular and hemodynamic effects of the conventional psychostimulants. It has been studied in fatigue associated with sleep deprivation and in multiple sclerosis. Randomized controlled trials of modafinil(Drug information on modafinil) for cancer-related fatigue are ongoing.
- Paroxetine-Based in part on the observation that cancer-related fatigue and depression frequently coexist, a common etiology involving synaptic 5-hydroxytryptamine levels has been hypothesized. This hypothesis was the basis for a randomized controlled trial of the selective serotonin-reuptake inhibitor (SSRI) antidepressant paroxetine(Drug information on paroxetine) in patients with cancer-related fatigue. Subjects reporting fatigue after two cycles of chemotherapy were randomly assigned to 20 mg of paroxetine daily (n = 277), or placebo (n = 272). Of note, 32% of patients in both arms had significant depressive symptoms at baseline, and half reported fatigue scores ≥ 5 on a 10-point scale. As in numerous other studies, depression and fatigue correlated strongly. After four cycles of chemotherapy, those receiving paroxetine had significantly lower mean levels of depression, but paroxetine had no impact-either positive or negative- on fatigue. Even in the subgroup of patients who were depressed, there was no improvement in fatigue with paroxetine. This study was one of a very few adequately powered, rigorously designed drug trials for cancer-related fatigue. The findings strongly support screening and treating cancer patients for depression, but the impact of this treatment on fatigue was disappointing. It is possible that the acute fatigue induced by chemotherapy is so overwhelming that it is not amenable to such an approach. Antidepressants might have a greater effect on fatigue in cancer survivors, in whom fatigue and depression are prevalent and frequently coexist.
- Miscellaneous Drugs-Other agents currently being investigated for treatment of cancer-related fatigue include L-carnitine and infliximab(Drug information on infliximab) (Remicade).
- Erythropoiesis-Stimulating Agents-There is little doubt that recombinant human erythropoietin(Drug information on erythropoietin) (rHuEPO, epoetin alfa(Drug information on epoetin alfa) [Epogen, Procrit] and epoetin beta(Drug information on epoetin beta)) and the related compound darbepoetin (Aranesp) stimulate erythropoiesis and reduce the likelihood of red blood cell transfusion in cancer patients undergoing chemotherapy. The impact of these treatments on quality of life and symptoms (including cancer-related fatigue) is less well established. A trend toward increased survival with recombinant erythropoietin has been reported in some studies; however, several recent reports have raised concerns about increased thrombotic events, accelerated tumor growth, and decrements in survival in cancer patients treated with such agents.[46-49]
The effects of any treatment on symptoms or quality of life are optimally assessed in randomized, controlled studies. At least nine such studies of recombinant erythropoietin have been published; several others remain unpublished. All but one of the published randomized controlled trials were small and reported very limited data on quality of life, fatigue, or other symptoms. Their applicability is generally quite low. Only one adequately powered, double-blind, randomized, placebocontrolled trial of the effects of rHuEPO on fatigue and quality of life in patients undergoing chemotherapy has been published. Patients were randomly assigned to receive placebo (n = 124) or rHuEPO (n = 251) subcutaneously three times per week. Patients were required to have a hemoglobin level ≤ 10.5 g/dL, or 10.6 to 12 g/dL with a decline of ≥ 1.5 g/dL per cycle of chemotherapy. The primary end point in the trial was the proportion of patients transfused after 4 weeks. Quality of life was assessed using the Functional Assessment of Cancer-Anemia (FACT-An) instrument (which contains a fatigue subscale), the Medical Outcomes Short Form-36 (SF-36), and a Linear Analog Scale Assessment. Hematologic response to rHuEPO was consistent with the results of prior studies: The proportion of patients requiring transfusion after day 28 was reduced by 14.8% (P = .0057). The mean increase in hemoglobin from baseline to final measurement was 2.2 g/dL in the rHuEPO arm, and 0.5 g/dL in the placebo arm, despite greater use of transfusion in those receiving placebo. It should be noted, however, that the change in hemoglobin was not reported on an intentionto- treat basis. The investigators found a statistically significant difference in the mean change on the fatigue subscale of the FACT-An, favoring rHuEPO over placebo. Significant advantages for rHuEPO were also observed in scores on the FACT-G, Linear Analog Scores for Energy, Ability to do Daily Activities, and Overall QOL scales. Nonsignificant trends in quality-of-life improvement favoring rHuEPO were observed on the SF-36. This study is frequently cited as providing the most robust evidence for an impact of recombinant erythropoietin on fatigue and quality of life. However, its conclusions are open to question for a number of reasons. A higher proportion of patients in the placebo arm had received transfusions within 3 months prior to starting the study (36% vs 28%), suggesting that they may have had impaired hematopoiesis, despite similar mean baseline hemoglobin levels in the two groups. Baseline data on important prognostic factors for quality of life, such as performance status and weight loss, were not reported. Hence, one cannot assume that the two arms were well balanced for these factors. Indeed, no baseline data was reported on any of the quality-of-life measures used in the study; the results were reported only as changes from baseline. Transfusions were permitted at the discretion of the physician but were to be avoided unless the hemoglobin was < 8 g/dL. Thus, patients were allowed to become significantly anemic. More reasonable transfusion support may have resulted in improved quality of life and less fatigue in the placebo group, and consequently attenuated the observed differences between placebo and treatment. In addition, the analysis of symptoms and quality of life was not based on an intention to treat. Twenty-six patients were excluded from this analysis because of missing data. FACTAn and SF-36 data were missing for an additional 54 patients because validated versions of these instruments were not available in the languages they spoke. The minimum differences in quality-of-life measures considered clinically significant were not defined prospectively. Finally, there was a trend toward prolonged survival in the rHuEPO arm (17 vs 11 months, P = .13), which the investigators speculated might be attributable to higher hemoglobin levels. However, it is also possible that the prolongation in survival was due to imbalances favoring the rHuEPO arm in important, but unreported, prognostic factors for both survival and quality of life, such as burden of disease, number of prior treatments, response to chemotherapy, performance status, and weight loss.
- Question of Increased Mortality- Concerns about the use of erythropoietic agents in cancer patients have been heightened by recent reports of two randomized, doubleblind, placebo-controlled trials in which subjects receiving recombinant erythropoietin had statistically significant decrements in survival. Henke et al randomized 351 anemic head and neck cancer patients undergoing radiation therapy to epoetin or placebo. Those treated with rHuEPO had a much higher hematologic response rate (82% vs 15%). However, in an intention-to-treat analysis taking into account other known prognostic factors, locoregional progression- free survival was poorer with rHuEPO (relative risk [RR] = 1.62; 95% CI = 1.22-2.14; P = .0008). Overall survival was also worse (RR = 1.39; 95% CI = 1.5-1.84; P = .02). There were no obvious study design issues that could have accounted for these findings. In a second randomized, doubleblind study, 939 nonanemic patients with metastatic breast cancer were assigned to recombinant erythropoietin or placebo in order to prevent anemia. The trial was designed to assess the effects of 12 months of rHuEPO treatment on survival. It was terminated early, at the recommendation of an independent data monitoring committee, due to a higher mortality rate in the rHuEPO arm. The 12-month survival rate was 76% in the placebo arm and 70% in the erythropoietin arm (P = .0117), mainly as a result of increased mortality in the first 4 months (41 deaths in the rHuEPO group and 16 in the placebo group). Most of the deaths were due to disease progression or thrombotic events. The investigators could not exclude the possibility that the survival difference was due to imbalances between the treatment groups, rather than to the treatment itself, although a multivariate analysis incorporating available prognostic factors reached results similar to the univariate analysis. The basis for increased mortality in patients treated with recombinant erythropoietin in these studies is unclear. The studies have been analyzed by their industry sponsors and by the US Food and Drug Administration (FDA) to determine if factors other than rHuEPO treatment could have contributed to the observed outcomes. At present, no firm conclusions can be drawn. The outcomes of these clinical studies have also refocused attention on the complex and pleiotropic biologic effects of erythropoietin. The erythropoietin receptor is widely expressed- in the central nervous system, the gastrointestinal tract, on vascular endothelium, and elsewhere.[ 56] Expression of the erythropoietin receptor has also been found in many tumor types, including 80% of the head and neck tumors from the study by Henke.[46,56] Although the clinical significance of these findings is unknown, erythropoietin has been found to accelerate the proliferation of some cancer cells in vitro, to render them resistant to chemotherapy, and to promote angiogenesis.[ 56] In vivo animal studies also support the hypothesis that erythropoietin is a growth factor for some types of tumors.
- Thrombosis Risk-As with the breast cancer study cited above, an increase in the risk of venous thrombosis with rHuEPO was suggested by a retrospective, case-control study of women with cervical or vaginal cancer. The rate of symptomatic thrombosis in those receiving rHuEPO was 22.6% (17 events in 75 patients), compared with 2.8% (2/72) in those not receiving the drug , even though the groups were similar in terms of other known risks for thrombosis. In multivariate analysis, the odds ratio for thrombosis with rHuEPO was 15.3 (95% CI = 3.1-76.7). Recently, a randomized trial of recombinant erythropoietin to reduce fatigue in patients with metastatic breast cancer and mild anemia was terminated after 4 out of 14 patients treated with the agent developed thrombotic events. None of the 13 patients in the control arm had a thrombotic event. Among 15 other randomized controlled trials of rHuEPO included in the AHRQ evidence report, only 6 reported rates of thrombotic events. The total number of subjects in these studies was 580; 4.7% of those treated with rHuEPO and 2.5% of controls had deep-vein thromboses or thromboembolism (P = .41). Thus, the number of cancer patients in whom the risk of thrombosis has been assessed may not be large enough to exclude a small but significant increased risk associated with rHuEPO, especially in cancer patients otherwise predisposed to thrombotic events. Due to the concerns raised by these studies, several other ongoing clinical trials of epoetin were terminated early. A common factor in the studies in which increased adverse events or excess mortality were reported was the targeting of hemoglobin to higher levels than recommended in the prescribing information for these drugs. Rapid increases in hemoglobin levels may also be associated with excess cardiovascular and thrombotic complications.
- Data Reconsidered-The possibility that recombinant erythropoietin has an adverse effect on survival under some circumstances, and the data on thrombotic risks, were reviewed on May 4, 2004, by an Oncologic Drugs Advisory Committee of the FDA's Center for Drug Evaluation and Research. Extensive unpublished data and analyses were presented in an FDA briefing document, and by industry representatives. Interested readers are referred to the briefing document and transcript of the meeting, available on the Internet (www.fda.gov/ohrms/dockets/ac/04/ briefing/4037B2_04_Aranesp- Procrit.doc, and www.fda.gov/ohrms/ d o c k e t s / a c / 0 4 / t r a n s c r i p t s / 4037T2.htm). Aside from these concerns, rHu- EPO appears to be generally well tolerated, with a low incidence of serious adverse events. Pure red cell aplasia due to anti-erythropoietin antibodies has been reported in more than 200 patients, most of whom were treated with epoetin alfa. Although rare, this syndrome should be considered when a patient treated with erythropoietin develops worsening anemia.