Canadian supportive care recommendations for the management of anemia in patients with cancer.

Anemia is a common finding in cancer patients, most often as a result of chemotherapy. Management of anemia requires a comprehensive approach of appropriate diagnosis, exclusion of reversible causes, use of erythropoiesis-stimulating agents (ESAS), and iron supplementation. Recently, consensus guidelines on the management of chemotherapy-induced anemia were published in Europe and the United States. The present review is intended to be a practical guide for Canadian physicians, based on published guidelines, but specifically tailored to the Canadian environment. Recommendations for the use of ESAS are presented, including initiation, target hemoglobin, dosing and adjustments, monitoring, and re-initiation. Issues of safety are also addressed, including thromboembolic risk, impact on survival, and tumour progression. The importance of iron metabolism and the use of iron supplementation (both oral and parenteral) is discussed.


INTRODUCTION
Anemia is common in cancer 1 and may be the result of many factors, including chemotherapy, marrow infiltration, radiation, nutritional deficiencies, and blood loss. Anemia in cancer adversely affects the patient by producing fatigue, cognitive impairment, a need for blood transfusions, and possible disruption of therapy. It is therefore incumbent on the treating physician to address anemia in cancer.
Easily reversible causes such as nutritional deficiencies, hypothyroidism, and iron deficiency should be corrected immediately; however, these causes ac-count for only a small proportion of anemias in cancer patients.
One method of treating anemia (in addition to transfusion) is the use of an erythropoiesis-stimulating agent (ESA). The rationale for the use of ESAs in chemotherapy-induced anemia (CIA) has been twofold: to improve the patient's hemoglobin level and to reduce the need for transfusions. A multitude of clinical trials and meta-analyses have validated the use of ESAs for these two purposes. In addition, the data have also demonstrated improvements in quality of life contributing to the indication for the use of ESAs in Canada [2][3][4][5][6] .
The purpose of the present article is to use published data and guidelines and consensus opinion from Canadian physicians to provide recommendations for the treatment of CIA in Canada. The recommendations provided here focus specifically on the use of ESAs for CIA. A medical guideline aims to guide decisions and criteria in specific areas of health care by examining and synthesizing the best current evidence. Guidelines can have a significant economic impact in health care and can also provide appropriate access to new medications. Patients benefit from guidelines by receiving timely treatment based on supporting evidence and also better information that facilitates treatment choice. Health care professionals receive new recommendations based on current clinical evidence, which also supports quality improvement activities. As a result of guidelines, health care systems may improve their efficiency and optimize value for money.

Existing Guidelines
The American Society of Clinical Oncology / American Society of Hematology (ASCO/ASH), the European Organization for Research and Treatment of Cancer (EORTC), and the National Comprehensive Cancer Canadian supportive care recommendations for the management of anemia in patients with cancer CURRENT ONCOLOGY-VOLUME 14, NUMBER 5 Network (NCCN) have all published guidelines for the use of ESAs in patients with cancer [7][8][9] . Variations between the guidelines can be attributed primarily to the scientific literature available at the time of guideline development 10

ESA Treatment Goals and Targets
The goals of ESA management of symptomatic anemic patients with malignancy undergoing systemic chemotherapy are to maximize quality of life and minimize transfusion requirements while using the lowest dose of ESA necessary to achieve a target hemoglobin level. International guidelines vary slightly in their recommendations for target hemoglobin level. The ASCO/ ASH and NCCN hemoglobin target is 120 g/L; EORTC recommends a target range of 120-130 g/L. Product monographs for epoetin alfa and darbepoetin alfa both indicate that target hemoglobin should not exceed 120 g/L, with a rate of rise not greater than 10 g/L over a 2-week period 12,13 .
Notably, a hemoglobin target is a designated value not to be exceeded, rather than a target to be aimed for. Hemoglobin levels higher than 130 g/L are not recommended, because randomized clinical trials have observed an adverse impact on overall survival at such levels 14 .

ESA Dosing and Adjustments
Optimal ESA dosing and titration for both epoetin alfa and darbepoetin alfa for patients with cancer receiving systemic chemotherapy have been well established through randomized trials.

Initiating ESA Dose
In adults, epoetin alfa and darbepoetin alfa each have two recommended initiation dosing schedules (Table I). Titration graphs il-lustrate the mean change in hemoglobin concentration from initiation of epoetin alfa ( Figure 1) and mean hemoglobin concentration levels for darbepoetin alfa ( Figure 2) using various dosing regimens 15,16 .

Initiating Hemoglobin Level
The Canadian standard has been to initiate ESA therapy at a hemoglobin level below 110 g/L in symptomatic patients. As with recommendations on hemoglobin targets, recommendations on ESA initiation vary slightly between published guidelines (Table II) The guideline from NCCN 9 recommends consideration of an ESA when the patient's hemoglobin level is below 110 g/L. The guideline from EORTC 8 initially recommended ESAs for either asymptomatic or symptomatic cancer patients with hemoglobin levels of 90-110 g/L who are receiving chemotherapy. In a 2006 update, EORTC refined the guideline by stating that "ESAs may be considered in asymptomatic patients with hemoglobin levels below 119 g/L to prevent a further decline according to individual characteristics" 11 . The guideline from ASCO/ ASH 7 recommends ESAs for patients with hemoglobin levels below 100 g/L who are receiving chemotherapy and, if symptomatic, for patients with hemoglobin levels below 120 g/L.
Some common principles emerge from the various guidelines-for example, a consideration of patient characteristics such as baseline hemoglobin level, rate of decline of hemoglobin, number of chemotherapy cycles remaining, and clinical symptomatology when initiating ESA therapy.
It is important to note that initiation of ESAs should not be delayed. A meta-analysis by Couture et al. 17 showed that early treatment with epoetin alfa could reduce the number of red blood cell transfusions in patients receiving chemotherapy. Quirt et al. 18 further demonstrated that the initiation of ESAs at a hemoglobin level below 100 g/L resulted in a higher risk of blood transfusions. Finally, a systematic review by Lyman and Glaspy 19 found a reduction in the relative risk of transfusion and hemoglobin below 100 g/L after early intervention (p < 0.0001). Collectively, these findings suggest that optimal benefit from ESA treatment of CIA may be achieved through early intervention.
Monitoring Maintenance therapy requires regular monitoring of hemoglobin levels. For patients who are responding, treatment can continue as long as that response is maintained, tolerance is ongoing, and systemic chemotherapy is continuing (Table III).

Dose De-escalation
If the rate of rise or the absolute level of hemoglobin achieved exceeds given values on ESA therapy, dose de-escalation should be initiated (Table IV).

Non-response and Dose Escalation
Non-response can be defined as the failure to achieve a hemoglobin increase of ≥10 g/L in 4 weeks with weekly epoetin alfa, or in 6 weeks with once-weekly or every-3-weeks darbepoetin alfa. Dose escalation of the ESA should be predicated on evidence of non-response and ongoing tolerance to treatment (Table V).
Discontinuation Treatment of anemia with an ESA should be discontinued if response is lost or if acute drug toxicities arise. Discontinuation of the ESA may be considered in patients who experience a thrombotic event. Considerable controversy remains regarding the optimal continuation time for an ESA following completion of chemotherapy. It is reason-able to continue ESA therapy for a duration beyond chemotherapy completion equivalent to 1-2 further cycles of therapy-for example, a further 6 weeks for an every-3-weeks regimen). If, in the epoetin alfa 3-times-weekly dosing schedule, the hemoglobin increase is less than 10 g/L and the reticulocyte count increase is less than 40,000 cells/µL above baseline after a further 4 weeks of therapy at 300 IU/kg, response is unlikely, and treatment should be discontinued. Similarly, for onceweekly dosing, if a satisfactory response to the increased weekly dose of 60,000 IU has not been obtained after a further 4 weeks, then the patient is unlikely to respond, and treatment should be discontinued 12 .

Re-initiation
Re-initiation of anemia treatment with an ESA can be recommended if systemic chemotherapy for malignant disease is again delivered and is accompanied by an observed anemia meeting the same criteria for initiation of ESA as during previous chemotherapy cycles. The patient should have no prior history of adverse events associated with ESA treatment. It is suggested that the regular criteria for ESA initiation, dose adjustment, and discontinuation be applied to all subsequent decisions for ESA re-initiation.

Safety of ESAs
Recent data have signalled safety issues involving the use of ESAs. These concerns-including thromboembolic risk, survival, and tumour proliferationare reviewed in this subsection. CURRENT ONCOLOGY-VOLUME 14 This difference became apparent within the first 4 months of therapy initiation and may have been secondary to a higher incidence of thrombotic cardiovascular events.
The ENHANCE trial randomized patients with headand-neck cancer to placebo or epoetin beta with a goal of achieving hemoglobin levels of 145-150 g/L 22 . Patients were not anemic at randomization, and they received concomitant radiation therapy but not chemotherapy. As in the BEST study, a significantly decreased survival rate was observed in the ESA-treated group (HR: 1.39; 95% CI: 1.07 to 1.74). Patients randomized to placebo survived a median of 928 days as compared with 605 days in the epoetin beta group (p = 0.09). Epoetin alfa 12 Darbepoetin alfa Three times weekly and Weekly dosing schedule Every-3-weeks dosing schedule weekly dosing schedules

Maintenance therapy requires
The Canadian standard has been to monitor Hb, Monitor Hb, prior to dosing, at least every weekly Hb monitoring prior to dosing, on a weekly basis until the 3 weeks until the target Hb (120 g/L) target Hb (120 g/L) is reached, and every is reached, and regularly thereafter 13 1-3 weeks thereafter Hb = hemoglobin. The EPO-CAN 20 trial was a randomized, placebocontrolled trial involving anemic patients with nonsmall-cell lung cancer who were not on active chemotherapy 23 . Because of evolving safety concerns regarding ESA use in patients not receiving chemotherapy, an unplanned interim safety analysis was performed after 70 patients had been randomized. Median survival rates were higher in the placebo group (HR: 1.84), resulting in study closure-although conclusions were unclear because of the high early mortality rate and the very small sample size.
Amgen's 20010103 study was a randomized, placebo-controlled trial of darbepoetin alfa for anemic patients with cancer but not receiving chemotherapy ]. The primary end point of transfusion reduction was not met, and an increase in mortality was seen in the ESA arm (HR: 1.29). In this preliminary report, median follow-up was only 4.3 months. As in the EPO-CAN 20 study, the use of ESAs in patients not on chemotherapy has not been common practice and is now discouraged.
Summarizing these four trials, BEST and ENHANCE treated patients to higher-than-recommended target hemoglobin levels, and the EPO-CAN 20 and Amgen 20010103 trials used ESAs in cancer patients not on active chemotherapy treatment. The results of these four trials have contributed to the generally accepted recommendations regarding a more modest hemoglobin target of 120 g/L and a restriction of ESA use to cancer patients with anemia who are receiving chemotherapy.
No Survival Disadvantage Despite the contrary evidence, some large randomized controlled trials of ESA therapy have failed to demonstrate a negative impact on overall survival. Studies involving patients receiving chemotherapy for non-small-cell lung cancer (early versus late intervention) with epoetin alfa 24 and placebo-controlled trials of darbepoetin 25 did not reveal any survival decrement. Another randomized double-blind placebo-controlled trial of patients with  If the rate of Hb increase is >10 g/L in a 2-week If the rate of Hb increase is >15 g/L per any 2-week period or if period or >15 g/L in a 3-week period, or when 3-week period or when Hb>120 g/L, Hb>120 g/L, the dose should Hb>120 g/L, the dose should be the dose should be reduced by 40% be reduced by 25% reduced by 25% of the previous dose If Hb>130 g/L, the dose If Hb>130 g/L, dosing should be temporarily If Hb>130 g/L, dosing should be should be temporarily withheld until Hb falls to 120 g/L; at that time, temporarily withheld; once Hb falls to withheld until Hb<120 g/L therapy should be reinitiated at a dose 25% 120 g/L, therapy should be reinitiated at and reinitiated at a dose 25% below the previous dose a dose 40% below the previous dose below the previous dose Hb = hemoglobin. CURRENT ONCOLOGY-VOLUME 14,NUMBER 5 small-cell lung cancer demonstrated similar survival rates in epoetin alfa and placebo arms 26 .
Recently, another trial (Amgen study 20010145) designed to explore a potential survival advantage when patients with extensive-stage small-cell lung cancer undergoing platinum-plus-etoposide chemotherapy were given darbepoetin alfa was completed. This randomized phase III placebo-controlled study showed no survival advantage in the darbepoetin arm, with a median survival in both arms of 40 weeks, and a hazard ratio (darbepoetin/placebo) of 0.93 (95% CI: 0.78 to 1.11) 27 .
The first Cochrane meta-analysis on this topic included trials through December 2001 and reported inconclusive evidence on the potential impact of ESA therapy on overall survival (HR: 0.81; 95% CI: 0.67 to 0.99) 28 . The updated Cochrane meta-analysis, incorporating trials up to 2005 (including the BEST, EN-HANCE, and EPO-CAN 20 studies), again concluded with uncertainty regarding "whether and how epoetin or darbepoetin affects overall survival" (HR: 1.08; 95% CI: 0.99 to 1.18) 6 . Tumour Progression Tumour progression secondary to the activation of erythropoietin receptors has been questioned 29 . The BEST and ENHANCE trials both showed reduced tumour control in the setting of targeting hemoglobin levels higher than current usual practice. Another study, DAHANCA, was modeled after the ENHANCE study to address some of the latter study's methodologic issues 30 .
The DAHANCA Study was a prospective trial in patients with squamous cell carcinoma of the head and neck who were undergoing treatment with definitive radiotherapy. Patients were randomized to darbepoetin or placebo with a primary endpoint of locoregional control rate at 5 years. An interim analysis of 484 patients demonstrated a 10% increase in the locoregional failure rate among darbepoetintreated patients (p = 0.01). Overall survival was not significantly different, but appeared to trend towards shorter survival in the ESA arm (p = 0.08). There were limitations to this study, including a lack of uniform imaging assessment at baseline or at recurrence, and a lack of confirmation of recurrent disease by tissue biopsy. The final study analysis will be reported to the U.S. Food and Drug Administration in late 2008.
The relationship between the presence of erythropoietin receptors and tumour proliferation because of exogenous erythropoietin with the use of ESAs has not yet been firmly established. Likewise, the function of erythropoietin receptors is not well understood. The in vitro studies to date vary in their findings and have been called into question based on assessment methodology 31 .

Safety Conclusions
Safety issues surrounding ESAs for the treatment of cancer-related anemia include thromboembolic risk, survival, and tumour proliferation. The risk of thromboembolic events are increased with ESAs, and careful monitoring of hemoglobin levels in patients is mandatory. In May 2004, the U.S. Food and Drug Administration's Oncologic Drugs Advisory Committee (ODAC) revised product labelling to include warnings against maintaining hemoglobin levels above 120 g/L. The negative survival studies are difficult to interpret in the context of current practice because of their higher-than-recommended hemoglobin targets, higher-than-recommended doses used, methodologic limitations in the individual trials, and the enrolment of patients not on active chemotherapy. The trials evaluating disease control rates suffer from similar limitations. Because

Iron
Despite the availability and ease of administration of ESAs to treat CIA in cancer patients, only about 60% of patients experience an adequate response to such agents 6 . Given that one of the pathologic mechanisms in anemia of cancer is the sequestration of iron out of the available circulating pool, and that ample experience exists in nephrology regarding administration of parenteral iron to improve hemoglobin levels in patients with chronic renal failure who are receiving ESAs, attempts have been made to improve the response to ESAs in oncology by supplementing with iron.
A few studies have tested iron supplementation in anemic cancer patients receiving chemotherapy and ESAs. In a study by Auerbach et al. 33 , cancer patients on chemotherapy who were anemic and receiving weekly epoetin alfa 40,000 IU were randomized to no iron supplementation, oral iron supplementation, total dose intravenous infusion of iron dextran, or weekly intravenous iron dextran. Patients receiving parenteral iron supplementation experienced a greater improvement in their hemoglobin level, and more patients experienced a hemoglobin response (hemoglobin at least 120 g/L or an increase of at least 20 g/L). Quality-of-life analyses suggested that patients treated with parenteral iron experienced improvements in energy, activity, and overall quality of life.
In another study by Vandebroek et al. 34 , patients receiving chemotherapy for non-myeloid malignancies and being treated with darbepoetin alfa 500 µg subcutaneously every 3 weeks were randomized to oral iron, no iron, or iron sucrose 200 mg intravenously every 3 weeks. Fewer patients required transfusions when treated with intravenous iron. Hemoglobin lev-els and clinically meaningful improvements in scores on the Functional Assessment of Cancer Therapy-Fatigue appeared higher, but differences were not statistically significant. Iron supplementation was not associated with any significant toxicities. This study is currently available only in abstract form.
In a recently published study by Henry et al. 35 , patients treated with epoetin alfa at a weekly dose of 40,000 IU were randomized to weekly intravenous sodium ferric gluconate 125 mg, oral ferrous sulphate 325 mg three times daily, or no iron. Hemoglobin increase was significantly better in the intravenous group as compared with the oral iron and no iron groups (24 g/L vs. 16 g/L vs. 15 g/L respectively; p = 0.0092 for intravenous as compared with oral iron). Hemoglobin response was seen in 73% of intravenous iron-treated patients as compared with 45% of patients treated with oral iron and 41% receiving no iron.
In another study by Bellet et al. 36 , cancer patients receiving chemotherapy and ESAs were randomized to intravenous iron sucrose or to no iron after 8 weeks of ESA treatment. All patients continued with ESA therapy for up to 12 weeks. Regardless of the response to the initial 8-week course of ESA alone, patients receiving iron sucrose experienced a greater hemoglobin rise than did patients not receiving iron.
It appears from the published literature that supplementation with parenteral iron is associated with benefits in patients treated with ESAs. Hemoglobin level and hemoglobin response both improve, and the frequency of transfusions declines. Parenteral iron appears to be well tolerated, particularly the iron sucrose and sodium ferric gluconate formulations.
Despite the benefits, outstanding questions remain regarding iron supplementation in patients with CIA who are receiving ESAs. Most of the studies have initiated iron supplementation along with an ESA, except for one study that tested iron after 8 weeks of ESA therapy. No study has compared upfront iron supplementation with no iron supplementation in only poor ESA responders. The studies screened patients and did not enrol patients who were felt to be iron-deficient (defined as a serum ferritin level below 100 ng/L and a transferrin saturation below 15%). Patients who are iron-deficient should initially be treated with iron rather than an ESA. Regarding the potential for inducing a state of iron overload, the two fully published studies of iron supplementation accepted patients with ferritin levels of up to 675 pmol/L 33 and 900 ng/mL (approximately 2025 pmol/L) 35 .
All of the available studies had rather limited periods of follow-up. Potential long-term toxicities of parenteral iron supplementation in cancer patients on chemotherapy have not been reported. No comparative data have been developed to inform the most effective schedule or iron formulation for use in CIA patients on an ESA. Information is also lacking regarding the potential value of new oral iron com-pounds (heme iron polypeptide with folic acid or polysaccharide iron, vitamin B complex, and folic acid) in this group of patients.
Despite the limited available evidence, the provision of parenteral iron to patients with CIA appears to improve response to the ESA and represents a reasonable treatment option for physicians to consider for their patients, even for those patients who experience a poor response to an ESA. Figure 3 summarizes iron management in patients with CIA.

SUMMARY
Anemia is a common finding in cancer patients, most often as a result of chemotherapy. The management of anemia requires a comprehensive approach of appropriate diagnosis, exclusion of reversible causes, and use of ESAs and iron supplementation. The ultimate goal of treating CIA is to increase hemoglobin levels to reduce the need for transfusions and improve quality of life for the patient. Recommendations for the use of ESAs include initiation at a hemoglobin level below 110 g/L, a target hemoglobin of 120 g/L, and dose adjustments to stay within appropriate levels. Hemoglobin should be regularly monitored, and rationales for discontinuation of ESAs or re-initiation should be observed.
Safety issues have recently been raised, namely thromboembolic risk, impact on survival, and tumour progression. Although these mechanisms are not completely understood, strategies to enhance safety include thrombosis prophylaxis when indicated and targeting to appropriate hemoglobin levels. Iron metabolism is critical to anemia, and the use of iron CURRENT ONCOLOGY-VOLUME 14,NUMBER 5 supplementation (especially parenteral iron) may improve response to ESAs, further reducing the need for transfusions.

ACKNOWLEDGMENTS
Amgen Canada Inc. provided an unrestricted educational grant; Klick Communications Inc. provided editorial and project management services. The authors also thank manuscript reviewer Dr. Pierre Laneuville of the Royal Victoria Hospital, McGill University Health Centre, Montreal, Quebec.