SEARCH

SEARCH BY CITATION

Keywords:

  • fatigue;
  • brain tumor;
  • performance status;
  • sex difference;
  • disease status

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

BACKGROUND:

In addition to neurologic symptoms, fatigue is commonly reported in patients with primary brain tumors during radiation therapy and in long–term survivors of low–grade brain tumors. Other factors have not been explored. The aim of this study was to identify demographic and clinical factors that predict fatigue severity and to evaluate the association of fatigue with other symptoms throughout the disease trajectory.

METHODS:

Two hundred one patients with primary brain tumors completed the M. D. Anderson Symptom Inventory–Brain Tumor Module and a demographic checklist. Clinical data, including treatment, tumor grade, and performance status, were also collected. Correlations among fatigue and other recorded symptoms were evaluated. Logistic regression modeling was performed to evaluate factors associated with fatigue severity.

RESULTS:

Fatigue severity was associated with symptoms including pain, drowsiness, distress, difficulty sleeping, and weakness as well as overall symptom severity and interference. Poor performance status (Karnofsky scale) (odds ratio [OR], 5.73; P = .001), female sex (OR, 2.48; P = .005), and disease status (OR, 2.20; P = .013) were the strongest predictors of fatigue. Severity of fatigue for women was primarily predicted by disease status (OR, 3.33; P = .01) For men, antidepressant use (OR, 4.43; P = .013) in addition to opioids (OR, 3.46; P = .017) and performance status (OR, 12.47; P = .0001) predicted fatigue severity.

CONCLUSIONS:

Fatigue should not be considered a solitary symptom with 1 root cause, but a complex symptom related to the severity of other symptoms and potentially having various etiologies. Future studies should consider these factors in planning interventions and assessing response. Cancer 2010. © 2010 American Cancer Society.

Primary brain tumors are a heterogeneous group of neoplasms that are thought to arise from the constituent cells of the brain. Despite their malignant phenotype, they rarely spread outside the central nervous system. Median survival time varies from approximately 1 year for high-grade (grade 4) tumors, such as glioblastoma, to 5 to 15 years for low-grade (grade 2) tumors. Regardless of tumor grade, patients often suffer from neurologic symptoms such as headaches, weakness, and seizures at the time of diagnosis and throughout the disease trajectory.1, 2 For most patients, initial therapy consists of a surgical procedure to obtain tissue for diagnosis and remove tumor if medically feasible.

Tumor cell infiltration into surrounding brain parenchyma is characteristic of all malignant tumors (grade 2 and higher), and even resection of all visible tumor is unlikely to eradicate all malignant cells.3, 4 Therefore, other therapies are often needed after surgical resection as the initial tumor treatment. Other therapies include radiation therapy, chemotherapy, or a combination of both. The most common treatment approach is radiation therapy. These therapies and other concomitant medications, such as corticosteroids and anticonvulsants, often cause additional symptoms that further limit functional status and overall quality of life (QOL).2

Similar to patients with other solid tumors, fatigue is among the most common and most troublesome symptoms for primary brain tumor patients throughout the disease trajectory.5, 6 In a survey evaluating QOL of patients at various times in the trajectory of illness and with a variety of primary brain tumors, 42% reported “quite a bit low” or “very low” energy levels.7, 8 Radiation therapy is the most common treatment modality for all tumor grades. Standard treatment with radiation therapy for grade 2 and 4 tumors is 60 grays (Gy) delivered in 2–Gy fractions over a 6-week period to the gross total tumor volume plus a 2- to 3-cm margin. Lovely and colleagues reported that >80% of primary brain tumor patients report fatigue during radiation therapy.9 Fatigue has been reported to occur as early as within 1 week of the first radiation treatment, and tends to increase with the number of radiation fractions.2

Fatigue that occurs during radiation therapy may continue into the postradiation period. Faithfull and Brada reported on the occurrence of a somnolence syndrome in the immediate postradiation period.10 This syndrome included fatigue, excessive drowsiness, feeling clumsy, and inability to concentrate. In this study, patients were observed during the immediate postradiation period. After completion of radiation therapy, the reported symptoms had a cyclical pattern, with increased severity during Day 1 to 21 and then Day 30 to 35 after treatment. Fatigue has been reported to persist for 1 to 3 months after the completion of treatment, but may be more chronic for some individuals.11 The applicability of these findings to current treatment are limited by the small sample size (n = 19), lack of baseline measurement, and changes in delivery and radiation field determination used with modern treatment.

Clinically, a variety of other factors may contribute to the frequency and intensity of the fatigue. These include concomitant medications such as anticonvulsants and corticosteroids, metabolic disturbances, and psychosocial issues such as depression and anxiety. Most patients require corticosteroids to treat brain edema and anticonvulsants for seizure management. These medications have been reported to contribute to fatigue in this patient population.2, 12 Depression and anxiety have been reported to occur in 16% to 50% of patients during the early stages of the disease.13, 14 Although not evaluated in relation to fatigue in the primary brain tumor population, depression, anxiety, and distress have been reported as being associated with fatigue in other solid tumor patients.15-17

Fatigue may also persist for years after diagnosis and completion of therapy. A recent report explored the occurrence of fatigue in patients with low-grade gliomas who were at least 3 years from completion of tumor therapy.18 In this study, 39% of patients reported severe fatigue >8 years after completion of therapy. The only variable found to predict fatigue was continued anticonvulsant use. Findings of this study are limited by the small sample size, lack of information on type of anticonvulsant prescribed, seizure activity, and use of chemotherapy in this patient population.

In summary, there are limited studies to date exploring the occurrence of fatigue and evaluating associated variables in patients with primary brain tumors. In patients with other solid tumors, fatigue is often the most common and severe symptom associated with the disease and treatment.15, 19 Fatigue has been demonstrated to cluster with other symptoms, including pain, distress, insomnia, and depression, and to influence outcomes such as perceived health and functional status.16, 20 In the limited studies to date in patients with primary brain tumors, fatigue has been identified as a common symptom occurring in patients with both low- and high-grade tumors. Fatigue occurs during radiation therapy and also occurs in long-term survivors of low-grade brain tumors. Most studies are limited to descriptive reports exploring fatigue in relation to a particular treatment,10, 21 tumor type,9 or stage of disease.18 None of these reports explored the occurrence of fatigue in relation to other symptoms and throughout the illness trajectory to uncover clinical factors that may be associated with the severity of fatigue in primary brain tumor patients. Evaluating these clinical factors is a critical first step to model the occurrence of fatigue longitudinally and to explore potential biologic correlates that may be associated with these factors.

The aim of this study was to determine individual demographic characteristics and clinical characteristics that may be associated with moderate to severe fatigue in the primary brain tumor population. Such patient characteristics include demographics such as age and sex as well as clinical characteristics such as tumor grade, disease status, tumor size, concomitant medications, cancer treatment, and performance status. We also wanted to identify the association of moderate to severe fatigue with the occurrence of other symptoms and the interference that symptoms have in daily life. Identification of risk factors related to fatigue severity may allow for individualized approaches to symptom management and determine key variables to explore in elucidating the biologic bases of fatigue in this patient population.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Patients

Two hundred one patients diagnosed with a primary brain tumor at various stages in the disease trajectory participated in this study. These patients were undergoing treatment at the Brain and Spine Center at The University of Texas M. D. Anderson Cancer Center, were at least 18 years of age, were able to speak and read English, and did not have cognitive deficits as determined by evaluation by the treating physician that would preclude self-report. We elected to exclude patients who were unable to self-report symptoms because of the discrepant results when a proxy was used to report symptoms for patients with cancer, and the lack of published data regarding use of proxy reports with the M. D. Anderson Symptom Inventory–Brain Tumor Module.22-24 Only 3 patients who were otherwise eligible were not approached about participation because of significant cognitive deficits. Patients were participants in the validation study of the M. D. Anderson Symptom Inventory–Brain Tumor Module.25 Once informed consent was obtained, patients completed a demographic questionnaire and the M. D. Anderson Symptom Inventory–Brain Tumor Module at 1 point in time. Clinical data were extracted from the medical record at the time of the visit and recorded on a clinician checklist. The study was approved by the institutional review board of The University of Texas M. D. Anderson Cancer Center, and University of Texas Health Science Center at Houston. Data collection occurred between September 2004 and February 2006.

Measures

Data collection tools included a patient-completed demographic data sheet that included age, sex, education level, income, religion, and ethnicity and an investigator-completed clinical checklist including diagnosis, tumor grade, tumor location, treatment type and history, concomitant medications, performance status, disease status, and tumor size based on imaging at the time of completion of the M. D. Anderson Symptom Inventory–Brain Tumor Module. Clinicians rated performance status using the Karnofsky performance score (KPS) as part of the routine clinical assessment.26 Treatment data were collected as the type of treatment (surgery, radiation, or chemotherapy), and type and history of treatment, including whether the patient was currently on treatment. Tumor size was determined using bidirectional measurement and calculation of the surface area on imaging performed the day the patient completed the questionnaire. Disease status was categorized as newly diagnosed (postsurgery only), stable disease, and progressive disease based on imaging findings.

The patients completed the M. D. Anderson Symptom Inventory–Brain Tumor Module, which included 22 symptoms and 6 interference items rated on an 11–point scale (0-10) to indicate the presence and severity of each individual item in the last 24 hours. Patients completed the M. D. Anderson Symptom Inventory–Brain Tumor Module at the time of clinical evaluation but before being told the results of imaging studies. Thirteen symptoms were part of the initial M. D. Anderson Symptom Inventory, a symptom-inventory instrument that was developed and validated in patients with other solid tumors. The M. D. Anderson Symptom Inventory includes symptoms common across cancer (pain, fatigue, nausea, disturbed sleep, distress, shortness of breath, trouble remembering, lack of appetite, drowsiness, dry mouth, sadness, vomiting, and numbness and tingling). These symptoms are considered the core items. Nine items comprising the brain tumor module (weakness on 1 side of the body, difficulty understanding, difficulty speaking, difficulty concentrating, change in vision, seizures, alteration in bowel pattern, irritability, and change in appearance) were added to the core items. The final 6 interference items allowed the patients to report how the symptoms interfered with daily life activities (general activity, work, mood, relations with others, ability to walk, and general activity). For each symptom and interference item, a 0 is ‘not present’ and a 10 is “as bad as you can imagine” in the last 24 hours. The M. D. Anderson Symptom Inventory–Brain Tumor Module has demonstrated content and discriminant validity and reliability (Cronbach α = .91) for use in symptom report in the primary brain tumor population.25

Statistical Analysis

All statistical analyses were conducted with SPSS Statistics, version 17 (SPSS Inc., Chicago, Ill). For the purposes of data analysis, tumors were classified as either low grade (grade 1 and 2) or high grade (grade 3 and 4). Performance status, reported using the KPS, was categorized as poor (KPS 30-70) or good (KPS 80 or higher). We categorized disease status as stable disease or active disease (included those patients who were newly diagnosed or with recurrence). On the basis of previous work examining pain and fatigue in patients with cancer, fatigue was categorized into low (0-4) or moderate/severe (5-10), as reported by the patient. Fatigue was dichotomized into these groups based on previous work with both pain and fatigue, which supported the clinical significance of patient report of moderate and severe fatigue. Studies have demonstrated that QOL and patient functional status for those with low fatigue severity levels were not different from patients without fatigue, but both moderate and severe levels of fatigue were associated with worse QOL and functional status.27, 28 In addition, recent reports indicate that once patients experience a moderate to severe level of fatigue, there is a nonlinear relationship to functional status, indicating that those with moderate fatigue severity may experience equal or greater impact on function than those with severe levels.29 Demographic and clinical data were initially analyzed with descriptive statistics. Comparisons of mean symptom scores by fatigue category were conducted using a t test for independent samples.

Univariate associations between self-reported fatigue and clinical and demographic factors were tested with a chi-square test. Fisher exact test was used for expected cell counts <5. Variables with P < .25 in the univariate analysis were retained for inclusion in logistic regression to determine those factors predictive of fatigue.30 An overall model was determined first using all of the subjects. On the basis of the results of the overall model, subsequent subgroup analysis models were developed by KPS category and sex. Odds ratios (ORs) and 95% confidence intervals (CIs) were adjusted for the other variables in the model. The level of significance for all logistic regression models was P < .05.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Patient Characteristics

Of the 214 patients approached, 7 patients refused participation, and 207 patients consented to participate in the study. Six patients did not fully complete the assessment form. Therefore, 201 patients fully participated in this study. More men (57%) than women (43%) participated in the study. Age ranged from 18 to 84 years, with a mean age of 45.8 years. Patients were primarily white (84%), married (72%), and made more than $50,000 per year (61%). Most patients had high-grade tumors (66%), had a good performance status (71%), had been treated with radiation therapy (57%), and were on anticonvulsants (67%). Clinical and demographic data are further presented in Tables 1 and 2.

Table 1. Demographic Characteristics of the Participants
CharacteristicNo.%
Sex  
 Men11557
 Women8643
Age, y  
 Range18-84 
 Mean45.8 
 Aged 18-45 52
 Aged 45-84 48
Years of education  
 Range7-17 
 Mean14.5 
Employment status  
 Employed part- or full-time7945
 Unemployed as a result of tumor4522
Marital status  
 Retired3115
 Married14472
 Divorced2010
 Separated31
 Single3015
 Widowed42
Ethnicity  
 White, non-Hispanic16683
 Hispanic126
 Black, non-Hispanic115
 Other126
Gross annual family income  
≥$50,00011861
$30,000-49,9993517
<$30,0004021
Table 2. Clinical Characteristics of the Participants
CharacteristicNo.%
  1. KPS indicates Karnofsky performance score.

Performance status  
 Good (KPS=90 or 100)14371
 Poor (KPS=80 or below)5829
Patient status  
 Outpatient15175
 Inpatient5025
Treatment history  
 Surgery19899
 Chemotherapy10251
 Radiation11557
Current treatment  
 Surgery only15477
 Chemotherapy only6030
 Radiation + chemotherapy63
 Radiation alone31.5
Tumor grade  
 110.5
 26833.8
 35728.4
 47537
Disease status  
 Newly diagnosed3015
 Stable disease11356
 Recurrent tumor5829
Tumor size  
 No measureable disease6733
 Measureable disease13457
 Range, cm0.48-5 
 Median/mean, cm1.143/0.815 
Concurrent medications  
 Corticosteroids8442
 Anticonvulsants13467
 Antidepressants3819
 Stimulants137
 Opioid analgesics4321

Prevalence and Severity of Fatigue

Fatigue of any severity level was reported in 153 (73%) of the patients enrolled. By using the analog scale of 0 to 10, the mean fatigue severity rating was 3.79, with a median of 3. Fatigue was the most severe symptom reported by this patient group. Eighty-one (40%) reported moderate to severe fatigue (5 or higher) in this sample, with 46 (23%) reporting severe fatigue (7 or higher). The prevalence and severity of fatigue for various clinical characteristics are presented in Table 3.

Table 3. Prevalence and Severity of Fatigue
VariableLow FatigueModerate/ Severe Fatigue (%)
  1. KPS indicates Karnofsky performance score.

Ethnicity  
 Minority1916 (46)
 Nonminority10165 (39)
Tumor grade  
 Low (1 and 2)3831 (45)
 High (3 and 4)8250 (38)
Sex  
 Women4541 (48)
 Men7540 (35)
Disease status  
 Newly diagnosed/recurrent4345 (51)
 Stable disease7736 (32)
Antidepressants  
 Yes1721 (55)
 No10360 (37)
Anticonvulsants  
 Yes7856 (42)
 No4225 (37)
Corticosteroids  
 Yes4341 (49)
 No7740 (34)
Opioids  
 Yes2023 (54)
 No10058 (37)
Stimulants  
 Yes76 (46)
 No11375 (40)
KPS  
 Poor618 (75)
 Good11463 (36)

Relationship of Fatigue to Other Symptoms

To test the relationship between the levels of fatigue and self–report of other symptoms, comparisons were made between those with low (0-4) and moderate to severe fatigue (5-10). Patients reporting moderate to severe fatigue showed a higher score for overall mean symptom severity (0.904 vs 3.00, P < .01), mean core symptom severity (1.01 vs 3.39, P < .01), mean brain tumor-specific symptom severity (0.84 vs 2.44, P < .01), and mean interference (1.49 vs 4.39, P < .01), compared with patients reporting low fatigue. In addition, the occurrence of moderate to severe fatigue was positively associated with moderate to severe reports of pain (P < .001), distress (P < .001), drowsiness (P < .001), and weakness (P < .002).

Predictors of Fatigue

As shown in Table 4 patient characteristics that significantly predicted the occurrence of moderate to severe fatigue included female sex (OR, 2.48; 95% CI, 1.32-4.65; P = .005), active disease (OR, 2.20; 95% CI, 1.18-4.10; P = .01), and poor performance status (OR, 5.73; 95% CI, 2.08-15.82; P = .001). Performance status was the strongest predictor of fatigue severity, with those patients with poor performance status being almost 6× as likely to report moderate to severe fatigue. Women were 2.5× as likely as men to indicate they had moderate to severe fatigue. Patients with active disease were also 2.2× as likely to report moderate to severe fatigue as those with stable disease on imaging. Because of the small number of patients with KPS of 70 or lower, we also performed a regression analysis including patients with a KPS of 80 in the poor performance group, arguing that a KPS of 80 does signify a significant decline in functional status, with patients requiring effort to conduct normal activities and having symptoms associated with the underlying disease. The results of this second analysis were similar in that performance status (OR, 2.98; 95% CI,1.51-5.88; P = .002), female sex (OR, 2.48; 95% CI, 1.31-4.67; P = .005), and having active disease (OR, 2.13; 95% CI, 1.14-3.97; P = .02) remained the strongest predictors of fatigue severity, confirming the demonstrated impact of performance status on fatigue severity and supporting our initial analysis. Patient age, treatment type, whether currently on treatment, treatment history, and tumor size were not associated with more severe fatigue.

Table 4. Predictors of Moderate/Severe Fatigue
VariableOdds Ratio (95% CI)P
  1. CI indicates confidence interval; KPS, Karnofsky performance score.

KPS5.73 (2.08-15.82).001
Sex2.48 (1.32-4.65).01
Disease status2.20 (1.18-4.10).01

On the basis of the results of this overall model, we then evaluated for factors associated with moderate to severe fatigue based on performance status category and sex. For the logistic regression model using only patients with poor KPS, no multivariate model solution could be estimated, as all of these patients were predicted to have moderate to severe fatigue, further underscoring the significant relationship between poor performance status and the occurrence of severe fatigue. However, the number of patients with a poor KPS score was small (n = 24), and this may also have impacted the ability to determine a valid model. Antidepressant use approached significance in the univariate analysis (P = .07), with all poor KPS patients on antidepressants reporting moderate to severe fatigue.

Subgroup analysis for patients with a good KPS score indicated that female sex (OR, 2.78; 95% CI, 1.45-5.34; P = .002) and having active disease (OR, 2.26; 95% CI, 1.17-4.36; P = .02) were associated with higher fatigue severity. Among patients with a good KPS score, women were almost 3× as likely to report moderate to severe fatigue, and those with active disease were more than twice as likely as those with stable disease to report fatigue.

For women, low-grade tumor diagnosis (P = .06), being on steroids (P = .03), and active disease status (P = .01) were associated with fatigue severity on univariate analysis. However, in the logistic regression model for women, the primary determinant of fatigue was disease status, with those with active disease being 3.3× as likely to report moderate to severe fatigue (OR, 3.33; 95% CI, 1.31-8.47; P = .01) (Table 5). There was a significant association between being on steroids and active disease status, with 73% of female patients with active disease being on corticosteroids, compared with only 21% of those with stable disease. This association may have resulted in corticosteroid use being dropped from the final model. For men, in addition to a poor KPS score (OR, 12.47; 95% CI, 3.51-44.37; P = .0001), the use of antidepressants (OR, 4.43; 95% CI, 1.37-14.31; P = .01) and opioids (OR, 3.46; 95% CI, 1.25-9.54; P = .02) was associated with moderate to severe fatigue (Table 6). Male patients with a poor KPS were more than 12× as likely to report moderate to severe fatigue.

Table 5. Predictors of Moderate/Severe Fatigue in Women
VariableOdds Ratio (95% CI)P
  1. CI indicates confidence interval.

Disease status3.33 (1.31-8.47).01
Table 6. Predictors of Moderate/Severe Fatigue in Men
VariableOdds Ratio (95% CI)P
  1. CI indicates confidence interval; KPS, Karnofsky performance score.

KPS12.47 (3.51-44.37).0001
Antidepressant use4.43 (1.37-14.31).01
Opioid use3.46 (1.25-9.54).02

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Cancer-related fatigue is defined as a “distressing, persistent, subjective sense of tiredness or exhaustion related to cancer or cancer treatment that is not proportional to recent activity and that interferes with usual functioning” (Mock et al,31 page 1). Cancer-related fatigue has been recognized as having multiple causes in patients with solid tumors, including the direct effect of the cancer and cancer treatment, comorbid conditions (such as anemia, infection, malnutrition, and cardiac disease), psychological factors (depression and anxiety), and concurrent symptoms (pain and sleep disturbances). Furthermore, the severity of fatigue varies among individuals. In this study, 73% of patients who were diagnosed with brain tumors reported having fatigue, with 40% reporting fatigue as moderate to severe.

Among the 22 reported symptoms, fatigue was the most prevalent symptom in this patient group, similar to reports in patients with other solid tumors. Patients reporting moderate to severe fatigue also showed significantly higher overall symptom severity and more interference with daily life activities than did patients reporting low fatigue. These findings support the prevalence and significance of fatigue in the primary brain tumor population as well.

The findings of this study support that there is an important relationship between fatigue and other concurrent symptoms such as pain, distress, drowsiness, and weakness in patients with brain tumors. In addition, worse overall symptom severity and the burden of symptoms on daily life activities such as mood, work, ability to walk, and interactions with others were demonstrated for those with more severe fatigue. Studies in patients with other solid tumors have also indicated that symptoms can occur concurrently and be multiplicative with other symptoms. It has been postulated that symptoms such as pain, fatigue, and insomnia may have a shared biologic mechanism, such as the sickness behavior seen in relation to cytokine administration or production.32, 33 Further studies are needed to evaluate the concordance of these symptoms over time and the relationship to shared biologic mechanisms in this patient population.

In patients with primary brain tumors, prior studies have identified that fatigue occurs commonly in patients undergoing radiation therapy and that chronic fatigue may occur in patients with low–grade tumors. In this study, patient performance status, sex, and having active disease were the strongest predictors of fatigue. Performance status was the single strongest predictor of fatigue severity and may represent the impact of neurologic impairment on the occurrence of fatigue in this unique cancer population. Overall, patients with a poor performance status were almost 6× as likely to report moderate to severe fatigue. When sex difference was considered, for men, poor KPS was associated with 12× the risk of moderate to severe fatigue. For women, KPS did not remain in the model, but this might have been the result of having only a small number of women with a poor KPS (n = 5). The strong association of fatigue and performance status has not been previously reported. The relationship among neurologic disability, depression, and fatigue has been reported in patients with other more chronic neurologic diseases, including multiple sclerosis,34, 35 muscular disease,36 and Parkinson disease.37 As noted, performance status, which is primarily associated with neurologic disability in the primary brain tumor patient population, was the strongest predictor of more severe fatigue in primary brain tumor patients overall. This may reflect a different etiology for fatigue in primary brain tumor patients as compared with those with other solid tumor malignancies, or alternatively a higher susceptibility to otherwise minor inciting factors in patients with neurologic disability.

Cancer-related fatigue has been shown to have a temporal relationship to treatment and variability during the course of treatment in other solid tumor malignancies.19 This has also been identified in primary brain tumor patients undergoing radiation therapy.9, 10 However, most studies have not identified a relation of increased fatigue with longer disease duration, stage, and recurrence.18 We identified an association with active disease status and the occurrence of more severe fatigue, with those with active disease being 2.2× as likely to report moderate to severe fatigue, and those with active disease and a good KPS 2.3× as likely. Why this association occurs is not known, but it may be related to underlying brain neurochemical changes that are associated with the disease and treatment, and related brain injury resulting in disability. In patients with systemic cancer, fatigue also has been shown to be associated with the production of proinflammatory cytokines, such as interleukin 6 and tumor necrosis factor-α, which are thought to trigger sickness behaviors through bidirectional communication with the brain.33-37 In patients with primary brain tumors, the disease and treatment may centrally impact the production of cytokines or other hormones such as melatonin, resulting in fatigue in patients with active disease.

As with other solid tumor malignancies, sex was also associated with the occurrence of moderate to severe fatigue. Women were 2.5× as likely as men to indicate that they had moderate to severe fatigue and 2.8× as likely when they had a good KPS. Additional predictors of fatigue for women included disease status, indicating the importance of active disease in predicting fatigue severity Corticosteroid use and having a low-grade tumor appeared significant on univariate analysis, but were not in the final model. The strong relationship between corticosteroid use and disease status may have resulted in steroid use not being included in the final model. Opioid and anticonvulsant use was also associated with report of fatigue for men, with those on opioids nearly 3.5× as likely to report moderate to severe fatigue and those on antidepressants over 4× as likely to experience moderate to severe fatigue. Currently, it is not known why the variability in fatigue severity occurs in patients in relation to sex.38 Studies in patients with other solid tumor malignancies have also been limited, primarily reporting on differences in prevalence of fatigue based on sex. The different factors found to be related to the severity of fatigue based on sex in this study support the multidimensional and multifactorial nature of fatigue, and further evaluation in a longitudinal study is warranted to validate these findings, explore potential biologic correlates, and development interventions to reduce the severity of fatigue in both women and men.

This study provides a comprehensive review of demographic, functional, and clinical factors related to the severity of fatigue in the primary brain tumor patient population, and is among the first studies to systematically evaluate fatigue in neuro-oncology. These findings may be limited by the cross-sectional sample, which results in the lack of baseline measurements and evaluation of change in fatigue severity over time, and the small number of patients with a poor KPS who were also undergoing radiation therapy. The comparative impact of the disease versus the impact of treatment and concomitant medications in the individual patient during the course of therapy cannot be adequately assessed in this cross-sectional sample.

Results from this study can assist the care provider in assessment and evaluation of fatigue in this patient population. This is the first report to identify the importance of sex, performance, and active disease status on the severity of fatigue in primary brain tumor patients. Additional methods to assess for the occurrence of fatigue in clinical care are warranted in these groups. Further studies are needed to explore the long-term trajectory of fatigue in this patient population and should include factors indentified in this report in the evaluation of fatigue severity. In addition, interventional trials to reduce fatigue severity should consider stratifying patients based on performance status, sex, disease status, and use of concomitant medications, such as opioids, corticosteroids, antidepressants, and anticonvulsants in research design and data analyses. Fatigue should not be considered a solitary symptom with a unidimensional cause, but a complex symptom related to the severity of other symptoms and having multidimensional etiologies. For example, improving functional status may be an effective way for reducing fatigue severity in patients with poor performance status, whereas modification of concurrent medications may be a more effective way for reducing fatigue severity in others. Furthermore, concurrent evaluation of potential biomarkers of fatigue will further advance the understanding of fatigue and the effectiveness of different interventions in reducing fatigue in this unique and vulnerable patient population.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

The original study of the development of the M. D. Anderson Symptom Inventory–Brain Tumor Module was supported by grants from the Oncology Nursing Society and Pfizer Pharmaceuticals.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES