As noted in the first half of this article, which appeared in the November issue of ONCOLOGY, the hematopoietic growth factors have transformed the fields of nephrology, hematology, and oncology over the past 2 decades. In part 1 of the article, we focused on the myeloid growth factors-particularly recombinant human (rh) granulocyte colonystimulating factor (G-CSF, filgrastim(Drug information on filgrastim) [Neupogen]) and granulocyte-macrophage colony-stimulating factor (GM-CSF, sargramostim(Drug information on sargramostim) [Leukine])- and their role in decreasing the duration of chemotherapy-induced neutropenia in cancer patients. Part 2 focuses on the use of recombinant erythropoietic agents, and concludes with a brief discussion of other hematopoietic growth factors (ie, thrombopoietic and progenitor cell-mobilizing factors) in oncologic practice. Erythropoietic Factors Biology, Clinical Pharmacology, and Background
Erythropoietin(Drug information on erythropoietin) is the major regulator of erythropoiesis in man. Unlike G-CSF, erythropoietin is not a mitogenic growth factor; it exerts its effects by inhibiting the programmed cell death of committed erythroid precursors. (For this reason, recombinant human erythropoietin [rhEPO] therapy given synchronously with myelosuppressive chemotherapy is not associated with the increased marrow toxicity that may occur when rhG-CSF is given with chemotherapy.) Erythropoietin deficiency is the major or sole cause of the hypoplastic anemia seen in patients with renal failure. In patients with chronic illnesses such as cancer, a relative erythropoietin (EPO) deficiency[1] is coupled with a decreased responsiveness to EPO mediated by the effects of inflammatory cytokines on the marrow and on ferrrokinetics, leading to a high incidence of anemia. When these patients receive chemotherapy, the incidence and severity of anemia increases. Anemia is frequently encountered in cancer patients and is a major concern in oncology practice. When human erythropoietin was cloned, it was found that its in vivo activity was severely impaired in the absence of posttranslational glycosylation. Therefore, rhEPO used in clinical practice is expressed in mammalian cells rather than bacteria. Two rhEPO preparations are currently in clinical use: epoetin alfa(Drug information on epoetin alfa) (Epogen, Procrit) and, unavailable in the United States, epoetin beta(Drug information on epoetin beta) (Marogen). These two preparations differ in the original source of the purified protein used for cloning, slightly in their isoform composition (the proportion of the prepa- ration that contains a given number of sialic acid molecules), and in the stabilizers added to the product. No clear difference in the efficacy of the two preparations has been demonstrated, and the data for both products will be included in the discussion of the effects of rhEPO. Based on the observation that the in vivo potency of a given erythropoietin isoform is directly related to the amount of posttranslational glycosylation, as manifested by total sialic acid content, the erythropoietin molecule has been modified through site-directed mutagenesis, changing five amino acids to add two additional glycosylation sites. The resulting glycoprotein, termed darbepoetin alfa(Drug information on darbepoetin alfa) (Aranesp), has enhanced in vivo potency, which is related to an approximately threefold prolongation of its half-life compared to epoetin alfa.[2] Both rhEPO and darbepoetin alfa have been shown to be effective in the treatment of the anemia that occurs in patients with renal failure on chronic hemodialysis. Because of obvious abnormalities in blood volume homeostasis in this setting, a rapid rise in hemoglobin level and red cell mass can be associated with complications including hypertension (with convulsions in severe cases) and thrombosis. Accordingly, a standard has evolved in which the anemia of dialysis is corrected gradually until the target hemoglobin is achieved. Over the last 15 years, nephrology studies have demonstrated a clear link between hemoglobin level, and both the quality of life and functional status of these patients. Dialysis physicians have been commendable in rapidly and diligently integrating quality-of-life data into their treatment standards and the target hemoglobin levels they set for patients. Treatment Paradigms and Transfusion Avoidance in Cancer Patients Prior to the introduction of rhEPO into oncology practice, the only treatment available for the frequently encountered anemia was red cell transfusion. Because of the risks associated with transfusion, a well-ingrained practice developed in which only severe anemia was recognized as warranting treatment. Understandably, when rhEPO was introduced, oncologists used this agent primarily to reduce the risk of transfusion, and the inclination was to focus on relatively severe degrees of anemia. Although the field has moved on to a focus on enhanced quality of life and treatment of mild and moderate degrees of anemia, for statistical reasons, a reduction in the risk of transfusion is still used as a patient benefit end point in phase III studies. Initial clinical trials of rhEPO in cancer patients demonstrated that patients with anemia often respond to such therapy with an increase in hemoglobin level, regardless of whether they are receiving chemotherapy. These early data indicated that patients not receiving chemotherapy were, if anything, more responsive to rhEPO.[3] Unfortunately, the design of the randomized, placebo-controlled trial involved a relatively low dose and a treatment period of only 8 weeks, which was insufficient to demonstrate a statistically significant impact on the risk of transfusion for patients not receiving chemotherapy.[4] This historical oversight has had a lasting impact on the field; at present, no erythropoietic agent has been approved in any country for the treatment of cancerrelated anemia in patients not receiving chemotherapy, despite its apparent safety and efficacy. Moreover, much less data is available on the optimal dose or effects on quality of life of these agents in this important subset of anemic cancer patients. Optimal Dosing of rhEPO
In the early trials of rhEPO for anemia during chemotherapy, doses of 25, 50, 100, 200, or 300 U/kg five times weekly for 4 weeks or a total dose of approximately 10,000 to 120,000 U/wk were studied.[5,6] The data suggested a relationship between dose and the proportion of patients in whom a given increase in hemoglobin concentration was observed. Pilot randomized trials comparing low doses (2,000 to 3,000 U) to higher doses (6,000 to 10,000 U) administered three times weekly suggested that the higher doses were more effective in terms of increasing hemoglobin levels.[7,8] Perhaps because of cost considerations or difficulty in administering higher doses of rhEPO given current preparations, no definitive study has followed- up on the early observation that high response rates were observed with rhEPO doses of 100,000 U/wk or greater. The doses of rhEPO taken forward into phase III studies became the current standard, but they were not necessarily the optimal doses for cancer chemotherapy patients. In the pivotal randomized, placebo- controlled trials of rhEPO for cancer chemotherapy patients, a starting dose of 150 U/kg or approximately 10,000 U three times weekly for 12 weeks, with a dose increase to 300 U/kg three times weekly in nonresponding patients, was shown to be associated with a reduction in transfusion risk.[4,9,10] Similar results were reported in trials that did not include the dose increase,[11] and no randomized trial has documented the efficacy of dose increases of the magnitude used in these phase III studies. Given that the response to rhEPO at doses of 10,000 U three times weekly can take 10 weeks or more to occur, it remains possible that responses observed after dose doubling would have occurred had the dose not been doubled. The findings of the phase III trials have been confirmed in two large, uncontrolled, community-based studies in which a decrease in transfusion rate was observed during therapy.[12,13] More recently, a large, uncontrolled community- based study that used a fixed starting dose of 40,000 U/wk, with a dose increase in nonresponding patients to 60,000 U/wk has suggested that similar results in terms of hemoglobin increase and transfusion requirements can be achieved with weekly dosing.[14] Although this approach has not been compared to placebo or to three-timesweekly dosing in randomized trials- and this dose and schedule therefore does not appear on the label-it is used in clinical practice in the United States for chemotherapy patients. Most clinicians believe that it produces results similar to the historical experience with three-times-weekly dosing. Recently, a large, well-designed randomized trial conducted in patients with lymphoid malignancies compared rhEPO starting doses of 10,000 U three times a week to a single weekly dose of 30,000 U.[15] The results suggest that the two approaches produce equivalent results in this subset of oncology patients, and call into question the unproven assumptions that underlie the use of an increased weekly dose as employed in the community-based study. This is an important point that requires confirmation and has major cost implications. If rhEPO given weekly at a dose of 30,000 U produces results indistinguishable from a dose of 10,000 U administered three times a week, we may be able to achieve our current level of patient benefit with 25% less cost. Darbepoetin Alfa
Darbepoetin alfa has been compared to placebo in two randomized placebo-controlled clinical trials, in patients with lung cancer who were receiving platinum-based chemotherapy[ 16] and in patients with lymphoid malignancy receiving chemotherapy.[ 17] Darbepoetin at 2.25 μg/kg/wk, with dose doubling permitted in nonresponders, was associated with an increase in hemoglobin and a significant reduction in transfusion requirements. As with rhEPO, the contribution of the dose increase to patient benefit has not been demonstrated. In a phase II, placebo-controlled trial in patients with lymphoid malignancy, three doses of darbepoetin alfa (1, 2.25, and 4.5 μg/wk) appeared to be superior to placebo in terms of hemoglobin increase and transfusion risk.[18] The phase III lung cancer trial was the basis for the current Food and Drug Administration approval of darbepoetin alfa for the treatment of cancer chemotherapy patients. In a recent large phase II study in cancer patients receiving chemotherapy, the dose-response relationships for darbepoetin alfa administered every 1 or 2 weeks were characterized.[ 19,20] A clear relationship was evident between dose and the magnitude of mean increase in hemoglobin in each cohort until a dose of 4.5 μg/kg/wk or 9 μg/kg every 2 weeks was reached; above these "optimal" doses, further increases in efficacy were not observed. These trials were randomized and included a control group treated with epoetin alfa administered at starting doses of either 150 U/kg three times weekly or 40,000 U/wk, with dose increases permitted for nonresponding patients. To date, this is the only randomized trial to compare various fixed doses of darbepoetin alfa with the current standard for the use of rhEPO. In this randomized trial, the lowest dose of darbepoetin alfa to produce an increase in hemoglobin indistinguishable from that achieved with 40,000 U/wk of rhEPO was 3 μg/kg (~200 μg per dose) given every 2 weeks. A subsequent larger, communitybased study has demonstrated similar hemoglobin responses at this dosage, with an increase to 5 μg/kg every 2 weeks in nonresponding patients.[21] Although this approach has not been compared to placebo in a randomized clinical trial and is not the label dose and schedule, it is the darbepoetin regimen currently used in the United States in chemotherapy patients. The lack of a significant difference in the efficacy of this dosing of darbepoetin alfa vs standard dosing of epoetin alfa is the subject of an ongoing large randomized trial. Darbepoetin alfa can be administered as infrequently as every 3 weeks with hemoglobin responses noted in chemotherapy patients, although the every-3-week dose, which produces responses similar to those achieved with current rhEPO dosing, has not been established in a randomized trial. The data linking mild anemia to fatigue and a decreased quality of life, and moderate and severe anemia to a worsening of these symptoms and an increased risk of transfusion are compelling.[ 22] They support a practice policy of early intervention in patients receiving chemotherapy with either rhEPO at a weekly dose of 40,000 U (an alternative being 10,000 U thrice weekly) or darbepoetin at a dose of 200 μg administered every 2 weeks (an alternative being 100 μg weekly). Improving Schedules to Enhance Quality of Life and Functional Status in Cancer Patients The most important advance in anemia management over the past 5 years has been the recognition-initially resisted due to deeply ingrained prejudices and clinical habits-that mild and moderate degrees of anemia (hemoglobin concentrations of 10 to 12 g/dL) are associated with significant levels of fatigue and compromise quality of life and functional status in cancer patients. More importantly, it has become clear that treatment of anemia is associated with significant improvement in fatigue and quality of life.[9,10,13,14,22,23] In retrospect, this should not have been surprising; the relationship between hemoglobin and symptom status in humans had been well characterized in the setting of dialysis a decade earlier, and there were no grounds to assume that cancer patients would be different. Recently, data from more than 4,000 cancer chemotherapy patients treated with rhEPO three times weekly have been analyzed to characterize the relationship between hemoglobin level and quality of life in this population (Figure 1).[24] The results demonstrate that as the hemoglobin level rises from 8 to 12 g/dL, quality of life and energy level improve; the greatest gain occurs as the hemoglobin level rises from 11 to 12 g/dL. Stated another way, the levels of anemia that we are most prone to ignore in oncology practice are those associated with the greatest symptoms and the greatest opportunity for benefit, if treated. Despite these data, standards similar to those developed by nephrologists for dialysis patients have not been developed in oncology, and many, if not most, chemotherapy patients with hemoglobin levels less than 10 g/dL receive no hematopoietic treatment, with milder degrees of anemia even more likely to go untreated. Tailoring Therapy in the Cancer Setting
One possible explanation for the lower level of enthusiasm for erythropoietic therapy in the oncology community may be that, unlike the situation in nephrology, the optimal approach to treatment for the cancer patient has not been developed. Cancer chemotherapy patients are usually treated for 20 weeks or less, and in cancer patients the erythron is relatively resistant to therapy. The dose-finding studies of darbepoetin alfa demonstrated that a clear relationship exists between dose and the rapidity of response and that when an intent-to-treat analysis is performed on the data, the median time to response (defined as a hemoglobin increase of 2 g/dL not due to a red cell transfusion) with the doses of rhEPO or darbepoetin alfa currently employed is approximately 10 weeks. When optimal doses of darbepoetin alfa were given, the median time to response was 7 weeks, and the overall proportion of responding patients was greater.[20] Because no known complications are associated with rapid increases in hemoglobin in patients with adequate renal function, the treatment of anemia in the setting of cancer chemotherapy may require higher, more effective initial doses of either rhEPO or darbepoetin alfa than have been used, with doses decreased once symptomatic relief and an increase in hemoglobin level have been achieved (Figure 2). A pilot randomized trial of this "front-loaded" dosing paradigm, tailored to drug efficacy and safety profiles and the unique requirements of the anemic, symptomatic oncology patient has been published, with results suggesting that this approach will be superior; a large, randomized controlled trial is in progress.[25] Unlike dialysis patients, chemotherapy patients are treated with erythropoietic agents for weeks, not years; an approach tailored to their needs should be characterized by both a high proportion of responders and a rapid response, especially when patients are significantly symptomatic. Another equally rational strategy for optimizing the quality of life and functional status of cancer patients would be to initiate erythropoietic agents much earlier in the course of anemia, before significant symptoms develop (ie, hemoglobin levels between 11.5 and 12 g/dL) and when lower doses of agent administered less frequently are likely to arrest the decline in hemoglobin level. The current practice, in many centers, of withholding therapy until patients are compromised and hemoglobin levels are less than 10 g/dL, is clearly not best for patients. Paradoxically, this "late-intervention" approach, which is often defended as being resource-conserving, may be misguided and actually increase costs by necessitating more aggressive dosing of erythropoietic agents to rescue severely symptomatic patients at imminent risk for transfusion. Early intervention trials with both rhEPO and darbepoetin alfa are in progress. Myelodysplastic Syndromes Chronic, hypoplastic anemia associated with myelodysplastic syndrome (MDS) is a common and difficult problem in oncology practice. Patients with this disorder are frequently elderly, tolerate anemia poorly, become transfusion dependent, and develop complications of transfusion including alloimmunization and iron overload. Sufficient data have accumulated regarding rhEPO therapy in this setting to demonstrate that it is safe, and not associated with an increased risk of progression to acute leukemia or with lineage steal. Unfortunately, rhEPO therapy is frequently ineffective, failing to increase hemoglobin levels in approximately 60% of patients.[26] Recombinant human erythropoietin doses of 200 to 3,000 U/kg/wk administered intravenously, and 150 to 2,000 U/kg administered subcutaneously have been studied, with hemoglobin increases noted mainly at doses of at least 60,000 U/wk and primarily in the subset of MDS patients with refractory anemia and sideroblastic anemia.[27-35] As noted above, the addition of myeloid growth factor to rhEPO therapy appears to be safe and may increase the erythropoietic response in patients with MDS. For patients with symptomatic refractory anemia without excess blasts, it is reasonable for the clinician to initiate an 8-week trial of rhEPO therapy, with or without myeloid growth factor therapy, utilizing the highest dose that is feasible in the particular reimbursement environment.[36,37] For responding patients, therapy can be continued and doses adjusted to maintain an asymptomatic hemoglobin level. No data are available for darbepoetin alfa in this setting. Safety of Erythropoietic Agents in Hematologic Oncology In cancer patients who do not also have renal failure, rhEPO and darbepoetin alfa therapy have been extraordinarily well tolerated. In placebo- controlled studies, a significant increase in hypertension, convulsions, or thrombosis has not been noted, and in studies published to date, there is no suggestion of any association between toxicity and a rapid rise in hemoglobin level. These observations are important, as they point the way to the development of treatment paradigms tailored to the need of cancer patients. In most studies to date, rhEPO and darbepoetin alfa therapy have been associated with injection site pain. In randomized trials, no difference between the safety profiles of rhEPO and darbepoetin alfa has been observed. Pure Red Cell Aplasia
Recent reports from Europe have described an increase in the incidence of pure red cell aplasia (PRCA) in dialysis patients undergoing erythropoietic therapy.[38] This syndrome, which is associated with the development of antibodies that neutralize the effects of endogenous erythropoietin, rhEPO, and darbepoetin alfa, had previously occurred with a very low frequency in patients receiving rhEPO. The increase in incidence observed over the past 3 years has been traced to a particular new preparation of epoetin alfa, which had been developed to eliminate the need for human albumin(Drug information on human albumin) as a stabilizing agent and to address the theoretical risk of prion-mediated disease transmission in Europe. It is important to note that no increase has occurred in the incidence of PRCA associated with epoetin beta or the epoetin alfa preparation used in the United States and that PRCA has not been documented in any cancer patients with any preparation. To date, PRCA has not been reported in any patient treated with darbepoetin alfa who had not also received the new epoetin alfa preparation.