A new form of Filgrastim with sustained duration in vivo and enhanced ability to mobilize PBPC in both mice and humans
Article Outline
Abstract
Granulocyte colony-stimulating factor (G-CSF) has proven effective in the prophylaxis of chemotherapy-induced neutropenia and as a mobilizer of peripheral blood progenitor cells. The longevity of G-CSF action is limited by its removal from the body by two mechanisms. The first is thought to be mediated via receptors (receptor mediated clearance [RMC]) predominantly on neutrophils, the second process is likely the result of renal clearance. With the intention of developing a novel form of Filgrastim (r-met HuG-CSF) with a sustained duration of action in vivo, a new derivative named SD/01 has been made by association of Filgrastim with poly(ethylene glycol). The desired properties of this new agent would include a prolonged duration of action sufficient to cover a complete single course of chemotherapy. SD/01 is shown here to sustain significantly elevated neutrophil counts in hematopoietically normal mice for 5 days. In neutropenic mice effects were noted for at least 9 days, accompanying a significant reduction in the duration of chemotherapy induced neutropenia. Normal human volunteers showed higher than baseline ANC for around 9 to 10 days after a single injection of SD/01. Data from these normal volunteers also indicate that mobilization of CD34+ cells and progenitors may occur in a more timely manner and to around the same absolute numbers as with repeated daily injections of unmodified Filgrastim. These data indicate that SD/01 represents an efficacious novel form of Filgrastim with actions sustained for between one and two weeks from a single injection.
Keywords: Filgrastim SD/01, PBPC mobilization, Neutropenia
Introduction
Filgrastim (r-met Hu-G-CSF) is a recombinant growth factor that has gained widespread clinical use for enhancing both the recovery of neutrophils after chemotherapy and the mobilization of transplantable hematopoietic progenitor populations to the blood [1]. Filgrastim stimulates the production of neutrophils not only in humans [2], but also in several other species including mice [3].
Filgrastim is normally administered by daily injections repeated over several days; the pharmacokinetic profile indicates the half life from a single infusion to be around 160 minutes in humans, though a little shorter in rodents and reducing following repeated daily injections ([4], reviewed in [5]). A form of Filgrastim with a sustained duration measured in days rather than minutes would possibly have fewer scheduling issues associated with its use, fewer issues of compliance (perhaps over weekend treatment periods), and could benefit patients undergoing chemotherapy in terms of a reduced number of injections per cycle of therapy [6]. Uninterrupted treatment may also prove more beneficial in terms of ANC recovery after chemotherapy or transplant and also in terms of PBPC yield. A further possibility offered by a sustained duration form of Filgrastim is that fluctuations in neutrophil counts throughout the day could be minimized by maintaining effective concentrations of Filgrastim in the blood at all times.
Studies have shown that modification of various proteins by the chemical addition of poly(ethylene glycol) can alter the pharmacokinetic and pharmacodynamic properties of the protein to significantly increase the time the modified protein remains effective in the circulation [7]. We have developed a form of Filgrastim modified by the addition of PEG that we show here to have a sustained duration of action in vivo. This material has been named Filgrastim (sustained duration/01) or SD/01. SD/01 is a homogeneous preparation of Filgrastim, which has been modified by the addition of a linear PEG molecule of 20 kD covalently bound to the N-terminus.
Because Filgrastim-SD/01 differs from Filgrastim only by the addition of a 20kD PEG, it was studied in systems in which Filgrastim has been shown to be effective. These systems are: (1) the elevation of neutrophil counts; and (2) the mobilization of PBPC to the blood. The comparison was made between repeated daily injections of unmodified Filgrastim and a single injection of Filgrastim-SD/01 in splenectomized mice. These data were then confirmed in normal human volunteers. To gain insight into the usefulness of this sustained duration form of Filgrastim in a model of chemotherapy-induced neutropenia it was also administered to mice, which had been treated with 5-fluorouracil.
The data presented below show that Filgrastim-SD/01 has the same effects as Filgrastim (elevation of ANC, mobilization of progenitor cells, reduction in duration of chemotherapy-induced neutropenia), but that these effects can be sustained for several days from a single injection of Filgrastim-SD/01 when multiple daily injections of Filgrastim would be required to obtain the same effect. It is also shown that more PBPC can be harvested from mice at an earlier time after a single injection of Filgrastim-SD/01 than five daily injections of unmodified Filgrastim. In humans, CD34+ cells and GM-CFC progenitor cells are also mobilized effectively (in terms of both numbers and time of peak mobilization) when human volunteers are treated with SD/01.
Overall, these data suggest an activity profile of SD/01 which will make it useful in many of the settings where Filgrastim has become the established treatment of choice, but with improved convenience and perhaps also increased efficacy.
Materials and methods
Growth factor administration
SD/01 was administered as a single subcutaneous (SC) injection to mice or humans at doses between 30 and 300 μg/kg in humans and up to 1000 μg/kg in mice. Doses were given as the equivalent amount of protein only; the molecular weight of the PEG is factored into the calculations. Filgrastim was injected at between 50 and 2500 μg/kg in various injection regimens outlined in the Results section.
Mice
Male C57Bl/6 × DBA/2 (BDF1) mice that were 12- to 14-weeks-old and had been surgically splenectomized at least 6 weeks previously, were obtained from Charles River Laboratories and maintained in the Immunocompromized Suite in the vivarium at Amgen Inc. for the duration of the studies. Mice were allowed standard rodent chow and water ad libitum throughout. Mice were injected SC between the scapulae with study drug in a standardized injection volume of 200 μL of normal saline supplemented with 0.1% bovine serum albumin (saline/BSA). Some mice were injected for 5 consecutive days; others were injected only once. Control mice were treated with carrier solution only (saline + 0.1% BSA). Filgrastim was administered at a dose of 200 or 300 μg/kg body weight per day in various experiments as indicated in the Results section.
The effect of SD/01 on recovery from chemotherapy-induced neutropenia was studied in mice given a single IV injection of 5-fluorouracil (FU; 150 mg/kg) on day 0. Two days thereafter, mice were randomized into treatment groups that received either a single SC injection of SD/01 (500 or 1000 μg/kg), daily SC injections of Filgrastim (300 μg/kg), or no further treatment. Daily blood samples were obtained from four mice per treatment group on days 3–12, and the absolute neutrophil count was determined as described below.
Terminal (day 5) bleeding and preparation of mononuclear cell fractions
Mice were euthanized by carbon dioxide inhalation, their chest fur was wetted with 70% isopropanol and blood was withdrawn directly from the heart with a 1-mL syringe fitted with a 25-gauge needle. Blood was collected from groups of ten mice and pooled in tubes containing EDTA as anticoagulant (Vacutainer 6536; Becton Dickinson, Rutherford, NJ). Aliquots of 6 mL of blood were then layered carefully over 25-mL cushions of Accudenz density separation medium (Accurate Chemical and Scientific Corporation, Westbury, NY). The cells were then centrifuged at 800 g for 20 minutes and buoyant density cells collected. These mononuclear cells (MNC) were then washed twice, resuspended, and counted on a Technicon H-1E automated blood analyzer. The count was confirmed manually with a hemocytometer and cells were resuspended at 2 × 106/mL. MNC obtained in this way were diluted and 500,000 cells injected intravenously into groups of ten mice which had been lethally irradiated (12 Gy, given as a split dose of 2 × 6 Gy, 4 hours apart on a GammaCell 137Cs irradiator).
Healthy human volunteers
Following IRB approval of the protocol, and signing of the written informed consent, cohorts of eight healthy volunteers were sequentially enrolled in a study to receive single dose SD/01 at doses of 30, 60, 100, and 300 μg/kg. Eligibility criteria included: normal organ functions, normal weight (± 15% of normal range), and age between 18 and 45 years. Baseline blood samples were drawn, and Filgrastim-SD/01 was administered at a constant volume across dose levels. Blood samples for CBC were drawn at 0.25, 0.5, 0.75, 1, 2, 3, 4, 8, 12, 16, 24, 36, and 48 hours after injection of Filgrastim-SD/01. Thereafter, daily samples were drawn for a total of 15 days. Daily blood samples were drawn for assessment of CD34+ levels in the peripheral blood. Peripheral blood was collected in 3-mL Vaccutainers (Becton Dickinson Corp.) containing EDTA and shipped to our laboratory on wet ice overnight. All analyses were performed within 24 hours of collection. Volunteers were monitored for adverse events on an ongoing basis.
Daily blood sampling
Mice were anesthetized by inhalation anesthetic (Aerrane; Ohmeda Carbide Inc., Guayama, Peurto Rico) and blood was withdrawn from the retro-orbital sinus via heparinized glass capillary tubes into tubes containing EDTA anticoagulant. For daily measurements of PBPC, blood cells were not further processed and were plated as outlined below in the presence of 2.5 ng/mL rmIL-3 and 100 ng/mL rrSCF. In experiments where blood sampling was required every 6 hours, 8 groups of 5 mice were set up and any group of 5 mice bled every 48 hours. No animal was bled more frequently than this to minimize the effects of repeated sampling. Analysis of red blood cells showed no adverse effects on blood parameters due to the sampling frequency (data not shown).
Blood counts
Blood samples obtained as outlined above were counted on a fully automated Technicon H-1E blood analyzer (Technicon Instruments Corp., Tarrytown, NY), which had been calibrated for mouse blood. Human blood samples obtained as outlined above were counted on a Technicon H-1E blood analyzer, which had been calibrated for human blood.
Progenitor assay
MousePreviously determined volumes of whole anticoagulated blood were immobilized in Difco Agar Noble (0.33% final concentration; Difco, Detroit, MI) supplemented by the addition of Iscoves Modified Dulbeccos Medium (Life Technologies, Grand Island, NJ), pretested fetal bovine serum (Cansera, Rexdale, Ontario, Canada), antibiotics, 100 ng/mL recombinant rat SCF, 2.5 ng/mL recombinant murine IL-3 (both Amgen, Inc), and excess L-glutamine. One milliliter of this mixture was added to each of triplicate 35-mm petri dishes (Becton Dickinson) and the dishes were incubated at 37°C in a fully humidified atmosphere of 90% N, 5% O2, and 5% CO2 for 7–10 days. Plates that contained whole blood were flooded with 300 μL 0.03% acetic acid to lyse excess erythrocytes prior to counting and all plates were examined by eye under a dissecting microscope. Colonies of more than 50 granulocyte/macrophage cells were counted.
HumanHuman peripheral blood (5 μL) was added to each tube containing 5-mL Methocult GF+ (Stem Cell Technologies Inc., Vancouver, Canda), vortexed, and allowed to stand for 10 minutes. Each tube contained the following five cytokines all at 100 ng/mL: rhSCF; rhIL-3; rhIL-6; rhG-CSF; and rhGM-CSF. One milliliter was added to each of three, 35-mm petri dishes and incubated for 14 days at 37°C in 5% CO2 in 100% humidified air. At the end of the incubation period, GM-CFC derived colonies were scored under a dissecting microscope at 20× and identified as colonies when containing more than 50 granulocyte-macrophage cells.
CD34 analysis in humansCD34 analysis was performed with the ProCOUNT progenitor cell enumeration kit (Becton Dickinson Immunocytometry Systems [BDIS], San Jose, CA) and fluorescence activated cell sorting (FACS) analysis according to the manufacturer's directions on a FACS Calibur (BDIS). Briefly, 50 μL of each sample of peripheral blood from each patient was added to each of two TruCOUNT tubes (BDIS). Each tube contains a bead of fluorescent particles used for the determination of absolute cell numbers. Then, 20-μL ProCOUNT control (nucleic acid dye, IgG1, and CD45) and test (nucleic acid dye, CD34, and CD45) reagents were added to each respective tube and incubated in the dark at room temperature for 15 minutes with vortexing at 5-minute intervals. After the incubation period, each tube received 450 μL of 1× FACS lysing solution (BDIS) and was allowed to stand in the dark for at least 15 minutes before acquiring on the FACS Caliber. Cell Quest software (BDIS) was utilized for analysis of each tube according to the manufacturer's recommendations. Results were calculated by subtracting control results from test results.
Statistical methods
Where appropriate, the differences between treatment groups were tested for significance using Student's t-test, part of the SigmaStat statistical analysis software package (Jandel Scientific, San Rafael, CA).
Results
Treatment of mice
Neutrophil responseThe neutrophil numbers increased in response to a single dose of 0-, 100-, 500-, 1000-, or 2500-μg Filgrastim/kg body weight (Fig. 1A). The degree of granulocytosis was dependent on the dose of Filgrastim. The duration of the response was a maximum of 2 days. Neutrophil counts returned to normal between days 2 and 3. Total doses of 0-, 500-, 1000-, or 2500-μg Filgrastim/kg body weight were also split into five equal treatments and administered by daily injection over 5 days (a daily dose of 100-, 200-, or 500-μg/kg per day). This resulted in neutrophil counts that increased over the injection period, culminating in a broadly dose-dependent granulocytosis of greater magnitude than the response seen in mice treated with the same dose as a bolus injection (Fig. 1B). The majority of cells recruited by Filgrastim were neutrophils whether the injections were given only once or repeated over the injection period (data not shown).
Figure 1.
(A) Neutrophil counts in splenectomized mice treated with a single injection of Filgrastim (G-CSF) on day 0. Each point represents the mean of 5–10 individual mice assessed separately ± SEM. Data are shown from a single representative experiment. The complete experiment was performed three times. No errors are given at day 5 because the blood from five donors was pooled and a single measurement was made. Where data differ significantly from carrier *p < 0.005 and **p < 0.0001. (B) Leukocyte counts in splenectomized mice treated with five daily injections of Filgrastim (G-CSF) on days 0–4. Each point represents the mean of 5–10 individual mice assessed separately ± SEM. Data are shown from a single representative experiment. The complete experiment was performed three times. No errors are given at day 5 because the blood from five donors was pooled and a single measurement made. Where data differ significantly from carrier *p < 0.005 and **p < 0.0001. (C) Neutrophil counts in splenectomized mice treated with a single injection of SD/01 on day 0 compared with mice that received five daily injections of unmodified Filgrastim on days 0–4. Each point represents the mean of 5–10 individual mice assessed separately ± SEM. Data are shown from a single representative experiment. The complete experiment was performed three times. No errors are given at day 5 because the blood from five donors was pooled and a single measurement was made. Where data differ significantly from carrier *p < 0.005 and **p < 0.0001, where data differ significantly from the response in Filgrastim-treated mice bp < 0.005 and bbp < 0.0001. (D) Single injection of SD/01 (1000 μg/kg) at day 0 compared to daily injections of 200 μg/kg per day Filgrastim (total dose = 1000 μg Filgrastim/kg). Six hourly neutrophil counts from groups of 10 normal (nonsplenectomized) mice; → = Filgrastim injection
SD/01 was administered as a single injection on day 0 of these studies. The changes in neutrophil counts were compared to carrier treated animals and those injected daily for 5 days with Filgrastim at 200 μg/kg per day (Figs. 1B and 1C). SD/01 was injected at doses between 50- and 1000-μg/kg body weight. One day after the injection of SD/01, neutrophils were elevated to around 10 × 103/μL irrespective of the dose. At this time, mice that had been injected with Filgrastim had around 8 × 103 neutrophils per μL of blood. Thereafter, the response to various doses of SD/01 diverge with the higher doses leading to a greater peak neutrophil number and an extended duration of leukocytosis (see Fig. 1C). Filgrastim stimulated a time-dependent increase in neutrophils to a value of 50 × 103/μL by day 5; broadly comparable with the peak resulting from a single injection of 1000 μg/kg SD/01, but occurring 2–3 days later.
In another series of experiments detailed measurements of the neutrophil response at six hourly intervals after SD/01 injection were examined. Data are shown in Fig. 1D. Fluctuations in neutrophil count through the day were minimized after SD/01 injection in comparison to daily Filgrastim injections.
Platelet numbers decreased throughout the observation period in response to daily Filgrastim or a single injection of SD/01. The magnitude of this response was comparable for both forms of Filgrastim (Fig. 2).
Figure 2.
Platelet numbers fell during treatment of splenectomized mice with human G-CSF preparations. Five individual mice per point (mean ± SEM). The results of a single experiment are shown, the experiment was done twice. Where data differ significantly from carrier *p < 0.005 and **p < 0.0001, where data differ significantly from the response in Filgrastim-treated mice bp < 0.005 and bbp < 0.0001
Mobilization of PBPC was assessed by measuring the number of IL-3/SCF responsive GM-CFC in the blood of treated donor animals. Predictably, repeated injections of Filgrastim at 200 μg/kg per day caused the accumulation of high numbers (100-fold increased over baseline) of PBPC in the blood over the course of the 5-day treatment. SD/01, on the other hand, caused a more rapid relocation of PBPC to the peripheral blood compartment when administered as a single subcutaneous injection of 1000 μg/kg at day 0. This early peak in PBPC numbers persisted for 2 to 3 days, and showed higher numbers (300-fold above baseline) than Filgrastim even at the peak of the Filgrastim response (day 4–5), see Fig. 3. PBPC from donor mice treated with G-CSF or SD/01 were also transplanted into irradiated recipient animals. The dose of G-CSF-mobilized PBPC was chosen from previous experiments to be sufficient to rescue approximately half the animals. The actual number of mice surviving to 30 days in the G-CSF-PBPC group was 60%. PBPC mobilized with carrier, 50 μg/kg SD/01 or 100 μg/kg SD/01 failed to rescue mice whereas the same cell dose (500,000 cells) from 500 or 1000 μg/kg SD/01 rescued 90%–100% of the animals, see Fig. 4.
Figure 3.
GM-CFC mobilization to the blood by repeated injections of Filgrastim (days 0–4) or a single injection of Filgrastim-SD/01 at day 0. Five individual mice per point (mean ± SEM). The results of a single experiment are shown, the experiment was done twice. Where data differ significantly from carrier *p < 0.005 and **p < 0.0001, where data differ significantly from the response in Filgrastim-treated mice bp < 0.005 and bbp < 0.0001
Figure 4.
The survival of mice transplanted with PBPC harvested at day 5 from donors treated with either G-CSF or Filgrastim SD/01. Ten mice were transplanted in each group. The experiment was performed twice
Neutropenia was induced in 160 female BDF1 mice by a single IV injection of 5-fluorouracil (5-FU) at a dose of 150 mg/kg. The mice were then randomized into treatment groups as follows: a single SC injection of Filgrastim-SD/01 (500 or 1000 μg/kg) on day 2 after 5-FU treatment; daily SC injections of Filgrastim (300 μg/kg) on days 2 through 11; or no further treatment after 5-FU. Beginning on day 3, four mice in each group were euthanized by carbon dioxide inhalation, blood samples were taken by cardiac puncture, and absolute neutrophil counts (ANCs) were recorded. The 5-FU-treated mice became neutropenic on day 5, as indicated by mean ANCs, and mice not subsequently treated with Filgrastim-SD/01 or Filgrastim remained neutropenic until day 12. The suppressed neutrophil counts returned to pretreatment values on day 7 in mice given a single 1000-μg/kg injection of Filgrastim-SD/01, and in mice receiving daily injections of Filgrastim (Fig. 5). Neutrophil counts returned to pretreatment values on day 8 in mice given a single injection of Filgrastim-SD/01 at 500 μg/kg.
Figure 5.
Normal human volunteers were injected with SD/01 at zero time. Analysis of peak ANC and duration of response are given in Table 2
Treatment of normal human volunteers with SD/01
Neutrophil responseA total of 32 normal volunteers (8 per dose level) were treated at a single US site with doses of SD/01 ranging from 30- to 300-μg/kg body weight administered as a single injection. This was a sequential dose escalation study. Table 1 shows that the volunteers were well balanced with respect to age, gender and race. Neutrophil numbers increased in a dose- and time-dependent manner (see Fig. 6). The peak ANC attained and the longevity of response were dose related (see Fig. 6 and Table 2). There was no evidence of a rebound neutropenia.
Table 1. Demographic summary for the study of SD/01 in normal volunteers
| SD/01 | |||||
|---|---|---|---|---|---|
| 30 mg/kg (n = 8) | 60 mg/kg (n = 8) | 100 mg/kg (n = 8) | 300 mg/kg(n = 8) | Total (n = 32) | |
| Gender | |||||
| Female | 5 | 5 | 5 | 5 | 20 |
| Male | 3 | 3 | 3 | 3 | 12 |
| Ethnicity | |||||
| Other | 0 | 0 | 1 | 0 | 1 |
| Black | 0 | 1 | 1 | 0 | 2 |
| Hispanic | 3 | 2 | 0 | 1 | 6 |
| Native American | 0 | 0 | 0 | 0 | 0 |
| White | 5 | 5 | 6 | 7 | 23 |
| Age (years) | |||||
| Mean | 26.9 | 26.3 | 27.8 | 25.3 | 26.5 |
| Median | 24.0 | 25.0 | 27.5 | 24.0 | 25.0 |
| S.D. | 7.0 | 6.6 | 7.0 | 3.9 | 6.0 |
| Range | 20.0–37.0 | 18.0–36.0 | 20.0–41.0 | 22.0–33.0 | 18.0–41.0 |
| Baseline ANC (x 109/L) | |||||
| Mean | 4.71 | 4.15 | 4.86 | 4.79 | 4.63 |
| Median | 4.9 | 4.3 | 5.0 | 4.4 | 4.7 |
| S.D. | 0.9 | 1.6 | 0.9 | 1.0 | 1.1 |
| Range | 2.9–6.0 | 1.8–6.1 | 3.7–6.0 | 3.4–6.2 | 1.8–6.2 |
| SD/01 = Filgrastim-SD/01. | |||||
Figure 6.
Mobilization of CD34+ cells into the blood by a single injection of Filgrstim SD/01 at ascending doses
Table 2. Median (range) parameter values in normal volunteers following single SC administration of 30, 60, 100, and 300 mg/kg Filgrastim-SD/01
| Filgrastim-SD/01 | ||||
|---|---|---|---|---|
| Parameter | 30 mg/kg (n = 8) | 60 mg/kg(n = 8) | 100 mg/kg (n = 8) | 300 mg/kg (n = 8) |
| ANCmax (x106/mL) | 30.4 | 31.2 | 36.7 | 50.8 |
| (21.0–35.2) | (24.1–36.0) | (29.3–51.1) | (30.0–96.0) | |
| ANCTmax (days) | 2.50 | 2.75 | 3.00 | 5.00 |
| (1.5–2.50) | (2.50–4.00) | (2.00–5.00) | (4.00–6.00) | |
| Duration of response (days) | 5.79 | 8.79 | 8.29 | 8.79 |
| (4.75–9.88) | (3.83–13.9) | (4.67–8.92) | (6.83–13.8) | |
| AOBEC | 101 | 125 | 141 | 223 |
| (x106 · day/mL) | (74–152) | (90–170) | (100–189) | (134–342) |
| ANCmax = Maximum observed absolute neutrophil counts; ANCTmax = time to ANCmax; AOBEC = area over baseline effect curve. | ||||
CD34+ cells were mobilized to the blood by SD/01 at the highest dose (300 μg/kg), a peak number of 120×103 CD34+ cells/μL was seen at day 4 (see Fig. 7). In this, cohort numbers had approached normal by day 12 or 13, while the other dose groups had peaked at the same time, but returned to the normal range by day 9.
Figure 7.
Mobilization of PBPC (IL-3/SCF responsive GM-CFC) by a single injection of SD/01 to normal volunteers
Correlating broadly with CD34+ mobilization, PBPC (measured as GM-CFC/mL) were also mobilized by SD/01 (see Fig. 8). Again, a broad dose dependence was noted, but now with a maximum effect seen at 100 μg/kg; nothing was gained in terms of GM-CFC/mL by further increasing the dose to 300 μg/kg.
Figure 8.
Dose of SD/01 versus number of GM-CFC in the blood of normal volunteers
Clinical adverse events observed in this study are summarized in Table 3 and included mild to moderate bone pain and headache. Mild transient thrombocytopenia was observed at the highest SD/01 dose levels. Transient elevations in serum enzymes (LDH, AP) and uric acid were also observed and had no clinical sequelae.
Table 3. Summary of adverse events that were recorded in normal volunteers treated with SD/01
| Filgrastim-SD/01 | |||||
|---|---|---|---|---|---|
| 30 μg/kg(n = 8) | 60 μg/kg (n = 8) | 100 μg/kg (n = 8) | 300 μg/kg (n = 8) | Overall (n = 32) | |
| Headache | 5 | 4 | 6 | 8 | 23 |
| Pain back | 6 | 3 | 3 | 2 | 14 |
| Myalgia | 2 | 2 | 5 | 3 | 12 |
| Arthralgia | 3 | 0 | 2 | 3 | 8 |
| Nausea | 2 | 2 | 0 | 3 | 7 |
| Pharyngitis | 0 | 3 | 1 | 3 | 7 |
| Pain chest | 0 | 1 | 2 | 0 | 3 |
| Vomiting | 1 | 0 | 1 | 1 | 3 |
| Pain skeletal | 1 | 0 | 1 | 1 | 3 |
| Sweating increased | 0 | 1 | 0 | 2 | 3 |
Discussion
A form of Filgrastim that is active for several days may have beneficial clinical properties such as reduced frequency of injection, better patient compliance, and, possibly, greater efficacy. We have shown here in mice that a single bolus dose of Filgrastim could not sustain elevated neutrophil counts beyond 24–48 hours irrespective of the dose injected. Thus, merely increasing the dose of Filgrastim will not in itself extend the duration of the resulting neutrophilia. It was confirmed, however, that maintaining a more modest treatment level for several days was much more effective in sustaining increased neutrophil counts. Current clinical management of chemotherapy-induced neutropenia and congenital neutropenia with Filgrastim reflects this requirement for repeated dosing with Filgrastim over several days [8]. The main aim of the study presented here was to evaluate the potential of SD/01, a novel form of Filgrastim, that was engineered to sustain greater longevity in vivo.
The experimental model chosen for some of these studies was the splenectomized mouse. Mice that lack a spleen generally have a better quantitative response to G-CSF [9] than do intact mice. The reasons for this effect are not clear, but it would appear that the rodent spleen, rather than contributing to the development of leukocytosis, actually dampens the effect in an attempt to re-establish homeostasis. The spleen does not play the same role in primates or humans. Because the splenectomized mouse lacks to a degree the ability to buffer excessive peripheral blood counts it represents a more sensitive and relevant model for this type of study.
The dose response data illustrated in Fig. 1B show that 200-μg Filgrastim/kg body weight/day is an optimum dose for increasing ANC in mice. Increasing the dose to 500-μg/kg per day does not result in a further increase in neutrophil counts. The doses of Filgrastim given to mice do not correspond directly with the doses given to humans. Based on these data, this human cytokine would appear to be around tenfold less active in mice than it is in humans. This is consistent with the required dosing in primates where 100 μg/kg is effective [10]. However, Fig. 1B indicates that by using 200-μg Filgrastim/kg per day we are obtaining a maximum response in ANC and it is, therefore, a fair baseline with which to compare the effects of SD/01. We have also confirmed that this is an optimum dose for PBPC mobilization in splenectomized mice (data not shown). Though this repeated injection regimen reflects what is done in current clinical practice we have shown previously that different treatment regimens, for instance continuous infusion, can produce a better response in ANC and PBPC mobilization [11].
A single injection of SD/01 into mice resulted in increased neutrophil numbers for several days. Indeed, when given in hematopoietic steady-state, the peak neutrophil count obtained was the same (at 50×103/μL) from a single injection of 1000 μg/kg of SD/01 as it was from five injections of Filgrastim spaced over 5 days (total Filgrastim dose 1000 μg/kg). The critical observation was the persistence of the peak neutrophil count until 4 days after a single injection of SD/01. Shortly after the single injection of SD/01 (at days 2, 3, and 4), the ANC count was significantly higher in the SD/01-treated mice than in the Filgrastim-treated cohort. By day 5 the effect of SD/01 was waning and the increasing effect of repeatedly injected Filgrastim caused the neutrophil count in the SD/01-treated mice to fall below that of the Filgrastim-treated group. Thus, at a dose of 1000 μg SD/01 per kg body weight there was a peak ANC response comparable to Filgrastim, but this peak was attained at least 3 days sooner and persisted until day 4 after injection. This profile indicates that SD/01 is an effective sustained duration form of Filgrastim. By increasing the sampling frequency in treated mice to every 6 hours, it was observed that throughout the day there were fluctuations in neutrophil counts related to the time of injection, which confirms previously published data [3]. Interestingly, though some diurnal variation in neutrophil counts still persisted after treatment with SD/01, the magnitude of the fluctuations was minimized and numbers above (an arbitrary level of) 5×103 neutrophils/μL were maintained for almost 5 days with no evidence of diurnal variation of sufficient magnitude to allow neutrophil counts to fall below this level. This profile may be of interest in various therapeutic areas.
It is also interesting to consider the effects of lower doses of SD/01; reducing the dose to 500-, 100-, or 50-μg/kg per day lowered not only the peak ANC attained, but also the longevity of the effect. It is likely that the combination of increased dose and sustained duration at the higher treatment level maintained an effective circulating concentration of SD/01 for several days longer than a single injection of unmodified Filgrastim [12]. The early response to SD/01 (days 1 and 2) was greater than the Filgrastim effect despite some treatment groups receiving only 50 or 100 μg/kg in comparison to 200 μg/kg of unmodified Filgrastim. In Fig. 1 it is shown that the short-term response to Filgrastim is independent of the dose and no significant difference from carrier-treated mice was seen at any treatment level by day 2. In contrast, doses of 50- or 100-μg SD/01/kg resulted in a significant neutrophilia at day 2, a phenomenon that can only be ascribed to the sustained duration of SD/01 and not to the size of the bolus dose.
The effectiveness of SD/01 in sustaining elevated neutrophil counts for several days is presumably linked to its pharmacokinetic properties 13, 14, 15. The material was injected subcutaneously in these studies and so it is possible that some of the effects are mediated via slower absorption to the circulation from the subcutaneous site. This would not appear to be a predominant factor as the early response to SD/01 is more marked than the response to similarly injected unmodified Filgrastim. At this stage we can only assume that SD/01 is cleared from the circulation by the same mechanisms as G-CSF, namely renal clearance [16], and destruction after internalization of G-CSF receptor/ligand complexes. It has been shown that PEGylation of proteins can extend their circulating half-lives; and in at least one case this has been shown to be due to reduced renal clearance [17], but little is known of the effects on neutrophil-borne G-CSF receptor-mediated clearance. Should renal clearance of SD/01 be reduced, as it would seem reasonable to assume, then clearance would depend in part on the number of neutrophils in the circulation. In neutropenia it is possible that SD/01 would persist longer than it does in normal or neutrophilic animals. It is likely, therefore, that in these studies in normal animals and humans the longevity of SD/01 is underestimated.
Mice that have been treated with 5-FU at 150 mg/kg show a marked neutropenia, which lasts for several days. Appropriately administered Filgrastim can reduce the duration of this neutropenia significantly. The data presented in Fig. 5 show that SD/01 is actually superior to daily injections of Filgrastim in reducing both the depth and the duration of neutropenia. Careful consideration of the doses used has revealed that a single injection of 1000 μg/kg is more effective than an accumulated dose of 1500 μg/kg Filgrastim given over 5 days. It would appear, therefore, that the hypothesis discussed above is not refuted by these data. In the neutropenic setting, SD/01 would indeed appear to be more effective (in terms of longevity) than would be predicted from the experiments in normal, non-neutropenic mice. This may be explained by the routes of elimination discussed above, i.e., when normal (or supranormal) neutrophil numbers exist then receptor mediated clearance (via neutrophil borne receptors) could predominate over renal based clearance [15]. Because SD/01 contains a PEG molecule of sufficient size to impair renal-mediated clearance [17], and neutrophil mediated clearance is minimized in neutropenic animals, the longevity of SD/01 is extended beyond that which may be observed in normals. Should this hypothesis be correct it raises the possibility that SD/01 could have a self-regulating aspect to its pharmacodynamic effects [17]. This would manifest itself as greater longevity and increased circulating levels of the drug during times of neutropenia, but then as neutropoiesis was stimulated by SD/01 and neutrophil numbers returned towards normal, clearance would be accelerated and the drug cleared more quickly.
Platelet numbers were reduced in this study in the splenectomized mice. This effect has been reported previously 18, 19 in mice, though how far it extends to the human situation is unclear [8]. Effects of G-CSF on the erythroid lineage have also been reported [20]. To date, the mechanism and consequences of these cross-lineage effects have not been defined.
The mobilization of PBPC by growth factors in mice has been shown in a number of studies to be predictive of the clinical effects of these factors (e.g., G-CSF, GM-CSF, and SCF 21, 22). Blood harvested from mice treated with Filgrastim has been shown previously to contain sufficient PBPC to protect mice from the lethal effects of irradiation [22].
SD/01 mobilized progenitor cells to the peripheral blood with markedly different kinetics than daily injections of Filgrastim (Fig. 3). The maximum number of PBPC reached was three times greater than that attained with daily Filgrastim. The peak number also appeared earlier; at day 3 with SD/01 compared with day 4–5 with Filgrastim. The progenitor population measured daily in these studies was granulocyte/macrophage colony-forming cells, which grew in response to SCF and IL-3. These cells are widely used as a marker of PBPC, but perhaps do not represent the cells responsible for radioprotection after transplantation to an ablated host. However, Fig. 4 shows that cells capable of radioprotection are also mobilized by SD/01. In human subjects it was shown that mobilization of CD34+ cells to the circulation by SD/01 shows some interesting differences from mobilization of CD34+ cells by Filgrastim. No Filgrastim-only treatment group was included in this phase I trial so it is not possible to make a direct comparison in this experiment. However, based on comparison with published data this would appear to be a very effective mobilizer of potentially transplantable progenitors in humans. Data on the timeliness of mobilization in the absence of chemotherapy are infrequently reported and most data rely on recovery of CD34+ cells in leukapheresis products. The data presented here were obtained using the ProCOUNT system which more easily allows enumeration of CD34+ cells from whole blood. Recently, Lane et al. [23] have reported that in a regimen of 5 days of Filgrastim treatment in normal donors the time of peak mobilization was 5.3 days. Data reported here for SD/01 compare favorably with this figure. However, a dose of 10-μg Filgrastim/kg for 5 days may not be comparable with a single dose of 300-μg/kg SD/01 and a peak of 67 ± 16 CD34+/μL (Lane et al. [23]) obtained with different methodology may not be comparable with the 50 and 120 CD34+/μL at day 4 after 100- or 300-μg SD/01/kg reported here. There exists in the data presented in this study a discrepancy between the mobilization of CD34+ cells and the progenitor population GM-CFC; the peak in CD34+ numbers and GM-CFC differs by 1 day. Though, in general, the data are comparable, there is greater variability in the GM-CFC data due, presumably, to the nature of the assay systems. With only eight volunteers per cohort it is not possible to discuss this divergence between the two assay systems, and confirmation will require a more extensive clinical program designed to investigate this issue. Divergence between the different assay systems has been shown before, especially when combined with chemotherapy [24].
In the phase I study in healthy volunteers, which has been completed, 32 subjects were treated (8 per dose cohort) and the consistency and dose-relatedness of the data make them noteworthy. Figure 6 shows that a dose- and time-dependent neutrophilia developed in response to a single injection of SD/01. Figure 7, Figure 8 also show that mobilization of potentially transplantable progenitors is easily obtainable with SD/01. This study also documented the adverse event profile following single dose administration of SD/01. The most frequently observed adverse event (AE) was mild to moderate headache and musculoskeletal pain. The AE profile was similar to that obtained with Filgrastim. The cause of bone pain after hematopoietic growth factor treatment is not well defined, and possibly those AEs that were noted in healthy volunteers may be less frequent when chemotherapy patients are treated.
The longevity of the neutrophil response in normal volunteers treated with SD/01 is longer than that seen in mice. Despite the similar kinetics of neutrophil production in mice and men 25, 26, and comparable peak ANCs in the murine and human studies presented here there is a discrepancy in the longevity of response between mice and humans. At present no explanation can be offered for this difference.
Considering the kinetics of neutrophil production, Fig. 6 shows that at the highest dose (300 μg/kg) the neutrophil count falls from a peak of 55×103/μL to 20×103/μL between 100 and 150 hours, when the rate of loss is at a maximum. Should de novo production of neutrophils have stopped at 100 hours, neutrophils could be expected to be eliminated at the rate of the half life of normal neutrophils, i.e., a 50% reduction every 8 hours or so. The data illustrated in Fig. 6 show that the descending curve actually reflects a lower rate of loss than this; indicating that either the half life of neutrophils is extended (which seems unlikely when Filgrastim-induced neutrophils have a normal residence time in the circulation) or that neutrophils continue to emerge into the circulation despite the declining neutrophil numbers. As the curve levels out after 192 hours, the rates of neutrophil production and loss are becoming equivalent. Thus, even at peak neutrophil numbers it would appear that neutrophil release into the circulation is not halted, only slowed. Then as neutrophil numbers fall, production gradually increases again. This would explain why we did not observe a rebound subnormal neutrophil number after the drug had been eliminated.
Recently, two preliminary reports have shown that a multi-PEGylated derivative of nartograstim (a human G-CSF mutein), RO 25-8215, results in a dose-dependent increase of CD34+, GM-CFC, and BFU-E in normal healthy human subjects with greater efficiency than Filgrastim and with a 4 day peak in response 27, 28. In the study presented here SD/01 is also shown to be effective in humans.
Overall, these studies indicate that a single injection of SD/01 is at least as effective as five repeated daily injections of unmodified Filgrastim when given in the non-neutropenic setting, and longer when given following myelosuppressive chemotherapy. This holds true for elevation of neutrophil numbers in splenectomized mice and mobilization of progenitor cells to the peripheral blood. The PBPC mobilized by SD/01 had progenitor content and radioprotection ability at least equivalent to a similar product harvested from mice receiving five daily injections of Filgrastim. Should these data be confirmed in patients (as is suggested by the presented data in normal volunteers) then it is possible that SD/01, a second generation Filgrastim molecule modified by the addition of poly(ethylene glycol), will allow sustained Filgrastim effects to be obtained from a single injection. Whether SD/01 has sufficient longevity in vivo to be a once-per-chemotherapy-cycle agent awaits confirmation of our data in a clinical chemotherapy setting, though preliminary data are encouraging in humans [15] and primates [29].
References
- Welte K, Gabrilove J, Bronchud MH, Platzer E, Morstyn G (1996) Filgrastim (r-metHuG-CSF): the first 10 years. Blood 88:1907
- . Phase I/II study of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemotherapy for small cell lung cancer. Br J Cancer. 1987;56:809
- . Kinetics and mechanisms of recombinant human granulocyte-colony stimulating factor-induced neutrophilia. Am J Pathol. 1988;133:630
- . Pharmacokinetics of recombinant human granulocyte colony-stimulating factor in the rat. Single and multiple dosing studies. Drug Metab Dispos. 1991;19:200
- . Pharmacology of filgrastim (r-metHuG-CSF). In: Morstyn G, Dexter TM, Foote MA editor. Filgrastim in clinical practice. 2nd edition. New York: Marcel Dekker; 1998;p. 51
- . Pharmacokinetics of recombinant human granulocyte colony-stimulating factor conjugated to poly(ethylene glycol) in rats. Cancer Res. 1991;51:3710
- Delgado C, Francis GE, Fisher D (1992) The uses and properties of PEG-linked proteins.
- (1998) In: G Morstyn, TM Dexter, MA Foote (eds) Filgrastim in clinical practice, 2nd edition. New York: Marcel Dekker.
- . A comparison of hematopoiesis in normal and splenectomized mice treated with granulocyte colony-stimulating factor. Blood. 1990;75:563
- . Recombinant human granulocyte-colony stimulating factor (in vivo effects on myelopoiesis in primates). Behring Inst Mitt. 1988;83:102
- . Continuous subcutaneous infusion of hematopoietic growth factors in mice. Exper Hematol. 1995;23:880
- . A phase I study of sustained-duration filgrastim in normal volunteers. Exper Hematol. 1998;26:709
- . A cytokinetic model describes the granulopoietic effects of r-metHuG-CSF-SD/01 (SD/01) and the homeostatic regulation of SD/01 clearance in normal volunteers. Clin Pharmacol Ther. 1999;65:196
- . Modelling of r-metHuG-CSF-SD/01 (SD/01) mediated granulopoiesis in normal animals with mathematical extrapolation to neutropenic settings. Blood. 1998;92(Suppl 1):379A
- . Cytokinetic model of rmethHuG-CSF-SD/01 (SD/01) mediated granulopoiesis and the and the “self-regulation” of SD/01 elimination in non-small cell lung cancer (NSCLC). Blood. 1998;92(Suppl):507A
- . Evaluation of the rapid plasma elimination of recombinant alpha1-proteinase (synthesis of poly(ethylene glycol) conjugates with improved therapeutic potential). J Lab Clin Med. 1990;116:58
- . Influence of renal and hepatic failure on the pharmacokinetics of recombinant human granulocyte colony-stimulating factor (KRN8601) in the rat. Cancer Res. 1990;50:6615
- . Megakaryocyte growth and development factor accelerates platelet recovery in peripheral blood progenitor cell transplant recipients. Blood. 1996;88:366
- . An analysis of the effects of combined treatment with rmGM-CSF and PEG-rHuMGDF in murine bone marrow transplant recipients. Stem Cells. 1997;15:43
- . Mutual inhibition of murine erythropoiesis and granulopoiesis during combined erythropoietin, granulocyte colony-stimulating factor, and stem cell factor administration (in vivo interactions and dose-response surfaces). Blood. 1994;84:4157
- . An analysis of the effects of combined treatment with rmGM-CSF and PEG-rHuMGDF in murine bone marrow transplant recipients. Stem Cells. 1997;15:1
- . Transplantation potential of peripheral blood stem cells induced by granulocyte colony-stimulating factor. Blood. 1990;76:2153
- . Mobilization of blood-derived stem and progenitor cells in normal subjects by granulocyte-macrophage- and granulocyte-colony-stimulating factors. Transplantation. 1999;39:39
- . Mobilisation kinetics of primitive haemopoietic cells following G-CSF with or without chemotherapy for advanced breast cancer. Ann Oncol. 1996;7:1051
- . Haemopoietic progenitor and myeloid cell kinetics in humans treated with interleukin-3 and granulocyte/macrophage colony-stimulating factor in combination. Int J Cancer. 1994;59:483
- . Myeloid cell kinetics in mice treated with recombinant interleukin-3,granulocyte colony-stimulating factor (CSF), or granulocyte-macrophage CSF in vivo. Blood. 1991;77:2154
- . A single injection of RO 25-8315, a long-acting pegylated G-CSF mutein, is safe and effective in mobilizing peripheral blood progenitor cells (PBPC) in healthy volunteers. A clinical pharmacology phase I study. Bone Marrow Transplant. 1997;19(Suppl 1):S12
- . Safety/tolerability and pharmacokinetics/pharmacodynamics (PK/PD) following administration of ascending subcutaneous doses of a long-acting (5-7 days), chemically modified G-CSF mutein (Ro 25-8315) to healthy subjects. Proc ASCO. 1997;16:87a
- . A single administration of r-metHuG-CSF-SD/01 (SD/01) significantly improves neutrophil recovery following autologous bone marrow transplantation. Blood. 1998;92(Suppl.):112A
PII: S0301-472X(99)00112-5
© 1999 International Society for Experimental Hematology. Published by Elsevier Inc All rights reserved.
