Weekly Epoetin Alfa Maintains Hemoglobin, Improves Quality of Life, and Reduces Transfusion in Breast Cancer Patients Receiving Chemotherapy
http://www.100md.com
▲還散笫雖悝◎
the Durham Regional Cancer Centre, Oshawa
Northeastern Ontario Regional Cancer Centre, Sudbury
Ortho Biotech Canada, Toronto, Ontario
Htel-Dieu de Levis, Levis
Htel-Dieu, Quebec City, Quebec, Canada
ABSTRACT
PURPOSE: Epoetin alfa administered at 40,000 U once weekly (qw) to anemic cancer patients receiving chemotherapy increases hemoglobin levels, improves quality of life (QOL), and reduces transfusions. The benefit of epoetin alfa in maintaining hemoglobin levels in cancer patients with hemoglobin less than 12 g/dL has not been evaluated.
METHODS: Breast cancer patients (N = 354) receiving chemotherapy were randomly assigned in 1:1 ratio to epoetin alfa (40,000 U qw) or standard of care (SOC). QOL was assessed at baseline and week 12. Hemoglobin responses, transfusion requirements, and prognostic factors for responses were measured.
RESULTS: At week 12, Functional Assessment of Cancer Therapy每Anemia (FACT每An; mean, 2.16 ㊣ 12.84 for epoetin alfa v 每4.43 ㊣ 13.42 for SOC) and FACT每An fatigue (mean, 1.85 ㊣ 10.52 for epoetin alfa v 每3.55 ㊣ 11.14 for SOC) change scores were significantly higher in the epoetin alfa group (P < .0001). Hemoglobin responses defined as mean hemoglobin 12 g/dL or a 2 g/dL increase compared with baseline were significantly higher in the epoetin alfa group versus SOC: 52.0% v 5.1% and 65.7% v 6.3%, respectively (P < .0001 for both comparisons). Percentage transfused was significantly lower in the epoetin alfa group compared with SOC (8.6% v 22.9%). More than 90% of patients did not require a dose increase and 28.7% had a dose reduction.
CONCLUSION: Epoetin alfa administered at 40,000 U qw is effective in improving QOL, maintaining hemoglobin level, and reducing transfusion requirements in breast cancer patients. The high effectiveness observed could be attributed in part to early treatment with epoetin alfa.
INTRODUCTION
There has been a declining death rate in breast cancer with a statistically significant annual percent change of 每2.4% in the United States.1 The decrease in death rate can be attributed to factors including earlier diagnosis, changing patterns of treatment, and new knowledge about the role of optimal therapy. Anemia is a problem frequently associated with the disease process in malignancy, and is exacerbated by myelotoxic chemotherapy. Anemia produces symptoms of fatigue, weakness, concentration deficits, respiratory distress, and chest pain, all of which can negatively affect well-being and overall quality of life (QOL). Patient preference in relationship to chemotherapy adverse effects has shifted such that fatigue, which is the most frequent manifestation of anemia, and other QOL concerns now predominate over historic concerns about nausea and vomiting.2 Although fatigue may be multifactorial, anemia is an important contributing factor in the breast cancer patient, and one that can be treated.
Anemia that develops during treatment with myelotoxic agents has traditionally been treated with the use of transfusions. However, given continuing concerns about transfusion, the developing blood shortage, and the increasing rate of transfusion used, alternatives are necessary. Transfusions are normally administered on the basis of clinical need, although they are not common at hemoglobin levels more than 9 g/dL. Epoetin alfa has been demonstrated consistently to be effective in treating anemia and reducing the need for transfusion.3-8 In studies with QOL parameters, epoetin alfa has been associated with improved energy level, improved functional capacity, and improved overall QOL.6,7
Previous studies have examined the impact of epoetin alfa when treatment is initiated after hemoglobin levels have decreased to 10.5 g/dL. The impact of earlier treatment with epoetin alfa on hemoglobin response, transfusion requirements, and QOL improvement has not been evaluated fully. The value of maintaining hemoglobin levels of more than 12 g/dL has been recognized in elderly oncology patients,9 and maintenance of hemoglobin more than 12 g/dL in these patients, as opposed to correction of anemia, is now part of the evidence-based guidelines in the United States.10 Research has demonstrated that patients with hemoglobin values more than 12 g/dL experience significantly reduced levels of fatigue, better physical and functional well-being, and higher overall QOL.11-13
A study by Cleeland et al14 showed that the largest improvements in QOL occur when hemoglobin values increase from 11 to 12 g/dL. This work suggests that a hemoglobin level below 12 g/dL may be an appropriate intervention point. The current study was undertaken to examine the impact on QOL with the introduction of treatment with epoetin alfa at a hemoglobin level of 12 g/dL with the purpose of maintaining hemoglobin levels during chemotherapy.
METHODS
Study Population and Design
This study was a phase III, randomized, controlled, multicenter, open-label trial in 354 breast cancer patients in Canada receiving or scheduled to receive myelotoxic chemotherapy for a planned minimum of 12 weeks of therapy, and was undertaken with local independent ethics committee approval. Female patients older than age 18 years with a confirmed diagnosis of breast cancer were considered for inclusion in the study. Study participation was open to both adjuvant breast cancer patients and those with metastatic disease. Patients were to be in good general health with a life expectancy of 6 months or longer. Patients with anemia resulting from causes other than cancer were excluded from study participation, as were patients with clinically significant comorbidities that were not attributable to the underlying breast cancer. Patients receiving dose-intensification regimens with bone marrow or stem-cell transplantation were ineligible for participation, as were patients participating in other investigational drug studies.
Patients who met the study entry criteria with hemoglobin of 15 g/dL when initiating chemotherapy provided informed consent and entered the screening phase with hemoglobin monitored weekly. Those patients with a hemoglobin value of 12 g/dL at the time of initial screening provided consent and were randomly assigned immediately. Those who entered the screening phase of the study with hemoglobin 15 g/dL were randomly assigned once hemoglobin decreased to 12 g/dL. Patients randomly assigned to the study drug started treatment with epoetin alfa at the beginning of the subsequent chemotherapy cycle. Epoetin alfa treatment was initiated at a dose of 40,000 U subcutaneously once a week for 16 weeks, or 4 weeks after the completion of the last chemotherapy cycle, whichever was longer (maximum of 28 weeks). The initial dose was maintained through the first on-study chemotherapy cycle of 4 weeks or through two cycles for 3-week chemotherapy cycles. If at the end of week 4 or 6, the patient's hemoglobin had decreased by more than 2 g/dL, the epoetin alfa dose was increased to 60,000 U weekly.
If hemoglobin increased above 14 g/dL, epoetin alfa was withheld until the hemoglobin decreased to 12 g/dL, then was resumed at 75% of the original dose. In addition, if hemoglobin was observed to be increasing by more than 2 g/dL per month, epoetin alfa dose was reduced by 25% to maintain the rate of hemoglobin increase to less than 2 g/dL/mo.
The transfusion of RBCs was allowed during the study, at the discretion of the investigator; every effort was made not to transfuse patients until hemoglobin had decreased below 8 g/dL. An oral daily dose of 200 mg of elemental iron was recommended to maintain iron availability and iron stores, as indicated by a transferrin saturation of more than 20%.
Evaluation of Efficacy and Safety
The primary efficacy end point of the study was QOL after 12 weeks of treatment, as measured by the Functional Assessment of Cancer Therapy每Anemia (FACT每An) anemia subscale scores, compared with baseline scores. Changes in the FACT每An subscale between baseline and 12 weeks were analyzed for all patients who completed 12 weeks of treatment. The FACT每An anemia subscale was administered on three occasions: baseline before study random assignment, after 12 weeks, and at study completion or early termination. The FACT每An anemia subscale is the 21-item anemia subscale, containing the 13 question FACT每fatigue scale.
In addition, the Linear Analog Scale Assessment, also known as the Cancer Linear Analog Scale (CLAS) was used to assess QOL. The scale consists of three linear analog scales, which measure level of energy, ability to do daily activities, and overall QOL as it relates to cancer symptoms. The CLAS was measured at baseline, at 12 weeks, and at study termination. Care was taken to ensure that the instruments were administered in the same order (FACT每An and CLAS) on each occasion. Each patient completed the QOL questionnaires before random assignment. Patients were unaware of any hematology or other clinical results before completing the questionnaires at scheduled visits.
Safety and tolerability of epoetin alfa were assessed through normal guidelines for reporting all adverse events (AEs) and serious AEs during the study period. Safety analysis includes all patients for whom safety information is available.
Statistical Analysis
The estimated sample size necessary to achieve the primary objective was calculated to be 350 breast cancer patients, who were receiving myelotoxic chemotherapy, or for whom chemotherapy was planned. Two study populations were designated for purposes of statistical analyses: total randomized, including all patients randomly assigned; and the efficacy (EFF) population, consisting of all randomly assigned patients with at least one postbaseline efficacy end point (FACT每An, hemoglobin levels, or transfusion incidences). Baseline comparability of the treatment groups was assessed using descriptive univariate statistics.
For patients terminating the study early, a common method, termed last observation carried forward (LOCF), was used to ensure vigor and minimize any bias due to early termination. The observation at visit L for the response variable X is defined as XLOCF = XL if the variable is measured at visit L, and XLOCF = Xk if the variable is not measured at visit L and the last observation before visit L occurred at visit k.
Analysis of the primary variable, the FACT每An, was carried out using an analysis of variance model with treatment group as the main factor and the baseline FACT每An anemia subscale as a covariate. Secondary analyses were carried out using simple t tests to compare the two treatment groups.
The hemoglobin level measured within 7 days of study randomization (at time of initiation of epoetin alfa treatment) was defined as the baseline hemoglobin level for analysis. Our primary definition of a responder for study results included patients with calculated average hemoglobin 12 g/dL from weeks 4 to 12. To allow comparisons with previous study data, an additional definition of responder was used whereby responders were defined as patients with a 2 g/dL hemoglobin increase without the benefit of transfusion within the previous 28 days.
Efficacy analyses were carried out using the EFF population (n = 350). Comparisons of the proportion of hemoglobin responders and percent transfusions were carried out using the Fisher's exact test. Safety analyses were carried out using all patients.
A comparison of proportions of patients reporting AEs was based on the Fisher's exact test. Comparisons of the treatment groups at baseline were based on simple t tests or Fisher's exact test depending on the type of variable. No adjustment for multiple comparisons was used to report P values.
RESULTS
A total of 354 patients were enrolled at 22 sites in Saskatchewan, Ontario, Quebec, and Atlantic Canada. Four patients (one in the epoetin alfa group and three in the SOC group) from the randomly assigned patients were excluded from EFF because they discontinued the study before completing one of the efficacy end points (QOL, transfusion, or hemoglobin response). Given that the EFF population of 350 (175 in the epoetin alfa group and 175 in the SOC group) is so close to the total recruited population of 354, it was the population used for all evaluations unless otherwise stated. Early withdrawal was defined as patients who did not complete the study (minimum of 16 weeks on treatment or 4 weeks after last chemotherapy to a maximum of 28 weeks) and was similar between the two groups, with 18.2% in the epoetin alfa group and 17.5% in the SOC group. Three patients in the epoetin alfa arm received only one dose of study drug because of early withdrawal from the study. The main reasons for withdrawal were patient withdrawal of consent or early discontinuation of chemotherapy.
Demographics and Baseline Characteristics
Demographics and other baseline characteristics of patients in the EFF population are listed in Table 1. Most women who entered onto the trial had adjuvant disease. Baseline characteristics were well balanced between the two groups. Patients with metastatic disease appeared to have lower baseline hemoglobin at entry and also had significantly higher level of serum ferritin, probably reflective of the inflammatory nature of the disease, specifically as seen in those with metastatic disease. Table 2 lists the major chemotherapeutic agents used in the study, of which the most commonly used regimen was the combination of cyclophosphamide, epirubicin, and fluorouracil (55.7 and 55.6% in the epoetin alfa and SOC group, respectively), followed by cyclophosphamide, doxorubicin, and fluorouracil (22.2% and 21.2% in epoetin alfa and SOC group, respectively). Interestingly, more cycles of chemotherapy were delivered in the epoetin alfa group than in the SOC group (mean, 5.0 v 4.6; P = .058), with a trend toward significance.
Hemoglobin Responses
Figure 1 illustrates the change in hemoglobin over time between the two groups. In the group treated with epoetin alfa, hemoglobin increased from a baseline of approximately 11.0 to more than 12.0 g/dL, whereas in the SOC group, hemoglobin decreased from 11.3 to approximately 10.5 g/dL. Because the objective was to maintain hemoglobin above 12 g/dL, dose adjustment was carried out to ensure that the hemoglobin level was kept between 12 and 14 g/dL. Hemoglobin response, prospectively defined as the mean hemoglobin from weeks 4 to 12 of 12 g/dL, is listed for the two groups in Table 3. Average hemoglobin was calculated as normalized area under the time-concentration curve from days 28 to 84. Fifty-two percent of the group treated with epoetin alfa met this rigorous criterion of maintaining a mean hemoglobin of 12 g/dL or higher during the study period and only 5.1% of the SOC group met the same standard, with the epoetin alfa group showing greater than a 10-fold higher percentage of responders. The percentage of responders is much higher for patients entering the study with baseline hemoglobin levels 11 versus less than 11 g/dL. To allow comparison with previous trials, which defined percentage of responders as patients with a 2 g/dL hemoglobin increase without transfusion in the previous 4 weeks, this alternative definition was retrospectively applied to the current study population. The percentage of responders according to this commonly used definition in this trial was 65.7% for epoetin alfa and 6.3% for SOC, which compared favorably with previous studies. The responder rate was similar between adjuvant and metastatic groups.
Transfusion Requirements
In addition to being able to maintain higher hemoglobin levels during the study period, the group treated with epoetin alfa also exhibited a lower transfusion rate, as listed in Table 4. An absolute reduction in transfusion of 14.3% and a relative reduction of 62.4% were accomplished with epoetin alfa treatment. Patients who entered the trial with baseline hemoglobin 11 g/dL had fewer transfusions in both arms. Both frequency of transfusion and total units transfused (49 units) were lower in the active epoetin alfa group versus the SOC group (148 units). Hemoglobin level at the time of transfusion was similar between the two groups.
QOL
The primary end point of this study was comparing the change in QOL between groups as measured by the FACT每An anemia subscale from baseline to week 12 between the group treated with epoetin alfa and the group receiving SOC without epoetin treatment. Of the 350 EFF patients, 338 patients (168 in the epoetin alfa group and 170 in the SOC group) had both baseline and at least one set of postbaseline QOL data collected and were therefore evaluated for QOL. As shown in Fig 2 the mean change scores in both FACT每An anemia subscale and FACT每An fatigue subscale after 12 weeks of epoetin alfa treatment were highly significant (P < .0001) when compared with the SOC group. Even in the FACT每An nonfatigue subscale, which mainly evaluates clinical signs and symptoms of anemia, mean change scores were significantly better in the epoetin alfa arm. The differences in the two groups were marked by improvement in the epoetin alfa group and significant deterioration in the SOC group. Change between end-of-study (anywhere between weeks 16 and 28) to baseline was also significantly better for the epoetin alfa group for both of the FACT每An subscales (data not shown)
Figure 3 shows results from CLAS assessment, and all three QOL variables: energy level, daily activities, and overall QOL were significantly better in the epoetin alfa group compared with the SOC group. In addition, and as observed with the FACT每An instrument, between-group differences for all three CLAS change scores from baseline to study end were also statistically significant, favoring epoetin alfa (data not shown).
Predictors of Responses
An exploratory logistic regression analysis was used to determine which baseline parameters (treatment; age; race; disease stage; menopausal status; adjuvant or metastatic; next chemotherapy cycle; time from first diagnosis; and baseline hemoglobin, neutrophils, platelets, and serum ferritin) were significant predictors of hemoglobin responses and incidence of transfusion. Table 5 shows the results for hemoglobin response. As expected, baseline hemoglobin was shown to be a strong predictor; higher hemoglobin at entry was the most robust predictor among all baseline parameters evaluated. Interestingly, older age was a significant predictor for hemoglobin response. For incidence of transfusion (Table 6), low baseline hemoglobin again was a strong predictor for subsequent transfusion. Lower platelets (an indication of bone marrow insufficiency) and higher serum ferritin levels (which could indicate an elevated inflammatory state) were predictive of higher transfusion requirements during chemotherapy.
Safety
Epoetin alfa was well tolerated by patients with breast cancer. The overall incidence of AEs was similar between the two groups (Table 7). The most common AEs were granulocytopenia, nausea, vomiting, and hot flushes, which were probably associated with chemotherapy treatment. Granulocytopenia, nausea, and anorexia were significantly higher in the epoetin alfa arm, possibly due to an observed higher intensity of chemotherapy delivered to the patients. Although the current study was not designed to examine dose-intensity, observed differences led to the calculation of relative dose-intensity, which showed differences between groups (P = .0383; data not shown). Incidences of fever and, as expected, anemia, were higher for the SOC group. Pure red-cell aplasia has not been observed in oncology patients, and there were no observed instances in the current study. Nineteen patients in the epoetin alfa group and 14 patients in the SOC group experienced thrombovascular events (Table 7). Of the 19 patients in the epoetin alfa group who experienced thrombovascular events, investigators attributed seven (3.9%) to epoetin alfa therapy. Two-year survival data are being collected and at the time of this writing, 24 patients had died in the epoetin alfa arm (seven adjuvant, 17 metastatic) and 27 had died in the SOC group (five adjuvant, 22 metastatic).
DISCUSSION
The therapeutic benefit of epoetin alfa in anemic cancer patients has been well documented with several large community and double-blind placebo controlled, randomized trials.3,5-8,15,16 Results from those studies consistently demonstrate that epoetin alfa significantly increased hemoglobin levels and decreased the incidence of blood transfusions. In studies in which QOL was assessed, strong association was observed between increase in hemoglobin and improvement in QOL.3 Although most of these studies focused on anemia correction (patients were recruited to the studies when hemoglobin decreased below 10.5 to 11 g/dL), the clinical benefits of anemia prevention have not been fully explored. In the study by Littlewood et al,3 a small group of patients entered the study with baseline hemoglobin between 10.5 and 12 g/dL and this subpopulation was compared with patients with entry hemoglobin less than 10.5 g/dL according to the prespecified hemoglobin strata. Both hemoglobin responses and transfusion reduction were numerically better in the higher hemoglobin stratum group. However, because of the small number of patients in the higher baseline hemoglobin stratum, statistical significance was not achieved.
When data from two large community-based studies were pooled to evaluate the correlation between hemoglobin change and QOL improvement, it was shown that a strong correlation exists but the relationship is not linear. An incremental analysis using regression method was used to explore the longitudinal relationship between incremental changes in hemoglobin and that of QOL scores.17 It was found that among all incremental increases in hemoglobin of 1 g/dL, the greatest gain in QOL occurred at the increase from 11 to 12 g/dL.
The objective of our study was to maintain a hemoglobin level at 12 g/dL throughout the chemotherapy cycle by maximizing the impact of epoetin alfa treatment. Hemoglobin response was predefined as the mean hemoglobin level being maintained at 12 g/dL over the evaluation period of 4 to 12 weeks after random assignment. This is a more rigorous definition of hemoglobin responses than that used in previous studies because it incorporates all hemoglobin values during the defined period in the calculation of response. Using this response definition, more than 50% percent of the patients (52%) in the epoetin alfa arm maintained a mean hemoglobin of more than 12 g/dL during the evaluation period, whereas only 5.1% of patients in the SOC group had such a response, with the epoetin alfa group showing more than a 10-fold increase in percentage of responders. Absolute hemoglobin difference between the two groups was slightly lower than 2 g/dL. This is because the mean hemoglobin in the SOC group never decreased below 10.5 g/dL and investigators were instructed to maintain hemoglobin just above 12 g/dL without going higher in the epoetin alfa arm. This is reflected by the higher frequency of dose reduction as compared with dose increase in the epoetin alfa arm (dose was stable for 62.1% of patients, dose was decreased or held for 28.7% of patients, and dose was increased for 9.2% of patients). Using a definition for hemoglobin response more consistent with that in previous trials ( 2 g/dL increase without benefit of transfusion), the response rate was 65.7% for the epoetin alfa group versus 6.5% for SOC, and this compares favorably with previous studies.3,8,18 The high entry hemoglobin (mean, > 11 g/dL) probably played an important role in the robust response compared with that in previous trials (baseline hemoglobin, < 10 g/dL), given that results similar to those in this study were observed in the higher hemoglobin stratum of the study by Littlewood et al.3
The incidence of transfusion was 22.5% in the SOC group, which was high considering that most patients were receiving adjuvant chemotherapy and no patient was receiving dose-dense or dose-intense regimens. The high percentage of patients receiving a CEF (cyclophosphamide, epirubicin, and fluorouracil) regimen, with its documented hematologic toxicity, might be one of the reasons for the relatively high incidence of transfusion. Treatment with epoetin alfa significantly reduced the transfusion rate to 8%, representing a 14.6% absolute reduction and 64% relative reduction〞the highest reduction of transfusion rate reported in the literature in a randomized setting with erythropoiesis stimulating factors. The mean transfusion threshold was slightly lower for the epoetin alfa group (7.73 ㊣ 1.83 g/dL) versus the SOC group (8.11 ㊣ 0.87 g/dL; P = .143); however, the range was similar between the two groups, suggesting that bias was not a major factor for the transfusion differences between groups. Despite the higher transfusion rate, the SOC group still had significantly fewer hemoglobin responders than the epoetin alfa group. The history of transfusion before study entry was minimal because of the high percentage of adjuvant patients. There was a slightly higher frequency of transfusion in the epoetin alfa arm than the SOC group (five v one, respectively). Improvement in transfusion reduction in this study was similar to that previously observed with the high hemoglobin stratum of the study by Littlewood et al3 (22% in placebo v 8% in epoetin alfa group), suggesting the benefit of higher hemoglobin levels at study entry.
The primary end point of this study was change of FACT每An anemia subscale over time between the two groups. Patients in the epoetin alfa arm showed improvement in QOL over baseline values, whereas QOL in the SOC group deteriorated during the course of chemotherapy. Changes in FACT每An anemia and FACT每An fatigue subscores were highly significant (P < .0001) between the two groups and even the FACT每An nonfatigue scores significantly favored the epoetin alfa group. This highly significant improvement in QOL further substantiates previous observations that the largest gain in QOL is achieved when hemoglobin increases from 11 to 12 g/dL (range, 11 to 13 g/dL). Furthermore, observations from the FACT每An instrument were well supported by the CLAS results. As has been shown in several previous studies, change in QOL correlated extremely well with change in hemoglobin (P value ranges from < .0001 to .0019 for six scales analyzed).
To avoid the burden of placebo injections in patients, an open-label design was selected to evaluate the benefit of early anemia management and hemoglobin maintenance on QOL. The validity of this design was predicated on the availability of objective measurements from other trials such as hemoglobin responses and transfusion reductions. It was stated in the protocol a priori that the correlation between hemoglobin changes and improvement in QOL would be analyzed and compared with previous studies to validate the study design. The hemoglobin responses in the epoetin alfa arm of this study (65.8%) compared well with the double-blind study by Littlewood et al3 (66.4% in the solid tumor stratum of the epoetin alfa arm using same definition for response). When correlation between change in hemoglobin levels and QOL were compared between the two trials, the correlation coefficient for the double-blind study3 for hemoglobin change and FACT每An fatigue subscale was 0.2879 (n = 273; P = .0002), whereas in our study, the correlation of the same variables was 0.2893 (n = 338; P < .0001). The comparable hemoglobin responses and the nearly identical correlation coefficients between change in hemoglobin and QOL score suggest that this open-label study mimics that of the double-blind study and did not adversely influence the outcome of the study. One problem with epoetin alfa studies is the difficulty of using a truly double-blind format for both physicians and patients because of continuous measurement of hemoglobin values. To minimize bias in this study hemoglobin levels were measured after patients had completed their QOL questionnaires.
Epoetin alfa treatment has been linked consistently to improvement in QOL in cancer patients. In this study, a highly statistically significant improvement in QOL as measured by FACT每An and CLAS was observed in patients treated with epoetin alfa. Cella et al19 conducted an Internet-based survey administering the FACT每An to a nationally representative sample of the United States population to better characterize the clinical significance of QOL changes associated with epoetin alfa in reference to the general population. Results from that study show that the FACT每An discriminated between respondents with no history of prespecified illnesses and those with a history of chronic conditions, such as anemia and cancer. Using FACT每An data from the normative sample, the authors concluded that treatment with epoetin alfa reported in the study by Littlewood et al3 leads to clinically relevant and important improvements in QOL.
The pattern of QOL difference between the epoetin alfa and SOC groups observed in the current study was strikingly similar to that observed in the study by Littlewood et al.3 Moreover, a substantial proportion of the QOL deficit observed between the respective study groups at baseline and the sample without any history of prespecified illnesses from the study by Cella et al19 was overcome through administration of epoetin alfa therapy, suggesting that the same findings of QOL clinical significance for the study by Littlewood et al3 should be applied to the current study.
Previous attempts have been made to identify predictors of hemoglobin responses. Serum erythropoietin levels, serum ferritin level, and change in hemoglobin or reticulocyte count within the first 4 weeks after initiation of therapy have all been shown to have some association with response to epoetin therapy; however, to date, the association demonstrated has been weak and not readily reproducible.20 In this study, using baseline variables as potential predictors of hemoglobin response, logistic regression suggests that baseline hemoglobin was a robust and significant predictor of subsequent hemoglobin response and transfusion reduction.
Epoetin alfa was well tolerated in the study. Several AEs reported more often in the epoetin alfa arm may be related to the higher intensity of chemotherapy delivered to patients in that arm (5.0 cycles) versus SOC (4.6 cycles; P = .058). In previous studies granulocytopenia was partly attributable to epoetin alfa treatment favoring erythropoiesis and leading to blunted white cell responses. However, this explanation appears unlikely in this study because of the higher incidences of other chemotherapy-associated AEs such as nausea and anorexia observed in the epoetin alfa arm.
The study results presented here demonstrate that early initiation of treatment with epoetin alfa is effective in and maintaining hemoglobin level, reducing transfusion requirements, and improving QOL in breast cancer patients throughout treatment with myelotoxic agents. With clinical recognition of the importance of achieving optimal chemotherapy dosing and the increasing use of more aggressive myelotoxic agents in the treatment of both adjuvant and metastatic disease, the treatment gains associated with epoetin alfa, which are reported here, are of great value and of significant clinical importance.
Appendix
List of Investigative Sites: Dr P. Goss, Princess Margaret Hospital; Dr R. Goel, Ottawa Regional Cancer Centre; Dr J. Chang, Lakeridge Health Oshawa; Dr M. Blackstein, Mount Sinai Hospital; Dr R. Haq, St Michael＊s Hospital; Dr J. Curtis/C. Hamm, Windsor Regional Cancer Centre; Dr B. Findlay, Hotel-Dieu St Catharines; Dr L. Yelle, CHUM Campus Notre-Dame; Dr G. Cantin, CHA ST-Sacrement; Dr D. Vergidis, Northwestern Ontario Regional Cancer Centre; Dr K. Laing, Dr H. Bliss Murphy Cancer Centre; Dr F. Couture, Htel-Dieu de Quebec and Htel-Dieu de Levis (2 separate sites); Dr J. Dufresne, Centre University Sante l＊Estrie-Fleur; Dr M. Jancewicz, Allan Blair Cancer Centre; Dr S. Young, Northeastern Ontario Regional Cancer Centre; Dr S. Fox, Hopital Charles LeMoyne; Dr S. Berger, Humber River Regional Hospital; Dr L. Zibdawi, Southlake Regional Health Centre; Dr J. Gapski, Trillium Health Centre; Dr R. Pahil, Cape Breton Cancer Centre; Dr R. Levesque, Quinte Health Care Corp, Belleville General Hospital.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Owns stock (not including shares held through a public mutual fund): Catherine Y. Lau, Ortho Biotech Canada Inc; Kara-Lee McWatters, Ortho Biotech Canada Inc. Acted as a consultant within the last 2 years: Jose Chang, Ortho Biotech Canada Inc. Received more than $2,000 a year from a company for either of the last 2 years: Catherine Y. Lau, Ortho Biotech Canada, Inc; Kara-Lee McWatters, Ortho Biotech Canada Inc; Jose Chang, Ortho Biotech Canada Inc.
NOTES
Research funding provided by Ortho Biotech Canada Inc.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
SEER Web site. Trends in SEER Incidence and US death rates by primary cancer site: 1992-2000. http://seer.cancer.gov/csr/1975_2001
Carelle N, Piotto E, Bellanger A, et al: Changing patient perceptions of the side effects of cancer chemotherapy. Cancer 95:155-163, 2002
Littlewood TJ, Bajetta E, Nortier W: Effects of epoetin alfa on hematologic parameters and QOL in cancer patients receiving nonplatinum chemotherapy: Results of a randomized, double blind, placebo-controlled trial. J Clin Oncol 19:2865-2874, 2001
Ludwig H, Fritz E: Anemia in cancer patients. Semin Oncol 25:2-6, 1998
Abels RL: Recombinant human erythropoietin in the treatment of the anaemia of cancer. Acta Haematol 87:4-11, 1992
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 15:1218-1234, 1997
Demetri GD, Gabrilove JL, Blasi MV, et al: Benefits of epoetin alfa in anemic breast cancer patients receiving chemotherapy. Clin Breast Cancer 3:45-51, 2002
Gabrilove JL, Cleeland CS, Livingston RB, et al: Clinical evaluation of once weekly dosing of epoetin alfa in chemotherapy patients: Improvements in hemoglobin and QOL similar to three times weekly dosing. J Clin Oncol 19:2875-2882, 2001
Balducci L, Extermann M: Cancer and aging: An evolving panorama. Hematol Oncol Clin North Am 14:1-16, 2000
Repetto L, Carreca I, Maraninchi D, et al: Use of growth factors in the elderly patient cancer: A report from the Second International Society for Geriatric Oncology (SIOG) 2001 meeting. Crit Rev Oncol Hematol 45:123-128, 2003
Cella D: The Functional Assessment of Cancer Therapy每Anemia (FACT-AN) scale: A new tool for the assessment of outcomes in cancer anemia and fatigue. Semin Haematol 34:13-19, 1997
Yellen SB, Cella DF, Webster K, et al: Measuring fatigue, and other anemia related symptoms with the Functional Assessment of Cancer Therapy (FACT) measurement system. J Pain Symptom Manage 13:63-74, 1997
Demetri GD: Anaemia and its functional consequences in cancer patients: Current challenges in management and prospects for improving therapy. Br J Cancer 84:31-37, 2001 (suppl)
Cleeland CS, Demetri GD, Glaspy J, et al: Identifying hemoglobin level for optimal quality of life: Results of an incremental analysis. Proc Am Soc Clin Oncol 18:574a, 1999 (abstr 2215)
Osterborg A, Brandberg Y, Molostova V, et al: Randomized, double-blind, placebo-controlled trial of recombinant human erythropoietin, epoetin beta, in hematological malignancies. J Clin Oncol 20:2486-2494, 2002
Dammacco F, Castoldi G, Rodjer S: Efficacy of epoetin alfa in the treatment of anemia of multiple myeloma. Br J Haematol 113:172-179, 2001
Crawford J, Cella D, Cleeland CS, et al: Relationship between changes in hemoglobin level and quality of life during chemotherapy in anemic cancer patients receiving epoetin alfa therapy. Cancer 95:888-895, 2002
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 tumour type: Results from a prospective community oncology study. J Clin Oncol 16:3412-3425, 1998
Cella D, Zagari MJ, Vandoros C, et al: Epoetin alfa treatment results in clinically significant improvements in quality of life in anemic cancer patients when referenced to the general population. J Clin Oncol 21:366-373, 2003
Littlewood TJ, Zagari M, Pallister C, et al: Baseline and early treatment factors are not clinically useful for predicting individual response to erythropoietin in anemic cancer patients. Oncologist 8:99-107, 2003(Jose Chang, Felix Couture)
Northeastern Ontario Regional Cancer Centre, Sudbury
Ortho Biotech Canada, Toronto, Ontario
Htel-Dieu de Levis, Levis
Htel-Dieu, Quebec City, Quebec, Canada
ABSTRACT
PURPOSE: Epoetin alfa administered at 40,000 U once weekly (qw) to anemic cancer patients receiving chemotherapy increases hemoglobin levels, improves quality of life (QOL), and reduces transfusions. The benefit of epoetin alfa in maintaining hemoglobin levels in cancer patients with hemoglobin less than 12 g/dL has not been evaluated.
METHODS: Breast cancer patients (N = 354) receiving chemotherapy were randomly assigned in 1:1 ratio to epoetin alfa (40,000 U qw) or standard of care (SOC). QOL was assessed at baseline and week 12. Hemoglobin responses, transfusion requirements, and prognostic factors for responses were measured.
RESULTS: At week 12, Functional Assessment of Cancer Therapy每Anemia (FACT每An; mean, 2.16 ㊣ 12.84 for epoetin alfa v 每4.43 ㊣ 13.42 for SOC) and FACT每An fatigue (mean, 1.85 ㊣ 10.52 for epoetin alfa v 每3.55 ㊣ 11.14 for SOC) change scores were significantly higher in the epoetin alfa group (P < .0001). Hemoglobin responses defined as mean hemoglobin 12 g/dL or a 2 g/dL increase compared with baseline were significantly higher in the epoetin alfa group versus SOC: 52.0% v 5.1% and 65.7% v 6.3%, respectively (P < .0001 for both comparisons). Percentage transfused was significantly lower in the epoetin alfa group compared with SOC (8.6% v 22.9%). More than 90% of patients did not require a dose increase and 28.7% had a dose reduction.
CONCLUSION: Epoetin alfa administered at 40,000 U qw is effective in improving QOL, maintaining hemoglobin level, and reducing transfusion requirements in breast cancer patients. The high effectiveness observed could be attributed in part to early treatment with epoetin alfa.
INTRODUCTION
There has been a declining death rate in breast cancer with a statistically significant annual percent change of 每2.4% in the United States.1 The decrease in death rate can be attributed to factors including earlier diagnosis, changing patterns of treatment, and new knowledge about the role of optimal therapy. Anemia is a problem frequently associated with the disease process in malignancy, and is exacerbated by myelotoxic chemotherapy. Anemia produces symptoms of fatigue, weakness, concentration deficits, respiratory distress, and chest pain, all of which can negatively affect well-being and overall quality of life (QOL). Patient preference in relationship to chemotherapy adverse effects has shifted such that fatigue, which is the most frequent manifestation of anemia, and other QOL concerns now predominate over historic concerns about nausea and vomiting.2 Although fatigue may be multifactorial, anemia is an important contributing factor in the breast cancer patient, and one that can be treated.
Anemia that develops during treatment with myelotoxic agents has traditionally been treated with the use of transfusions. However, given continuing concerns about transfusion, the developing blood shortage, and the increasing rate of transfusion used, alternatives are necessary. Transfusions are normally administered on the basis of clinical need, although they are not common at hemoglobin levels more than 9 g/dL. Epoetin alfa has been demonstrated consistently to be effective in treating anemia and reducing the need for transfusion.3-8 In studies with QOL parameters, epoetin alfa has been associated with improved energy level, improved functional capacity, and improved overall QOL.6,7
Previous studies have examined the impact of epoetin alfa when treatment is initiated after hemoglobin levels have decreased to 10.5 g/dL. The impact of earlier treatment with epoetin alfa on hemoglobin response, transfusion requirements, and QOL improvement has not been evaluated fully. The value of maintaining hemoglobin levels of more than 12 g/dL has been recognized in elderly oncology patients,9 and maintenance of hemoglobin more than 12 g/dL in these patients, as opposed to correction of anemia, is now part of the evidence-based guidelines in the United States.10 Research has demonstrated that patients with hemoglobin values more than 12 g/dL experience significantly reduced levels of fatigue, better physical and functional well-being, and higher overall QOL.11-13
A study by Cleeland et al14 showed that the largest improvements in QOL occur when hemoglobin values increase from 11 to 12 g/dL. This work suggests that a hemoglobin level below 12 g/dL may be an appropriate intervention point. The current study was undertaken to examine the impact on QOL with the introduction of treatment with epoetin alfa at a hemoglobin level of 12 g/dL with the purpose of maintaining hemoglobin levels during chemotherapy.
METHODS
Study Population and Design
This study was a phase III, randomized, controlled, multicenter, open-label trial in 354 breast cancer patients in Canada receiving or scheduled to receive myelotoxic chemotherapy for a planned minimum of 12 weeks of therapy, and was undertaken with local independent ethics committee approval. Female patients older than age 18 years with a confirmed diagnosis of breast cancer were considered for inclusion in the study. Study participation was open to both adjuvant breast cancer patients and those with metastatic disease. Patients were to be in good general health with a life expectancy of 6 months or longer. Patients with anemia resulting from causes other than cancer were excluded from study participation, as were patients with clinically significant comorbidities that were not attributable to the underlying breast cancer. Patients receiving dose-intensification regimens with bone marrow or stem-cell transplantation were ineligible for participation, as were patients participating in other investigational drug studies.
Patients who met the study entry criteria with hemoglobin of 15 g/dL when initiating chemotherapy provided informed consent and entered the screening phase with hemoglobin monitored weekly. Those patients with a hemoglobin value of 12 g/dL at the time of initial screening provided consent and were randomly assigned immediately. Those who entered the screening phase of the study with hemoglobin 15 g/dL were randomly assigned once hemoglobin decreased to 12 g/dL. Patients randomly assigned to the study drug started treatment with epoetin alfa at the beginning of the subsequent chemotherapy cycle. Epoetin alfa treatment was initiated at a dose of 40,000 U subcutaneously once a week for 16 weeks, or 4 weeks after the completion of the last chemotherapy cycle, whichever was longer (maximum of 28 weeks). The initial dose was maintained through the first on-study chemotherapy cycle of 4 weeks or through two cycles for 3-week chemotherapy cycles. If at the end of week 4 or 6, the patient's hemoglobin had decreased by more than 2 g/dL, the epoetin alfa dose was increased to 60,000 U weekly.
If hemoglobin increased above 14 g/dL, epoetin alfa was withheld until the hemoglobin decreased to 12 g/dL, then was resumed at 75% of the original dose. In addition, if hemoglobin was observed to be increasing by more than 2 g/dL per month, epoetin alfa dose was reduced by 25% to maintain the rate of hemoglobin increase to less than 2 g/dL/mo.
The transfusion of RBCs was allowed during the study, at the discretion of the investigator; every effort was made not to transfuse patients until hemoglobin had decreased below 8 g/dL. An oral daily dose of 200 mg of elemental iron was recommended to maintain iron availability and iron stores, as indicated by a transferrin saturation of more than 20%.
Evaluation of Efficacy and Safety
The primary efficacy end point of the study was QOL after 12 weeks of treatment, as measured by the Functional Assessment of Cancer Therapy每Anemia (FACT每An) anemia subscale scores, compared with baseline scores. Changes in the FACT每An subscale between baseline and 12 weeks were analyzed for all patients who completed 12 weeks of treatment. The FACT每An anemia subscale was administered on three occasions: baseline before study random assignment, after 12 weeks, and at study completion or early termination. The FACT每An anemia subscale is the 21-item anemia subscale, containing the 13 question FACT每fatigue scale.
In addition, the Linear Analog Scale Assessment, also known as the Cancer Linear Analog Scale (CLAS) was used to assess QOL. The scale consists of three linear analog scales, which measure level of energy, ability to do daily activities, and overall QOL as it relates to cancer symptoms. The CLAS was measured at baseline, at 12 weeks, and at study termination. Care was taken to ensure that the instruments were administered in the same order (FACT每An and CLAS) on each occasion. Each patient completed the QOL questionnaires before random assignment. Patients were unaware of any hematology or other clinical results before completing the questionnaires at scheduled visits.
Safety and tolerability of epoetin alfa were assessed through normal guidelines for reporting all adverse events (AEs) and serious AEs during the study period. Safety analysis includes all patients for whom safety information is available.
Statistical Analysis
The estimated sample size necessary to achieve the primary objective was calculated to be 350 breast cancer patients, who were receiving myelotoxic chemotherapy, or for whom chemotherapy was planned. Two study populations were designated for purposes of statistical analyses: total randomized, including all patients randomly assigned; and the efficacy (EFF) population, consisting of all randomly assigned patients with at least one postbaseline efficacy end point (FACT每An, hemoglobin levels, or transfusion incidences). Baseline comparability of the treatment groups was assessed using descriptive univariate statistics.
For patients terminating the study early, a common method, termed last observation carried forward (LOCF), was used to ensure vigor and minimize any bias due to early termination. The observation at visit L for the response variable X is defined as XLOCF = XL if the variable is measured at visit L, and XLOCF = Xk if the variable is not measured at visit L and the last observation before visit L occurred at visit k.
Analysis of the primary variable, the FACT每An, was carried out using an analysis of variance model with treatment group as the main factor and the baseline FACT每An anemia subscale as a covariate. Secondary analyses were carried out using simple t tests to compare the two treatment groups.
The hemoglobin level measured within 7 days of study randomization (at time of initiation of epoetin alfa treatment) was defined as the baseline hemoglobin level for analysis. Our primary definition of a responder for study results included patients with calculated average hemoglobin 12 g/dL from weeks 4 to 12. To allow comparisons with previous study data, an additional definition of responder was used whereby responders were defined as patients with a 2 g/dL hemoglobin increase without the benefit of transfusion within the previous 28 days.
Efficacy analyses were carried out using the EFF population (n = 350). Comparisons of the proportion of hemoglobin responders and percent transfusions were carried out using the Fisher's exact test. Safety analyses were carried out using all patients.
A comparison of proportions of patients reporting AEs was based on the Fisher's exact test. Comparisons of the treatment groups at baseline were based on simple t tests or Fisher's exact test depending on the type of variable. No adjustment for multiple comparisons was used to report P values.
RESULTS
A total of 354 patients were enrolled at 22 sites in Saskatchewan, Ontario, Quebec, and Atlantic Canada. Four patients (one in the epoetin alfa group and three in the SOC group) from the randomly assigned patients were excluded from EFF because they discontinued the study before completing one of the efficacy end points (QOL, transfusion, or hemoglobin response). Given that the EFF population of 350 (175 in the epoetin alfa group and 175 in the SOC group) is so close to the total recruited population of 354, it was the population used for all evaluations unless otherwise stated. Early withdrawal was defined as patients who did not complete the study (minimum of 16 weeks on treatment or 4 weeks after last chemotherapy to a maximum of 28 weeks) and was similar between the two groups, with 18.2% in the epoetin alfa group and 17.5% in the SOC group. Three patients in the epoetin alfa arm received only one dose of study drug because of early withdrawal from the study. The main reasons for withdrawal were patient withdrawal of consent or early discontinuation of chemotherapy.
Demographics and Baseline Characteristics
Demographics and other baseline characteristics of patients in the EFF population are listed in Table 1. Most women who entered onto the trial had adjuvant disease. Baseline characteristics were well balanced between the two groups. Patients with metastatic disease appeared to have lower baseline hemoglobin at entry and also had significantly higher level of serum ferritin, probably reflective of the inflammatory nature of the disease, specifically as seen in those with metastatic disease. Table 2 lists the major chemotherapeutic agents used in the study, of which the most commonly used regimen was the combination of cyclophosphamide, epirubicin, and fluorouracil (55.7 and 55.6% in the epoetin alfa and SOC group, respectively), followed by cyclophosphamide, doxorubicin, and fluorouracil (22.2% and 21.2% in epoetin alfa and SOC group, respectively). Interestingly, more cycles of chemotherapy were delivered in the epoetin alfa group than in the SOC group (mean, 5.0 v 4.6; P = .058), with a trend toward significance.
Hemoglobin Responses
Figure 1 illustrates the change in hemoglobin over time between the two groups. In the group treated with epoetin alfa, hemoglobin increased from a baseline of approximately 11.0 to more than 12.0 g/dL, whereas in the SOC group, hemoglobin decreased from 11.3 to approximately 10.5 g/dL. Because the objective was to maintain hemoglobin above 12 g/dL, dose adjustment was carried out to ensure that the hemoglobin level was kept between 12 and 14 g/dL. Hemoglobin response, prospectively defined as the mean hemoglobin from weeks 4 to 12 of 12 g/dL, is listed for the two groups in Table 3. Average hemoglobin was calculated as normalized area under the time-concentration curve from days 28 to 84. Fifty-two percent of the group treated with epoetin alfa met this rigorous criterion of maintaining a mean hemoglobin of 12 g/dL or higher during the study period and only 5.1% of the SOC group met the same standard, with the epoetin alfa group showing greater than a 10-fold higher percentage of responders. The percentage of responders is much higher for patients entering the study with baseline hemoglobin levels 11 versus less than 11 g/dL. To allow comparison with previous trials, which defined percentage of responders as patients with a 2 g/dL hemoglobin increase without transfusion in the previous 4 weeks, this alternative definition was retrospectively applied to the current study population. The percentage of responders according to this commonly used definition in this trial was 65.7% for epoetin alfa and 6.3% for SOC, which compared favorably with previous studies. The responder rate was similar between adjuvant and metastatic groups.
Transfusion Requirements
In addition to being able to maintain higher hemoglobin levels during the study period, the group treated with epoetin alfa also exhibited a lower transfusion rate, as listed in Table 4. An absolute reduction in transfusion of 14.3% and a relative reduction of 62.4% were accomplished with epoetin alfa treatment. Patients who entered the trial with baseline hemoglobin 11 g/dL had fewer transfusions in both arms. Both frequency of transfusion and total units transfused (49 units) were lower in the active epoetin alfa group versus the SOC group (148 units). Hemoglobin level at the time of transfusion was similar between the two groups.
QOL
The primary end point of this study was comparing the change in QOL between groups as measured by the FACT每An anemia subscale from baseline to week 12 between the group treated with epoetin alfa and the group receiving SOC without epoetin treatment. Of the 350 EFF patients, 338 patients (168 in the epoetin alfa group and 170 in the SOC group) had both baseline and at least one set of postbaseline QOL data collected and were therefore evaluated for QOL. As shown in Fig 2 the mean change scores in both FACT每An anemia subscale and FACT每An fatigue subscale after 12 weeks of epoetin alfa treatment were highly significant (P < .0001) when compared with the SOC group. Even in the FACT每An nonfatigue subscale, which mainly evaluates clinical signs and symptoms of anemia, mean change scores were significantly better in the epoetin alfa arm. The differences in the two groups were marked by improvement in the epoetin alfa group and significant deterioration in the SOC group. Change between end-of-study (anywhere between weeks 16 and 28) to baseline was also significantly better for the epoetin alfa group for both of the FACT每An subscales (data not shown)
Figure 3 shows results from CLAS assessment, and all three QOL variables: energy level, daily activities, and overall QOL were significantly better in the epoetin alfa group compared with the SOC group. In addition, and as observed with the FACT每An instrument, between-group differences for all three CLAS change scores from baseline to study end were also statistically significant, favoring epoetin alfa (data not shown).
Predictors of Responses
An exploratory logistic regression analysis was used to determine which baseline parameters (treatment; age; race; disease stage; menopausal status; adjuvant or metastatic; next chemotherapy cycle; time from first diagnosis; and baseline hemoglobin, neutrophils, platelets, and serum ferritin) were significant predictors of hemoglobin responses and incidence of transfusion. Table 5 shows the results for hemoglobin response. As expected, baseline hemoglobin was shown to be a strong predictor; higher hemoglobin at entry was the most robust predictor among all baseline parameters evaluated. Interestingly, older age was a significant predictor for hemoglobin response. For incidence of transfusion (Table 6), low baseline hemoglobin again was a strong predictor for subsequent transfusion. Lower platelets (an indication of bone marrow insufficiency) and higher serum ferritin levels (which could indicate an elevated inflammatory state) were predictive of higher transfusion requirements during chemotherapy.
Safety
Epoetin alfa was well tolerated by patients with breast cancer. The overall incidence of AEs was similar between the two groups (Table 7). The most common AEs were granulocytopenia, nausea, vomiting, and hot flushes, which were probably associated with chemotherapy treatment. Granulocytopenia, nausea, and anorexia were significantly higher in the epoetin alfa arm, possibly due to an observed higher intensity of chemotherapy delivered to the patients. Although the current study was not designed to examine dose-intensity, observed differences led to the calculation of relative dose-intensity, which showed differences between groups (P = .0383; data not shown). Incidences of fever and, as expected, anemia, were higher for the SOC group. Pure red-cell aplasia has not been observed in oncology patients, and there were no observed instances in the current study. Nineteen patients in the epoetin alfa group and 14 patients in the SOC group experienced thrombovascular events (Table 7). Of the 19 patients in the epoetin alfa group who experienced thrombovascular events, investigators attributed seven (3.9%) to epoetin alfa therapy. Two-year survival data are being collected and at the time of this writing, 24 patients had died in the epoetin alfa arm (seven adjuvant, 17 metastatic) and 27 had died in the SOC group (five adjuvant, 22 metastatic).
DISCUSSION
The therapeutic benefit of epoetin alfa in anemic cancer patients has been well documented with several large community and double-blind placebo controlled, randomized trials.3,5-8,15,16 Results from those studies consistently demonstrate that epoetin alfa significantly increased hemoglobin levels and decreased the incidence of blood transfusions. In studies in which QOL was assessed, strong association was observed between increase in hemoglobin and improvement in QOL.3 Although most of these studies focused on anemia correction (patients were recruited to the studies when hemoglobin decreased below 10.5 to 11 g/dL), the clinical benefits of anemia prevention have not been fully explored. In the study by Littlewood et al,3 a small group of patients entered the study with baseline hemoglobin between 10.5 and 12 g/dL and this subpopulation was compared with patients with entry hemoglobin less than 10.5 g/dL according to the prespecified hemoglobin strata. Both hemoglobin responses and transfusion reduction were numerically better in the higher hemoglobin stratum group. However, because of the small number of patients in the higher baseline hemoglobin stratum, statistical significance was not achieved.
When data from two large community-based studies were pooled to evaluate the correlation between hemoglobin change and QOL improvement, it was shown that a strong correlation exists but the relationship is not linear. An incremental analysis using regression method was used to explore the longitudinal relationship between incremental changes in hemoglobin and that of QOL scores.17 It was found that among all incremental increases in hemoglobin of 1 g/dL, the greatest gain in QOL occurred at the increase from 11 to 12 g/dL.
The objective of our study was to maintain a hemoglobin level at 12 g/dL throughout the chemotherapy cycle by maximizing the impact of epoetin alfa treatment. Hemoglobin response was predefined as the mean hemoglobin level being maintained at 12 g/dL over the evaluation period of 4 to 12 weeks after random assignment. This is a more rigorous definition of hemoglobin responses than that used in previous studies because it incorporates all hemoglobin values during the defined period in the calculation of response. Using this response definition, more than 50% percent of the patients (52%) in the epoetin alfa arm maintained a mean hemoglobin of more than 12 g/dL during the evaluation period, whereas only 5.1% of patients in the SOC group had such a response, with the epoetin alfa group showing more than a 10-fold increase in percentage of responders. Absolute hemoglobin difference between the two groups was slightly lower than 2 g/dL. This is because the mean hemoglobin in the SOC group never decreased below 10.5 g/dL and investigators were instructed to maintain hemoglobin just above 12 g/dL without going higher in the epoetin alfa arm. This is reflected by the higher frequency of dose reduction as compared with dose increase in the epoetin alfa arm (dose was stable for 62.1% of patients, dose was decreased or held for 28.7% of patients, and dose was increased for 9.2% of patients). Using a definition for hemoglobin response more consistent with that in previous trials ( 2 g/dL increase without benefit of transfusion), the response rate was 65.7% for the epoetin alfa group versus 6.5% for SOC, and this compares favorably with previous studies.3,8,18 The high entry hemoglobin (mean, > 11 g/dL) probably played an important role in the robust response compared with that in previous trials (baseline hemoglobin, < 10 g/dL), given that results similar to those in this study were observed in the higher hemoglobin stratum of the study by Littlewood et al.3
The incidence of transfusion was 22.5% in the SOC group, which was high considering that most patients were receiving adjuvant chemotherapy and no patient was receiving dose-dense or dose-intense regimens. The high percentage of patients receiving a CEF (cyclophosphamide, epirubicin, and fluorouracil) regimen, with its documented hematologic toxicity, might be one of the reasons for the relatively high incidence of transfusion. Treatment with epoetin alfa significantly reduced the transfusion rate to 8%, representing a 14.6% absolute reduction and 64% relative reduction〞the highest reduction of transfusion rate reported in the literature in a randomized setting with erythropoiesis stimulating factors. The mean transfusion threshold was slightly lower for the epoetin alfa group (7.73 ㊣ 1.83 g/dL) versus the SOC group (8.11 ㊣ 0.87 g/dL; P = .143); however, the range was similar between the two groups, suggesting that bias was not a major factor for the transfusion differences between groups. Despite the higher transfusion rate, the SOC group still had significantly fewer hemoglobin responders than the epoetin alfa group. The history of transfusion before study entry was minimal because of the high percentage of adjuvant patients. There was a slightly higher frequency of transfusion in the epoetin alfa arm than the SOC group (five v one, respectively). Improvement in transfusion reduction in this study was similar to that previously observed with the high hemoglobin stratum of the study by Littlewood et al3 (22% in placebo v 8% in epoetin alfa group), suggesting the benefit of higher hemoglobin levels at study entry.
The primary end point of this study was change of FACT每An anemia subscale over time between the two groups. Patients in the epoetin alfa arm showed improvement in QOL over baseline values, whereas QOL in the SOC group deteriorated during the course of chemotherapy. Changes in FACT每An anemia and FACT每An fatigue subscores were highly significant (P < .0001) between the two groups and even the FACT每An nonfatigue scores significantly favored the epoetin alfa group. This highly significant improvement in QOL further substantiates previous observations that the largest gain in QOL is achieved when hemoglobin increases from 11 to 12 g/dL (range, 11 to 13 g/dL). Furthermore, observations from the FACT每An instrument were well supported by the CLAS results. As has been shown in several previous studies, change in QOL correlated extremely well with change in hemoglobin (P value ranges from < .0001 to .0019 for six scales analyzed).
To avoid the burden of placebo injections in patients, an open-label design was selected to evaluate the benefit of early anemia management and hemoglobin maintenance on QOL. The validity of this design was predicated on the availability of objective measurements from other trials such as hemoglobin responses and transfusion reductions. It was stated in the protocol a priori that the correlation between hemoglobin changes and improvement in QOL would be analyzed and compared with previous studies to validate the study design. The hemoglobin responses in the epoetin alfa arm of this study (65.8%) compared well with the double-blind study by Littlewood et al3 (66.4% in the solid tumor stratum of the epoetin alfa arm using same definition for response). When correlation between change in hemoglobin levels and QOL were compared between the two trials, the correlation coefficient for the double-blind study3 for hemoglobin change and FACT每An fatigue subscale was 0.2879 (n = 273; P = .0002), whereas in our study, the correlation of the same variables was 0.2893 (n = 338; P < .0001). The comparable hemoglobin responses and the nearly identical correlation coefficients between change in hemoglobin and QOL score suggest that this open-label study mimics that of the double-blind study and did not adversely influence the outcome of the study. One problem with epoetin alfa studies is the difficulty of using a truly double-blind format for both physicians and patients because of continuous measurement of hemoglobin values. To minimize bias in this study hemoglobin levels were measured after patients had completed their QOL questionnaires.
Epoetin alfa treatment has been linked consistently to improvement in QOL in cancer patients. In this study, a highly statistically significant improvement in QOL as measured by FACT每An and CLAS was observed in patients treated with epoetin alfa. Cella et al19 conducted an Internet-based survey administering the FACT每An to a nationally representative sample of the United States population to better characterize the clinical significance of QOL changes associated with epoetin alfa in reference to the general population. Results from that study show that the FACT每An discriminated between respondents with no history of prespecified illnesses and those with a history of chronic conditions, such as anemia and cancer. Using FACT每An data from the normative sample, the authors concluded that treatment with epoetin alfa reported in the study by Littlewood et al3 leads to clinically relevant and important improvements in QOL.
The pattern of QOL difference between the epoetin alfa and SOC groups observed in the current study was strikingly similar to that observed in the study by Littlewood et al.3 Moreover, a substantial proportion of the QOL deficit observed between the respective study groups at baseline and the sample without any history of prespecified illnesses from the study by Cella et al19 was overcome through administration of epoetin alfa therapy, suggesting that the same findings of QOL clinical significance for the study by Littlewood et al3 should be applied to the current study.
Previous attempts have been made to identify predictors of hemoglobin responses. Serum erythropoietin levels, serum ferritin level, and change in hemoglobin or reticulocyte count within the first 4 weeks after initiation of therapy have all been shown to have some association with response to epoetin therapy; however, to date, the association demonstrated has been weak and not readily reproducible.20 In this study, using baseline variables as potential predictors of hemoglobin response, logistic regression suggests that baseline hemoglobin was a robust and significant predictor of subsequent hemoglobin response and transfusion reduction.
Epoetin alfa was well tolerated in the study. Several AEs reported more often in the epoetin alfa arm may be related to the higher intensity of chemotherapy delivered to patients in that arm (5.0 cycles) versus SOC (4.6 cycles; P = .058). In previous studies granulocytopenia was partly attributable to epoetin alfa treatment favoring erythropoiesis and leading to blunted white cell responses. However, this explanation appears unlikely in this study because of the higher incidences of other chemotherapy-associated AEs such as nausea and anorexia observed in the epoetin alfa arm.
The study results presented here demonstrate that early initiation of treatment with epoetin alfa is effective in and maintaining hemoglobin level, reducing transfusion requirements, and improving QOL in breast cancer patients throughout treatment with myelotoxic agents. With clinical recognition of the importance of achieving optimal chemotherapy dosing and the increasing use of more aggressive myelotoxic agents in the treatment of both adjuvant and metastatic disease, the treatment gains associated with epoetin alfa, which are reported here, are of great value and of significant clinical importance.
Appendix
List of Investigative Sites: Dr P. Goss, Princess Margaret Hospital; Dr R. Goel, Ottawa Regional Cancer Centre; Dr J. Chang, Lakeridge Health Oshawa; Dr M. Blackstein, Mount Sinai Hospital; Dr R. Haq, St Michael＊s Hospital; Dr J. Curtis/C. Hamm, Windsor Regional Cancer Centre; Dr B. Findlay, Hotel-Dieu St Catharines; Dr L. Yelle, CHUM Campus Notre-Dame; Dr G. Cantin, CHA ST-Sacrement; Dr D. Vergidis, Northwestern Ontario Regional Cancer Centre; Dr K. Laing, Dr H. Bliss Murphy Cancer Centre; Dr F. Couture, Htel-Dieu de Quebec and Htel-Dieu de Levis (2 separate sites); Dr J. Dufresne, Centre University Sante l＊Estrie-Fleur; Dr M. Jancewicz, Allan Blair Cancer Centre; Dr S. Young, Northeastern Ontario Regional Cancer Centre; Dr S. Fox, Hopital Charles LeMoyne; Dr S. Berger, Humber River Regional Hospital; Dr L. Zibdawi, Southlake Regional Health Centre; Dr J. Gapski, Trillium Health Centre; Dr R. Pahil, Cape Breton Cancer Centre; Dr R. Levesque, Quinte Health Care Corp, Belleville General Hospital.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Owns stock (not including shares held through a public mutual fund): Catherine Y. Lau, Ortho Biotech Canada Inc; Kara-Lee McWatters, Ortho Biotech Canada Inc. Acted as a consultant within the last 2 years: Jose Chang, Ortho Biotech Canada Inc. Received more than $2,000 a year from a company for either of the last 2 years: Catherine Y. Lau, Ortho Biotech Canada, Inc; Kara-Lee McWatters, Ortho Biotech Canada Inc; Jose Chang, Ortho Biotech Canada Inc.
NOTES
Research funding provided by Ortho Biotech Canada Inc.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
SEER Web site. Trends in SEER Incidence and US death rates by primary cancer site: 1992-2000. http://seer.cancer.gov/csr/1975_2001
Carelle N, Piotto E, Bellanger A, et al: Changing patient perceptions of the side effects of cancer chemotherapy. Cancer 95:155-163, 2002
Littlewood TJ, Bajetta E, Nortier W: Effects of epoetin alfa on hematologic parameters and QOL in cancer patients receiving nonplatinum chemotherapy: Results of a randomized, double blind, placebo-controlled trial. J Clin Oncol 19:2865-2874, 2001
Ludwig H, Fritz E: Anemia in cancer patients. Semin Oncol 25:2-6, 1998
Abels RL: Recombinant human erythropoietin in the treatment of the anaemia of cancer. Acta Haematol 87:4-11, 1992
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 15:1218-1234, 1997
Demetri GD, Gabrilove JL, Blasi MV, et al: Benefits of epoetin alfa in anemic breast cancer patients receiving chemotherapy. Clin Breast Cancer 3:45-51, 2002
Gabrilove JL, Cleeland CS, Livingston RB, et al: Clinical evaluation of once weekly dosing of epoetin alfa in chemotherapy patients: Improvements in hemoglobin and QOL similar to three times weekly dosing. J Clin Oncol 19:2875-2882, 2001
Balducci L, Extermann M: Cancer and aging: An evolving panorama. Hematol Oncol Clin North Am 14:1-16, 2000
Repetto L, Carreca I, Maraninchi D, et al: Use of growth factors in the elderly patient cancer: A report from the Second International Society for Geriatric Oncology (SIOG) 2001 meeting. Crit Rev Oncol Hematol 45:123-128, 2003
Cella D: The Functional Assessment of Cancer Therapy每Anemia (FACT-AN) scale: A new tool for the assessment of outcomes in cancer anemia and fatigue. Semin Haematol 34:13-19, 1997
Yellen SB, Cella DF, Webster K, et al: Measuring fatigue, and other anemia related symptoms with the Functional Assessment of Cancer Therapy (FACT) measurement system. J Pain Symptom Manage 13:63-74, 1997
Demetri GD: Anaemia and its functional consequences in cancer patients: Current challenges in management and prospects for improving therapy. Br J Cancer 84:31-37, 2001 (suppl)
Cleeland CS, Demetri GD, Glaspy J, et al: Identifying hemoglobin level for optimal quality of life: Results of an incremental analysis. Proc Am Soc Clin Oncol 18:574a, 1999 (abstr 2215)
Osterborg A, Brandberg Y, Molostova V, et al: Randomized, double-blind, placebo-controlled trial of recombinant human erythropoietin, epoetin beta, in hematological malignancies. J Clin Oncol 20:2486-2494, 2002
Dammacco F, Castoldi G, Rodjer S: Efficacy of epoetin alfa in the treatment of anemia of multiple myeloma. Br J Haematol 113:172-179, 2001
Crawford J, Cella D, Cleeland CS, et al: Relationship between changes in hemoglobin level and quality of life during chemotherapy in anemic cancer patients receiving epoetin alfa therapy. Cancer 95:888-895, 2002
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 tumour type: Results from a prospective community oncology study. J Clin Oncol 16:3412-3425, 1998
Cella D, Zagari MJ, Vandoros C, et al: Epoetin alfa treatment results in clinically significant improvements in quality of life in anemic cancer patients when referenced to the general population. J Clin Oncol 21:366-373, 2003
Littlewood TJ, Zagari M, Pallister C, et al: Baseline and early treatment factors are not clinically useful for predicting individual response to erythropoietin in anemic cancer patients. Oncologist 8:99-107, 2003(Jose Chang, Felix Couture)