- © 2002 by American Society of Clinical Oncology
Randomized, Controlled, Dose-Range Study of Ro 25-8315 Given Before and After a High-Dose Combination Chemotherapy Regimen in Patients With Metastatic or Recurrent Breast Cancer Patients
- P. Viens,
- C. Chabannon,
- P. Pouillard,
- M. Janvier,
- W. Brugger,
- J. Y. Blay,
- F. Oberling,
- R. Capdeville,
- C. Newman,
- V. Méresse,
- Z. X. Xu,
- E. Platzer,
- P. Van der Auwera and
- D. Maraninchi
- From the Department of Oncology, Institut Paoli-Calmettes, Marseille; Institut Curie and Quintiles, Paris; Centre René Huguenin, St Cloud; Centre Léon Bérard, Lyon; Hôpital Hautepierre and Quintiles SA, Strasbourg, France; Eberhard-Karls-Universität, Tübingen, and Amtsgartenweg 13, Badenweiler, Germany; and Hoffmann-La Roche Inc, Basel, Switzerland.
- Address reprint requests to Patrice Viens, MD, Department of Medical Oncology, Institut Paoli-Calmettes, 232 Blvd Sainte Marguerite, 13273 Marseille Cedex 9, France; email: viensp{at}marseille .fnclcc.fr.
Abstract
PURPOSE: To evaluate the safety, pharmacokinetics, and efficacy of three different dose levels of pegylated granulocyte colony-stimulating factor (Ro 25-8315) on progenitor cell mobilization and hematologic recovery in cancer patients.
PATIENTS AND METHODS: Breast cancer patients (n = 36) were randomly assigned to receive before (part I) and after (part II) chemotherapy either a single-dose injection of Ro 25-8315 (20 μg/kg, n = 9; 60 μg/kg, n = 9; 100 μg/kg, n = 10) or a standard daily dose of filgrastim (part I, 10 μg/kg/d; part II, 5 μg/kg/d) (control group, n = 8).
RESULTS: Overall, Ro 25-8315 was well tolerated. In part I, more progenitor cell mobilization was observed with Ro 25-8315 100 μg/kg. The peak of circulating CD34+ cells was obtained at day +5 in the four groups, and the absolute neutrophil count (ANC) returned to less than 20 × 109/L by day +15. In part II, high levels of circulating CD34+ cells (> 20 cells/μL) were obtained in all four groups. The chemotherapy-induced neutropenia (< 1 × 109/L) was similar in the four groups. Ro 25-8315 100 μg/kg was more effective than filgrastim in reducing the number of patients with an ANC less than 0.5 × 109/L on day +12 after chemotherapy.
CONCLUSION: A single injection of Ro 25-8315 100 μg/kg might be the optimal dose for steady-state peripheral-blood progenitor cell mobilization. A single injection of 20, 60, or 100 μg/kg could be as efficient as daily administration of filgrastim to correct chemotherapy-induced cytopenia. The optimal dose of Ro 25-8315 should be determined according to the planned chemotherapy regimen.
GRANULOCYTE COLONY-stimulating factor (G-CSF) is one of numerous cytokines that act on hematopoietic cells and cytokines by specific interactions via its membrane receptor. G-CSF is a hematopoietic growth factor for cells of the neutrophil lineage, although it also displays biologic effects at other stages of hematopoiesis. Produced using gene recombination technology, recombinant human G-CSF (rhG-CSF) was made available to physicians more than 10 years ago.1,2 Initially introduced to reduce chemotherapy (CT)-induced cytopenia, rhG-CSF quickly showed itself to be highly efficient in mobilizing progenitors into adult human peripheral blood.3 Thus, either alone or in association with selected CT regimens, rhG-CSF allowed substitution of autologous bone marrow by autologous apheresis collections, with demonstrated advantages in terms of hematopoietic recovery and transplantation costs.4,5 Administration of rhG-CSF alone at a dose of 10 μg/kg/d for 5 to 7 days or the combination of CT followed by rhG-CSF at a dose of 5 μg/kg/d currently represents the safest method of mobilizing peripheral-blood progenitor cells (PBPCs) for transplantation in cancer patients. Evaluation in the allogeneic setting is ongoing.6,7 Since its commercial introduction, the postmarketing safety surveillance has demonstrated a satisfactory rhG-CSF safety profile. Side effects, including injection site reaction and dose-dependant musculoskeletal pain, remain the most consistently observed adverse events (AEs) across all cancer patients in the major setting of rhG-CSF utilization.2
Because of their short half-lives, both subcutaneous (SC) and intravenous (IV) rhG-CSF require daily or twice-daily administration. Pegylation of a moiety results in an increase in its half-life, due to reduced degradation. Pegylated G-CSF (Ro 25-8315) is the long-acting polyethylene glycol conjugate of an rhG-CSF in which amino acids have been replaced at five positions of the N-terminal region of the human G-CSF molecule.8,9 Pegylation has resulted in a decrease in the systemic clearance of pegylated proteins, which consequently increases the circulation serum half-life of free proteins.10 Preliminary results in healthy volunteers indicate that the safety profile of Ro 25-8315 is similar to that of other G-CSFs and that Ro 25-8315 is effective in both the stimulation of neutrophil counts and the mobilization of CD34+ progenitor cells (Van der Auwera et al, manuscript submitted for publication).
We report here the results of a controlled, randomized, three-dose study of Ro 25-8315 in breast cancer patients receiving a high-dose CT combination regimen for PBPC mobilization.11 The main objective was to characterize the safety profile and efficacy of Ro 25-8315 in cancer patients before and after they received myelosuppressive CT in comparison with filgrastim (recombinant methionyl human G-CSF; Neupogen; Amgen-Hoffman La Roche, Thousand Oaks, CA). Another objective was to define the optimal dose of Ro 25-8315 for treatment of CT-induced neutropenia and the optimal doses for PBPC mobilization when Ro 25-8315 is given alone (steady-state mobilization regimen) and after CT (CT mobilization regimen). In addition, we determined the pharmacokinetics and pharmacodynamics of different doses of Ro 25-8315 and investigated tumor cell mobilization in a subset of patients.
PATIENTS AND METHODS
Patient Eligibility
The study was reviewed and approved by the relevant institutional ethics committees, and all patients gave written informed consent before study entry. Patients between 18 and 60 years of age and with histologically proven metastatic or locoregionally recurrent breast cancer were eligible. Eligible patients had to have a Karnofsky performance status score of 70 or higher and comply with phase I study requirements. At entry, an absolute neutrophil count (ANC) of 1.5 × 109/L, a platelet count of 150 × 109/L, and a hemoglobin level of at least 10 g/dL were required.
Patients were not included if they had received (1) CT within 6 months before study entry, (2) more than six cycles of prior CT in adjuvant or metastatic settings, (3) a cumulative dose of doxorubicin greater than 400 mg/m2 and epirubicin greater than 600 mg/m2, and/or (4) prior intensive CT supported by autologous PBPC transplantation. Patients should not have received prior radiotherapy to more than 50% of the bone marrow–active fields. Patients were not enrolled if they had clinically symptomatic CNS metastasis or major renal (serum creatinine > 1.5 times the upper normal limit [UNL]), liver (AST and ALT > 2.5 times UNL, bilirubin > 1.5 times UNL), lung, heart (left ventricular ejection fraction at rest < 50% at echocardiography or < 55% at isotopic measurement), or CNS dysfunction. Other exclusion criteria included active infection, human immunodeficiency virus seropositivity, known allergy to Escherichia coli–derived products, and history of previous malignancy with the exception of curatively treated basal cell carcinoma, squamous cell carcinoma, and in situ cervical carcinoma. Patients could not enter the study if they were pregnant, lactating, or of childbearing potential and not using effective contraception. Besides their use as part of an antiemetic multidrug regimen the day after CT administration, use of corticosteroids and nonsteroidal anti-inflammatory drugs was restricted to clinical situations with no other therapeutic alternatives, because their use was thought to interfere with the pharmacodynamic effect of Ro 25-8315. The concurrent use of cytokines, other hematopoietic growth factors, and prophylactic antibiotics was prohibited during the course of the study.
Study Design
This was an open-label, multicenter, randomized, four-parallel-groups, filgrastim-controlled study of three doses of Ro 25-8315. It consisted of two parts (Table 1). Patients were centrally randomized in cohorts of at least eight patients per dose level to receive either Ro 25-8315 (20 μg, 60 μg, or 100 μg/kg body weight) or filgrastim before CT (part I) and after CT administration (part II). Part II started from day 15 of part I, once the WBC counts had returned to less than 20 × 109/L. Patients randomized to the Ro 25-8315 treatment groups received one SC injection of Ro 25-8315 at least 2 weeks before CT (part I) and one SC injection 5 days (day +5) after the first day of CT (part II). Patients randomized to the filgrastim group received daily SC injections of filgrastim at a dose of 10 μg/kg/d for 7 days starting at least 2 weeks before CT (part I) and 5 μg/kg/d from day +5 after CT until neutrophil recovery (defined as ANC > 1 × 109/L for 2 consecutive days) (part II). All patients randomized to the Ro 25-8315 treatment groups who had persistent severe neutropenia in part II (defined as ANC < 0.5 × 109/L on day +12) were to receive a second injection of Ro 25-8315 at a dose of 30 μg/kg on day +12. The CT regimen consisted of doxorubicin (Adriblastine; Pharmacia, Paris, France) 75 mg/m2 body-surface area in a bolus IV injection followed immediately by cyclophosphamide (Cytoxan; Bristol-Myers Squibb Co, Princeton, NJ; Endoxan; Asta Médica, Paris, France) 3,000 mg/m2 body-surface area in a 1-hour infusion associated with an equidose of mesna (Uromitexan; Bristol-Myers Squibb) and hyperhydration. The decision to administer a subsequent cycle of CT was left up to the investigator; subsequent cycles should have been started from day +21 of part II as soon as the patient obtained an ANC greater than 1.5 × 109/L and a platelet count greater than 100 × 109/L, whichever occurred first.
Overall Study Design
Patients were monitored daily for adverse effects. Vital signs were measured and clinical examinations were performed at regular intervals and when clinically indicated.
Laboratory Studies
In part I, complete blood counts were obtained daily from day +1 to day +7, every other day from day +8 to day +14, on day 15, and then three times a week until the start of part II. In part II, counts were performed daily from day +5 to day +14 and then every other day until blood counts fell to within the normal range. Biochemistry (blood and urine) tests were performed at regular intervals. Serum was taken for assay of antibodies to Ro 25-8315 at day +1 (predose) of part I and day +21 of part II (end of study). Pharmacodynamic parameters were estimated from blood drawn in part I on days +1, +3 to +6, +8, +10, and +12 and in part II on days +1, +5, +9 to +12, +14, and +21. Assays included assessment of blood progenitor cell levels (CD34+: area under the serum concentration–time curve [AUC], time to peak value [in days], peak value [cells/L], and time [in days] to reach more than 20 CD34+ cells/L).
CD34+ cell enumeration was performed centrally (H. Johnsen, MD, Department of Hematology, Herlev Hospital, Copenhagen, Denmark) using a published flow cytometry procedure.12,13 The standard flow cytometry procedure was performed within 24 hours of the sample arriving in the laboratory, using a FACSan cytometer (Becton Dickinson, Mountain View, CA). The antibodies HPCA-2–phycoerythrin (anti-CD34), CD45-fluorescein, immunoglobulin G1–phycoerythrin, and immunoglobulin G1–fluorescein were all purchased from Becton Dickinson. Results were expressed as percentage of mononuclear cells and converted to number of CD34+ cells/mL of blood. The progenitor cell clonogenic assay (colony-forming unit–granulocyte macrophage [CFU-GM]/blast-forming unit, erythroid [BFU-E]) was performed locally at each participating center according to a standard defined procedure: mononuclear cells were separated by centrifugation over a density gradient (Ficoll-Hypaque gradient; Pharmacia, Uppsala, Sweden; d = 1.077). Mononuclear cells were plated in Methocult H4434 methylcellulose medium (Stem Cell Technology, Vancouver, Canada; the same lot number was used at all participating centers) at two cell concentrations (2 × 105 cells/mL and 5 × 105 cells/mL), in triplicate. Colonies were counted at day 14 ± 1, according to published criteria, and results are expressed per 105 mononuclear cells and per milliliter of blood.
Pharmacokinetic parameters were estimated on days +1 to +6 and on day +8 of part I (total of seven samples) and on days +1, +5 to +10, and +12 of part II (total of eight samples). Serum concentrations of Ro 25-8315 were measured by separate sandwich enzyme-linked immunosorbent assay methods (data on assay sensitivity, specificity, linearity, and reproducibility are on file). Pharmacokinetic parameters included maximum serum concentration (Cmax), time of maximum serum concentration (Tmax), area under the serum concentration–time curve (AUC0-168hr), and systemic clearance (Cl) and were estimated with a model-independent approach by using the WIN-NONLIN computer program (Pharsight, NC). Because of insufficient sampling time, the terminal-phase half-life of Ro 25-8315 was estimated on the basis of data only from part I.
Circulating breast cancer cells were enumerated on cytocentrifuge slides prepared from peripheral-blood samples from patients at two selected participating centers (Marseilles and Tubingen) by a central laboratory (Laboratory for Molecular Oncology, Micromet GmbH, Munich, Germany). Epithelial cells were identified as cytokeratin-positive cells using the A45-B/B3 monoclonal antibody and immunocytochemistry method (alkaline phosphatase-antialkaline phosphatase technique).14,15 Whenever possible, 4 × 106 cells were examined on cytospin slides.
Statistical Methods
The primary efficacy end points were the peak of circulating CD34+ cells in parts I and II and the duration of CT-induced neutropenia in part II (number of days with ANC < 1.0 × 109/L until the first of two consecutive measurements with ANC > 1.0 × 109/L). The secondary efficacy parameters included the neutrophil count kinetics in parts I and II (nadir of ANC, duration of ANC < 0.5 × 109/L, time to reach ANC 0.5 × 109/L), incidence and duration of febrile neutropenia (defined as fever > 38.5°C for one measurement or > 38°C for two consecutive measurements 2 hours apart [mouth measurement]) concomitant with a documented neutropenia with ANC less than 0.5 × 109/L, number of days under IV antibiotics, and delay in starting a subsequent cycle of CT (number of days beyond day 21 of part II).
Descriptive analyses were performed for all patients who received at least one dose of the study drug. Data are expressed as means and/or medians (ranges) for continuous data and as frequencies (percentages) for categorical data, unless otherwise specified. The study was not designed to perform comparative analyses within the four treatment groups.
RESULTS
Patients
The characteristics of the 36 randomized patients are listed in Table 2. The groups were well matched for age and disease characteristics. However, patients in the filgrastim and Ro 25-8315 60-μg groups had received slightly more previous CT and been treated for a longer period of time before they were entered onto the study. Twenty-eight patients were treated with Ro 25-8315 (10, nine, and nine patients in the 100-, 60-, and 20-μg groups, respectively), and eight patients were treated with filgrastim. All patients were entered onto the safety analysis. Thirty-five and 33 patients completed the efficacy analyses of parts I and II, respectively (Table 2): one patient (filgrastim group) developed progressive disease during part I and did not enter part II. One patient (Ro 25-8315 20-μg group) received 10 times the planned dose of Ro 25-8315 during part I. One patient (Ro 25-8315 60-μg group) received cytarabine on day +9 of part II. Eight patients (four patients, three patients, and one patient in the Ro 25-8315 20-, 60-, and 100-μg groups, respectively) received a further injection of Ro 25-8315 30 μg/kg for persistent severe neutropenia at day +12 and were censored at that time for the efficacy analysis of part II (subset 1 population analysis) (Table 2).
Demographics, Baseline Disease Characteristics, Patient Disposition, and Overview of the Analysis Population
Duration of CT-Induced Neutropenia in Part II
Ro 25-8315 stimulated neutrophil recovery after CT at all three doses of administration. Mean duration of CT-induced neutropenia (ANC < 1 × 109/L) in the censored subset population of patients was 4.7, 3.6, 4.1, and 4.0 days in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively (Table 3). The mean ANCs after CT for patients administered filgrastim and Ro 25-8315 are shown in Fig 1. Ro 25-8315 enhanced the peak ANC recovery in a dose-related manner, with a median peak ANC after nadir of 11.8 × 109/L, 11 × 109/L, 13.3 × 109/L, and 33.6 × 109/L in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively (Table 3). There was no difference in the depth or time of the nadir among any of the cohorts. Twelve (36%) of 33 patients experienced severe neutropenia at day +12 (Table 3): eight (31%) of 26 patients in the Ro 25-8315 groups and four (57%) of seven in the filgrastim group. The duration of neutropenia less than 0.5 × 109/L in patients receiving a second Ro 25-8315 injection after CT was between 5.94 and 7.02 days in the Ro 25-8315 20-μg group (four patients), 5.36 days in the Ro 25-8315 60-μg group (one patient), and 18.37 days in the Ro 25-8315 100-μg group (one patient).
CT-Induced Neutropenia After CT Administration Followed by Filgrastim or Ro 25-8315 Treatment (part II, subset 1 population analysis)
Fig 1. Mean ANCs after CT (part II, subset 1 population) for patients given Ro 25-8315 at 20 (group A), 60 (group B), and 100 μg/kg (group C) and filgrastim (group D).
Peak of Circulating CD34+ Cells in Parts I and II
The median time to reach the peak (maximum cell count per microliter) of circulating CD34+ cells in the blood after treatment with filgrastim and Ro 25-8315 without CT (part I) was the same (day +5) within all the treatment study groups (Table 4, Fig 2). Only patients treated with filgrastim and Ro 25-8315 100 μg reached a median level of CD34+ cells above 20 cells/μL. All patients who received one injection of Ro 25-8315 100 μg/kg achieved a peak value ≥ = 25 cells/uL” to “..peak value [greater than or equal to] 25 cells/uL”? If not, please clarify what is meant by “=” in the original text.‘ 25 cells/μL. When filgrastim or Ro 25-8315 was given after CT (part II), a median level of CD34+ cells above 20 cells/μL was obtained in the four treatment groups (Table 5). The kinetics of the CD34+ cells mobilized in the blood after CT failed to demonstrate a clear dose-efficacy relationship between filgrastim and the three doses of Ro 25-8315 (Fig 3), with a median AUC of CD34+ above 20 cells/μL ≥ = 100 in..” to “.. a median AUC of CD34+ above 20 cells/uL [greater than or equal to] 100 in..”? If not, please clarify what is meant by “=” in the original text.‘ 100 in three of the four study groups (the filgrastim and Ro 25-8315 60- and 100-μg groups) (Table 5).
CD34+ Cells in the Blood After Mobilization Regimen With Filgrastim or Ro 25-8315 Without Chemotherapy (part I, steady-state mobilization)
Fig 2. Part I: Mean number of CD34+ cells after filgrastim (D) and Ro 25-8315 administration (A, 20 μg/kg; B, 60 μg/kg; C, 100 μg/kg).
CD34+ Cells in the Blood After Mobilization Regimen With CT Followed by Treatment With Filgrastim or Ro 25-8315 (part II, post-CT mobilization) (subset 1 population analysis)
Fig 3. Part II: Mean number of CD34+ cells after CT followed by filgrastim (D) and Ro 25-8315 treatment (A, 20 μg/kg; B, 60 μg/kg; C, 100 μg/kg) (censored subset population).
Progenitor Cell Kinetics
Part I.
The CFU-GM and BFU-E cells depicted a dose and time-response curve parallel to that of the CD34+ cells, with a median peak count greater in the Ro 25-8315 100-μg group than in the filgrastim group and lower in the Ro 25-8315 20- and 60-μm groups than in the filgrastim group (data not shown). The interpatient variability in CD34+ cell results was most marked in the Ro 25-8315 100-μg group. Peak values were obtained at days +6, +4.5, +5, and +5 in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively. The median increase in CFU-GM counts over baseline was 27-, 21-, 33-, and 170-fold in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively.
Part II.
Day +5 CFU-GM and BFU-E counts were generally lower in the Ro 25-8315 treatment groups than in the filgrastim group. The peak CFU-GM cell counts were observed at days (median) +14, +12, +13, and +14 after CT in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively. The greatest increase in the CFU-GM count was observed in the Ro 25-8315 100-μg group (data not shown).
Secondary Efficacy End Points
Part I.
When filgrastim and Ro 25-8315 were given alone, ANCs were superior in the filgrastim and Ro 25-8315 100-μg groups, with median peak ANCs of 44 × 109/L and 45 × 109/L observed at median times of 6.5 and 4.5 days after the start of treatment administration (Table 6). In all patients but one (in the Ro 25-8315 60-μg group), the ANC returned to below 20 × 109/L by day 15 (Table 6, Fig 4). The time it took to drop to an ANC less than 10 × 109/L was longer in the Ro 25-8315 groups than in the filgrastim group (Table 6).
ANC After Treatment With Filgrastim or Ro 25-8315 (no CT administration, part I)
Fig 4. Mean ANC after treatment with filgrastim (D) and Ro 25-8315 at doses of (A) 20, (B) 60, and (C) 100 μg/kg.
Part II.
After CT administration, an ANC response to filgrastim and Ro 25-8315 was observed early at day +6 in all the treatment groups (Fig 1). The mean and median ANC peak values preceding the nadir were higher in the treatment groups than in the filgrastim group, with a positive dose-response correlation but greater interpatient variability in the treatment groups. The time to recovery to an ANC ≥ = 1 x 109/L” to “an ANC [greater than or equal to] 1 x 109/L”? If not, please explain what is meant by “=”.‘ 1 × 109/L after the post-CT nadir was longer in the filgrastim group than in the Ro 25-8315 groups (Table 3). The ANCs were higher in the Ro 25-8315 groups than in the filgrastim group, with a mean peak ANC after nadir more than three times higher in the Ro 25-8315 100-μg group than in the other treatment groups (Fig 1, Table 3). This difference was associated with a higher ANC at day +21 or beyond in the Ro 25-8315 100-μg group (median ANC, 14.6 × 109/L) than in the filgrastim and Ro 25-8315 20- and 60-μg groups (median ANCs, 3.0, 2.6, and 4.0 × 109/L, respectively) (Fig 1). Three (9%) of 33 patients in part II of the study experienced a febrile neutropenia episode (one patient in the Ro 25-8315 60-μg group and two patients in the Ro 25-8315 100-μg group) on days + 5, +2, and +4 after CT administration. In part II of the study, a higher proportion of patients was treated with IV antibiotics in the Ro 25-8315 groups (six [75%], four [50%], and six [60%] patients in the Ro 25-8315 20-, 60-, and 100-μg groups, respectively) than in the filgrastim group (four patients [50%]). Median duration of IV antibiotics use was 3.5 (range, 3 to 4), 6 (range, 2 to 12), 5 (range, 3 to 7), and 4.5 (range, 2 to 6) days in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively (all patients). There was no difference among the treatment groups with regard to the delay (median, 1 day in each cohort) in starting a subsequent cycle of CT.
Tumor Cell Circulation
Circulating epithelial cells in peripheral blood were present in seven of 22 analyzed patients in the four treatment groups (two of five, two of five, one of six, and two of six in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively), both in part I (maximum of three cells/106 mononuclear cells between days 0 and 8) and part II (maximum of 12 cells/106 mononuclear cells between days 18 and 28). There were no differences in the tumor cell circulation patterns among the four treatment groups, with frequencies of epithelial cells below one in 105 (with one exception), with positive samples either early or late after G-CSF administration, and with successive positive and negative samples in the same patient.
Safety
In part I, median cumulative doses of administered cytokines were 4,270, 1,380, 3,600, and 6,000 μg/kg in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively. In part II, the median cumulative doses were 2,700, 3,000, 3,300, and 6,000 in those groups. Overall, Ro 25-8315 was well tolerated and associated with a manageable toxicity level. The majority of observed AEs (n = 281) were mild or moderate in intensity, and distribution frequencies between and within body systems were similar in all treatment groups. There were no deaths during the course of the study, nor were there withdrawals due to study drug administration. The most frequent related AEs experienced during the study were bone pain and headache, each being reported by 20% of the patients. Among the 57 AEs considered to be related to the study drug treatment (10, 15, 13, and 19 in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively) (Table 7), five were considered severe. Three occurred during part I and two during part II of the study: bone pain (n = 2; part I, Ro 25-8315 60-μg group; part II, Ro 25-8315 100-μg group), injection site pain (n = 1; part I, Ro 25-8315 100-μg group), leukocytosis (n = 1; part I, Ro 25-8315 100-μg group), and leuko- and thrombocytopenia (n = 1; part II, Ro 25-8315 100-μg group). This severe thrombocytopenia and leukocytopenia occurred on days +21 and +22 after CT, respectively, for a duration of 8 and 4 days and resolved without sequelae. Four among the five severe AEs were observed in the Ro 25-8315 100-μg group. Overall, independent of causality and dose, the incidence and severity of AEs were higher in the Ro 25-8315 groups compared with the filgrastim group. Thrombocytopenia (all grades) was more frequent after CT (part II) in the highest Ro 25-8315 dose group and was observed in one (12.5%) of eight, two (22%) of nine, two (22%) of nine, and five (50%) of 10 patients of the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively. Evolution of platelet count in parts I and II of the study is shown in Fig 5. Twenty serious AEs (one in part I [hyperbilirubinemia], 19 in part II) unrelated to the study drug treatment were reported; 18 (54% of the patients) were hospitalizations for febrile neutropenia after CT in part II (three in the filgrastim group and five, four, and six in the Ro 25-8315 20-, 60-, and 100-μg groups, respectively). The baseline ANC was different among the treatment arms and between the two parts of the study. In part I, the median baseline ANC was 4.72 × 109/L, 4.02 × 109/L, 3.36 × 109/L, and 2.82 × 109/L in the filgrastim and Ro 25-8315 20-, 60-, and 100-μg groups, respectively. In part II, it was 2.85 × 109/L, 5.41 × 109/L, 5.64 × 109/L, and 9.84 × 109/L. For technical reasons, anti–Ro 25-8315 serum antibodies were not measured.
Frequency and Severity of AEs Considered Remotely, Possibly, or Probably Related to the Administration of Ro 25-8315 Part I and Part II
Fig 5. Mean platelet count after treatment with filgrastim (D) and Ro 25-8315 at doses of (A) 20, (B) 60, and (C) 100 μg/kg: (a) part I; (b) part II.
Pharmacokinetics
Pharmacokinetic parameters are summarized in Table 8 and concentration-versus-time graphs are shown in Figs 6 and 7⇓. Pharmacokinetic analyses were based on seven, seven, and nine patients for the Ro 25-8315 20-, 60-, and 100-μg/kg dose groups, respectively.
Summary of Pharmacokinetic Parameters of Ro 25-8315 After SC Injection in Breast Cancer Patients, part I (without CT) and part II (after CT)
Fig 6. Pharmacokinetic parameters: serum concentration of Ro 25-8315 × time—part I of the study.
Fig 7. Pharmacokinetic parameters: serum concentration of Ro 25-8315 × time—part II of the study.
Part I.
Serum concentrations of Ro 25-8315 increased rapidly and reached mean maximum concentrations of 162.7, 170.7, and 261.6 ng/mL between 6 and 72 hours after SC administration. Thereafter, the serum concentrations of Ro 25-8315 declined mainly in a monoexponential fashion. The mean time to reach the peak levels was 40, 24, and 33 hours at the doses of 20, 60, and 100 μg/kg, respectively. The pharmacokinetic parameters (Tmax, systemic clearance, half-life) of Ro 25-8315 were comparable in the 20-, 60-, and 100-μg/kg dose groups (Table 8). The Cmax and AUC were comparable in the 20- and 60-μg/kg dose groups and increased at the 100-μg/kg dose. In general, the pharmacokinetic parameters of part I of this study were different from the results obtained from healthy subjects, except for serum half-life (data on file).
Part II.
When given after CT administration, the increase in serum concentrations of Ro 25-8315 was slower in comparison with the results from part I and reached mean maximum values of 141, 379, and 590 ng/mL at the doses of 20, 60, and 100 μg/kg between 72 and 120 hours (Table 8, Fig 7). Thereafter, the serum concentrations of Ro 25-8315 declined mainly in a monoexponential fashion. The mean times to reach the peak levels were 110, 84, and 99 hours at the doses of 20, 60, and 100 μg/kg, respectively, and the Cmax and AUC increased proportionally with the dose at these dose ranges.
DISCUSSION
This controlled, randomized, phase I study examined the safety and biologic effects of three doses of Ro 25-8315 (20, 60, and 100 μg/kg) when given subcutaneously alone and after high-dose combination CT in breast cancer patients. The increase in the half-life of the study cytokine allows for the induction of biologic effects after a single dose injection, as evidenced by blood counts, numeration of CD34+ cells, and clonogenic progenitors in the peripheral blood of these patients. Two rhG-CSFs (filgrastim and lenograstim) with similar biologic properties have been approved for the last 10 years for clinical use in oncologic and hematologic settings. The major clinical applications of rhG-CSF are their use in accelerating neutrophil recovery after myelosuppressive CT, decreasing the myelotoxicity associated with sequential high-dose CT, and mobilizing PBPCs obtained by leukapheresis for reinfusion by autologous or allogenic transplantation.1,2,16,17
In this study, when filgrastim or Ro 25-8315 was administered after the doxorubicin-cyclophosphamide combination regimen, the duration of CT-induced neutropenia was similar within the four groups. Ro 25-8315 at 100 μg/kg was more efficient than Ro 25-8315 at 20 and 60 μg/kg and filgrastim in reducing the incidence of severe neutropenia (ANC < 0.5 × 109/L) at day +12 after CT. The incidence of hospitalization for febrile neutropenia was high in our study (overall incidence, 55%) and higher in the Ro 25-8315 groups (15 of 26 patients; 58%) compared with the filgrastim group (three of seven patients; 43%). The higher myelotoxicity could be due to the cumulative amount of previous CT received as well as the advanced stage of disease of our breast cancer population.18 Another explanation could be an early chemotoxicity on circulating progenitor cells at the time of CT administration in part II of the study. CT was administered to patients without a washout period between the two administrations of study drug, whereas the WBCs and progenitor cells had not returned to the steady-state baseline level. The levels of progenitor cells (CD34+, CFU-GM, and BFU-E cells) were lower at day +5 after CT in the Ro 25-8315 groups than in the filgrastim group. This could represent a biologic expression of the chemotoxicity effect on immature hematopoietic cells. Similarly, the higher incidence of CT-induced thrombocytopenia—also described in patients similarly treated with filgrastim in earlier studies2—observed in the Ro 25-8315 100-μg group in part II of the study could indicate a greater susceptibility to CT of committed hematopoietic progenitor cells after one injection of a long-acting Ro 25-8315 2 weeks earlier. This potential risk of cumulative myelotoxicity should be carefully studied in the context of multiple CT. The prolonged and dose-dependent duration of action of single doses of Ro 25-8315 is consistent with the pharmacodynamic effects of Ro 25-8315 observed in the healthy volunteers study (Van der Auwera et al, manuscript submitted for publication).19 In this study, administration of various dosages of Ro 25-8315 was safe and well tolerated, with a safety profile consistent with that observed after filgrastim treatment and similar to the AE profile observed in the Ro 25-8315 study with human volunteers (Van der Auwera et al, manuscript submitted for publication). The higher incidence of related AEs was observed in the higher dosing group (100 μg/kg) of Ro 25-8315 and was consistent with the dose-related safety profile of G-CSF. Dose-dependent musculoskeletal pain remains the most consistently observed AE associated with filgrastim administration across all cancer patient populations in multiple settings, including as an adjunct to CT and PBPC mobilization.2,20 Hyperleukocytosis has been described as a rare, dose-related, transient, and generally asymptomatic AE of G-CSF administration in patients and healthy donors,21 leading most frequently to the reduction or discontinuation of G-CSF administration. The potential risks of sustained hyperleukocytosis could, however, be a concern after a protracted pharmacodynamic effect of Ro 25-8315. Splenomegaly and splenic rupture have been described as exceptional AEs after G-CSF administration, and these risks could be increased after administration of Ro 25-8315.
We demonstrated the ability of Ro 25-8315 to mobilize progenitor cells (CD34+, CFU-GM, and BFU-E cells) in the peripheral blood of cancer patients in a dose-related fashion when given alone or after CT. In a steady-state mobilization regimen, a single injection of Ro 25-8315 administered alone at a dose of 100 μg/kg was able to mobilize a large number of peripheral blood CD34+ cells (> 25/μL) in 100% of the treated patients and was more efficient than seven daily injections of filgrastim given at 10 μg/kg/d. In clinical practice, collection of CD34+ cells by leukapheresis is usually started when more than 20 CD34+ cells/μL are monitored in peripheral blood. In our study’s steady-state mobilization regimen, an efficacy dose of Ro 25-8315 100 μg/kg allowed a threshold of 20 CD34+ cells/μL to be reached in 100% of patients. A combination of CT followed by an injection of a hematopoietic growth factor such as filgrastim can mobilize higher numbers of PBPCs compared with filgrastim alone.22 This also translated with Ro 25-8315 in our study. After CT, a single injection of Ro 25-8315 at the doses of 60 and 100 μg/kg seems to be as efficient or even more efficient than a daily filgrastim injection to mobilize progenitor cells; however, daily filgrastim is more efficient than one administration of Ro 25-8315 20 μg/kg. The high number of CD34+ cells mobilized with a single injection of Ro 25-8315 100 μg/kg in a steady-state regimen, or with CT followed by a single injection of Ro 25-8315 100 μg/kg and 60 μg/kg, may facilitate the collection of high numbers of PBPCs with a minimum number of leukaphereses. In our study, filgrastim was administered at the standard treatment dose. Ro 25-8315 could be evaluated further in comparison with a higher dosage of G-CSF for PBPC mobilization.23
The kinetics of mobilized progenitor cells was similar within the four treatment groups during the steady-state regimen as well as after CT administration. This consistent and predictable pharmacologic response to Ro 25-8315 is similar to that observed with filgrastim in cancer patients.24 Our study showed tumor cell contamination in the peripheral blood of patients receiving filgrastim or Ro 25-8315 at either dosage, when the hematopoietic growth factors were administered alone or after CT. It has been demonstrated that filgrastim-mobilized PBPC collections of various pathologic entities, including breast cancers, contain tumor cells,25-27 but there is no convincing evidence that the mobilization of hematopoietic progenitors into the peripheral blood is itself associated with the circulation of tumor cells. Moreover, it is not known to what extent these tumor cells are clonogenic and thus might be able to initiate tumoral disease if infused to the patients.
In this cancer patient population, most values for pharmacokinetic parameters (Tmax, Cmax, Cl/F, and AUC) of Ro 25-8315 administered in a steady-state regimen were different from the corresponding values found in healthy subjects (Van der Auwera et al, manuscript submitted for publication), indicating that the disease status may affect the pharmacokinetics of Ro 25-8315. After CT administration, the serum clearance of Ro 25-8315 decreased dramatically, as demonstrated by the increase of AUC0-168hr values in part II compared with part I of the study. Such a difference in pharmacokinetics before and after CT may be explained by the decrease in circulating neutrophils as a result of CT and consequently in reduced availability of serum neutrophil G-CSF receptors to which G-CSF binds. It is probable that, as for filgrastim pharmacokinetics,24 the clearance and half-life of Ro 25-8315 are dependent on dose and neutrophil count: G-CSF receptor-mediated clearance is saturated by a high concentration of G-CSF and is diminished by neutropenia. This may have led to a significant fraction of Ro 25-8315 remaining in the circulation in an unbound form rather than binding to neutrophils. However, the linearity of Ro 25-8315 is still maintained over the tested dose range. The increases in Cmax and AUC of Ro 25-8315 in part II of the study were proportional to the administered doses of Ro 25-8315. Therefore, it is possible that the CT may have influenced the disposition kinetics of Ro 25-8315 without having an impact on its linearity at the test dose range.
In conclusion, this dose-range study shows that Ro 25-8315 is a long-acting hematopoietic growth factor that enhances neutrophil recovery after intensive CT and mobilizes hematopoietic progenitor cells in the blood of cancer patients when given alone or after CT, in a dose-related manner. In addition, it can be administered safely. Further larger studies are warranted to define the optimal treatment dose according to the clinical indications and to define the best schedule of administration when associated with CT.
Acknowledgments
Supported in part by research funding from Hoffmann-La Roche Inc to P.VdA. and D.M.
ACKNOWLEDGMENT
We thank Françoise Beaujean, MD (Creteil), Gilles Bourguignon, MD (Lyon), Olivier Feugeas, MD (Strasbourg), Claire Mathiot, MD (Paris), and Alain Bourguignat, MD (Saint-Cloud, France), for their contribution to this study.
Footnotes
-
P.VdA. and R.C. declare a financial interest in the company whose product was studied in the present report. P.VdA. declares a financial interest in a competing company whose product was studied in the present report.
- Received September 14, 2000.
- Accepted August 2, 2001.
