Anders Österborg, MD, PhD, Department of Oncology, Karolinska Hospital, S-171 76 Stockholm, Sweden. E-mail: email@example.com
Summary. Darbepoetin alfa is a novel erythropoiesis-stimulating protein with a prolonged serum half-life. This randomized, double-blind, placebo-controlled, dose-finding study investigated the efficacy and safety of darbepoetin alfa in anaemic patients with lymphoproliferative malignancies who were receiving chemotherapy. Patients were randomized in a 1:2:2:1 ratio to receive darbepoetin alfa 1·0 μg/kg (n = 11), 2·25 μg/kg (n = 22), 4·5 μg/kg (n = 22) or placebo (n = 11), administered subcutaneously once weekly for 12 weeks. No dose increases were allowed during the study. A higher proportion of patients achieved a haemoglobin response (defined as a ≥ 2·0 g/dl increase from baseline) in the darbepoetin alfa 1·0 μg/kg (45%), 2·25 μg/kg (55%) and 4·5 μg/kg (62%) groups than in the placebo group (10%; P < 0·01). The mean change in haemoglobin from baseline to week 13 was 1·56 g/dl in the 1·0 μg/kg group, 1·64 g/dl in the 2·25 μg/kg group and 2·46 g/dl in the 4·5 μg/kg group, compared with a mean change of 1·00 g/dl in the placebo group. The overall safety profile of darbepoetin alfa in this study was similar to that of placebo. These results show that darbepoetin alfa effectively and safely increased haemoglobin concentrations in patients with lymphoproliferative malignancies. Confirmative studies at doses of 2·25 and/or 4·5 μg/kg/week in this population are warranted.
Anaemia is common in patients with lymphoproliferative malignancies, occurring in most patients during the course of their disease and its treatment (Casadevall, 1998; Groopman & Itri, 1999). The pathophysiology of anaemia in this population is multifactorial; the infiltration of bone marrow and the myelosuppressive effects of chemotherapy are often implicated, along with aspects of anaemia of chronic disease, such as deficient production of erythropoietin, impaired iron re-utilization and overproduction of erythropoiesis-inhibiting cytokines (Faquin et al, 1992; Moliterno & Spivak, 1996; Coiffier, 2000).
Recombinant human erythropoietin (rHuEPO) has been shown to increase haemoglobin levels, ameliorate the symptoms of anaemia, and reduce transfusion requirements in patients with solid tumours (Nowrousian, 1998) and lymphoproliferative malignancies (Cazzola et al, 1995; Österborg et al, 1996; Dammacco et al, 2001). Published data on the use of rHuEPO for the treatment of anaemia in lymphoproliferative malignancies (for review see Österborg, 2000) have consisted primarily of studies in patients with multiple myeloma (MM) (Ludwig et al, 1990; Barlogie & Beck, 1993; Garton et al, 1995; Silvestris et al, 1995; Mittelman et al, 1997; Dammacco et al, 2001). Relatively few randomized studies have specifically examined more heterogeneous populations, including patients with MM, lymphoma and chronic lymphocytic leukaemia (CLL) (Cazzola et al, 1995; Österborg et al, 1996). A dose-finding study in patients with MM or non-Hodgkin's lymphoma (NHL) demonstrated haemoglobin response rates of up to 62% with administration of rHuEPO at daily doses of 5000 U or 10 000 U for 8 weeks (Cazzola et al, 1995). In addition, a recent study in anaemic, transfusion-dependent patients with MM, low-grade NHL or CLL showed no significant differences in the efficacy of rHuEPO between these malignancy types. Currently, administration of rHuEPO is indicated three times per week, although once-weekly administration may be possible with an increase in overall dose (Gabrilove et al, 2001).
Research has indicated that greater sialic acid content on the carbohydrate portion of erythropoietin is associated with a greater serum half-life (Egrie & Brown, 2001). Darbepoetin alfa is a novel erythropoiesis stimulating protein (NESP) with two extra consensus N-linked carbohydrate addition sites compared with rHuEPO. Consequently, darbepoetin alfa has a longer serum half-life and greater in vivo biological activity than rHuEPO in animal models (Egrie & Browne, 2001). In patients with chronic renal failure, darbepoetin alfa has a threefold longer serum half-life than rHuEPO, allowing less frequent administration with a similar efficacy and safety profile (Macdougall et al, 1999; Macdougall, 2000). Darbepoetin alfa has also demonstrated a prolonged half-life in patients with cancer (Glaspy et al, 2000; Heatherington et al, 2001) and appears to be effective for the treatment of anaemia when administered every 1, 2 or 3 weeks to patients with solid tumours receiving cytotoxic chemotherapy (Kotasek et al, 2000; Glaspy et al, 2001; Pirker et al, 2001).
This randomized, double-blind, placebo-controlled study was designed to assess the safety and dose–response relationship of darbepoetin alfa administered subcutaneously at doses of 1·0, 2·25 and 4·5 μg/kg once weekly in patients with various types of lymphoproliferative malignancies who were receiving multicycle chemotherapy.
Patients and methods
Patients. Each institution's independent ethics committee approved the study and all patients gave written informed consent before participation. Patients with a diagnosis of lymphoproliferative malignancy (MM, low- and intermediate-grade NHL, Hodgkin's disease or CLL) were enrolled at 15 study centres in Europe and Australia. Patients were required to have a life expectancy of at least 6 months and an Eastern Cooperative Oncology Group (ECOG) performance status of 0–2. All patients were scheduled to receive chemotherapy for at least 12 additional weeks. Additional entry criteria included: anaemia (haemoglobin concentration≤11·0 g/dl); adequate iron stores (transferrin saturation ≥ 15% or ferritin ≥ 10 μg/l), normal serum vitamin B12 and folate concentrations; and adequate liver function (serum bilirubin≤1·5 times the upper limit of the normal range) and renal function (serum creatinine≤177 μmol/l). Exclusion criteria included high-grade NHL, myeloablative chemotherapy or radiotherapy for transplantation or chemotherapy regimens using investigational agents, primary or metastatic malignancy involving the central nervous system, clinical evidence of active infection or inflammatory disease, or other disorders that could potentially interfere with the response to darbepoetin alfa. Patients had not received two red blood cell (RBC) transfusions within 4 weeks of randomization or any RBC transfusion within 2 weeks of randomization.
Study design and treatment. This was a multicentre, randomized, double-blind, placebo-controlled study. All patients completed a screening period within the 4 d before randomization. Eligible patients were randomized in a 1:2:2:1 ratio to receive darbepoetin alfa at doses of 1·0, 2·25, or 4·5 μg/kg or a matched volume of placebo (sterile darbepoetin alfa vehicle), administered subcutaneously once weekly for 12 weeks. Randomization was performed using a central computerized system and was stratified to balance the treatment groups with respect to malignancy type (myeloma vs lymphoma). After 12 weeks of treatment, patients underwent a 4 week follow-up evaluation.
Patients received the first dose of study drug on the first day of their next cycle of chemotherapy after randomization, before chemotherapy and hydration. Weekly study drug doses were reduced by 50% for patients who had a ≥ 2·0 g/dl increase in haemoglobin during any 28 d period in the absence of RBC transfusion. Study drug was withheld for patients who had haemoglobin concentrations > 15·0 g/dl (for men) or > 14·0 g/dl (for women), and was reinstated at 50% of the previous weekly dose once haemoglobin concentrations decreased to≤13·0 g/dl. Red blood cell transfusions were recommended for patients with haemoglobin concentrations≤8·0 g/dl.
The objective of the study was to assess the dose–response relationship of darbepoetin alfa with respect to haemoglobin response. The proportion of patients achieving a haemoglobin response or a haematopoietic response and the overall change in haemoglobin during the treatment phase were evaluated. Haemoglobin response was defined as an increase in haemoglobin of ≥ 2·0 g/dl from baseline in the absence of RBC transfusions. Haematopoietic response was defined as either a haemoglobin response or an increase in haemoglobin concentration to ≥ 12·0 g/dl in the absence of RBC transfusions, which has been reported in recent studies of rHuEPO (Demetri et al, 1998; Gabrilove et al, 2001). Additionally, ‘sustained’ haemoglobin response was defined as a haemoglobin response that was maintained for 28 d or until the end of the treatment phase. The change in haemoglobin was measured after 4 weeks of treatment and after 12 weeks of treatment for values observed in the absence of a RBC transfusion in the previous 28 d. The incidence of RBC transfusions was assessed from week 5 to the end of the treatment phase, as data from studies of rHuEPO indicate that its effects on RBC transfusion requirements were not apparent until the second month of treatment (Abels, 1992; Gabrilove et al, 2001; Littlewood et al, 2001; Österborg et al, 2002). Haemoglobin concentrations were measured weekly throughout the study. During the treatment phase, measurements were taken on the day that the study drug was administered, before dosing. Information on RBC transfusions was monitored continually during the study.
The safety of darbepoetin alfa was evaluated by monitoring adverse events, excess increases in haemoglobin, changes in laboratory variables and vital signs, and antibody formation resulting from darbepoetin alfa administration.
Statistical analysis. All randomized patients who received at least one dose of study drug were included in the analyses of efficacy (as randomized; intent-to-treat analysis set) and safety. Patients who received placebo were combined to form a single placebo group.
Rates of haemoglobin and haematopoietic response were estimated using the Kaplan–Meier method. Logistic regression models were used to assess the treatment effect of darbepoetin alfa versus placebo, dose–response relationships and the effects of covariates. The effect of treatment was modelled using three orthogonal contrasts (Draper & Smith, 1981) representing the effect of darbepoetin alfa (darbepoetin alfa versus placebo), the linear dose effect of darbepoetin alfa and the quadratic dose effect of darbepoetin alfa. Effects of treatment and covariates were considered statistically significant if the P-value of the likelihood ratio test for the addition of the variable to the model was less than 0·05. The covariates evaluated included malignancy type (myeloma vs lymphoma), sex, baseline haemoglobin concentration (categorized as < 8·0 g/dl, 8·0 to < 9·0 g/dl, 9·0 to < 10·0 g/dl, 10·0 to < 11 g/dl or ≥ 11 g/dl), RBC transfusions during the 4 weeks before randomization and baseline serum endogenous erythropoietin concentration (categorized as≤100 U/l or > 100 U/l).
Sixty-six patients were randomized to receive darbepoetin alfa 1·0 μg/kg (n = 11), 2·25 μg/kg (n = 22), 4·5 μg/kg (n = 22) or placebo (n = 11). All patients received blinded study drug and were included in the efficacy and safety analyses.
Demographic characteristics were generally similar between groups, except for a higher proportion of women in the placebo group (Table I). Forty-eight patients (73%) had lymphoma (Hodgkin's disease, NHL or CLL) and 18 (27%) had MM. Patient baseline and disease characteristics were also, in general, well balanced (Table I). However, although absolute neutrophil counts and platelet counts for all treatment groups were within the reference range for this population, values were higher in the placebo group than in the darbepoetin alfa groups. Baseline serum ferritin and transferrin saturation values indicated that all patients were iron replete.
Two patients receiving darbepoetin alfa were withdrawn from the study because of a delay in chemotherapy and one patient randomized to receive placebo withdrew consent.
Effects on haemoglobin. The proportion of patients achieving a haemoglobin response increased with darbepoetin alfa dose, from 45% in the 1·0 μg/kg group to 55% in the 2·25 μg/kg group and 62% in the 4·5 μg/kg group (Fig 1). Only 10% of patients responded in the placebo group. The median time to response was observed at week 13 (starting study d 85; range: 1–13 weeks) in the darbepoetin alfa 2·25 μg/kg group and week 8 (starting study d 50; range: 3–11 weeks) in the 4·5 μg/kg group, but could not be estimated for the placebo or 1·0 μg/kg groups as less than 50% of patients responded. The proportion of patients who achieved a haemoglobin response was significantly higher for the darbepoetin alfa dose groups combined than for the placebo group (P < 0·01). Similarly, darbepoetin alfa was associated with a statistically significant increase in the proportion of patients achieving a sustained haemoglobin response or haematopoietic response compared with placebo (P < 0·01 for both parameters). Haematopoietic response rates also appeared to increase with increasing darbepoetin alfa dose (Table II). The dose–response trends observed for haemoglobin response and haematopoietic response were not found to be statistically significant.
Table II. Kaplan–Meier rates of haematopoietic response and red blood cell transfusions from week 5 to the end of the treatment phase.
Darbepoetin alfa (μg/kg/week)
RBC, red blood cell.
Number of patients
% Patients with haematopoietic response (95% CI)
% Patients with RBC transfusion (95% CI)
The mean change in haemoglobin from baseline at week 5 (starting study d 29) and week 13 (starting study d 85) is presented in Fig 2. At week 13, the mean change in haemoglobin increased from 1·56 g/dl (95% CI: −0·13, 3·25) in the 1·0 μg/kg group to 1·64 g/dl (95% CI: 1·05, 2·24) in the 2·25 μg/kg group and 2·46 g/dl (95% CI: 1·68, 3·24) in the 4·5 μg/kg group, compared with a mean change of 1·00 g/dl (95% CI: 0·55, 1·45) in the placebo group.
Effects on red blood cell transfusions. The proportion of patients receiving a RBC transfusion from week 5 (starting study d 29) to the end of the treatment phase is shown in Table II. During this period, 45% of patients in the placebo group, 27% in the darbepoetin alfa 1·0 and 2·25 μg/kg groups, and 15% in the 4·5 μg/kg group received a RBC transfusion. The treatment effect (darbepoetin alfa vs placebo) and dose–response effects were not statistically significant.
Effects of covariates. Patients who had received at least one RBC transfusion in the 4 weeks before randomization were less likely to achieve a haemoglobin response (P = 0·001) during treatment, and were more likely to receive a RBC transfusion during week 5 to the end of the treatment phase (P < 0·001), than those who had not received a transfusion before randomization. After adjusting for this covariate, the difference in the proportion of patients achieving a haemoglobin response between the darbepoetin alfa and placebo groups was still statistically significant (P < 0·01).
No statistically significant associations between baseline haemoglobin or baseline serum erythropoietin levels and haemoglobin response were observed; however, both covariates were associated with the proportion of patients receiving a RBC transfusion during week 5 to the end of treatment phase. Lower baseline haemoglobin and higher baseline serum erythropoietin level (> 100 U/l) were associated with a higher proportion of patients receiving a RBC transfusion (P < 0·01 and P = 0·01 respectively).
The stratification factor, malignancy type (myeloma versus lymphoma), was found to be associated with sustained haemoglobin response (P = 0·02). Patients with myeloma were more likely to achieve a sustained haemoglobin response compared with lymphoma patients. However, there was no evidence of malignancy type being associated with any of the other haemoglobin or transfusion endpoints.
Darbepoetin alfa administration. The proportion of patients requiring a dose reduction because of a > 2·0 g/dl increase in haemoglobin during any 28 d period in the absence of RBC transfusion was 27% in the 1·0 μg/kg group, 23% in the 2·25 μg/kg group and 41% in the 4·5 μg/kg group. The median weekly doses of darbepoetin alfa administered during the treatment period, adjusted for weight, were 0·98 μg/kg in the 1·0 μg/kg group, 2·20 μg/kg in the 2·25 μg/kg group and 4·40 μg/kg in the 4·5 μg/kg group.
The safety profile of darbepoetin alfa was similar to that of placebo. Fifty-two of 55 patients (95%) receiving darbepoetin alfa and 10 of 11 patients (91%) receiving placebo experienced at least one adverse event during the study, most of which were mild to moderate in severity. No relationship between darbepoetin alfa dose and the incidence or severity of adverse events was observed. The most frequently reported adverse events were nausea, fatigue, fever and vomiting (Fig 3). Most events were considered by the investigating physician to be attributable to the patients' cancer or chemotherapy and were consistent with those expected for a population of patients with lymphoproliferative malignancies. No deaths occurred during this study and there were no discontinuations of study medication as a result of adverse events.
Increases in haemoglobin concentrations were well controlled in patients receiving darbepoetin alfa. The maximum increase in haemoglobin during any 28 d period was 2·34 g/dl for patients receiving darbepoetin alfa and 2·85 g/dl for patients receiving placebo. Twenty-two patients (40%) receiving darbepoetin alfa and one patient (9%) receiving placebo experienced a protocol-defined rapid rise in haemoglobin (≥ 2 g/dl within a 28 d period). Additionally, two men (6%) and one woman (5%) in the darbepoetin alfa 4·5 μg/kg group exceeded the specified maximum haemoglobin concentration (15·0 g/dl for men and 14·0 g/dl for women) and had their study drug doses withheld. Haemoglobin concentrations for all patients remained within normal limits, and no clinical sequelae associated with a rapid rise in haemoglobin or maximum haemoglobin concentrations were observed.
Changes in laboratory measures and vital signs were similar between patients receiving darbepoetin alfa and placebo. The observed difference in absolute neutrophil counts and platelet counts between the placebo and darbepoetin alfa groups at baseline did not persist during the study, and values were more similar between groups at the end of treatment. No serum reactivity to darbepoetin alfa was observed in any of the individual serum samples collected at week 7 and week 16 from the 55 patients exposed to 633 patient weeks of darbepoetin alfa.
The results of this multicentre, randomized, double-blind, placebo-controlled, dose-finding study clearly indicated that darbepoetin alfa was effective and well tolerated when administered once weekly in patients with lymphoproliferative malignancies. A significantly higher proportion of patients receiving darbepoetin alfa achieved a haemoglobin response relative to patients receiving placebo, despite absolute neutrophil and platelet counts suggesting relatively greater myelosuppression in the darbepoetin alfa groups than in the placebo group at baseline. In addition, the need for RBC transfusions was lower in the darbepoetin alfa groups than in the placebo group, particularly for patients receiving 4·5 μg/kg. Darbepoetin alfa was also effective in achieving a sustained haemoglobin response. The haemoglobin response rates in this study ranged from 45% to 62% in the darbepoetin alfa dose groups and were similar to results of previous studies of rHuEPO in patients with lymphoproliferative malignancies, in which typical response rates of 50–60% were achieved (Cazzola et al, 1995; Glaspy et al, 1997; Dammacco et al, 2001; Österborg et al, 2002). Additionally, despite no dose increases for patients receiving darbepoetin alfa in this study, the haematopoietic response rates (ranging from 55% to 68%) appear to be comparable to published studies of rHuEPO, where rates of 61% and 68% have been reported when an increase from the original starting dose was allowed during treatment (Demetri et al, 1998; Gabrilove et al, 2001).
None of the dose trends that were observed for the haemoglobin and transfusion endpoints were found to be statistically significant; however, this may have been a result of the small number of patients in each of the darbepoetin alfa dose groups. The results suggested that the effects of darbepoetin alfa on haemoglobin response and other measures of efficacy (including reductions in RBC transfusions and changes in haemoglobin) increased with increasing doses, with the greatest effects observed at 4·5 μg/kg. These results are consistent with those observed in dose-finding studies of darbepoetin alfa in patients with solid tumours receiving chemotherapy, in which the proportion of patients achieving a haemoglobin response increased with increasing doses of 0·5–4·5 μg/kg/week (Glaspy et al, 2001) or 4·5–12·0 μg/kg once every 3 weeks (Kotasek et al, 2000). Additionally, a large, phase 3, randomized study in lung cancer patients receiving platinum-based chemotherapy showed that a darbepoetin alfa starting dose of 2·25 μg/kg/week had significantly greater efficacy compared with placebo for reducing the proportion of patients receiving a RBC transfusion and increasing the incidence of haemoglobin response and the change in haemoglobin from baseline (Pirker et al, 2001).
High haemoglobin response rates with rHuEPO have previously been observed in patients with myeloma, with response rates in excess of 75% reported in several studies (Barlogie & Beck, 1993; Ludwig et al, 1993; Silvestris et al, 1995; Dammacco et al, 1998). However, a direct comparison of response rates in patients with MM or other lymphoproliferative diseases (e.g. CLL or lymphoma) has been prospectively examined in only three published randomized trials (Cazzola et al, 1995; Österborg et al, 1996, 2002), all of which showed no statistically significant difference in response between these malignancy types. Additionally, a recent large phase 3 trial reported a haemoglobin response rate of 57·6% for patients with MM (Dammacco et al, 2001). Although our study was stratified to ensure a balance of patients with myeloma and lymphoma in each dose group, it was not designed to detect differences in response rates between these two malignancy types. Myeloma patients had an increased likelihood of achieving a sustained haemoglobin response, but no evidence of this effect was found for any of the other haemoglobin or transfusion endpoints.
Evaluation of the effects of other covariates in this study indicated that patients with low baseline haemoglobin or a high baseline endogenous erythropoietin level had an increased risk of transfusion while on study. These results confirm similar findings reported in previous studies of rHuEPO, in which baseline endogenous erythropoietin concentrations were suggested as a possible factor influencing haemoglobin response in haematological malignancies (Ludwig et al, 1990, 1994; Barlogie & Beck, 1993; Cazzola et al, 1995; Österborg et al, 1996). Indeed, most predictive models for achieving a haemoglobin response with rHuEPO therapy in patients with MM, NHL or CLL include baseline serum erythropoietin and haemoglobin levels as important prognostic factors (Cazzola et al, 1996; Beguin, 1998). Although the median baseline serum erythropoietin levels seen in our study were within the normal range expected for this population, levels exceeding 600 U/l were seen in three patients (all in the 2·25 μg/kg group), with values of 684, 925 and 1362 U/l. None of these patients achieved the haemoglobin response criteria, which affected response rates in this dose group. Future studies should be performed to define the characteristics at baseline and especially during early treatment that might identify patients likely to need higher doses of darbepoetin alfa or those who are unlikely to respond regardless of dose. Such an evaluation may enable the most cost-effective use of this agent.
The safety profile of darbepoetin alfa was consistent with that expected for a population of patients with lymphoproliferative malignancies who were receiving chemotherapy and did not suggest any safety concerns. The overall incidence of adverse events was similar between patients receiving darbepoetin alfa and placebo, and no relationship was evident between darbepoetin alfa dose and adverse event incidence or severity. No deaths occurred on study and no patient withdrew from the study because of an adverse event. Increases in haemoglobin were well controlled at all dose levels. No clinical sequelae were observed for patients who had doses withheld as a result of maximum haemoglobin concentrations, indicating no unwanted, sustained effects on erythropoiesis after the withdrawal of therapy. No antibody formation was detected for any patient in the study.
In summary, the results of this study indicated that darbepoetin alfa, administered once weekly at doses of 1·0, 2·25 and 4·5 μg/kg, was associated with greater effects on haemoglobin than placebo in patients with lymphoproliferative malignancies. An apparent increase in the effects of darbepoetin alfa on haemoglobin response rates and other efficacy measures was observed with increasing doses, with the greatest effects at 4·5 μg/kg. No safety concerns were associated with darbepoetin alfa in this population. Thus, darbepoetin alfa has demonstrated a beneficial treatment effect that supports its use in anaemic patients with lymphoproliferative malignancies who are receiving chemotherapy. Based on these results, confirmative studies of the efficacy and safety of darbepoetin alfa at doses of 2·25 and/or 4·5 μg/kg/week in this population are warranted.
This study was supported by Amgen Inc., Thousand Oaks, CA, USA. In addition to the authors, the Darbepoetin alfa 990114 Study Group includes the following individuals and study centres:
Nigel Russell, Department of Haematology, City Hospital, Nottingham, UK
Werner Linkesch, Medizinische Universitatsklinik Graz, Graz, Austria
Klaus Lechner, Universitatsklinik fur Inner Medizin 1, Wien, Austria
António Parreira, Instituto Portugues de Oncologia, Centro de Lisboa, Lisboa, Portugal
Ann Hunter, Department of Oncology, Leicester Royal Infirmary, Leicester, UK
Donald Milligan, Department of Haematology, Birmingham Heartlands Hospital, Birmingham, UK
Dieter Hoelzer, Klinikum der Johann Wolfgang Goethe, Universitaet Theodor-Stern-Kai 7, Frankfurt, Germany
Giorgio Lambertenghi, Unità Trapianti di Midollo Osseo, Ospedale Maggiore Policlinico, Milano, Italy
Martin Schipperus, University Hospital Rotterdam/Dijkzigt, Department of Haematology, Rotterdam, the Netherlands.