Initiation of epoetin-α therapy at a starting dose of 120,000 units once every 3 weeks in patients with cancer receiving chemotherapy

An open-label, multicenter study with randomized and nonrandomized treatment arms

Authors


Abstract

BACKGROUND:

Epoetin-α initiated once weekly, followed by once-every-3-weeks maintenance, was effective and well tolerated for chemotherapy-induced anemia. This study evaluated a starting dose of epoetin-α 120,000U once every 3 weeks for chemotherapy-induced anemia using early and late initiation regimens.

METHODS:

Patients with baseline hemoglobin 11.0-12.0 g/dL were randomly assigned to early intervention with immediate epoetin-α (n = 68) or to standard intervention with epoetin-α when hemoglobin decreased to <11 g/dL (n = 68). A third group of patients with baseline hemoglobin <11 g/dL (n = 50) were enrolled but not randomized; epoetin-α was initiated immediately. The primary endpoint was mean proportion of hemoglobin values within the target range (11.0-13.0 g/dL) among randomized patients.

RESULTS:

The mean proportion of hemoglobin values in range through week 16 was 60% in each randomized group. Mean hemoglobin by week showed similar increases over the study. Blood transfusions were administered in 9%, 8%, and 24% of patients in the early, standard, and nonrandomized groups. Mean epoetin-α doses were similar between treatment groups. Dose reductions and withholds were more common in the early intervention group. Adverse events (eg, diarrhea, fatigue, nausea) were consistent with the safety profile for epoetin-α . Clinically relevant thrombotic vascular events (regardless of relationship to study treatment) were reported for 9%, 12%, and 12% of patients in the early, standard, and nonrandomized groups.

CONCLUSIONS:

Early and standard intervention with epoetin-α, administered once every 3 weeks, increased and maintained hemoglobin levels within 11.0-13.0 g/dL in patients with chemotherapy-induced anemia. Cancer 2009. © 2009 American Cancer Society.

Anemia is a common complication of myelosuppressive chemotherapy that may affect patient function and quality of life.1 Recombinant human erythropoietin (epoetin-α) is indicated for the treatment of anemia in patients with nonmyeloid malignancies for which anemia is due to the effect of concomitantly administered chemotherapy.2, 3 Numerous clinical trials have confirmed the efficacy of epoetin-α for increasing hemoglobin (Hb) and decreasing blood transfusions in cancer chemotherapy patients.4, 5

Most clinical trials have evaluated the administration of epoetin-α 3 times weekly or once weekly (qw). However, an analysis of 4 observational studies of actual prescribing patterns for erythropoiesis-stimulating agents in the treatment of chemotherapy-induced anemia (CIA) reported that between 22% and 41% of patients were given epoetin-α once every 2 weeks (q2w) and 3% to 9% of patients were given epoetin-α once every 3 weeks (q3w).6 A prospective, randomized comparison of epoetin-α 80,000 U q2w versus 40,000 U qw in the management of CIA reported similar improvement in Hb in both treatment groups.7 The majority of patients in that study had a Hb increase of at least 2 g/dL or achieved a Hb level of at least 12 g/dL. Recent prospective trials reported that initiating epoetin-α at a dose of 40,000 or 60,000 U qw, and then extending the dose regimen to 80,000 or 120,000 U q3w, increased Hb and decreased transfusion requirements.8-10 Initiating therapy with a q3w dose may be more convenient for the patient than starting with qw dosing and then extending the interval. To our knowledge, studies of an initial dose of epoetin-α q3w in the setting of CIA have not been reported.

The objective of this study was to evaluate the efficacy, including maintenance of Hb between 11.0-13.0 g/dL, and safety of epoetin-α at an initial dose of 120,000 U q3w in cancer patients with CIA. Patients with baseline Hb of 11.0 to 12.0 g/dL were randomly assigned to either early intervention with immediate initiation of epoetin-α q3w, or standard intervention with initiation of epoetin-α q3w when Hb decreased to <11.0 g/dL. After the study was under way, it became evident that enrolling only patients with baseline Hb between 11.0 and 12.0 g/dL excluded several patients who might benefit from treatment. In addition, it was assumed that enrollment of patients with Hb less than 11 g/dL would be more reflective of actual clinical practice, based on published observational evidence that mean Hb levels are less than 10 g/dL before erythropoiesis-stimulating therapy for CIA.11 Therefore, the protocol was amended to allow enrollment of patients with Hb <11.0 g/dL and allocation of these patients to immediate initiation of treatment with epoetin-α q3w.

MATERIALS AND METHODS

Eligibility Criteria

This randomized, open-label study was conducted at 24 centers in the United States. Men and women ≥18 years of age were eligible for randomization if they had a histologically confirmed nonmyeloid malignancy and a baseline Hb of ≥11.0 g/dL and ≤12.0 g/dL. Patients who met the other study criteria but had a baseline Hb <11.0 g/dL were enrolled to a nonrandomized treatment group. All patients provided informed written consent to participate.

All patients were required to receive chemotherapy at week 1. Chemotherapy cycles were to be administered for a minimum of 9 weeks during the study. Patients were also required to have adequate hematologic, renal, and hepatic function; iron transferrin saturation (TSAT) >20%; life expectancy of ≥6 months; and Eastern Cooperative Oncology Group (ECOG) Performance Status score of 0-2.

Key study exclusion criteria were diagnosis of a myeloid malignancy or known history of myelodysplasia, planned nonpalliative radiation during the study, current anemia due to factors other than cancer/chemotherapy, treatment with an erythropoiesis-stimulating agent within the previous 3 months, untreated central nervous system metastases, uncontrolled hypertension, seizures, active second malignancy (except for basal cell carcinoma or carcinoma in situ), and transfusion of platelets or packed red blood cells within the prior 28 days. Patients also were excluded if they had a history of uncontrolled cardiac arrhythmias (within 6 months), pulmonary emboli, deep vein thrombosis, ischemic stroke, chronic hypercoagulable disorders, or other arterial or venous thrombotic events.

The study protocol and amendments were reviewed by an Institutional Review Board. Patients (or their legally acceptable representatives) signed an informed consent document before participation in the study. The study was conducted in accordance with the provisions of the Declaration of Helsinki and its amendments, as well as the International Conference on Harmonization Good Clinical Practice guidelines.

Interventions

A computer-generated randomization schedule was prepared by the sponsor before the study. The randomization was balanced using randomly permuted blocks. All study treatment was administered without concealment.

Epoetin-α was administered subcutaneously by a health care professional. The initial dosing regimen was 120,000 U q3w in each treatment group. Subsequent doses were to be administered within ±72 hours of the q3w interval. In the early intervention group, epoetin-α therapy was initiated at week 1 when Hb was 11.0 to 12.0 g/dL. In the standard intervention group, initiation of epoetin-α therapy was delayed until Hb decreased to <11.0 g/dL. In the nonrandomized group, all patients began study with Hb <11.0 g/dL and epoetin-α therapy was initiated at week 1.

If Hb rose to >12.0 g/dL or increased by >1.5 g/dL in a 3-week period during epoetin-α therapy, the dose of epoetin-α was reduced according to the following sequence: from 120,000 U q3w to 80,000 U q3w; from 80,000 U q3w to 60,000 U q3w; and from 60,000 U q3w to 40,000 U q3w. If at any scheduled visit during the study Hb was >13.0 g/dL, epoetin-α was withheld until Hb fell to <12.0 g/dL, at which time epoetin-α could be resumed according to the dose-reduction schedule defined above. Transfusion-related increases in Hb were not to be used as the basis for an epoetin-α dose decrease, but epoetin-α was always to be held for Hb >13.0 g/dL, regardless of transfusion. (The study was designed and conducted before recent label changes recommending initiation of therapy in patients with chemotherapy-induced anemia only if Hb is <10 g/dL.2, 3, 12)

If Hb dropped by ≥1 g/dL after dose reduction, the previous dose of epoetin-α was restarted. If Hb at any scheduled visit was <10.0 g/dL after at least 1 dose of epoetin-α, the patient was considered to have failed q3w therapy and was treated with epoetin-α 40,000 U qw starting at that visit. If Hb had not risen by ≥1 g/dL after 4 weeks at this dose, the dose of epoetin-α was increased to 60,000 U qw. If Hb had not increased by ≥1 g/dL from baseline after 4 weeks at 60,000 U qw, the patient was considered to have failed qw therapy.

Treatment with epoetin-α was to continue until week 16 (ie, the last dose could not be administered after week 15), or through the final cycle of chemotherapy, whichever came first. All patients were to be followed weekly for Hb, hematocrit, and blood pressure measurements until 3 weeks after the last dose of epoetin-α, or up to a maximum of 16 weeks on study, whichever came first. Patients who received the minimum 9 weeks of chemotherapy were followed weekly for 3 weeks after chemotherapy to be considered completers.

Unless there was a contraindication to iron therapy, patients were to receive ferrous sulfate 325 mg orally once daily, an equivalent oral formulation of iron as tolerated, or an equivalent intravenous formulation of iron. Transfusions could be administered throughout the study when Hb fell below a “transfusion trigger” that was predefined for each site per local transfusion policy.

Data Collection

The entire study period was up to 22 weeks, with a screening phase lasting up to 2 weeks, study treatment for up to 16 weeks, and safety follow-up for 4 weeks. Blood pressure, Hb, and hematocrit were measured once weekly. A quality-of-life assessment, the Functional Assessment of Cancer Therapy-Fatigue (FACT-F) scale,13 was administered at study baseline, just before the first dose of epoetin-α (in the standard intervention group only), and at weeks 7 and 16. Assessments performed at baseline and week 16 (or the time of early study completion) included iron indices, folate, and serum vitamin B12. Transfusion data recorded included number of units, date, type, pretransfusion and posttransfusion Hb, and reason for transfusion. Adverse event information was recorded at weekly study visits.

Statistical Analysis

Demographics and baseline characteristics were analyzed for all enrolled patients. Analyses of efficacy endpoints were performed using the modified intent-to-treat (mITT) population, which included all patients who had at least 1 postbaseline Hb value. Safety results were summarized for the safety population, which included all patients who received at least 1 study dose.

There were no statistical comparisons between treatment groups. Continuous variables were summarized by treatment group using descriptive statistics. Categorical variables were summarized by treatment group using frequency statistics.

The primary efficacy endpoint was “percent values in range,” defined as the mean proportion of weekly Hb values during chemotherapy treatment that were ≥11.0 g/dL and ≤13.0 g/dL among randomized patients. The primary efficacy endpoint was not evaluated in the nonrandomized group. The secondary efficacy endpoint of “complete success” was defined as maintenance of all weekly Hb values during epoetin-α treatment between 11.0 and 13.0 g/dL, beginning at week 1 in the early intervention group and once Hb was ≥11.0 g/dL in the standard intervention group. Another secondary efficacy endpoint was the proportion of patients whose mean weekly Hb level was between 11.0 and 13.0 g/dL.

Secondary efficacy endpoints included mean Hb values by week, mean scores on the FACT-F scale, and blood transfusion data after week 4 of epoetin-α treatment, including the proportion of patients who received at least 1 transfusion, the number of units transfused, and the pretransfusion Hb. Summaries of weekly Hb values were performed using observed data as well as the last-observation-carried-forward (LOCF) method to estimate missing values and to impute values for 28 days after a transfusion. For both methodologies, Hb values collected from patients after a switch in dosing from q3w to qw treatment were excluded.

Adverse events were summarized by severity and possible relationship to study treatment, as well as discontinuation of study treatment due to adverse event. Each patient contributed only once for multiple occurrences of the same event. Other safety endpoints included the proportion of patients with an Hb increase of ≥1 g/dL within a 2-week period and an increase of ≥1.5 g/dL within a 2-week period. Special attention was given to all thrombotic vascular events and thrombotic vascular events considered clinically relevant; the incidences of these events were determined overall, among patients with rapid Hb rise (ie, ≥1 g/dL in a 2-week period), and among patients with elevated Hb (ie, >13 g/dL).

RESULTS

Study Participants

The study was conducted between June 2005 and August 2006. Disposition of study participants is summarized in Figure 1. Of the 364 potential participants who were screened, 186 were enrolled. There were 136 patients who had baseline Hb between 11.0 and 12.0 g/dL and were randomly assigned to the early intervention group (n = 68) or the standard intervention group (n = 68), and 50 patients who had baseline Hb <11.0 g/dL and were enrolled directly to the nonrandomized group. The mITT population included 68 (100%) patients in the early intervention group, 67 (99%) patients in the standard intervention group, and 49 (98%) patients in the nonrandomized group who had at least 1 Hb value recorded after baseline.

Figure 1.

Patient disposition. mITT, modified intent-to-treat.

Seventeen (25%) patients in the standard intervention group never received epoetin-α because their Hb remained 11.0 g/dL or greater throughout the study. The remaining 169 study participants received at least 1 dose of epoetin-α and were included in the safety analyses. There were 12 (18%) patients in the early intervention group, 15 (22%) patients in the standard intervention group, and 18 (36%) patients in the nonrandomized group who discontinued study treatment early. The leading reasons for early discontinuation were patient request, adverse event, noncompliance with the dosing regimen, and stopping chemotherapy early.

Demographic and clinical characteristics of the safety population were similar between the treatment groups at baseline (Table 1). Mean age was between 60 and 62 years in each group. Patients were predominantly female and white. The most common tumor types were breast, gastrointestinal, and lung. Most patients had ECOG performance status of 0 or 1 at baseline, and the majority had metastases. The mean Hb level at baseline was 11.5 g/dL in each of the randomized treatment groups and 10.1 g/dL in the nonrandomized group.

Table 1. Patient Demographics (All Patients)
 Early InterventionStandard InterventionNonrandomized
  1. ECOG indicates Eastern Cooperative Oncology Group; Hb, hemoglobin; TSAT, transferrin saturation.

No. of patients686850
Age, y   
 Mean (SD)60.5 (12.8)61.3 (15.4)61.9 (13.4)
 (Min, Max)(23.0, 85.0)(23.0, 85.0)(33.0, 89.0)
Sex, no. (%)   
 Women45 (66)49 (72)33 (66)
 Men23 (34)19 (28)17 (34)
Race, no. (%)   
 White56 (82)50 (74)35 (70)
 Hispanic/Latino7 (10)11 (16)9 (18)
 Black3 (4)3 (4)4 (8)
 Asian2 (3)4 (6)2 (4)
Primary malignancy, no. (%)   
 Breast17 (25)29 (43)12 (24)
 Gastrointestinal15 (22)10 (15)14 (28)
 Lung9 (13)9 (13)7 (14)
 Hematologic7 (10)7 (10)10 (20)
 Gynecologic12 (18)5 (7)3 (6)
 Genitourinary2 (3)5 (7)1 (2)
 Sarcoma2 (3)0 (0)0 (0)
 Head and neck1 (2)1 (2)0 (0)
 Other3 (4)2 (3)3 (6)
ECOG Performance Status, no. (%)   
 029 (43)21 (31)17 (34)
 136 (53)40 (59)29 (58)
 23 (4)7 (10)4 (8)
Metastases present, no. (%)41 (60)34 (50)26 (52)
Baseline Hb (g/dL), mean (SD)11.5 (0.3)11.5 (0.4)10.1 (0.6)
Baseline ferritin (μg/L), mean (SD)298.1 (293.4)272.8 (226.1)425.4 (384.1)
Baseline TSAT (%), mean (SD)0.28 (0.165)0.31 (0.217)0.28 (0.208)

Treatments Administered

There were 16 (24%) patients in the early intervention group, 17 of 51 (33%) treated patients in the standard intervention group, and 20 (40%) patients in the nonrandomized group who were not responsive to q3w treatment and switched to qw dosing. Among the patients who received at least 1 dose of epoetin-α, the mean (SD) cumulative dose of both q3w and qw doses combined was 421,177 (147,936) U in the early intervention group, 415,686 (148,395) U in the standard intervention group, and 442,000 (187,453) U in the nonrandomized group.

Ninety-three percent of study participants received some form of concomitant iron supplementation, most commonly bivalent oral iron preparations (90% of patients) such as ferrous sulfate (55% of patients); no patient received intravenous iron therapy.

For patients who switched from q3w dosing to qw dosing, postswitch data were included in the analyses of dose reductions and withholds. One or more dose reductions was required during the study for 30 (44%) patients in the early intervention group, 13 of 51 (25%) treated patients in the standard intervention group, and 12 (24%) patients in the nonrandomized group. Likewise, withholds tended to occur more commonly in the early intervention group than in the other groups. Epoetin-α was withheld at least once for 31 (46%) patients in the early intervention group, 17 of 51 (33%) treated patients in the standard intervention group, and 17 (34%) patients in the nonrandomized group.

Hemoglobin

Analysis of weekly Hb values using observed data (excluding values for 28 days after transfusion) showed a gradual increase in Hb values over time (Fig. 2A,B). Similar increases were also seen with LOCF methodology.

Figure 2.

Mean hemoglobin level by study week–observed-case approach (modified intent-to-treat [mITT] Population). (A) Randomized patients by treatment group; (B) nonrandomized patients. CI, confidence interval; Grp, group.

The mean values for the primary study endpoint—the proportion of weekly Hb values during chemotherapy treatment in the target range of 11.0 to 13.0 g/dL—were 60% (95% confidence interval [CI], 53% to 67%) in the early intervention group and 60% (95% CI, 53% to 67%) in the standard intervention group. Complete success, which was defined as maintaining all Hb values between 11.0 and 13.0 g/dL during treatment with epoetin-α q3w alone, was achieved by 4 (6%) patients in the early intervention group. In the standard intervention group, 41 patients achieved at least 1 Hb value ≥11.0 g/dL during epoetin-α q3w therapy, and 5 (12%) of these patients had complete success thereafter.

There were 49 (72%) patients in the early intervention group who had a mean Hb level between 11.0 and 13.0 g/dL. Of the 41 standard intervention patients who received epoetin-α therapy and achieved Hb ≥11.0 g/dL, 28 (68%) patients maintained a mean Hb level between 11.0 and 13.0 g/dL during epoetin-α therapy.

Mean Hb remained between 11.0 and 12.0 g/dL at every week of study in each of the randomized groups, and between 10.0 and 10.5 g/dL at every week in the nonrandomized group. Among the 51 patients whose Hb fell below 11.0 g/dL in the standard intervention group, Hb decreased to a mean of 10.4 g/dL before initiation of epoetin-α and the subsequent mean increase in Hb was 0.7 g/dL at the final visit on q3w therapy. Secondary analyses of Hb were performed separately for patients who did not switch to qw dosing during the study. Using observed data in this population, mean Hb increased significantly (P < .05) from baseline at weeks 3 through 16 in the early intervention group (n = 52), at weeks 11 through 16 in the standard intervention group (n = 50, including patients who received only q3w dosing and those who never received epoetin-α), and at weeks 2 through 16 in the nonrandomized group (n = 29).

Blood Transfusions

Use of blood transfusions is summarized in Table 2. At least 1 blood transfusion was given after 4 weeks of treatment for 6 (9%) of 68 treated patients in the early intervention group, 4 (8%) of 51 treated patients in the standard intervention group, and 12 (24%) patients in the nonrandomized group. The mean number of units transfused after 4 weeks of treatment was 2.3 U in the early intervention group, 6.0 U in the standard intervention group (1 of these patients received 12 U of blood), and 3.3 U in the nonrandomized group. Mean Hb before transfusion after 4 weeks on treatment was 8.2 g/dL in the early intervention group, 8.3 g/dL in the standard intervention group, and 7.8 g/dL in the nonrandomized group.

Table 2. Transfusion of Packed Red Blood Cells (mITT Population) After 4 Weeks on Treatment
 Early InterventionStandard InterventionNonrandomized
  • mITT indicates modified intent-to-treat; Hb, hemoglobin.

  • *

    Excludes patients who did not receive epoetin-α.

No. of patients6851*49
Patients with ≥1 transfusion, no. (%)6 (9)4 (8)12 (24)
No. of units transfused, mean (SD)2.3 (0.8)6.0 (4.9)3.3 (2.0)
Hb (g/dL) prior to transfusion, mean (SD)8.2 (0.2)8.3 (0.6)7.8 (0.7)

Quality of Life

Mean (± SD) values for FACT-F at baseline were 33.5 ± 13.2 in the early intervention group, 27.8 ± 12.0 in the standard intervention group, and 29.3 ± 12.4 in the nonrandomized group; at the final visit, the mean (± SD) values were 32.0 ± 13.2, 30.4 ± 11.7, and 28.1 ± 14.1, respectively.

Safety

All 169 patients who received epoetin-α were included in the safety analyses. An Hb increase of at least 1.0 g/dL within a 2-week period was observed in 57 (84%) patients in the early intervention group, 45 (88%) patients in the standard intervention group, and 34 (68%) patients in the nonrandomized group. An increase of 1.5 g/dL or more within a 2-week period was observed in 38 (56%) patients in the early intervention group, 31 (61%) patients in the standard intervention group, and 21 (42%) patients in the nonrandomized group. There were 28 (41%) patients in the early intervention group, 15 (29%) in the standard intervention group, and 11 (22%) in the nonrandomized group who had at least 1 Hb greater than 13.0 g/dL.

Most patients (97% in the early intervention group, 98% in the standard intervention group, and 98% in the nonrandomized group) experienced at least 1 adverse event during the study, regardless of association with study treatment. Adverse events experienced by at least 10% of patients are displayed in Table 3.

Table 3. Adverse Events Experienced by at Least 10% of Patients in Any Treatment Group (Safety Population)
 Early Intervention No. (%)Standard Intervention No. (%)Nonrandomized No. (%)
No. of patients685150
Adverse event   
Diarrhea27 (40)13 (26)18 (36)
Fatigue18 (26)13 (26)22 (44)
Nausea20 (29)18 (35)15 (30)
Neutropenia19 (28)18 (35)13 (26)
Vomiting14 (21)12 (24)13 (26)
Peripheral edema12 (18)12 (24)13 (26)
Pyrexia14 (21)8 (16)13 (26)
Constipation15 (22)12 (18)12 (24)
Cough14 (21)9 (13)8 (16)
Thrombocytopenia12 (18)7 (10)9 (18)
Abdominal pain11 (16)6 (9)7 (14)
Insomnia10 (15)12 (18)5 (10)
Dehydration10 (15)9 (13)8 (16)
Pain10 (15)4 (6)3 (6)
Anorexia9 (13)11 (16)9 (18)
Rash9 (13)8 (12)5 (10)
Arthralgia9 (13)6 (9)4 (8)
Depression9 (13)6 (9)2 (4)
Pharyngolaryngeal pain9 (13)5 (7)3 (6)
Dyspnea8 (12)11 (16)6 (12)
Pain in extremity8 (12)6 (9)1 (2)
Neuropathy peripheral8 (12)3 (4)2 (4)
Headache7 (10)5 (7)3 (6)
Paraesthesia7 (10)5 (7)2 (4)
Shoulder pain7 (10)1 (2)4 (8)
Dizziness6 (9)12 (18)8 (16)
Leukopenia6 (9)7 (10)7 (14)
Asthenia6 (9)7 (10)5 (10)
Upper respiratory tract infection5 (7)11 (16)2 (4)
Hypokalemia5 (7)6 (9)6 (12)
Bone pain5 (7)6 (9)5 (10)
Hypotension5 (7)5 (7)6 (12)
Hypomagnesemia5 (7)1 (2)5 (10)
Hypoesthesia4 (6)8 (12)2 (4)
Back pain3 (4)11 (16)7 (14)
Alopecia3 (4)9 (13)5 (10)
Non-cardiac chest pain3 (4)7 (10)4 (8)
Anemia3 (4)2 (3)8 (16)
Urinary tract infection2 (3)7 (10)4 (8)
Dyspepsia2 (3)5 (7)6 (12)
Peripheral sensory neuropathy1 (2)0 (0)5 (10)

Of the 30 drug-related adverse events reported, 22 were considered possibly related to study drug, 4 were considered probably related to study drug, and 4 (pain, injection-site pain, bone pain, and deep vein thrombosis) were considered very likely related to study drug. Three adverse events of disease progression resulted in discontinuation of epoetin-α treatment, but no treatment-related adverse event resulted in discontinuation.

Nine patients died during the study, including 2 (3%) in the early intervention group, 2 (4%) in the standard intervention group, and 5 (10%) in the nonrandomized group (3 progressive cancer, 1 sepsis syndrome, 1 pneumonia). None of the deaths was considered related to study treatment. Table 4 outlines causes of death for those patients who died.

Table 4. Causes of Death
Age, ySexTumor TypeCumulative Epoetin Alfa Dose (U)Cause of Death
Early intervention group    
 61WomenGastric280,000Progressive disease
 75MenColorectal320,000Acute respiratory failure
Standard intervention group    
 59MenColorectal360,000Gastrointestinal hemorrhage
 63MenNon-small cell lung240,000Cardiopulmonary arrest
Nonrandomized group    
 80MenPancreatic380,000Progressive disease
 50MenPancreatic240,000Progressive disease
 59MenMultiple myeloma160,000Pneumonia
 61MenNon-small cell lung120,000Progressive disease
 80MenNon-Hodgkin lymphoma120,000Sepsis

There were 15 (22%) patients in the early intervention group, 15 (29%) patients in the standard intervention group, and 15 (30%) patients in the nonrandomized group who had at least 1 serious adverse event. Only 1 serious adverse event (pulmonary artery thrombosis in the early intervention group) was considered related to study treatment.

There were 6 (9%) patients in the early intervention group, 6 (12%) patients in the standard intervention group, and 6 (12%) patients in the nonrandomized group who had at least 1 clinically relevant thrombotic vascular event (Table 5). Three patients had thrombotic vascular events (2 with deep vein thrombosis and 1 with both deep vein thrombosis and pulmonary artery thrombosis) that were considered possibly, probably, or very likely related to study drug. No thrombotic vascular event led to study withdrawal.

Table 5. Clinically Relevant Thrombotic Vascular Adverse Events (Safety Population)
Patients Experiencing Adverse EventEarly Intervention No. (%)Standard Intervention No. (%)Nonrandomized No. (%)
  1. NOS indicates not otherwise specified.

No. of patients685150
At least 1 clinically relevant thrombovascular event6 (9)6 (12)6 (12)
Myocardial infarction0 (0)0 (0)2 (4)
Pulmonary artery thrombosis1 (1)0 (0)0 (0)
Deep vein thrombosis3 (4)3 (6)4 (8)
Pulmonary embolism1 (1)3 (6)0 (0)
Thrombosis NOS1 (1)1 (2)0 (0)
Jugular vein thrombosis1 (1)0 (0)0 (0)

Table 6 presents the incidence of clinically relevant thrombotic vascular events among patients with a rapid Hb rate of rise (ie, ≥1 g/dL in a 2-week period) or an elevated Hb (ie, >13 g/dL) All events that occurred during the study or within 4 weeks of study completion or withdrawal were included in this analysis, regardless of temporal or causal association with a rapid Hb rate of rise (n = 114) or an elevated Hb (n = 54). There was no evidence of a consistent effect of Hb elevation on the incidence of clinically relevant thrombotic vascular events in any treatment group (early, standard, nonrandomized) among patients with a rapid Hb rate of rise (12%, 5%, 12%), without a rapid Hb rate of rise (0%, 18%, and 12%), with an elevated Hb (4%, 7%, 9%), or without an elevated Hb (13%, 17%, 13%).

Table 6. Incidence of Clinically Relevant Thrombotic Vascular Events Among Patients with a Rapid Hb Rate of Rise or an Elevated Hb* (Safety Population)
 Early Intervention No. (%)Standard Intervention No. (%)Nonrandomized No. (%)
  • HB indicates hemoglobin; CRTVE, clinically relevant thrombotic vascular event.

  • *

    Hb changes could have occurred either before or after the event and the event could have occurred up to 4 weeks after study completion/withdrawal.

No. of patients685150
Rapid Hb rate of rise (≥1g/dL in a 2-wk period)   
 With CRTVE / rate of rise ≥1g/dL6/51 (12%)2/38 (5%)3/25 (12%)
 With CRTVE / rate of rise <1g/dL0/17 (0%)4/13 (18%)3/25 (12%)
Elevated Hb (>13 g/dL)   
 With CRTVE / Hb >13 g/dL1/28 (4%)1/15 (7%)1/11 (9%)
 With CRTVE / Hb ≤13 g/dL5/40 (13%)6/36 (17%)5/39 (13%)

DISCUSSION

This investigational study was the first prospective evaluation of an initial dose of epoetin-α on a q3w dosing schedule in the management of CIA. Randomization to early intervention (initiation of epoetin-α q3w when Hb was between 11.0 and 12.0 g/dL) or standard intervention (initiation of epoetin-α q3w when Hb decreased below 11.0 g/dL) resulted in similar safety and efficacy among patients receiving epoetin-α 120,000 U q3w. In both groups, a mean proportion of 60% of Hb values were in the study-specified target range of 11.0 to 13.0 g/dL during chemotherapy treatment and both approaches resulted in maintenance of mean Hb within this range every week of study. On-study average Hb values during epoetin-α therapy were within this range for most patients. Some patients were able to achieve Hb within this range throughout treatment with epoetin-α q3w.

Weekly Hb values increased over time when all observed data were included in the analysis. Mean Hb did not increase significantly from baseline when the data were analyzed with LOCF methodology, but significant improvement was observed in the subgroup of patients who never switched to qw dosing.

Given the low rates of transfusion in the randomized patients, consistent control of Hb in this study was achieved with epoetin-α q3w, as reported previously in studies of qw dosing followed by q3w dosing8-10 and studies that used more frequent administration of epoetin-α throughout treatment.4, 5 Blood transfusion in the present study was more common among the nonrandomized patients (24%) than the randomized patients (8% to 9%). This finding was consistent with the baseline Hb levels in each group, which were approximately 1.5 g/dL lower in the nonrandomized group than in the randomized groups. One could argue that the higher rate of transfusion in the nonrandomized patients was expected based on the study design because it isolated a group of patients who were more anemic even before they received chemotherapy on study. Heavy pretreatment with chemotherapy before enrollment or other reasons to be anemic, such as advanced disease with severe inflammation, may have contributed to more severe baseline anemia. However, the sample size is too small to be definitive. The patients who were randomized to the standard intervention group were followed closely and epoetin-α therapy was begun when their Hb level dropped below 11 g/dL, whereas those in the nonrandomized group were initiated on epoetin-α therapy with an Hb level already below 11 g/dL.

The safety of epoetin-α in the present study was consistent with product labeling for erythropoiesis-stimulating agents in cancer chemotherapy patients.2, 3, 12 Epoetin-α was generally well tolerated with an initial q3w dose. Three patients had thrombotic vascular events that were considered related to study treatment, including 1 serious treatment-related event, but no patient discontinued therapy due to a thrombotic vascular event.

An analysis of thrombotic vascular events among patients with rapid rise in Hb (≥1 g/dL in a 2-week period) or a single elevated Hb level (≥13 g/dL) did not demonstrate a consistent association between elevations in Hb and the risk of thrombotic vascular events in this study. These findings are consistent with previous studies that documented no increased risk of thrombotic events with the use of erythropoiesis-stimulating agents for chemotherapy-induced anemia when Hb is <11 g/dL.4 However, data from other clinical trials have reported the potential risks of increased tumor progression and decreased overall survival associated with higher target Hb levels for therapy with erythropoiesis-stimulating agents in anemic cancer patients.2, 3, 12 The present study was designed and conducted before much of this research was published. Progression and survival were not assessed as safety endpoints in the present study. Numerically, there were more deaths in the nonrandomized arm than the other 2 arms. As discussed above, this finding may reflect that the patients in the nonrandomized arm were different from the patients in the randomized groups, in that they were sicker at baseline given their lower baseline hemoglobin and therefore inherently more likely to have an adverse outcome over time. The lack of a randomized comparator arm and a small sample precludes us from drawing any conclusions about the deaths in this study.

Potential study limitations included the lack of a double-blind comparator group or comparative statistical analyses, and the amendment to the study design after some of the patients had been enrolled. A double-blind comparator group and comparative statistical testing were not included in the study design because previous studies had demonstrated the hematologic efficacy of epoetin-α relative to placebo in the management of chemotherapy-induced anemia.4, 5 This study was amended after enrollment had begun when it was determined that many patients who might benefit from treatment were being excluded from enrollment because they to Hb levels <11 g/dL at screening. The addition of this third treatment group provided the study with much greater applicability to clinical practice because chemotherapy-induced anemia often presents with Hb levels <11 g/dL, but it also may have introduced some bias by including a group of patients who might have been inherently different as described above.

This study demonstrated that early and standard intervention with epoetin-α every 3 weeks increased and maintained Hb levels within 11.0-13.0 g/dL in anemic patients with cancer receiving concomitant chemotherapy. Patients were treated to a target Hb range of 11.0-13.0 g/dL based on guidelines in place at the time of protocol development.14, 15 Recently revised product labeling for erythropoiesis-stimulating agents in the United States now specifies that therapy should not be initiated in patients with Hb ≥10 g/dL and doses should be titrated for each patient to achieve and maintain the lowest Hb level sufficient to avoid the need for blood transfusion.2, 3, 12 In addition, ESA administration should be discontinued following the completion of a chemotherapy course.

Acknowledgements

The authors wish to thank Jonathan N. Latham and Linda Packard Nash for their assistance in the preparation of this manuscript. The study investigators were as follows (listed alphabetically): Heather J. Allen, MD, Las Vegas, NV; John Anagnost, MD, Wilmington, NC; John Jay Baker, MD, Lewisburg, WV; Linda Bosserman, MD, La Verne, CA; Robert Carroll, MD, Gainesville, FL; Veena Charu, MD, Anaheim, CA; Sheldon Davidson, MD, Northridge, CA; Robert A. Dichmann, MD, Santa Maria, CA; Peter Eisenberg, MD, Greenbrae, CA; Craig Englund, MD, Inverness, FL; Sunil Gandhi, MD, Lecanto, FL; George Geils, MD, Charleston, SC; John Glaspy, MD, Los Angeles, CA; Lucio N. Gordan, MD, Gainesville, FL; Eddie Hu, MD, Alhambra, CA; Frederic C. Kass, MD, Santa Barbara, CA; Andre Liem, MD, Long Beach, CA; Richard E. Lloyd, MD, Fullerton, CA; Rosemary E. McIntyre, MD, Oxnard, CA; Ravi Patel, MD, Bakersfield, CA; Attique Samdani, MD, Richmond, VA; Christopher Seidler, MD, Worcester, MA; Mukund Shah, MD, Lancaster, CA; and Julie A. Taguchi, MD, Santa Barbara, CA. Planned interim results of this study were presented at the 2006 American Society of Hematology Annual Meeting (abstract 3774). Final results were presented at the 2007 American Society of Clinical Oncology Annual Meeting (abstracts 19564 and 19589).

Conflict of Interest Disclosures

The study was supported by Centocor Ortho Biotech Services, LLC (Study PR04-27-018). This protocol is identified on the http://www.clinicaltrials.gov Web site as NCT00255749.

Ancillary