- © 2000 by American Society of Clinical Oncology
Mobilization of Peripheral-Blood Stem Cells by Concurrent Administration of Daniplestim and Granulocyte Colony-Stimulating Factor in Patients With Breast Cancer or Lymphoma
- John F. DiPersio,
- Michael W. Schuster,
- Camille N. Abboud,
- Jane N. Winter,
- Vicki R. Santos,
- Denise M. Collins,
- Jeffrey W. Sherman and
- Charles M. Baum
- From the Division of Bone Marrow Transplantation and Stem Cell BiologyWashington University School of Medicine, St Louis, MO; the Division of Hematology and Medical Oncology, Weill Medical College of Cornell University, the New York Presbyterian Hospital, New York; the Hematology/Oncology Unit, Department of Medicine, the University of Rochester Medical Center, Rochester, NY; the Division of Hematology-Oncology, Northwestern University Medical School, Chicago; and Searle A. Monsanto Co, Skokie, IL.
- Address reprint requests to John F. DiPersio, MD, PhD, Division of Bone Marrow Transplantation and Stem Cell Biology, Washington University School of Medicine, 660 S Euclid, Box 8007, St Louis, MO 63110-1093; email jdipersi{at}imgate.wustl.edu
Abstract
PURPOSE: To evaluate the safety and hematopoietic activity of daniplestim administered concurrently with granulocyte colony-stimulating factor (G-CSF) for peripheral-blood stem-cell (PBSC) mobilization.
PATIENTS AND METHODS: In the initial dose-escalation phase, 25 patients with adenocarcinoma of the breast (AB; 13 patients) or lymphoma (12 patients) were given daniplestim at doses ranging from 0.1 to 3.75 μg/kg/d plus G-CSF 10 μg/kg/d. In the randomized phase, 52 patients with AB (27 patients) or lymphoma (25 patients) were randomized within disease categories to the daniplestim dose chosen in the dose-escalation phase plus G-CSF 10 μg/kg/d (D+G) or placebo plus G-CSF 10 μg/kg/d (P+G) for up to 7 days.
RESULTS: A daniplestim dose of 2.5 μg/kg/d was chosen for further study because it was hematopoietically active and had an acceptable side-effect profile. In the randomized phase, in patients with AB, D+G was associated with a higher probability (P = .0696) of collecting ≥ 2.5 × 106 CD34+ cells/kg and significantly higher circulating CD34+ cell counts (P = .0498) on days 6 through 9 after the initiation of dosing. The target level was more likely to be reached with additional leukaphereses in the patients given D+G. Patients given P+G did not benefit from additional leukaphereses beyond the first procedure. The type of mobilization did show a trend toward a shorter duration of neutropenia in the D+G group. The adverse events with D+G consisted largely of mild to moderate flu-like symptoms, including headache and fever, and occurred more frequently than with P+G.
CONCLUSION: Daniplestim administered at 2.5 μg/kg/d is tolerable and active when combined with G-CSF, and the combination may prove more effective than G-CSF alone in promoting the collection of adequate numbers of CD34+ cells for PBSC infusion in patients with AB.
AUTOLOGOUS PERIPHERAL-blood stem-cell (PBSC) transplantation is effective in restoring hematopoiesis in patients who undergo myeloablative therapy for a wide variety of tumors. It has become standard practice to administer a hematopoietic growth factor (HGF) before chemotherapy to increase the numbers of CD34+ cells in the peripheral blood and thus ensure a sufficient harvest of CD34+ cells by leukapheresis. It has been observed that infusing an adequate number of CD34+ cells is necessary for early and sustained hematopoietic recovery. Furthermore, recent data suggest that there is a dose-response relationship between the number of CD34+ cells infused and hematopoietic recovery.1 Several investigators have found that the infusion of at least 2.5 × 106 CD34+ cells/kg resulted in timely hematopoietic recovery.2 In addition, more recently it was observed that the infusion of at least 5.0 × 106 CD34+ cells/kg is consistently associated with more predictable and rapid recovery, particularly of platelets.1
Granulocyte colony-stimulating factor (G-CSF) or granulocyte-macrophage colony-stimulating factor (GM-CSF) is commonly used for mobilizing stem cells.1 The hematopoietic effect of either can be increased by sequential use with recombinant human interleukin-3 (rhIL-3).3,4 With G-CSF, this increase may be partly due to the fact that rhIL-3 stimulates the replication and differentiation of early progenitor cells,5,6 and G-CSF acts primarily on more-mature cells.7 However, the clinical utility of rhIL-3 has been limited by significant dose-limiting toxic effects that may be related to its potency in stimulating the release of inflammatory leukotrienes and histamine.8-10
Daniplestim, a novel genetically engineered agonist of the IL-3 receptor complex, is 20 times more potent than rhIL-3 in stimulating the maturation of CD34+ cells into colony-forming units (CFU) in vitro. Daniplestim is approximately twice as potent as rhIL-3 in stimulating the release of histamine and leukotrienes in vitro. Therefore, daniplestim is predicted to have a 5- to 10-fold greater therapeutic index than rhIL-3.11
Mobilization studies in rhesus monkeys showed that the concurrent use of daniplestim and G-CSF for 10 days produced higher and more persistent levels of circulating CD34+ cells than did a 10-day course of G-CSF alone. Concurrent dosing with daniplestim plus G-CSF resulted in a greater than four-fold increase in the mean circulating level of CD34+ cells by day 5 of dosing, which remained at approximately this level for the remaining 5 days. G-CSF administered alone resulted in a three-fold increase by day 3 and a drop to baseline by day 5, with no subsequent increase. The increase in the total number of circulating colony-forming cells was similar with both treatments, but the kinetics of mobilization occurred more quickly and were maintained throughout dosing with daniplestim plus G-CSF.12
We now report on a phase I/II study that was designed to evaluate the safety, tolerability, and hematopoietic activity of daniplestim used concurrently with G-CSF for PBSC mobilization in patients with breast cancer or lymphoma. In the first part of the study, a dose-escalation strategy was used to determine a suitable dose of daniplestim for further study. In the second part of the study, patients were randomized to either the dose of daniplestim determined suitable for further study plus G-CSF 10 μg/kg/d (D+G) or placebo plus G-CSF 10 μg/kg/d (P+G). It was assumed that although the biologic responses to the HGF could differ between patients with breast cancer and patients with lymphoma, tolerability would not. Therefore, the disease categories were combined for the analysis in the dose-escalation phase and for safety parameters in the randomized phase but were stratified for the analysis of efficacy parameters in the randomized phase.
PATIENTS AND METHODS
Patient Selection
Patients eligible for the study were at least 18 years old and had histologically confirmed chemotherapy-responsive stage II to IV adenocarcinoma of the breast with the involvement of at least 10 axillary nodes or had biopsy-proven lymphoma. Eligibility for patients with lymphoma required one of the following conditions: complete remission not having been achieved after at least one chemotherapy induction regimen, relapse within 1 year after complete remission with an initial combination therapy, or relapse more than 1 year after complete remission with a doxorubicin-based regimen. The eligibility requirements for all patients included an absolute neutrophil count (ANC) ≥ 1,500/μL, platelet count ≥ 100,000/μL, total WBC count ≥ 3,000/μL, hematocrit ≥ 30%, untransfused hemoglobin level ≥ 8.5 g/dL, and Karnofsky performance status ≥ 70.
Patients were excluded if they had been treated with more than two previous chemotherapy regimens for breast cancer, with the exception of adjuvant or hormonal therapy, or more than three previous chemotherapy regimens for lymphoma; had been treated with mitomycin or a nitrosourea; had previously undergone PBSC or autologous or allogeneic bone marrow transplantation; had myeloid malignancy or biopsy-proven bone marrow involvement that affected more than 25% of the marrow cells; or had known hypersensitivity to any Escherichia coli protein. Each patient signed an informed consent form in accordance with federal and institutional guidelines.
Study Design
The study, conducted at four centers, consisted of (1) an initial open-label dose-escalation phase to determine a dose of daniplestim that was suitable for further study and (2) a subsequent double-blind placebo-controlled randomized phase.
Dose-escalation phase.
In the dose-escalation phase, successive cohorts of at least three patients each were to be given subcutaneous daily doses of daniplestim (GD Searle & Co, Skokie, IL) for up to 7 days at dose levels of 0.1, 0.25, 0.5, 1, 1.75, 2.5, 3.75, 5, or 7.5 μg/kg. Dose escalation within an individual patient was not permitted. Escalation could occur if no patient at a dose level experienced a dose-limiting toxic effect, defined as any toxicity other than leukocytosis, eosinophilia, or thrombocytosis that was grade 3 or higher by the criteria of the Southwest Oncology Group,13 was related to treatment, and necessitated removing the patient from the study. Grade 3 or 4 toxicity that could be controlled with acetaminophen or other standard agents were not considered dose-limiting. G-CSF (filgrastim; Amgen Inc, Thousand Oaks, CA) 10 μg/kg/d was given concurrently to all patients. All patients were to undergo daily leukaphereses beginning on day 5 of HGF dosing and continuing for up to 5 days or until ≥ 4 × 106 CD34+ cells/kg had been collected. The dose-escalation phase was to be stopped when the principal investigator judged one of the doses to be suitable for further study on the basis of tolerability and hematopoietic activity.
Randomized phase.
Patients with breast cancer or lymphoma were separately randomized in a ratio of three who were given D+G for every one who was given P+G. Patients were treated for up to 7 days, and on day 5 leukapheresis was initiated as described in the previous paragraph for the dose-escalation phase. After the completion of HGF dosing and leukaphereses (mobilization period), patients were treated with myeloablative chemotherapy followed by PBSC infusion and subcutaneous G-CSF 5 μg/kg/d until their ANC was ≥ 1,500/μL or for up to 21 days (engraftment period; Fig 1).
Fig 1. Design of the randomized phase of the study. Abbreviation: HDCT, high-dose chemotherapy.
Mobilization period of randomized phase.
Treatment with D+G or P+G began on day 1 of the mobilization period. Prophylactic oral medication for potential adverse effects was permitted after the administration of the first dose of the study medication, if deemed necessary. On the basis of the evaluation of prophylactic regimens in the dose-escalation phase and in some patients in the randomized phase, treatment with prednisone 10 to 20 mg every 12 hours, acetaminophen 650 mg every 6 hours, and diphenhydramine hydrochloride 50 mg every 6 hours was initiated 24 hours before the initiation of HGF treatment and was continued throughout dosing in 22 patients in the randomized phase of the study. Equal percentages of patients in each treatment group received this regimen (43% in the D+G group, 42% in the P+G group).
Leukapheresis was performed for up to 5 days, beginning on day 5 of the mobilization period. For each leukapheresis, 10 L of blood was to be processed over a period of approximately 3 hours. All leukapheresis products were cryopreserved, frozen, and stored according to standard procedures at each study center.
Engraftment period of randomized phase.
Patients were to be given myeloablative chemotherapy within 21 days after the completion of leukapheresis and subsequently were to be infused with the PBSC product that had been collected. Supportive therapy with G-CSF 5 μg/kg/d was to begin 4 hours after the PBSC infusion and was to be continued for up to 21 days or until the ANC was ≥ 1,500/μL for 3 consecutive days and the platelet count was ≥ 20,000/μL without transfusions. Patients were to be assessed at 3 to 5 days after the discontinuation of G-CSF or at the time of premature withdrawal.
Progenitor-cell analysis.
CD34+ cell counts were determined by a local laboratory using modifications of standard methods14,15 at the site of patient treatment as well as by the central laboratory (Cytometry Associates, San Diego, CA) using a standard method.16,17 The decision of when to discontinue PBSC collection was made at each site on the basis of its local laboratory results.
Only the central laboratory, using a standard method,16,17 assayed for granulocyte-macrophage CFUs (CFU-GM) in the peripheral blood and leukapheresis product. The CFU-GM were counted after 14 days of incubation in HCC-4434 methylcellulose medium (Stem Cell Technologies, Vancouver, British Columbia, Canada) at 37°C in a humidified incubator.
Criteria for hematopoietic activity.
The parameters used to evaluate hematopoietic activity during the mobilization period included the percentage of patients from whom ≥ 2.5 × 106 and ≥ 4.0 × 106 CD34+ cells/kg were collected, the number of leukaphereses necessary to obtain these counts, the cumulative CD34+ cell yields, and the daily counts of CD34+ cells and CFU-GM in the peripheral blood. The parameters used to evaluate hematopoietic effects in the engraftment period included the time to neutrophil engraftment (the first day on which the ANC was ≥ 500/μL, with all subsequent counts ≥ 500/μL), the time to platelet engraftment (the first day of a platelet count of ≥ 20,000/μL in the absence of transfusions, with all subsequent counts ≥ 20,000/μL), and the durations of neutropenia and thrombocytopenia.
Safety and tolerability.
Adverse events were graded according to the four-step toxicity scale of the Southwest Oncology Group.13 An assay for antibodies to daniplestim was performed on serum samples that were obtained at four times during the study. The detection method was an enzyme-linked immunosorbent assay that used a coating of daniplestim on microtiter plates to capture any antibodies present in the serum.
Criteria for discontinuation from the study.
The criteria for withdrawing patients from the study included a grade 3 or 4 allergic reaction, worsening of a laboratory parameter to grade 3 or 4, a serious adverse event (eg, one that was life-threatening or permanently disabling), or the inability to collect ≥ 1.0 × 106 CD34+ cells/kg after five leukaphereses (mobilization failure). Patients who were withdrawn because of mobilization failure were included in the analyses of safety and hematopoietic effects during the mobilization period.
Statistical Methods
All randomized patients were included in statistical intent-to-treat (ITT) analyses regardless of adherence to the protocol. Efficacy comparisons between the D+G group and the P+G group were performed separately within the breast cancer and the lymphoma cohorts. All statistical tests were two-sided. Demographic and baseline variables were analyzed by using Fisher’s exact test or the Kruskal-Wallis test, depending on whether the variable was discrete or continuous, respectively. The proportion of patients from whom ≥ 2.5 × 106 CD34+ cells/kg and ≥ 4.0 × 106 CD34+ cells/kg were obtained was evaluated by using a Cochran-Mantel-Haenszel test stratified by investigator to control for factors that may have resulted in differences at the study sites. A Breslow-Day test was used for determining whether an investigator-by-treatment interaction had occurred. The numbers of leukaphereses necessary to obtain these cell numbers, as well as the times to neutrophil and platelet engraftment, the times to the nadirs of the ANC and platelet count, and the durations of neutropenia and thrombocytopenia, were analyzed by using a time-to-event analysis, and the two treatment groups were compared by using a log-rank test. The median and the number censored (those for whom the event of interest was not observed) were used for characterizing the data. The ANC and platelet nadirs were compared by using a t statistic without assuming equal variance.
RESULTS
Dose-Escalation Phase
Twenty-five patients, 13 with breast cancer and 12 with lymphoma, were enrolled onto the dose-escalation phase of the study, and 20 completed this phase. Cohorts of three patients each were given daniplestim at dose levels of 0.1, 0.25, 0.5, 1, 1.75, and 2.5 μg/kg/d, and seven patients were given 3.75 μg/kg/d. Two patients in the dose-escalation phase had entry criteria violations in that they did not meet the criteria for hematologic parameters. (One patient who received 3.75 μg/kg/d had a hematocrit of 26.9 g/dL and a WBC count of 1,800/μL; one patient who was given 1.75 μg/kg/d had pretreatment values of 1,500/μL, 450/μL, and 83,000/μL for WBC count, ANC, and platelet count, respectively. All of these values exceeded the entry criteria before the first dose of study drug.) All patients had had one or two prior chemotherapy regimens, with the exception of two patients (one each in the 0.1 and 2.5 μg/kg/d dose groups) who had had three prior chemotherapy regimens. Of the five patients (20%) who discontinued the study prematurely, one each had been assigned to 0.1, 0.5, and 1.75 μg/kg/d and two had been assigned to 3.75 μg/kg/d. The reasons for withdrawal were inability to achieve an adequate CD34+ cell collection in four patients and inability to return for study procedures in one patient.
In the dose-escalation phase, the majority of adverse events that occurred during dosing and leukapheresis and before chemotherapy were flu-like symptoms; however, no patients withdrew because of adverse events or dose-limiting toxicities. The most frequently reported events were headache (64%), fever (64%), nausea (32%), flushing (32%), and rigors (28%). Table 1 lists the adverse events by dose level for all events that were reported in at least five patients (20%). The incidence of adverse events increased with increasing daniplestim dose, particularly with the 3.75 μg/kg/d dose. Headache (20%) was the only grade 3 event that occurred in at least two patients.
Adverse Events Most Frequently Reported During Dosing and Leukapheresis and Before Chemotherapy in the Dose-Escalation Phase
One of the 25 patients who was treated with daniplestim 3.75 μg/kg/d plus G-CSF in the dose escalation phase had a serious adverse event (fever) that was considered to be related to the study drug. A temperature of 39.7°C developed in this patient on day 5, and the patient was hospitalized for treatment with antibiotics; the fever resolved by the following day and all cultures were negative. No patient discontinued treatment because of a study drug–related adverse event.
Doses greater than 3.75 μg/kg/d were not evaluated because hematopoietic activity was judged to be adequate at this and the 2.5 μg/kg/d level; lower doses were not considered to be sufficiently active. In addition, more CD34+ cells were collected in the 2.5 and 3.75 μg/kg/d dose groups than in the lower-dose groups, with the highest cumulative cell counts obtained in the 2.5 μg/kg/d group (Fig 2). The lack of greater hematopoietic effect and the occurrence of more side effects at the 3.75 μg/kg/d dose dictated the choice of 2.5 μg/kg/d for the randomized phase.
Fig 2. Median cumulative CD34+ cell counts in the leukapheresis product after the first (light gray bar) and last (dark gray bar) day of leukapheresis with the daniplestim dosages used in the dose-escalation phase of the trial. All patients were given concurrent G-CSF 10 μg/kg/d. One patient who withdrew after having been given only a single dose of the study drug was excluded from this analysis.
Randomized Phase
Patient characteristics.
Fifty-two patients were entered onto the randomized phase of the study, 27 with breast cancer and 25 with lymphoma (Table 2). There were no statistically significant differences between the treatment groups in any baseline parameter in either the breast cancer or the lymphoma cohort.
Number of Prior Chemotherapy Regimens by Treatment Group and Diagnosis
In the randomized phase of the study, 11 patients (nine who were given D+G and two who were given P+G) had violations in entry criteria during the screening period. Five of these patients had hematology-related violations involving laboratory values that were outside of the required ranges before the first dose of the study drug. One breast cancer patient who received D+G had received hormonal therapy (anastrozole) within 4 weeks of leukapheresis. Of the four lymphoma patients who received D+G, one did not fulfill the criteria for hematology (WBC of 2,300/μL), two had received carmustine (one of these also required treatment with warfarin sodium), and one had received five prior chemotherapy regimens (three were allowed). One patient with lymphoma who received P+G had biopsy-proven bone marrow involvement that affected more than 25% of the marrow. Therefore, 46 patients (89%) met the entry criteria for the study before dosing.
Patient disposition.
Of these 52 patients, 27 (52%) completed the study and 25 (48%) withdrew before completion. A large proportion (17 [68%] of 25) of the withdrawals occurred after the dosing period and were due to inability to achieve an adequate CD34+ cell collection, rendering these patients ineligible to be treated with ablative chemotherapy as defined in the protocol. Three patients with breast cancer (all D+G) were withdrawn because they refused the study drug, and another was given an overdose of the study drug. One patient with lymphoma was also withdrawn because she refused the study drug (D+G). None of the 22 premedicated patients withdrew from the study because of adverse events, and three of the 30 patients who were not premedicated withdrew because of adverse events.
Adverse Events During Dosing and Leukapheresis and Before Chemotherapy
All 52 randomized patients were included in the analysis of safety. At least one adverse event occurred in all 40 patients who were given D+G (data not shown). The four most common adverse events in this group that were judged to be related to the study drug, with a statistically significant difference between the two groups, are listed in Table 3.
Adverse Events with Probable Relation to Study Drug Reported During Dosing and Leukapheresis and Before Chemotherapy in the Randomized Phase
Two patients were inadvertently given 10-fold overdoses of daniplestim (25 μg/kg/d) at the same site on the same day. One patient experienced a headache that was considered to be related to the study medication, and this patient continued in the study after being informed of the medication error. The second patient, who had a history of medically controlled hypertension, experienced a brief episode of hypotensive syncope followed by headache and fever. The hypotension that led to syncope was judged to be related to the study medication. This patient chose to withdraw from the study after being informed of the error.
Three patients with breast cancer and one with lymphoma had serious adverse events that were considered to be related to the study drug. Of the patients with breast cancer, one was withdrawn with severe headache, vomiting, and nausea and two were given 10 times the assigned dose of medication, as described in the previous paragraph. The patient with lymphoma had mild vomiting that lasted less than a day.
Seventeen (42%) of the 40 patients in the D+G group and five (42%) of the 17 patients in the P+G group were premedicated with diphenhydramine, prednisone, and acetaminophen in an attempt to minimize adverse events. There was no impact on the frequency of the adverse events in either treatment group. No clinically significant laboratory abnormalities were observed, and no antibody titers to daniplestim were detected at any time in any patient.
Hematopoietic Effects in the Mobilization Period: CD34+ Cell Counts
On the basis of the hematopoietic and tolerability data in the dose-escalation phase of the study, the 2.5-μg/kg/d dose was chosen for use in the randomized phase of the study to compare D+G with P+G. The numbers of patients in whom the target CD34+ cell collections were achieved are presented in Table 4.
Number of Patients Who Achieved 2.5 × 106 or 4.0 × 106 CD34+ cells/kg in the Leukapheresis Product (all randomized patients)*
Among patients with breast cancer in the ITT analysis, a higher proportion of those treated with D+G achieved the target numbers of CD34+ cells than did those treated with P+G. This difference was marginally significant (P = .0895, Cochran-Mantel-Haenszel test) in the ITT analysis for the ≥ 2.5 × 106 CD34+ cells/kg target. There were no significant differences between the two treatment groups in the lymphoma cohort.
Among patients with breast cancer in the ITT analysis, the probability of collecting ≥ 2.5 × 106 CD34+ cells/kg with additional leukaphereses beyond one was greater in those given D+G than in those given P+G (P = .0696, log-rank test; Fig 3). Only one (17%) of six patients treated with P+G achieved ≥ 2.5 × 106 cells/kg (achieved with one leukapheresis). In contrast, three (14%) of 21 patients given D+G achieved ≥ 2.5 × 106 cells/kg after a single leukapheresis, and eight (38%) achieved ≥ 2.5 × 106 cells/kg after two to five leukaphereses. Combining both disease cohorts and treatment groups, 20% of patients (three of 15) who had three to seven previous regimens were able to mobilize the target number of cells, compared with 47% of patients (17 of 36) who had one or two previous regimens. The sample size was too small to detect differences within disease cohorts and to test for treatment differences within a disease cohort after controlling for the number of previous regimens.
Fig 3. Kaplan-Meier probabilities for the number of leukaphereses necessary to obtain ≥ 2.5 × 106 CD34+ cells/kg with D+G (
) and P+G (- - -) in the breast cancer (A) and lymphoma (B) cohorts.
In the lymphoma cohort, the number of leukaphereses required to collect ≥ 2.5 × 106 CD34+ cells/kg did not differ significantly between the two treatment groups (Fig 3). The number of previous chemotherapy regimens had a marginally significant impact (P = .072) on the ability to mobilize CD34+ cells.
The mean CD34+ cell counts in the peripheral blood peaked on day 5 in the patients with breast cancer who were treated with P+G. There was a sustained peak on days 5 and 6 in patients treated with D+G, with a gradual decline in the counts on days 7, 8, and 9. This finding may explain why the daily yield of CD34+ cells in the leukapheresis product increased or remained the same in patients who were given D+G (Fig 4). There were statistically significantly greater counts on days 6, 7, 8, and 9 in patients with breast cancer who were treated with D+G. These differences were not apparent in patients with lymphoma. The pattern of mean CFU-GM counts in the peripheral blood was similar to that of the CD34+ cell counts, with significantly higher (P < .05) counts in the D+G group on days 6 and 7 in patients with breast cancer and on day 8 in patients with lymphoma (data not shown).
Fig 4. Median cumulative number of CD34+ cells after 1 or more days of leukapheresis in patients treated with D+G (dark gray bar) and P+G (light gray bar) in the breast cancer (A) and lymphoma (B) cohorts.
Hematopoietic Effects in the Engraftment Period
After PBSC infusion, the median times to neutrophil and platelet engraftment did not differ significantly between the two treatment groups in either the breast cancer or the lymphoma cohort. Ninety percent of patients in the D+G group had platelet counts that were ≥ 20,000/μL by day 25 post-PBSC infusion; however, 90% of the patients in the P+G group had not reached this level by the time of the follow-up visit.
There were no significant differences between treatment groups in the nadirs of the ANC and platelet count, the times to these nadirs, the duration of thrombocytopenia, or the number of platelet or RBC transfusions required in either the breast cancer or the lymphoma cohort. The median duration of absolute neutropenia (ANC < 100/μL) was 2 days less in patients with breast cancer who were treated with D+G than in those who were treated with P+G (P = .0565; Table 5).
Duration of Thrombocytopenia and Neutropenia in Breast Cancer Patients After PBSC Infusion
DISCUSSION
This study shows that daniplestim 2.5 μg/kg/d can be safely used concurrently with G-CSF 10 μg/kg/d for mobilizing PBSCs before high-dose chemotherapy and PBSC infusion. The results from the study suggest that the addition of daniplestim to G-CSF may (P = .0895) increase the mobilization of PBSCs in patients with breast cancer.
Transplantation of autologous PBSCs has replaced bone marrow transplantation as the preferred method of hematopoietic recovery after myeloablative chemotherapy.1 Its advantages include a relative ease of collection that leads to larger harvests of CD34+ cells, more rapid hematopoietic reconstitution with faster recovery of platelet and neutrophil levels, less need for antibiotics, decreased transfusion requirements, shorter hospital stays, and decreased costs.18-21 In addition, the frequency and extent of tumor contamination may be lower in the PBSC harvest than when bone marrow is used.22-25
Alternatives to a single HGF for mobilization have been studied for their potential to reduce both the number of leukaphereses required to obtain an adequate PBSC product and the associated costs, inconvenience, and risks of the procedure. The sequential use of cytotoxic chemotherapy and G-CSF or GM-CSF for mobilization has increased stem-cell harvests, but this approach poses a risk of neutropenic infection and has not been found to hasten hematopoietic recovery when compared with G-CSF alone.26-28
The case for using a multilineage HGF to increase the collection of PBSCs for autologous transplantation was recently supported by three clinical trials that evaluated the mobilization of G-CSF with stem-cell factor (SCF),29-31 a cytokine with activity that includes the stimulation of pre–lineage-committed hematopoietic progenitor cells.32 In all three studies,29-31 patients were randomized to SCF plus G-CSF or G-CSF alone before high-dose chemotherapy. A study by Moskowitz et al29 in 38 patients with non-Hodgkin’s lymphoma showed that the use of G-CSF with SCF was associated with greater yields of CD34+ cells in the leukapheresis product; the benefit, however, was confined to patients with extensive prior chemotherapy, and the study was not powered to detect statistical significance. In 215 patients with high-risk breast cancer, Glaspy et al30 found that this combination therapy was associated with significantly greater yields of CD34+ cells in the leukapheresis product. In a study of 62 patients with breast cancer and no history of prior chemotherapy, Basser et al31 found that combination therapy was associated with significantly higher yields of CD34+ cells and that the collection of ≥ 5 × 106 CD34+ cells/kg was associated with significantly fewer leukaphereses. In the clinical trials with SCF, the most commonly reported adverse event was injection-site reaction, followed by respiratory symptoms. Overall, 4% of patients had severe systemic reactions that resulted from the mast cell–mediated release of histamine.
The evaluation of mobilization efficacy is complicated by conflicting data regarding the number of CD34+ cells/kg that are required to support engraftment after PBSC infusion.2,33-45 A review of the literature by Demirer et al28 concluded that (1) the criteria used to define adequate PBSC infusions vary considerably, (2) the minimum number of PBSCs required for successful engraftment is not known but is probably between 2.0 × 106 and 3.0 × 106 CD34+ cells/kg, and (3) prolonged platelet recovery is seen with infusions of fewer than 2.0 × 106 CD34+ cells/kg.
In the present study, ≥ 2.5 × 106 CD34+ cells/kg were collected in the leukapheresis product in a greater percentage of patients who were treated with D+G (P = .0895). More patients achieved this count in the D+G group because of sustained levels of CD34+ cells in the peripheral blood, which resulted in more cells being collected with additional leukaphereses. Sustained counts were not observed in the P+G group.
This advantage was not observed in the lymphoma cohort. A possible explanation for this difference is the extent of previous cytotoxic chemotherapy among the patients with lymphoma. Eleven (44%) of 25 patients with lymphoma but only five (19%) of 27 patients with breast cancer had been treated with more than two previous chemotherapy regimens before participating in this study. In analyzing the data on 243 patients with a variety of malignancies to determine the factors that affect the collection of PBSCs before transplantation and the rate of engraftment after transplantation, Bensinger et al2 found that fewer previous cycles of chemotherapy was associated with larger harvests of CD34+ cells. Several other reports have confirmed this finding.33,37,42,46,47 Another possible explanation for the unexpected results with D+G in the patients with lymphoma is that the number of patients was too small for detecting differences between the treatment groups.
The literature generally reports that neutrophil engraftment (ANC ≥ 500/μL) after the reinfusion of PBSCs that had been mobilized with HGFs occurs at a median of 9 to 14 days, and platelet engraftment (≥ 20,000/μL) occurs at a median of 10 to 17 days.2,21,29,30,37,39,42,45,48 In our study, the times to neutrophil and platelet engraftment did not differ significantly between the two treatment groups in the breast cancer or lymphoma cohorts and were similar to those reported by other investigators. The ANC and platelet nadirs, the times to these nadirs, and the durations of neutropenia and thrombocytopenia were also comparable between the treatment groups in our study. However, the median duration of absolute neutropenia (ANC < 100/μL) was shorter (by 2 days) in the patients with breast cancer who had been mobilized with D+G (P = .0565). Given the size of this study, the lack of statistically significant differences between the treatment groups for the engraftment parameters is not unexpected. Similarly, the addition of SCF to G-CSF for PBSC mobilization in patients with non-Hodgkin’s lymphoma29 or breast cancer30,31 was not associated with shorter times to neutrophil or platelet engraftment. Bensinger et al2 found that the number of days to neutrophil engraftment or platelet-transfusion independence after PBSC infusion did not differ significantly between patients who were given chemotherapy plus G-CSF or GM-CSF for mobilization and those who were given G-CSF alone.
In general, the adverse events and the severity of these events were not unexpected, given the patient population in this study. The administration of daniplestim was associated with mild to moderate flu-like symptoms, including fever, headache, and flushing. Unlike recent clinical evaluations of mobilization regimens that include SCF,29-31 which can cause mast cell–mediated reactions,49-51 our study did not exclude patients with histories of asthma or other significant hypersensitivities that are mediated by immunoglobulin E, nor was routine medical prophylaxis against mast cell–mediated reactions required. Nonetheless, there were no allergic-type adverse events or anaphylactoid reactions that were attributable to the use of daniplestim.
In conclusion, daniplestim 2.5 μg/kg/d is hematopoietically active and can be safely used in combination with G-CSF for mobilizing PBSCs in preparation for high-dose chemotherapy and PBSC transplantation. On the basis of the results of this phase II study, a larger randomized, double-blind, phase III study has been initiated to further characterize the ability of daniplestim to mobilize PBSCs in breast cancer patients.
- Received March 23, 1999.
- Accepted March 16, 2000.
