Up to 50% of Americans will develop cancer, and >90% of patients will experience cancer-related fatigue (CRF) at some point in their disease course. Patients describe CRF as devastating to many life domains, degrading their vocational, familial, and societal roles. This article describes current best efforts to define CRF and to characterize its epidemiology using these definitions. In addition, the article will outlines assessment tools, proposed mechanisms, associated symptoms, and treatment strategies. Readers are offered highly more comprehensive references relating to specific aspects of CRF throughout.
Cancer-related fatigue (CRF) is a prevalent and morbid etiologic puzzle. Its relevance is demonstrated by the fact that between one-half and one-third of Americans will develop cancer, and >90% of patients will experience CRF at some point in their disease course. Patients describe CRF as devastating to many life domains, degrading their vocational, familial, and societal roles. Patients rate CRF as more distressing than any other cancer- or treatment-related symptom, including pain, with symptoms persisting for years after discontinuation of cancer treatment. CRF is perhaps the single most challenging barrier to effective rehabilitation of cancer patients, because it severely constrains patients’ ability to actively engage in functional restoration and adhere to the long-term programs required for success. CRF may become so severe that patients interrupt or abandon cancer treatment.
Despite accord that CRF is an important issue with wide-reaching public health and fiscal consequences, a pithy, simplistic definition that distinguishes CRF from related symptoms and syndromes (eg, chemotherapy-induced anemia, “chemo-brain”) remains elusive. Most experts concur that fatigue is a multidimensional experience with physical and cognitive dimensions. Whether and how often physical and/or cognitive features persist when all remediable factors have been definitively addressed (eg, endocrinopathies, depression) is uncertain, hence the ongoing debate regarding the unique features of CRF.
Dishearteningly, despite extensive research, effective primary and secondary preventive strategies remain limited. This lack reflects the investigative challenge of differentiating and measuring fatigue in populations with many relevant and changing characteristics (eg, disease and treatment status, elapsed time since treatment).
Definition
Several definitions have been proposed. All are useful, and the subtle differences between them illustrate extant controversies in the field. The National Consortium of Cancer Centers (NCCN) defines fatigue as, “an unusual persistent subjective sense of tiredness related to cancer or cancer treatment that interferes with usual functioning.” Diagnosis of CRF according to the International Classification of Disease (ICD)-10 requires a known tumor and daily persistence of the symptom for ≥2 weeks plus 6 of the following 11 complaints: diminished energy, increasing need for rest, limb heaviness, diminished ability to concentrate, decreased interest in engaging in normal activities, sleep disorder, inertia, emotional lability due to fatigue, perceived problems with short-term memory, and postexertional malaise exceeding several hours ( Table 1 ).
| Diminished energy |
| Increasing need for rest |
| Limb heaviness |
| Diminished ability to concentrate |
| Decreased interest in engaging in normal activities |
| Sleep disorder |
| Inertia |
| Emotional lability due to fatigue |
| Perceived problems with short-term memory |
| Postexertional malaise exceeding several hours |
Although not included in most formal definitions, additional, accepted characteristics of CRF include tiredness disproportionate to the intensity of patients’ exertional level, which is not relieved by rest or sleep and subjective weakness. Experts concur that fatigue reduces the mental capacity and psychological resilience of cancer patients. Psychological complaints may include reduced motivation, capacity to attend, and concentration. Patients may also experience difficulty with new learning.
Epidemiology
Fatigue occurs most commonly and severely during administration of anticancer therapies. CRF is experienced by up to 99% of patients receiving chemotherapy with complaints generally being most intense during the first 3 days after chemotherapy administration and improving gradually over the subsequent week. Over 40% of patients rate their fatigue as “severe” or ≥7 on an 11-point numerical rating scale. CRF incidence rates in the clinical trial setting range from 70% to 80%, with variations being explained by differences in cancer and treatment type. CRF is frequently present during diagnosis, increases throughout treatment, and commonly persists for years after the completion of therapy. CRF continues to be problematic even when survivors’ cancer is undetectable. Reported incidence rates of CRF among cancer survivors range as high as 81%, with 17% to 38% reporting high levels of fatigue during 6 months after treatment. Fatigue is associated with decreased disease-free and overall survival among cancer patients.
Epidemiology
Fatigue occurs most commonly and severely during administration of anticancer therapies. CRF is experienced by up to 99% of patients receiving chemotherapy with complaints generally being most intense during the first 3 days after chemotherapy administration and improving gradually over the subsequent week. Over 40% of patients rate their fatigue as “severe” or ≥7 on an 11-point numerical rating scale. CRF incidence rates in the clinical trial setting range from 70% to 80%, with variations being explained by differences in cancer and treatment type. CRF is frequently present during diagnosis, increases throughout treatment, and commonly persists for years after the completion of therapy. CRF continues to be problematic even when survivors’ cancer is undetectable. Reported incidence rates of CRF among cancer survivors range as high as 81%, with 17% to 38% reporting high levels of fatigue during 6 months after treatment. Fatigue is associated with decreased disease-free and overall survival among cancer patients.
Assessment
Symptom characterization is the prime goal of fatigue assessment. However, comprehensive evaluation should address cancer treatment history, medical comorbidities, and associated symptoms. Physical examination may reveal remediable etiologies. An approach to fatigue assessment similar to that endorsed for pain evaluation has been proposed, including use of 11-point numeric rating scales (NRS) to characterize symptom intensity “at worse,” “at best,” and “on average.” NRS can also be used to query patients regarding fatigue interference with work, relations with others, recreational activities, and so on. The responsiveness of fatigue NRS to changes in symptom intensity has been established.
Fatigue is frequently assessed in epidemiologic studies and therapeutic trials using validated multidimensional instruments. Many such instruments are available, and they vary considerably in length. Examples of patient report assessment instruments include the Functional Assessment of Cancer Treatment-Fatigue scale (FACT-F), Profile of Mood States (POMS) vigor and fatigue subscales, Piper Fatigue Scale, and Multidimensional Fatigue Inventory. The length and complex scoring algorithms of many of these instruments limit their utility in clinical practice. However, short forms have been developed for some instruments, for example, POMS Short-Form fatigue subscale. Coadministration with depression and pain assessment tools can help clinicians to unravel the frequent puzzle of concurrent symptoms.
Associated symptoms
Fatigue frequently occurs in association with pain, depression and anxiety, dyspnea, and insomnia, among other symptoms. The strength of the associations varies depending on the study population, whether patients are receiving active treatment, and patients’ cancer stage. The fact that these associations have not been prospectively or longitudinally studied in a broad range of cohorts severely limits any conclusions that can be drawn regarding causal relationships or common etiologies. Clinicians have long appreciated the inciting and sustaining roles that symptoms such as pain can play in the development and persistence of fatigue. Intense, persistent, or refractory symptoms unquestionably engender fatigue. Symptoms may reciprocally intensify each other. For example, depression and fatigue may aggravate each other though arising from different sources. Conversely, in many cases a common pathology, for example, tumors, may produce local symptoms of effects on adjacent bones and nerves while simultaneously elaborating biological response modifiers that produce fatigue.
Most reported associations between CRF and other symptoms are the product of regression modeling of cross-sectional data gleaned from cohorts defined by disease, stage, and treatment status (eg, breast cancer survivors). Few studies have followed cohorts longitudinally. Therefore, little is known about the inter-relationships of symptoms over time. Approximately 19% of patients with CRF are clinically depressed. An estimated 30% to 35% of the variance in fatigue can be attributed to concurrent symptoms, with pain being the most relevant. Dyspnea associates strongly with fatigue in advanced cancer, and along with pain and psychological distress explained 56% of fatigue variance in 1 report. In a different study, pain accounted for 40% of fatigue variance.
The construct of “symptom clusters” has featured prominently in the cancer literature in recent years, with fatigue appearing among many proposed clusters. Investigations are underway to identify genetic factors that may explain cluster patterns; however, such efforts have yet to generate clinically applicable information. Although fatigue clearly associates with adverse symptoms, evidence is inconsistent as to whether symptom control ameliorates CRF. Nonetheless, a comprehensive symptom assessment is warranted in the evaluation of CRF, and effective treatment can be reasonably anticipated to improve fatigue.
Remediable contributors
A discrete source and are of of fatigue can be identified in some patients, leading to effective treatment and symptom reversal. Commonly, many potentially contributory mechanisms can be identified and are of uncertain relative importance. Abetting conditions include endocrinopathies (hypogonadism, hypothyroidism, adrenal insufficiency), blood dyscrasia (anemia), degraded sleep quality (obstructive sleep apnea), centrally acting medications, steroid myopathy, and cachexia ( Table 2 ). Each of these possibilities should be examined in cancer patients presenting with fatigue; though anemia often receives disproportionate attention.
| Endocrinopathies |
| Hypogonadism |
| Hypothyroidism |
| Adrenal insufficiency |
| Blood dyscrasia |
| Anemia |
| Degraded sleep quality |
| Obstructive sleep apnea |
| Centrally acting medications |
| Steroid myopathy |
| Cachexia |
The historic focus on anemia is understandable, since hemoglobin diminishes in patients receiving antineoplastic therapy. Roughly 50% of patients with solid tumors are anemic at diagnosis. Hematologic malignancies are associated with higher prevalences; for example, 60% to 70% of patients with non-Hodgkin’s lymphoma are anemic at the time of diagnosis. The relevance of anemia to CRF has received less emphasis of late. Initial interest stemmed from reports that fatigue severity paralleled reductions in serum hemoglobin. More recently, it has been appreciated that the time course of fatigue onset and improvement differs substantially from fluctuations in blood counts. Normalization of hemoglobin levels through blood transfusion or erythropoietin administration inconsistently alleviates fatigue. Further, physician practices with respect to threshold hemoglobin levels for treatment are inconsistent, ranging from 7.5 g/dL to 10.7 g/dL. This inconsistency reflects the fact that no specific decrement or increment in hemoglobin value has been definitively linked to quantitative changes in quality of life (QOL).
Apart from anemia’s role in CRF, no one disputes that even mild anemia can significantly compromise subjective QOL. Hence, proactive treatment with recombinant erythropoietin is widely accepted. The Anemia Guidelines Development Group recommends starting epoetins in patients with prechemotherapy baseline hemoglobin of <10 g/dL, symptomatic baseline anemia, or a drop of 1 to 2 g/dL per chemotherapy cycle. The American Society of Clinical Oncology (ASCO)/American Society of Hematology (ASH) guidelines also use 10 g/dL as the threshold hemoglobin value. Patients who have poor responses to epoetin therapy, intensely symptomatic anemia, hemoglobin levels ≤9 g/dL, or economic constraints to epoetin access may require red blood cell transfusion.
Endocrinopathies should be sought due to the frequency of underdiagnosis and the ready availability of replacement therapies. Disruption of the adrenal axis, thyroid gland, testes, and ovaries by chemical ablation, surgical resection, or irradiation can contribute to fatigue. Appropriate serologies will facilitate rapid identification of deficiencies.
Other remediable contributors can be expeditiously detected through appropriate assessments. Reports of poor sleep may indicate the need for a sleep study if elimination of daytime napping and use of soporifics are unhelpful. Centrally acting pharmaceuticals should be eliminated or replaced by less problematic alternatives whenever possible. A reduction or withdrawal trial of nonessential drugs can identify those producing fatigue. Although steroid myopathy may be inescapable due to the need for co-administration of steroids with chemotherapy, its timely recognition may permit steroid dose reduction and accurate prognostication, which is appreciated by patients.
Deconditioning related to inactivity is common among cancer patients. If not an instigating factor, deconditioning can aggravate fatigue from other causes. Aerobic exercise, as discussed later, features prominently in the management of CRF. Additionally, exercise contributes to the primary and secondary prevention of breast and colon cancers and perhaps other malignancies as well. Few cancer patients receive formal, practical guidance in how to begin an exercise program. For these reasons, rehabilitation clinicians should provide each patient with concrete, detailed recommendations for an exercise regimen specifying type, intensity, frequency, and plans for program advancement.
Proposed Mechanisms
The pathophysiological processes underlying CRF are incompletely understood. Most proposed mechanisms reflect efforts to account for associations between the intensity of CRF and markers of disease burden, treatment toxicity, symptom burden, and patient behaviors. The elaboration of biological response modifiers (eg, cytokines) by tumors and by the body in response to tumors or treatments was among the earliest explanations. Support derived principally from a well-characterized association between the administration of biological response modifiers for therapeutic purposes and the onset of severe fatigue. Equally suggestive were reports that tumor necrosis factor-α and interleukin-6 are elevated in some patients with chronic fatigue syndrome and that synthetic antibodies directed at proinflammatory cytokines reduce fatigue in patients with rheumatoid arthritis. However, efforts to correlate levels of circulating cytokines with CRF have not succeeded. Biological response modifiers, therefore, figure less prominently in current mechanistic discussions, particularly regarding CRF that far outlasts the administration of cancer treatments.
Serotonin (5-HT) dysregulation has been examined as a potential contributor to CRF largely due to mounting evidence that 5-HT plays an important role in disparate fatigue states. 5-HT may contribute to fatigue experienced by healthy subjects during vigorous exercise. Tryptophan is a precursor of 5-HT, and brain tryptophan levels increase significantly during normal exercise. Animal models demonstrate an inverse dose relationship between treadmill endurance and 5-HT. Further, circulating tryptophan levels are elevated in chronic fatigue syndrome. Despite suggestive findings implicating 5-HT, central 5-HT concentrations do not correlate with the presence or intensity of CRF.
Data directly link aberrant hypothalamic-pituitary-adrenal (HPA) axis function to CRF, among other sources of fatigue. Breast cancer survivors with CRF have reduced waking serum cortisol levels, relative to their unaffected counterparts. Exposing CRF-afflicted breast cancer survivors to a controlled stressor produced blunted stress responses as reflected in low salivary cortisol levels ( Fig. 1 ). The investigators question whether irregularities in diurnal cortisol regulation may ultimately prove more relevant to CRF than overall cortisol levels. The HPA axis has been proposed to be the means by which cytokines and 5-HT may affect CRF, since cortisol, cytokines, and 5-HT levels cross-regulate one another.
