Extended Activity in Cynomolgus Monkeys of a Granulocyte ColonyStimulating Factor Mutein Conjugated With High Molecular Weight Polyethylene Glycol JAMES F. ELIASON,a,b ANTHONY GREWAY,c NADINE TARE,d TOMOAKI INOUE,e SHARON BOWEN,d MARILYN DAR,c MOTOO YAMASAKI,f MASAMI OKABE,g IKUO HORIIe
a
Dept. of Internal Medicine, Wayne State School of Medicine, Detroit, Michigan, USA; bBarbara Ann Karmanos Cancer Institute, Detroit, Michigan, USA; cDrug Metabolism and Pharmacokinetics; dDrug Discovery, Roche Research Center, F. Hoffmann-La Roche, Inc., Nutley, New Jersey, USA; eDept. of Toxicology, Nippon Roche Research Center, Kamakura, Japan; fTokyo Research Laboratories; g Pharmaceutical Research & Development Division, Kyowa Hakko Kogyo Co. Ltd., Tokyo, Japan Key Words. G-CSF · Neutrophils · In vivo · Mutein · Monkeys · Pharmacokinetics
A BSTRACT The activity of a granulocyte colony-stimulating factor (G-CSF) mutein (nartograstim; [NTG]) conjugated with an average of two polyethylene glycol (PEG) chains per protein molecule was examined in cynomolgus monkeys following a single s.c. injection. Groups of monkeys were given 10 µg/kg, 30 µg/kg, or 100 µg/kg. For comparison, one group of monkeys was given 5 µg/kg of recombinant human G-CSF (rHuG-CSF) daily for six days. In monkeys given 100 µg/kg of PEG-NTG, neutrophil levels reached a peak one day after injection approximately 20-fold higher than baseline levels. Neutrophil numbers in these animals were still significantly elevated six days after injection. In contrast, peak neutrophil levels in monkeys given six injections of rHuG-CSF reached a peak only on day 6 and were approximately the same as that in monkeys given a single dose of PEG-NTG six days before. Pharmacokinetics of PEG-NTG in these monkeys indicated that the area
under the plasma concentration time curve (AUC) increased with increasing the dose from 497 ng•h/ml at 10 µg/kg, 6,140 ng•h/ml at 30 µg/kg to 27,900 ng•h/ml at 100 µg/kg. In a separate study, the effects of single doses of 100 µg/kg of PEG-NTG, rHuG-CSF, and unmodified NTG were compared. In this experiment, peak numbers of neutrophils were reached two days after injection in animals receiving PEG-NTG and one day after in animals given unmodified proteins. The pharmacokinetic parameters demonstrated increased exposure for PEGNTG relative to the unmodified proteins with an AUC0∞ of 21,012 ng•h/ml compared with 5,492 ng•h/ml for rHuG-CSF and 5,153 ng•h/ml for NTG. These results demonstrate that conjugation of a G-CSF mutein with high molecular weight PEG results in a preparation that can induce prolonged elevation of neutrophils in normal nonhuman primates following a single injection. Stem Cells 2000;18:40-45
INTRODUCTION First identified as an activity in leukemia cell conditioned medium that stimulated formation of pure granulocyte colonies when added at low concentrations to bone marrow cell cultures [1], G-CSF was purified from murine [2] and human [3] sources in the early 1980s. Cloned human G-CSF [4] entered
clinical trials in 1986 and is beneficial in situations where circulating levels of neutrophils are depressed, as in patients receiving chemotherapy with anticancer drugs and in chronic congenital neutropenia [5]. We have previously described an approach for creating G-CSF preparations with longer in vivo half-lives, so that the
Correspondence: James F. Eliason, Ph.D., Karmanos Cancer Institute, HWCRC 724, 110 E. Warren Avenue, Detroit, Michigan 48201 USA. Telephone: 313-966-7858; Fax: 313-966-7558; e-mail:
[email protected] Accepted for publication December 10, 1999. ©AlphaMed press 1066-5099/2000/$5.00/0
STEM CELLS 2000;18:40-45
www.StemCells.com
Eliason, Greway, Tare et al. inconvenience of daily or twice-daily injections of the compound could be decreased [6]. Starting with the mutein nartograstim (NTG; [7]), having five amino acid replacements (Thr-1, Leu-3, Gly-4, Pro-5, and Cys-17 replaced by Ala, Thr, Tyr, Arg, and Ser, respectively), various conjugates were prepared with different amounts of polyethylene glycol (PEG). Results of studies on the activities of these conjugates in normal mice demonstrated that in vivo activity increased with increasing total molecular mass of the conjugate. This was in contrast to the activity in vitro, which decreased with increasing PEG. We now report the results of studies on these compounds in normal cynomolgus monkeys, which demonstrate that the relationship between molecular mass and in vivo activity holds for this species as well. The compound having the greatest activity in our preliminary studies, with an average of two 20 kDa PEG units per protein backbone (PEG-NTG), was examined in more detail. The results indicate that significantly elevated neutrophil levels are reached in monkeys following a single administration with doses of PEG-NTG that may be applicable in the clinic. MATERIAL AND METHODS Growth Factor Preparations The PEG conjugates of the G-CSF mutein NTG were prepared as previously described [8]. The average molecular weight of the PEG moieties was 5 kDa, 10 kDa, 12 kDa, and 20 kDa. The conditions used for the PEGylation reaction resulted in an average of two PEG units per NTG molecule. The preparation chosen for detailed studies in monkeys was predominantly the di-PEGylated isoform with 20 kDa PEG (PEG-NTG). Nartograstim, the PEG conjugates, and bacterial recombinant native G-CSF (rHuG-CSF, filgrastim; Amgen; Thousand Oaks, CA) were diluted in phosphate-buffered saline supplemented with 1% normal cynomolgus monkey serum. Concentrations of the stock solutions were calculated on the basis of protein weight rather than total weight, so that doses of the different preparations are directly comparable. Monkey Studies Cynomolgus monkeys (Macaca fascicularis) of either sex were used for these studies. The males weighed 3.0-6.5 kg and the females 2.0-4.5 kg at the start of the experiments. To establish the baseline values for blood cell counts in the experiment testing different doses of PEG-NTG and comparing them with filgrastim, blood was drawn from limb veins on days -7 and 0 before administration of the compounds. Total leukocytes, platelets, and red cells were determined using a Technicon H*1 autoanalyzer (Bayer Sankyo; Tokyo, Japan). Differential counts of neutrophils, eosinophils,
41
basophils, monocytes, and lymphocytes were determined, and absolute numbers of cells of each type were calculated by multiplying the leukocyte numbers by the percentages of each. The PEG-NTG was injected s.c. in a volume of 0.5 ml/kg on day 0 at doses of 10 µg/kg to five monkeys, 30 µg/kg to three monkeys, and 100 µg/kg to three monkeys. Vehicle was also administered to three monkeys. The rHuG-CSF preparation was administered s.c. on days 0-5 to three monkeys. Blood samples were obtained after 8 h and on days 1-7, 10, 14, 21, and 28. In a separate experiment designed to compare the pharmacokinetics of PEG-NTG, NTG, and rHuG-CSF, two monkeys per group were injected with 100 µg/kg s.c. of each factor or vehicle. Blood was drawn before agents were administered (0 h) and after 2, 4, 8, 24, 30, 48, 72, and 96 h. Cell counts were performed on the 0, 24, 48, 72, and 96 h samples. Statistical Analysis Results were analyzed using SigmaStat software (SPSS; Chicago, IL). One-way analysis of variance was used to determine if there were significant differences, and individual comparisons were made using Dunnett’s Method for multiple pairwise comparisons. Differences were considered significant if p values were less than 0.05. Bioassay The bioassay utilizes the G-NFS-60 cell line derived from a murine myeloid leukemia that has been shown to proliferate in the presence of G-CSF [6]. Briefly, 5 × 105 cells per well in a microtiter plate were exposed to sera containing PEG-NTG for 42 h. The cells were then pulsed for 6 h with 1 Ci [3H]-thymidine per well. The cells were harvested, washed and the amount of [3H] incorporation into DNA determined by liquid scintillation counting. The lower limit of quantitation in this assay was 200 pg/ml. Enzyme-Linked Immunosorbent Assay (ELISA) Rabbits were immunized with PEG-NTG, and an ELISA assay was established with the immune serum. Briefly, PEG-NTG in cynomolgus monkey sera was captured on a microtiter plate precoated with a murine monoclonal antibody raised against NTG, and detected with an HRP-conjugated rabbit anti-PEG-NTG polyclonal antibody. The lower limit of quantitation in this assay was 100 pg/ml. Pharmacokinetic Analysis Following s.c. administration of PEG-NTG, pharmacokinetic analyses of the serum concentration data were performed in a noncompartmental approach using DMLIMS (RS1; BBN Software, Cambridge, MA). The area under the serum concentration-time curve (AUC) from 0 to t, was
42
determined using the linear trapezoidal rule. The apparent terminal elimination rate constant, ke, was obtained by least squares analysis. The apparent terminal elimination half-life was calculated from ln(2)/ke. The AUC0→t was extrapolated to infinity, AUC0→∞, by adding the term Ct/ke. The systemic clearance, CL/F was calculated from Dose/AUC0→∞. The absolute bioavailability, when applicable, Fabs, was calculated from ((AUCsc *Doseiv) / (AUCiv *Dosesc)). RESULTS Preliminary studies were performed by administering a single dose of NTG preparations conjugated with different molecular weight PEG moieties to groups of two monkeys each. The total molecular weights of these conjugates were between 30 kDa and 70 kDa [6]. Doses of 3.8 mg/kg or 1.3 mg/kg were given on day 0. These doses were equivalent to those giving elevated neutrophil numbers in mice for up to one week following a single injection. Because monkeys are much more responsive to G-CSF than are mice, these were, in fact, very high doses and resulted in neutrophilia for at least 14 days (data not shown). No clinical signs of toxicity were evident in the monkeys, either systemically or at the site of injection. These results confirmed those obtained with mice, demonstrating that in vivo activities of PEGylated NTG increased
PEGylated G-CSF in Monkeys with increasing total molecular weight of the conjugates [6]. Therefore, the preparation having an average of two units of 20 kDa PEG (PEG-NTG) was selected for further study. The doses of PEG-NTG for the dose-response study were chosen to bracket the total dosage of rHuG-CSF used clinically (5 µg/kg/d) given over a six-day period. Monkeys, thus, were given 10, 30, or 100 µg/kg of PEG-NTG once on day 0, and blood was sampled at various times through day 28. The results are depicted in Figure 1 for WBCs (A), neutrophils (B), monocytes (C), and lymphocytes (D). No significant changes were seen in erythrocytes or platelets (data not shown). Increases in WBCs, neutrophils and monocytes were evident as early as 8 h after injection. The increase in WBCs was about threefold greater than controls in monkeys given 100 µg/kg, primarily reflecting the increases in neutrophils, which were 14-fold elevated, and monocytes, which were increased about threefold. After 24 h, the neutrophil counts in the animals receiving the highest dose reached a peak level about 20-fold above the initial levels and decreased gradually during the next three days. Similar patterns were observed in animals given the lower doses, with the neutrophil levels near baseline by day 4. At the high dose, neutrophil levels began to fall only after the second day and were still above control levels at day 10. The differences in neutrophil levels of monkeys receiving 100
Figure 1. Effect of different doses of PEG-NTG given as single s.c. injections to cynomolgus monkeys on white blood cells (A), neutrophils (B), monocytes (C), and lymphocytes (D). The results are expressed as percentage of initial, where initial is calculated from the mean of the two preinjection values at days -7 and 0.
Eliason, Greway, Tare et al. µg/kg of PEG-NTG compared with monkeys receiving vehicle alone were statistically significant at days 1-4 and day 6. Monocyte levels were decreased in all dose groups after 24 h but remained elevated through day 5 in all monkeys given PEG-NTG. Lymphocyte levels were elevated 8 h after injection in monkeys given the vehicle and were significantly decreased between 30% and 70% of control levels in monkeys given PEG-NTG. Lymphocyte numbers remained depressed at 24 h after injection in monkeys given 30 µg/kg and 100 µg/kg of PEG-NTG but returned to normal thereafter. For comparison, rHuG-CSF was administered on days 0-6 at a dose of 5 µg/kg/d. The results from these monkeys are shown in Figure 2 and are compared with the results from monkeys given 100 µg/kg of PEG-NTG. As with the PEGNTG preparation, neutrophil (Fig. 2A) and monocyte (Fig. 2B) levels were elevated 8 h after the first injection. Lymphocyte levels were depressed at this time by about 40% relative to controls in monkeys given rHuG-CSF (data not shown). In contrast to the situation in monkeys treated with PEG-NTG,
43
neutrophil and monocyte numbers were decreased at the 24 h time point following the first administration of rHuG-CSF and then began to increase with repeated injections. At day 6, the increases in neutrophil numbers were similar to those in monkeys given a single injection of 100 µg/kg PEG-NTG, falling off more rapidly in the rHuG-CSF-treated animals than in the monkeys receiving PEG-NTG. The extended in vivo activity of PEG-NTG suggests that it is cleared less rapidly than the unmodified protein. Therefore, studies were performed to investigate the pharmacokinetics of this preparation. The concentration of PEGNTG in sera from the monkeys used for the experiment depicted in Figures 1 and 2 were determined. As shown in Figure 3, the Tmax was achieved by 8 h post administration in all three dose groups (this was the earliest time point measured). Exposure (AUC) increased more than dose-proportionally with dose-normalized AUCs of 49.7, 205, and 279 ng•h/ml/µg/kg, in monkeys injected with 10, 30, and 100 µg/kg, respectively. The elimination t1/2 was 9 h following the 30 µg/kg dose and 11 h following 100 µg/kg (Table 1). At a dose of 10 µg/kg, the serum concentration of PEG-NTG
Figure 3. Serum pharmacokinetics of PEG-NTG in cynomolgus monkeys following a single s.c. administration of 10, 30, and 100 µg/kg. Serum concentrations were determined by bioassay.
Table 1. Single-dose pharmacokinetics of PEG-NTG in the monkey Dose (µg/kg)
Figure 2. Comparison of the activity of a single dose of PEG-NTG given once to cynomolgus monkeys with that of rHuG-CSF given on days 0-5. Panel A depicts the results for neutrophils and Panel B those for monocytes. The results are expressed as percentage of initial, where initial is calculated from the mean of the two preinjection values at days -7 and 0.
Parameter
10
30
100
t1/2 (h)
ND
9.2
11.0
CL/F (ml/kg/h)
ND
4.9
3.6
AUC (ng•h/ml)
497
6,140
27,900
Cmax (ng/ml)
23.5
341
752
8
8
8
Tmax (h) ND = not determined.
PEGylated G-CSF in Monkeys
44
was at or below the lower limit of quantitation at several time points; therefore, a full pharmacokinetic analysis was not possible. A single-dose PK/PD comparison of PEG-NTG, NTG, and rHuG-CSF was conducted following s.c. administration of 100 µg/kg to two female cynomolgus monkeys for each preparation. This study was performed to investigate the PK/PD profiles and confirm that the PEGylation of NTG did lead to improvements in the disposition and activity profiles. Neutrophil counts were also monitored at 24 h intervals (Fig. 4). The maximum increase in neutrophils was on day 1 for NTG and rHuG-CSF, but 24 h later for PEG-NTG. Return to baseline counts was reached by day 3 for NTG and rHuG-CSF, but was still elevated by two to threefold four days after administration of PEG-NTG. The pharmacokinetics of PEG-NTG demonstrated a much longer time in serum than NTG or rHuG-CSF (Fig. 5). Both the Cmax and the time to reach Cmax were increased as a result of PEGylation. The exposure of PEG-NTG was increased approximately fourfold relative to the non PEGylated G-CSFs (Table 2). The t1/2,app was three to fourfold longer. DISCUSSION We have confirmed in nonhuman primates our previous results in mice that nartograstim preparations having various molecular weight PEG moieties have increasing in vivo activity with increasing overall molecular weight [6]. Our results also demonstrated that prolonged neutrophilia at times in excess of one week can be achieved following a single injection, provided high enough doses are given. The conjugate with an average of two moieties of 20 kDa PEG was selected for further study because it had the highest in vivo activity. The doses of PEG-NTG that were studied in detail were selected to be within a range that should be clinically useful, based upon current recommended doses of non PEGylated G-CSF. As seen previously in murine studies, the duration of neutrophilia and the peak levels of neutrophils were related to the dose of PEG-NTG [6]. The highest dose tested in these normal monkeys, 100 µg/kg, resulted in significant neutrophilia six days after injection. It is well documented that addition of PEG to various proteins, including G-CSF, increases their in vivo half-lives [8-10]. This is due, at least
Figure 4. Neutrophil counts in cynomolgus monkeys following a single s.c. administration of 100 µg/kg PEG-NTG, rHuG-CSF, or unmodified NTG.
Figure 5. Serum pharmacokinetics of PEG-NTG, rHuG-CSF, or unmodified NTG in cynomolgus monkeys following a single s.c. administration of 100 µg/kg. Serum concentrations were determined by ELISA.
in part, to decreased removal by the kidney in a manner proportional to the molecular weight of the compound [11]. Our pharmacokinetic determinations confirmed that the presence of PEG-NTG in vivo is prolonged compared with rHuG-CSF or unmodified NTG. The clearance was about threefold lower, the half-life threefold longer, and the AUC
Table 2. Pharmacokinetic parameters of PEG-NTG, rHuG-CSF, and NTG in female cynomolgus monkeys following a single s.c. administration of 100 µg/kg G-CSF
Tmax (h)
T1/2 (h)
PEG-NTG
8.0
10.1
rHuG-CSF
2.0
3.45
NTG
4.0
3.57
a
t = 48 h
Cmax (ng/ml)
AUC(0-t)a (ng•h/ml)
AUC0-∞ (ng•h/ml)
834
19,730
21,012
18.2
485
5,491
5,492
19.4
387
5,152
5,153
CL/F (ml/kg/h) 4.76
Eliason, Greway, Tare et al. nearly fourfold greater than for either unmodified protein. In addition, the Cmax was approximately twice as high. Interestingly, the Tmax occurred much later, perhaps because of the slower diffusion of the higher molecular weight molecule from the site of injection into the blood stream. Together, these differences in pharmacokinetic parameters result in measurable levels of PEG-NTG 96 h after injection of 100 µg/kg, whereas the unmodified proteins are no longer detectable after 48 h. This compound was not examined in neutropenic monkeys, but it can be expected that the effects would be further prolonged because mature neutrophils and monocytes represent an important clearance pathway for G-CSF [12-15]. Tushinski and Stanley [16] first demonstrated that CSF-1 is selectively destroyed by macrophages in receptor-mediated endocytosis. In a similar process, G-CSF is also destroyed by bone marrow cells in a saturable process [13, 14]. Thus, when numbers of neutrophils are low, clearance should be decreased. Our pharmacokinetic study suggested a saturable effect resulting in greater than dose-proportional increases in AUC and Cmax with increasing dose. Our previous results demonstrated that the activity of the PEGylated NTG molecules in vitro are inversely related to their molecular weights, with the PEG-NTG tested in the
45
current study having only 5% of the activity of the unmodified protein [6]. Although improved pharmacokinetics can provide a mechanism by which this compound can have better in vivo activity, the AUC was increased only fourfold relative to the unmodified protein. Clearly, more work is required at the molecular level to fully understand the disparity between the two assays. One possibility is that binding of the PEGylated molecule to the extracellular matrix in the bone marrow microenvironment may keep it in more prolonged contact with receptors on the target cells than occurs when it is in solution in the tissue culture medium. This could be important if it has a rapid off-rate when bound to the receptor. On another level, it could be that optimal neutrophil production requires a prolonged exposure to G-CSF concentrations above a threshold rather than being strictly dose proportional to total AUC. Whatever the mechanism, PEG-NTG has prolonged activity in monkeys following a single injection. If this translates to the clinical situation, it would provide a form of G-CSF that would require fewer injections for the patients. ACKNOWLEDGMENTS We thank Drs. Erich Platzer, Ueli Gubler, and Gubler William Benjamin for their advice and discussion during the various phases of this work.
R EFERENCES 1 Williams N, Eger RR, Moore MA et al. Differentiation of mouse bone marrow precursor cells into neutrophil granulocytes by an activity separation from WEHI-3 cell-conditioned medium. Differentiation 1978;11:59-63. 2 Nicola NA, Metcalf D, Matsumoto M et al. Purification of a factor inducing differentiation in murine myelomonocytic leukemia cells. Identification as granulocyte colony-stimulating factor. J Biol Chem 1983;258:9017-9023. 3 Welte K, Platzer E, Lu L et al. Purification and biochemical characterization of human pluripotent hematopoietic colonystimulating factor. Proc Natl Acad Sci USA 1985;82:1526-1527. 4 Souza LM, Boone TC, Gabrilove J et al. Recombinant human granulocyte colony-stimulating factor: effects on normal and leukemic myeloid cells. Science 1986;232:61-65. 5 Welte K, Gabrilove J, Bronchud MH et al. Filgrastim (r-metHuG-CSF): the first 10 years. Blood 1996;88:1907-1929. 6 Bowen S, Tare N, Inoue T et al. Relationship between molecular mass and duration of activity of polyethylene glycol conjugated granulocyte colony-stimulating factor mutein. Exp Hematol 1999;27:425-432. 7 Kuga T, Komatsu Y, Yamasaki M et al. Mutagenesis of human granulocyte colony-stimulating factor. Biochem Biophys Res Commun 1989;159:103-111. 8 Yamasaki M, Asano M, Okabe M et al. Modification of recombinant human granulocyte colony-stimulating factor (rhG-CSF) and its derivative ND 28 with polyethylene glycol. J Biochem Tokyo 1994;115:814-819. 9 Tanaka H, Satake-Ishikawa R, Ishikawa M et al. Pharmacokinetics of recombinant human granulocyte colony-stimulating
factor conjugated to polyethylene glycol in rats. Cancer Res 1991;51:3710-3714. 10 Keating MJ, Holmes R, Lerner S et al. L-asparaginase and PEG asparaginase—past, present and future. Leuk Lymphoma 1993;10:153-157. 11 Yamaoka T, Tabata Y, Ikada Y. Distribution and tissue uptake of poly(ethylene glycol) with different molecular weights after intravenous administration to mice. J Pharmaceut Sci 1994;83:601-606. 12 Kuwabara T, Uchimura T, Takai K et al. Saturable uptake of a recombinant human granulocyte colony-stimulating factor derivative, nartograstim, by the bone marrow and spleen of rats in vivo. J Pharmacol Expl Ther 1995;273:1114-1122. 13 Kuwabara T, Kobayashi S, Sugiyama Y. Kinetic analysis of receptor-mediated endocytosis of G-CSF derivative, nartograstim, in rat bone marrow cells. Am J Physiol 1996;271:E73-E84. 14 Kuwabara T, Uchimura T, Kobayashi H et al. Receptor-mediated clearance of G-CSF derivative nartograstim in bone marrow of rats. Am J Physiol 1995;269:E1-E9. 15 Kuwabara T, Kato Y, Kobayashi S et al. Nonlinear pharmacokinetics of a recombinant human granulocyte colony-stimulating factor derivative (nartograstim): species differences among rats, monkeys and humans. J Pharmacol Exp Ther 1994;271:15351543. 16 Tushinski RJ, Oliver IT, Guilbert LJ et al. Survival of mononuclear phagocytes depends on a lineage-specific growth factor that the differentiated cells selectively destroy. Cell 1982;28:71-81.