The longitudinal relationship of hemoglobin, fatigue and quality of life in anemic cancer patients: results from five randomized clinical trials

D. Cella1,*, J. Kallich2, A. McDermott3 and X. Xu3

1 Northwestern University and Evanston Northwestern Healthcare, Evanston, IL; 2 Amgen, Inc., Thousand Oaks, CA; 3 Covance Health Economics and Outcomes Services, Inc., Gaithersburg, MD, USA

Received 21 October 2003; revised 10 February 2004; accepted 11 February 2004


    ABSTRACT
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Background:

Anemia is common in cancer and has been associated with fatigue and reduced health-related quality of life (HRQOL). We report the association between hemoglobin and fatigue and the impact of reducing fatigue on several domains of HRQOL.

Patients and methods:

These analyses were based on five randomized trials. Patients completed the Functional Assessment of Cancer Therapy (FACT) Anemia scales and numeric rating scales of Energy, Activity and Overall Health at baseline and after the 12-week treatment period. t-tests and linear regression models were used to evaluate associations. Analyses were stratified into three groups: solid tumor chemotherapy patients, lymphoproliferative malignancy chemotherapy patients and non-chemotherapy patients.

Results:

Adjusted mean differences (95% CI) in FACT Fatigue change scores between hemoglobin responders (≥2 g/dl increase) and non-responders were 3.0 (1.2, 4.7), 2.8 (0.6, 5.0) and 5.8 (2.2, 9.5) among the solid tumor, lymphoproliferative malignancy and non-chemotherapy groups, respectively. Significantly greater improvements (P <0.01) were observed in the FACT well-being scales for patients with meaningful improvement in fatigue (FACT Fatigue change score ≥3 points). After controlling for other factors, patients whose fatigue improved reported substantially greater improvements in energy, ability to perform usual activities and overall health (P <0.0001).

Conclusions:

Across five trials of cancer patients on and off chemotherapy, hemoglobin response was associated with meaningful improvements in fatigue, which, in turn, was associated with improved physical, functional, emotional and overall well-being.

Key words: anemia, fatigue, neoplasms, quality of life


    Introduction
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Anemia is common among patients with cancer, particularly among those receiving myelosuppressive chemotherapy. Depending upon the chemotherapy treatment, anemia rates can be as high as 100% [1]. The incidence of anemia also varies depending upon the type of underlying malignancy, the stage and duration of disease, the regimen and intensity of tumor therapy and possibly the occurrence of intercurrent infections or surgery [2].

Whether related to chemotherapy or other factors, anemia appears to have a significant adverse impact on the health-related quality of life (HRQOL) of people with cancer [36]. Cancer-related anemia has been associated with increased fatigue, poorer self-perceived overall quality of life and reduced ability to work productively [7]. In a large, household survey of cancer patients, those experiencing fatigue at least a few days per month during their chemotherapy reported a 45% reduction in their ability to accomplish usual activities due to fatigue [8]. In another observational study, 25% of cancer patients with hemoglobin values ≤12 g/dl reported not being able to work at all, compared with 8% of patients with hemoglobin values >12 g/dl [3].

However, whereas fatigue is often attributed to low hemoglobin levels, anemia is not the only cause of fatigue in cancer patients, and the precise relationship between hemoglobin level and fatigue is not well understood [9]. Therefore, despite much attention to the impact of anemia and anemia correction on fatigue and HRQOL, the longitudinal relationship among these factors remains unclear. Better understanding of this complex relationship over time may help oncology providers to prioritize anemia management in the larger context of cancer care.

In community-based, uncontrolled trials, increasing hemoglobin has appeared to reduce fatigue and improve overall HRQOL [10, 11]. This has also been shown in randomized controlled trials [1217]. The impact of hemoglobin on overall HRQOL is likely mediated by reductions in fatigue that in turn lead to improvements in physical functioning, resumption of everyday activities and improved mood. Despite its logic and supporting evidence, this hypothesis has not been formally evaluated with pre-specified testing.

The recent completion of five randomized controlled trials, all using similar eligibility criteria, questionnaires and assessment plan, provided an opportunity to evaluate the relationship between hemoglobin, self-reported fatigue and HRQOL over time in a large sample of anemic cancer patients. Specifically, we report the association between hemoglobin and fatigue and evaluate the impact of reduced fatigue on specific HRQOL outcomes observed over a 12-week period.


    Patients and methods
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
Patients and trials
These analyses were conducted on data from five randomized clinical trials of darbepoetin alfa (Aranesp®; Amgen Inc., Thousand Oaks, CA): two open-label, active-controlled trials among patients receiving chemotherapy for solid tumors [12, 18]; a multinational, double-blind, placebo-controlled trial among lung cancer patients receiving platinum-based chemotherapy [19]; a double-blind, placebo-controlled trial among patients with non-myeloid malignancy who were not receiving chemotherapy [20]; and a multinational double-blind, placebo-controlled trial among patients receiving chemotherapy for lymphoproliferative malignancy [21]. The two multinational trials were conducted in Australia, Canada and Europe. The remaining trials were conducted in the USA. The institutional review boards of the participating centers approved the protocols, and all patients gave written informed consent before any study-specific procedures were carried out.

Inclusion criteria were similar across trials. To be eligible, patients had to be ≥18 years of age and have the study-defined cancer type and chemotherapy status. All patients were anemic (hemoglobin ≤11 g/dl) primarily due to their cancer treatment or to the cancer itself. Patients with anemia due to iron, folate or B12 deficiency, hemolysis, bleeding, or active infection, and those with transferrin saturation <15% or serum ferritin concentration <10 ng/ml, were not eligible for the trials. Furthermore, patients receiving more than two red blood cell transfusions within 4 weeks prior to the study, any red blood cell transfusion within 2 weeks or treatment with epoetin alfa within 8 weeks also were not eligible.

In all of the trials, patients received study treatment for 12 weeks. Across the five trials, patients were randomly assigned to receive darbepoetin alfa, recombinant human erythropoietin (rHuEPO) or placebo. The analyses presented here are pooled across study treatment (darbepoetin alfa, rHuEPO or placebo).

We combined the solid tumor [12, 18] and lung cancer [19] trials into a single ‘solid tumor with chemotherapy’ group. Since the trials included various dosing regimens of active therapies and one trial included placebo, we are more likely to observe a variety of hemoglobin responses in the combined data. This variation enhances the ability to evaluate fatigue and other HRQOL responses across a wider range of hemoglobin change. We note that the direction of the association between hemoglobin and fatigue was consistent across all three solid tumor trials. The remaining two trials were analyzed separately to report associations in these different patient samples, namely the ‘lymphoproliferative malignancy with chemotherapy’ and ‘non-myeloid malignancy without chemotherapy’ groups. We also refer to the latter group as the ‘non-chemotherapy’ group, since it is the only group of patients included in these analyses that is not receiving concurrent chemotherapy. Finally, we report the findings from all five trials combined, referred to here as the ‘all study groups combined’ group.

In each trial, patients completed a HRQOL questionnaire at baseline, during interim study visits and at the end of the 12-week treatment period. These analyses were limited to patients in the trials who completed at least 4 weeks of study treatment and provided responses to the HRQOL questionnaire at baseline and at least once after 4 weeks of treatment. Baseline and last-reported HRQOL values were used to assess change over time. The last-reported value is referred to as the end of treatment period (EOTP).

Study variables
The HRQOL questionnaire included the FACT Anemia assessment and global ratings of energy, ability to do desired activities and overall health. The FACT Anemia assessment measures general HRQOL areas and fatigue- and anemia-specific concerns [3]. The general measures (FACT-G) includes four subscales (Physical, Social, Emotional and Functional Well-Being) and an Overall scale. The specific measures include the FACT Fatigue and FACT Anemia subscales. Finally, the FACT Anemia Total scale is a summary measure of all of these general and specific scales.

The FACT Fatigue subscale was used in these analyses as the primary measure of fatigue [3]. It includes 13 items, such as ‘I feel fatigued’ and ‘I feel weak all over’. Patients respond to each item on a scale from 0 (not at all) to 4 (very much) based on their experience of fatigue during the past 2 weeks. The scale score is computed by summing the item scores, after reversing those items that are worded in the negative direction. When there are missing item scores, the subscale score is computed by summing the non-missing item scores, multiplying by 13 (the total number of items in the scale) and dividing by the number of non-missing items. The latter rule applies only when at least half of the items (seven or more) are non-missing. FACT Fatigue subscale scores range from 0 to 52, where higher scores represent less fatigue.

The remaining FACT general and specific subscales are scored in the same manner as the FACT Fatigue subscale described above. The FACT-G Overall scale score is based on the sum of the four general well-being subscale scores. The FACT Anemia subscale includes the 13 FACT Fatigue subscale items plus an additional seven anemia-related items. The FACT Anemia Total scale score is computed as the sum of the FACT-G Overall scale score and the FACT Anemia subscale score.

The HRQOL questionnaire also included three global ratings of energy, ability to do desired activities and overall health. Patients evaluated each global rating on an 11-point numeric rating scale (NRS) ranging from 0 (worst possible) to 10 (perfect health). In these analyses, each NRS was transformed to a 0 to 100 scale. Use of the NRS has been shown to be more accurate than visual analog scales, which offer a large number of response options and may confuse respondents [22]. All HRQOL scales are scored such that higher scores represent positive HRQOL outcomes (such as, better physical well-being).

Hemoglobin (g/dl) was measured weekly during the clinical trials. We defined hemoglobin response as an improvement of at least 2 g/dl from baseline to the EOTP.

The following variables were analyzed as potential covariates when evaluating the relationships between hemoglobin and fatigue: age, gender, primary tumor type (for the solid tumor with chemotherapy and non-myeloid malignancy without chemotherapy groups), malignancy type (either lymphoma or myeloma for the lymphoproliferative malignancy with chemotherapy group), baseline Eastern Cooperative Oncology Group (ECOG) performance status (0, 1 or 2+), prior history of chemotherapy (for the non-myeloid group), the occurrence of any hospitalizations during the study, whether or not a red blood cell transfusion was received within 4 weeks of the EOTP and disease progression reported by the investigator at the EOTP. Tumor type was categorized as breast, lung, gastrointestinal, gynecologic, genitourinary, and other in the solid tumor group. An additional category, lymphoid malignancy, was included for the non-myeloid malignancy group. Disease progression was rated as a complete or partial response, stable disease, or progressive disease. These same variables were evaluated as potential covariates of the relationship between fatigue and other HRQOL outcomes, with the exception of red blood cell transfusion status at the EOTP, which was not hypothesized to directly affect this relationship.

Statistical methods
To describe the study groups, we computed means, standard deviations (SDs) and ranges of continuous variables and frequency distributions of categorical variables. To evaluate the association between hemoglobin and fatigue, we compared mean changes in FACT Fatigue subscale scores between hemoglobin responders (hemoglobin improvement of at least 2 g/dl) and non-responders using t-tests. To further quantify the magnitude of the association, we computed effect sizes. The effect size is the mean change in the FACT Fatigue subscale score observed among those with a hemoglobin response divided by the SD of the score at baseline [23]. Effect sizes of 0.2, 0.5 and 0.8 are considered small, moderate and large, respectively [24]. All analyses were exploratory, and no adjustments were made for multiple comparisons.

We also compared mean changes in the general FACT subscales, the FACT-G Overall scale, and the NRS Energy, Activity and Overall Health scores by level of fatigue improvement using t-tests and effect sizes. Clinically meaningful improvement in fatigue was defined as a FACT Fatigue subscale score improvement of ≥3 points [25].

To explore potential covariates of these associations, we analyzed linear regression models. For the relationship between hemoglobin and fatigue in each study group, we modeled change in FACT Fatigue subscale scores as a function of hemoglobin response, age, gender, baseline FACT Fatigue subscale score, tumor or malignancy type, baseline ECOG rating, history of chemotherapy (only in the non-myeloid without chemotherapy group), hospitalization during the study, red blood cell transfusion within 4 weeks of the EOTP, and EOTP disease progression. For the ‘all study groups combined’ model, we excluded those variables that are not consistent across the study groups (tumor or malignancy type and history of chemotherapy). Similarly, for the relationship between fatigue and other HRQOL outcomes, we modeled change in each HRQOL outcome as a function of clinically meaningful improvement in fatigue, age, gender, baseline HRQOL scale score, tumor or malignancy type, baseline ECOG rating, history of chemotherapy (only in the non-chemotherapy group), hospitalization during the study and EOTP disease progression. Parameter estimates were evaluated using t-tests. We reported the parameter estimates and associated 95% confidence intervals (CIs) and the estimated variation explained by each model (R2). We note that change in the FACT Fatigue subscale and other HRQOL outcomes was relatively normally distributed, and that an analysis of regression diagnostics did not reveal any major violations of the underlying statistical assumptions of the models.


    Results
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
These analyses are based on clinical trial patients who received at least 4 weeks of study treatment and who provided baseline and at least one follow-up HRQOL assessment post-week 4. For the solid tumor with chemotherapy group, these analyses include 687 (80%) of 858 randomized trial patients; for the lymphoproliferative malignancy with chemotherapy group, these analyses include 306 (89%) of 344 trial patients; and finally, for the non-myeloid without chemotherapy group, these analyses include 159 (85%) of 188 trial patients. In the groups receiving concurrent chemotherapy, patients who were excluded from these analyses reported poorer baseline FACT Fatigue subscale scores relative to those who were included. The differences in mean baseline FACT Fatigue subscale scores were 5.2 and 6.6 points lower among the excluded patients in the solid tumor and lymphoproliferative malignancy groups, respectively. Patients excluded from the solid tumor group analyses also tended to be older (65.3 versus 61.3 years).

The demographic and clinical characteristics of the patients included in these analyses are summarized in Tables 1 and 2. At baseline, mean (SD) hemoglobin values were 10 (1) g/dl and 9.6 (1.2) g/dl in the solid tumor and lymphoproliferative malignancy groups who were receiving concurrent chemotherapy. Mean (SD) hemoglobin was 9.9 (1) g/dl in the non-chemotherapy group. Mean (SD) FACT Fatigue subscale scores were 27.2 (11.8), 30.9 (10.8) and 27.2 (12.8), in the solid tumor, lymphoproliferative malignancy and non-chemotherapy groups, respectively. These baseline FACT Fatigue subscale scores are 1.5–2 SDs below those observed in the general population [26].


View this table:
[in this window]
[in a new window]
 
Table 1. Baseline characteristics of the study groups
 

View this table:
[in this window]
[in a new window]
 
Table 2. Study period and EOTP clinical indicators of the study groups
 
Mean changes in FACT Fatigue subscale scores were significantly associated with hemoglobin response in each of the study groups (Figure 1). Patients with a hemoglobin improvement of at least 2 g/dl reported significantly greater increases in FACT Fatigue subscale scores relative to those who did not achieve this level of hemoglobin response. In the solid tumor and lymphoproliferative malignancy with chemotherapy groups, mean changes in FACT Fatigue subscale scores were 3.8 (P = 0.0001) and 3.6 (P = 0.0030) points greater among hemoglobin responders. The magnitude of these differences is clinically meaningful, and associated with effect sizes of 0.43 and 0.39, respectively. In the non-chemotherapy group, the observed difference was substantially larger (7.8; P <0.0001), and associated with an effect size of 0.85.



View larger version (26K):
[in this window]
[in a new window]
 
Figure 1. Mean change in FACT Fatigue subscale scores by hemoglobin respondera status. Error bars indicate 95% confidence intervals.

 
After controlling for clinical and demographic characteristics, hemoglobin response remained significantly associated with improved FACT Fatigue scores. Adjusted mean differences (95% CI) in FACT Fatigue change scores between hemoglobin responders and non-responders were 3.0 (1.2, 4.7), 2.8 (0.6, 5.0) and 5.8 (2.2, 9.5) in the solid tumor, lymphoproliferative malignancy and non-chemotherapy groups, respectively. Across all study groups, the adjusted mean difference (95% CI) was 3.1 (1.8, 4.4). In the non-chemotherapy group, hemoglobin response and the covariates explained half of the variation in FACT Fatigue change scores (R2 = 0.51). In the groups where concurrent chemotherapy was administered, hemoglobin response and the covariates explained ~30% of the variance (solid tumor: R2 = 0.31; lymphoproliferative malignancy: R2 = 0.27). With the exception of the baseline FACT Fatigue score, few of the covariates were significantly associated with change in FACT Fatigue subscale scores, and significance varied across study groups. In the all study groups combined regression analysis, the following covariates were associated with less improvement in FACT Fatigue subscale scores: increasing age (P = 0.0044), higher baseline FACT Fatigue subscale scores (P <0.0001), hospitalization during the study (P = 0.0204), and the presence of a red blood cell transfusion (P = 0.0002) or progressive disease (P <0.0001) at the EOTP.

Mean changes in the FACT general well-being subscales and in the FACT-G Overall scale were significantly associated with fatigue improvement. Table 3 reports mean change scores for the FACT general scales by fatigue improvement status for all study groups combined. Findings were similar for each study group, although the Social subscale did not reach statistical significance in the lymphoproliferative and non-chemotherapy groups. Since the FACT scales have different potential ranges, it is helpful to examine the effect sizes, since they are expressed in SDs. Effect sizes ranging from 0.53 to 0.77 were observed among patients with improved fatigue for the FACT Physical and Functional subscales and for the FACT-G Overall scale across the study groups (Table 3). Even after controlling for variation in clinical and demographic characteristics, patients experiencing clinically meaningful improvement in fatigue reported significantly greater improvements in all areas of HRQOL except the FACT Social subscale (Figure 2).


View this table:
[in this window]
[in a new window]
 
Table 3. Mean change in HRQOL scale scores by level of improvement of FACT Fatigue subscale scores (all study groups combined)
 


View larger version (18K):
[in this window]
[in a new window]
 
Figure 2. Adjusteda mean differencesb in health-related quality of life (HRQOL) change scores associated with fatigue improvement (all study groups combined). Error bars indicate 95% confidence intervals.

 
Similarly, mean changes in the NRS Energy, Activity and Overall Health scores were significantly higher (P <0.0001) among those with meaningful fatigue improvement relative to those who did not achieve this level of improved fatigue (Table 3). The NRS scores range from 0 to 100 and, as such, can be expressed as percentages. After controlling for clinical and demographic characteristics (Figure 2), patients whose fatigue improved reported 17% and 19% greater improvements on NRS Energy and Activity scores, respectively.


    Discussion
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
This analysis, combining data across five randomized clinical trials of cancer patients on and off chemotherapy to create a large sample size, consistently demonstrated a significant and positive relationship between hemoglobin rise and reduction in fatigue. Patients with a hemoglobin improvement of at least 2 g/dl, meeting our pre-set criterion for hemoglobin response, reported significantly greater increases in FACT Fatigue subscale scores relative to those who did not achieve this level of hemoglobin response. In the solid tumor and lymphoproliferative malignancy with chemotherapy groups, mean changes in FACT Fatigue subscale scores were 3.8 and 3.6 points greater among hemoglobin responders. While the magnitude of these differences is clinically meaningful [25], they were smaller in magnitude when compared with the patients with non-myeloid malignancy that were not receiving chemotherapy, where the observed difference in change scores was about twice as large (7.8 points). One might speculate that this greater magnitude of benefit among this third group of patients is because they were not experiencing fatigue due to treatment. A better understanding of the way in which anemia affects patients with cancer may facilitate informed decisions regarding the appropriate management of anemia.

The association between hemoglobin and fatigue could not be explained by other clinical and demographic characteristics in our model, which included potential confounding variables such as baseline fatigue, age, presence of transfusion and progressing disease. While the multivariate model attenuated the magnitude of the association, hemoglobin response remained significantly associated with fatigue improvement after controlling for other potential explanatory variables. Previous reports of the association between hemoglobin and fatigue have not generally controlled for other factors [1016].

The second purpose of these analyses was to evaluate the impact of a change in fatigue upon other HRQOL dimensions such as physical, mental and social well-being. To do this, we used a 3-point improvement as an indication that an individual patient had a fatigue improvement, and then evaluated HRQOL dimension improvement of the fatigue-improved group compared with the others. Clinically meaningful improvements in fatigue were significantly associated with moderate improvements in physical, emotional and functional well-being, large improvements in energy and activity levels and perceived overall health. This relationship was observed consistently across the three study groups, and remained present after controlling for potentially confounding patient characteristics. These findings are consistent with previous reports from the solid tumor studies, where clinically meaningful improvements in fatigue were associated with improvements in emotional well-being as measured by the Brief Symptom Inventory Anxiety and Depression subscales [14], and with self-reported gains in productive time and reductions in required caregiving time [26].

This study has some limitations. First, the analyses are based on data obtained from anemic cancer patients participating in clinical trials and may not represent the broader anemic cancer patient population. We note, however, that the association between hemoglobin and fatigue observed in these analyses is comparable to that observed in community-based studies [10, 11]. Since the community-based studies were limited to cancer patients receiving chemotherapy and did not evaluate the association between fatigue and other HRQOL outcomes, further confirmation of these findings is warranted. Secondly, it may be difficult to interpret the findings for all study groups combined given the heterogeneity of cancers and concurrent chemotherapy status that it represents. As such, we presented study group-specific findings when there were notable differences, namely with respect to the magnitude of the association between hemoglobin response and reduced fatigue. The association between reduced fatigue and other HRQOL outcomes was similar across study groups, and pooling the data allowed us to examine this relationship in a broad clinical trial sample of anemic cancer patients. Finally, we note that patients who were excluded from these analyses, because either they did not complete 4 weeks in the trials or they did not provide follow-up HRQOL assessments, had poorer baseline fatigue than those who were included. Patients excluded from the solid tumor group analyses also tended to be older (65.3 versus 61.3 years). The possibility that increasing age may confer some added risk of non-response (or higher fatigue) should be investigated in future studies. While the longitudinal data to evaluate the impact of these exclusions are not available, we note that the excluded patients in these analyses represented only 17% of the total number of patients enrolled across the five trials.


    Acknowledgements
 
This research was funded by Amgen Inc., Thousand Oaks, CA, USA.


    FOOTNOTES
 
* Correspondence to: Dr D. Cella, Evanston Northwestern Healthcare, 1001 University Place – Suite 100, Evanston, IL 60201, USA. Tel: +1-847-570-7370; Fax: +1-847-570-8033; E-mail: d-cella{at}northwestern.edu Back


    REFERENCES
 Top
 ABSTRACT
 Introduction
 Patients and methods
 Results
 Discussion
 REFERENCES
 
1. Groopman JE, Itri LM. Chemotherapy-induced anemia in adults: incidence and treatment. J Natl Cancer Inst 1999; 91: 1616–1634.[Abstract/Free Full Text]

2. Mercadante S, Gebbia V, Marrazzo A et al. Anaemia in cancer: pathophysiology and treatment. Cancer Treat Rev 2000; 26: 303–311.[CrossRef][ISI][Medline]

3. Cella D. The Functional Assessment of Cancer Therapy-Anemia (FACT-An) Scale: a new tool for the assessment of outcomes in cancer anemia and fatigue. Semin Hematol 1997; 34: 13–19.[ISI][Medline]

4. Demetri GD, Kris M, Wade J et al. Quality-of-life benefit in chemotherapy patients treated with epoetin alfa is independent of disease response or tumor type: results from a prospective community oncology study. J Clin Oncol 1998; 16: 3412–3425.[Abstract]

5. Glaspy J, Bukowski R, Steinberg D et al. Impact of therapy with epoetin alfa on clinical outcomes in patients with nonmyeloid malignancies during cancer chemotherapy in community oncology practice. J Clin Oncol 1997; 15: 1218–1234.[Abstract]

6. Yellen SB, Cella DF, Webster KA et al. Measuring fatigue and other anemia-related symptoms with the Functional Assessment of Cancer Therapy Measurement System. J Pain Symptom Manage 1997; 13: 63–74.[CrossRef][ISI][Medline]

7. Cella D. Factors influencing quality of life in cancer patients: Anemia and fatigue. Semin Oncol 1998; 25 (3 Suppl 7): 43–46.[ISI][Medline]

8. Curt G, Breitbart W, Cella D et al. Impact of cancer-related fatigue on the lives of patients: new findings from the Fatigue Coalition. Oncologist 2000; 5: 353–360.[Abstract/Free Full Text]

9. Stone P, Richards M, A’Hern R et al. A study to investigate the prevalence, severity and correlates of fatigue among patients with cancer in comparison with a control group of volunteers without cancer. Ann Oncol 2000; 11: 561–567.[Abstract]

10. Crawford J, Cella D, Cleeland C et al. Relationship between changes in hemoglobin level and quality of life during chemotherapy in anemic cancer patients receiving epoetin alfa therapy. Cancer 2002; 95: 888–895.[CrossRef][ISI][Medline]

11. Gabrilove J, Cleeland C, Livingston R et al. Clinical evaluation of once-weekly dosing of epoetin alfa in chemotherapy patients: Improvements in hemoglobin and quality of life are similar to three-times-weekly dosing. J Clin Oncol 2001; 19: 2875–2882.[Abstract/Free Full Text]

12. Glaspy J, Jadeja J, Justice G et al. Darbepoetin alfa given every 1 or 2 weeks alleviates anaemia associated with cancer chemotherapy. Br J Cancer 2002; 87: 268–276.[CrossRef][ISI][Medline]

13. Johansson JE, Wersall P, Brandberg Y et al. Efficacy of epoetin beta on hemoglobin, quality of life, and transfusion needs in patients with anemia due to hormone-refractory prostate cancer. Scand J Urol Nephrol 2001; 35: 288–294.[CrossRef][ISI][Medline]

14. Kallich J, Tchekmedyian NS, Damiano A et al. Psychological outcomes associated with anemia-related fatigue in cancer patients. Oncology 2002; 16 (9 Suppl): 117–124.[Medline]

15. Littlewood T, Bajetta E, Nortier JWR et al. Effects of epoetin alfa on hematologic parameters and quality of life in cancer patients receiving nonplatinum chemotherapy: results of a randomized, double-blind, placebo-controlled trial. J Clin Oncol 2001; 19: 2865–2874.[Abstract/Free Full Text]

16. Osterborg A, Brandberg Y, Molostova V et al. Randomized, double-blind, placebo-controlled trial of recombinant human erythropoietin, epoetin beta, in hematologic malignancies. J Clin Oncol 2002; 20: 2486–2494.[Abstract/Free Full Text]

17. Cella D, Dobrez D, Glaspy J. Control of cancer-related anemia with erythropoietic agents: a review of evidence for improved quality of life and clinical outcomes. Ann Onc 2003; 14: 511–519.[CrossRef][ISI]

18. Glaspy J, Jadeja J, Justice G et al. A randomized, active-control, pilot trial of front-loaded dosing regimens of darbepoetin-alfa for the treatment of patients with anemia during chemotherapy for malignant disease. Cancer 2003; 97: 1312–1320.[CrossRef][ISI][Medline]

19. Vansteenkiste J, Pirker R, Massuri B et al. Double-blind, placebo-controlled, randomized phase III trial of darbepoetin alfa in lung cancer patients receiving chemotherapy. J Natl Cancer Inst 2002; 94: 1211–1220.[Abstract/Free Full Text]

20. Smith RE, Tchekmedyian NS, Chan D et al. A dose- and schedule-finding study of darbepoetin alfa for the treatment of chronic anaemia of cancer. Br J Cancer 2003; 88: 1851–1858.[CrossRef][ISI][Medline]

21. Hedenus M, Adriansson M, San Miguel J et al. Efficacy and safety of darbepoetin alfa in anaemic patients with lymphoproliferative malignancies: a randomized, double-blind, placebo-controlled study. Br J Haematol 2003; 122: 394–403.[CrossRef][ISI][Medline]

22. Downie WW, Leatham PA, Rhind VM et al. Studies with pain rating scales. Ann Rheum Dis 1978; 37: 378–381.[Abstract]

23. Deyo RA, Diehr P, Patrick DL. Reproducibility and responsiveness of health status measures – statistics and strategies for evaluation. Control Clin Trials 1991; 12: 142S–158S.[Medline]

24. Cohen J. Statistical Power Analysis for the Behavioral Sciences, 2nd edition. Hillsdale: Lawrence Erlbaum Associates 1988.

25. Cella D, Eton D, Lai J et al. Combining anchor and distribution based methods to derive minimal clinically important differences on the Functional Assessment of Cancer Therapy (FACT) Anemia and Fatigue Scales. J Pain Symptom Manage 2002; 24: 547–561.[CrossRef][ISI][Medline]

26. Cella D, Lai JS, Chang CH et al. Fatigue in cancer patients compared with fatigue in the general United States population. Cancer 2002; 94: 528–538.[CrossRef][ISI][Medline]

27. Berndt E, Kallich J, Xu X et al. Reductions in anemia and fatigue are associated with improvements in productivity. Blood 2002; 100: 876a (Abstr).